665 files changed, 14581 insertions, 6698 deletions

diff --git a/Documentation/DocBook/sh.tmpl b/Documentation/DocBook/sh.tmpl
index d858d92cf6d9..4a38f604fa66 100644
--- a/Documentation/DocBook/sh.tmpl
+++ b/Documentation/DocBook/sh.tmpl
@@ -79,10 +79,6 @@
       </sect2>
     </sect1>
   </chapter>
-  <chapter id="clk">
-    <title>Clock Framework Extensions</title>
-!Iinclude/linux/sh_clk.h
-  </chapter>
   <chapter id="mach">
     <title>Machine Specific Interfaces</title>
     <sect1 id="dreamcast">
diff --git a/Documentation/DocBook/uio-howto.tmpl b/Documentation/DocBook/uio-howto.tmpl
index 4d4ce0e61e42..b4665b9c40b0 100644
--- a/Documentation/DocBook/uio-howto.tmpl
+++ b/Documentation/DocBook/uio-howto.tmpl
@@ -16,7 +16,7 @@
         </orgname>
 
         <address>
-           <email>hjk@linutronix.de</email>
+           <email>hjk@hansjkoch.de</email>
         </address>
     </affiliation>
 </author>
@@ -114,7 +114,7 @@ GPL version 2.
 
 <para>If you know of any translations for this document, or you are
 interested in translating it, please email me
-<email>hjk@linutronix.de</email>.
+<email>hjk@hansjkoch.de</email>.
 </para>
 </sect1>
 
@@ -171,7 +171,7 @@ interested in translating it, please email me
 <title>Feedback</title>
         <para>Find something wrong with this document? (Or perhaps something
         right?) I would love to hear from you. Please email me at
-        <email>hjk@linutronix.de</email>.</para>
+        <email>hjk@hansjkoch.de</email>.</para>
 </sect1>
 </chapter>
 
diff --git a/Documentation/arm/OMAP/DSS b/Documentation/arm/OMAP/DSS
index 0af0e9eed5d6..888ae7b83ae4 100644
--- a/Documentation/arm/OMAP/DSS
+++ b/Documentation/arm/OMAP/DSS
@@ -255,9 +255,10 @@ framebuffer parameters.
 Kernel boot arguments
 ---------------------
 
-vram=<size>
-        - Amount of total VRAM to preallocate. For example, "10M". omapfb
-          allocates memory for framebuffers from VRAM.
+vram=<size>[,<physaddr>]
+        - Amount of total VRAM to preallocate and optionally a physical start
+          memory address. For example, "10M". omapfb allocates memory for
+          framebuffers from VRAM.
 
 omapfb.mode=<display>:<mode>[,...]
         - Default video mode for specified displays. For example,
diff --git a/Documentation/development-process/2.Process b/Documentation/development-process/2.Process
index 97726eba6102..911a45186340 100644
--- a/Documentation/development-process/2.Process
+++ b/Documentation/development-process/2.Process
@@ -154,7 +154,7 @@ The stages that a patch goes through are, generally:
    inclusion, it should be accepted by a relevant subsystem maintainer -
    though this acceptance is not a guarantee that the patch will make it
    all the way to the mainline.  The patch will show up in the maintainer's
-   subsystem tree and into the staging trees (described below).  When the
+   subsystem tree and into the -next trees (described below).  When the
    process works, this step leads to more extensive review of the patch and
    the discovery of any problems resulting from the integration of this
    patch with work being done by others.
@@ -236,7 +236,7 @@ finding the right maintainer.  Sending patches directly to Linus is not
 normally the right way to go.
 
 
-2.4: STAGING TREES
+2.4: NEXT TREES
 
 The chain of subsystem trees guides the flow of patches into the kernel,
 but it also raises an interesting question: what if somebody wants to look
@@ -250,7 +250,7 @@ changes land in the mainline kernel.  One could pull changes from all of
 the interesting subsystem trees, but that would be a big and error-prone
 job.
 
-The answer comes in the form of staging trees, where subsystem trees are
+The answer comes in the form of -next trees, where subsystem trees are
 collected for testing and review.  The older of these trees, maintained by
 Andrew Morton, is called "-mm" (for memory management, which is how it got
 started).  The -mm tree integrates patches from a long list of subsystem
@@ -275,7 +275,7 @@ directory at:
 Use of the MMOTM tree is likely to be a frustrating experience, though;
 there is a definite chance that it will not even compile.
 
-The other staging tree, started more recently, is linux-next, maintained by
+The other -next tree, started more recently, is linux-next, maintained by
 Stephen Rothwell.  The linux-next tree is, by design, a snapshot of what
 the mainline is expected to look like after the next merge window closes.
 Linux-next trees are announced on the linux-kernel and linux-next mailing
@@ -303,12 +303,25 @@ volatility of linux-next tends to make it a difficult development target.
 See http://lwn.net/Articles/289013/ for more information on this topic, and
 stay tuned; much is still in flux where linux-next is involved.
 
-Besides the mmotm and linux-next trees, the kernel source tree now contains
-the drivers/staging/ directory and many sub-directories for drivers or
-filesystems that are on their way to being added to the kernel tree
-proper, but they remain in drivers/staging/ while they still need more
-work.
-
+2.4.1: STAGING TREES
+
+The kernel source tree now contains the drivers/staging/ directory, where
+many sub-directories for drivers or filesystems that are on their way to
+being added to the kernel tree live.  They remain in drivers/staging while
+they still need more work; once complete, they can be moved into the
+kernel proper.  This is a way to keep track of drivers that aren't
+up to Linux kernel coding or quality standards, but people may want to use
+them and track development.
+
+Greg Kroah-Hartman currently (as of 2.6.36) maintains the staging tree.
+Drivers that still need work are sent to him, with each driver having
+its own subdirectory in drivers/staging/.  Along with the driver source
+files, a TODO file should be present in the directory as well.  The TODO
+file lists the pending work that the driver needs for acceptance into
+the kernel proper, as well as a list of people that should be Cc'd for any
+patches to the driver.  Staging drivers that don't currently build should
+have their config entries depend upon CONFIG_BROKEN.  Once they can
+be successfully built without outside patches, CONFIG_BROKEN can be removed.
 
 2.5: TOOLS
 
diff --git a/Documentation/fb/00-INDEX b/Documentation/fb/00-INDEX
index a618fd99c9f0..30a70542e823 100644
--- a/Documentation/fb/00-INDEX
+++ b/Documentation/fb/00-INDEX
@@ -4,33 +4,41 @@ please mail me.
                                     Geert Uytterhoeven <geert@linux-m68k.org>
 
 00-INDEX
-        - this file
+        - this file.
 arkfb.txt
         - info on the fbdev driver for ARK Logic chips.
 aty128fb.txt
         - info on the ATI Rage128 frame buffer driver.
 cirrusfb.txt
         - info on the driver for Cirrus Logic chipsets.
+cmap_xfbdev.txt
+        - an introduction to fbdev's cmap structures.
 deferred_io.txt
         - an introduction to deferred IO.
+efifb.txt
+        - info on the EFI platform driver for Intel based Apple computers.
+ep93xx-fb.txt
+        - info on the driver for EP93xx LCD controller.
 fbcon.txt
         - intro to and usage guide for the framebuffer console (fbcon).
 framebuffer.txt
         - introduction to frame buffer devices.
-imacfb.txt
-        - info on the generic EFI platform driver for Intel based Macs.
+gxfb.txt
+        - info on the framebuffer driver for AMD Geode GX2 based processors.
 intel810.txt
         - documentation for the Intel 810/815 framebuffer driver.
 intelfb.txt
         - docs for Intel 830M/845G/852GM/855GM/865G/915G/945G fb driver.
 internals.txt
         - quick overview of frame buffer device internals.
+lxfb.txt
+        - info on the framebuffer driver for AMD Geode LX based processors.
 matroxfb.txt
         - info on the Matrox framebuffer driver for Alpha, Intel and PPC.
+metronomefb.txt
+        - info on the driver for the Metronome display controller.
 modedb.txt
         - info on the video mode database.
-matroxfb.txt
-        - info on the Matrox frame buffer driver.
 pvr2fb.txt
         - info on the PowerVR 2 frame buffer driver.
 pxafb.txt
@@ -39,13 +47,23 @@ s3fb.txt
         - info on the fbdev driver for S3 Trio/Virge chips.
 sa1100fb.txt
         - information about the driver for the SA-1100 LCD controller.
+sh7760fb.txt
+        - info on the SH7760/SH7763 integrated LCDC Framebuffer driver.
 sisfb.txt
         - info on the framebuffer device driver for various SiS chips.
 sstfb.txt
         - info on the frame buffer driver for 3dfx' Voodoo Graphics boards.
 tgafb.txt
-        - info on the TGA (DECChip 21030) frame buffer driver
+        - info on the TGA (DECChip 21030) frame buffer driver.
+tridentfb.txt
+        info on the framebuffer driver for some Trident chip based cards.
+uvesafb.txt
+        - info on the userspace VESA (VBE2+ compliant) frame buffer device.
 vesafb.txt
-        - info on the VESA frame buffer device
+        - info on the VESA frame buffer device.
+viafb.modes
+        - list of modes for VIA Integration Graphic Chip.
+viafb.txt
+        - info on the VIA Integration Graphic Chip console framebuffer driver.
 vt8623fb.txt
         - info on the fb driver for the graphics core in VIA VT8623 chipsets.
diff --git a/Documentation/feature-removal-schedule.txt b/Documentation/feature-removal-schedule.txt
index d8f36f984faa..6c2f55e05f13 100644
--- a/Documentation/feature-removal-schedule.txt
+++ b/Documentation/feature-removal-schedule.txt
@@ -554,3 +554,13 @@ Why:    This is a legacy interface which have been replaced by a more
 Who:    NeilBrown <neilb@suse.de>
 
 ----------------------------
+
+What:   i2c_adapter.id
+When:   June 2011
+Why:    This field is deprecated. I2C device drivers shouldn't change their
+        behavior based on the underlying I2C adapter. Instead, the I2C
+        adapter driver should instantiate the I2C devices and provide the
+        needed platform-specific information.
+Who:    Jean Delvare <khali@linux-fr.org>
+
+----------------------------
diff --git a/Documentation/filesystems/configfs/configfs_example_explicit.c b/Documentation/filesystems/configfs/configfs_example_explicit.c
index d428cc9f07f3..fd53869f5633 100644
--- a/Documentation/filesystems/configfs/configfs_example_explicit.c
+++ b/Documentation/filesystems/configfs/configfs_example_explicit.c
@@ -89,7 +89,7 @@ static ssize_t childless_storeme_write(struct childless *childless,
         char *p = (char *) page;
 
         tmp = simple_strtoul(p, &p, 10);
-        if (!p || (*p && (*p != '\n')))
+        if ((*p != '\0') && (*p != '\n'))
                 return -EINVAL;
 
         if (tmp > INT_MAX)
diff --git a/Documentation/gpio.txt b/Documentation/gpio.txt
index 9633da01ff46..792faa3c06cf 100644
--- a/Documentation/gpio.txt
+++ b/Documentation/gpio.txt
@@ -617,6 +617,16 @@ and have the following read/write attributes:
                 is configured as an output, this value may be written;
                 any nonzero value is treated as high.
 
+                If the pin can be configured as interrupt-generating interrupt
+                and if it has been configured to generate interrupts (see the
+                description of "edge"), you can poll(2) on that file and
+                poll(2) will return whenever the interrupt was triggered. If
+                you use poll(2), set the events POLLPRI and POLLERR. If you
+                use select(2), set the file descriptor in exceptfds. After
+                poll(2) returns, either lseek(2) to the beginning of the sysfs
+                file and read the new value or close the file and re-open it
+                to read the value.
+
         "edge" ... reads as either "none", "rising", "falling", or
                 "both". Write these strings to select the signal edge(s)
                 that will make poll(2) on the "value" file return.
diff --git a/Documentation/hwmon/lm93 b/Documentation/hwmon/lm93
index ac711f357faf..7a10616d0b44 100644
--- a/Documentation/hwmon/lm93
+++ b/Documentation/hwmon/lm93
@@ -11,7 +11,7 @@ Authors:
         Mark M. Hoffman <mhoffman@lightlink.com>
         Ported to 2.6 by Eric J. Bowersox <ericb@aspsys.com>
         Adapted to 2.6.20 by Carsten Emde <ce@osadl.org>
-        Modified for mainline integration by Hans J. Koch <hjk@linutronix.de>
+        Modified for mainline integration by Hans J. Koch <hjk@hansjkoch.de>
 
 Module Parameters
 -----------------
diff --git a/Documentation/hwmon/max6650 b/Documentation/hwmon/max6650
index 8be7beb9e3e8..c565650fcfc6 100644
--- a/Documentation/hwmon/max6650
+++ b/Documentation/hwmon/max6650
@@ -8,7 +8,7 @@ Supported chips:
     Datasheet: http://pdfserv.maxim-ic.com/en/ds/MAX6650-MAX6651.pdf
 
 Authors:
-    Hans J. Koch <hjk@linutronix.de>
+    Hans J. Koch <hjk@hansjkoch.de>
     John Morris <john.morris@spirentcom.com>
     Claus Gindhart <claus.gindhart@kontron.com>
 
diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index 92e83e53148f..cdd2a6e8a3b7 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -2385,6 +2385,11 @@ and is between 256 and 4096 characters. It is defined in the file
                         improve throughput, but will also increase the
                         amount of memory reserved for use by the client.
 
+        swapaccount[=0|1]
+                        [KNL] Enable accounting of swap in memory resource
+                        controller if no parameter or 1 is given or disable
+                        it if 0 is given (See Documentation/cgroups/memory.txt)
+
         swiotlb=        [IA-64] Number of I/O TLB slabs
 
         switches=       [HW,M68k]
diff --git a/Documentation/power/opp.txt b/Documentation/power/opp.txt
index 44d87ad3cea9..cd445582d1f8 100644
--- a/Documentation/power/opp.txt
+++ b/Documentation/power/opp.txt
@@ -37,6 +37,9 @@ Typical usage of the OPP library is as follows:
 SoC framework   -> modifies on required cases certain OPPs      -> OPP layer
                 -> queries to search/retrieve information       ->
 
+Architectures that provide a SoC framework for OPP should select ARCH_HAS_OPP
+to make the OPP layer available.
+
 OPP layer expects each domain to be represented by a unique device pointer. SoC
 framework registers a set of initial OPPs per device with the OPP layer. This
 list is expected to be an optimally small number typically around 5 per device.
diff --git a/Documentation/sh/clk.txt b/Documentation/sh/clk.txt
deleted file mode 100644
index 114b595cfa97..000000000000
--- a/Documentation/sh/clk.txt
+++ /dev/null
@@ -1,32 +0,0 @@
-Clock framework on SuperH architecture
-
-The framework on SH extends existing API by the function clk_set_rate_ex,
-which prototype is as follows:
-
-    clk_set_rate_ex (struct clk *clk, unsigned long rate, int algo_id)
-
-The algo_id parameter is used to specify algorithm used to recalculate clocks,
-adjanced to clock, specified as first argument. It is assumed that algo_id==0
-means no changes to adjanced clock
-
-Internally, the clk_set_rate_ex forwards request to clk->ops->set_rate method,
-if it is present in ops structure. The method should set the clock rate and adjust
-all needed clocks according to the passed algo_id.
-Exact values for algo_id are machine-dependent. For the sh7722, the following
-values are defined:
-
-        NO_CHANGE       = 0,
-        IUS_N1_N1,      /* I:U = N:1, U:Sh = N:1 */
-        IUS_322,        /* I:U:Sh = 3:2:2        */
-        IUS_522,        /* I:U:Sh = 5:2:2        */
-        IUS_N11,        /* I:U:Sh = N:1:1        */
-        SB_N1,          /* Sh:B = N:1            */
-        SB3_N1,         /* Sh:B3 = N:1           */
-        SB3_32,         /* Sh:B3 = 3:2           */
-        SB3_43,         /* Sh:B3 = 4:3           */
-        SB3_54,         /* Sh:B3 = 5:4           */
-        BP_N1,          /* B:P   = N:1           */
-        IP_N1           /* I:P   = N:1           */
-
-Each of these constants means relation between clocks that can be set via the FRQCR
-register
diff --git a/MAINTAINERS b/MAINTAINERS
index 88b74a75d932..b3be8b3d0437 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -945,7 +945,7 @@ M:	Magnus Damm <magnus.damm@gmail.com>
 L:      linux-sh@vger.kernel.org
 W:      http://oss.renesas.com
 Q:      http://patchwork.kernel.org/project/linux-sh/list/
-T:      git git://git.kernel.org/pub/scm/linux/kernel/git/lethal/genesis-2.6.git
+T:      git git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6.git rmobile-latest
 S:      Supported
 F:      arch/arm/mach-shmobile/
 F:      drivers/sh/
@@ -1359,7 +1359,7 @@ F:	include/net/bluetooth/
 
 BONDING DRIVER
 M:      Jay Vosburgh <fubar@us.ibm.com>
-L:      bonding-devel@lists.sourceforge.net
+L:      netdev@vger.kernel.org
 W:      http://sourceforge.net/projects/bonding/
 S:      Supported
 F:      drivers/net/bonding/
@@ -1829,6 +1829,13 @@ W:	http://www.chelsio.com
 S:      Supported
 F:      drivers/net/cxgb4vf/
 
+STMMAC ETHERNET DRIVER
+M:      Giuseppe Cavallaro <peppe.cavallaro@st.com>
+L:      netdev@vger.kernel.org
+W:      http://www.stlinux.com
+S:      Supported
+F:      drivers/net/stmmac/
+
 CYBERPRO FB DRIVER
 M:      Russell King <linux@arm.linux.org.uk>
 L:      linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
@@ -2008,6 +2015,7 @@ F:	drivers/hwmon/dme1737.c
 DOCBOOK FOR DOCUMENTATION
 M:      Randy Dunlap <rdunlap@xenotime.net>
 S:      Maintained
+F:      scripts/kernel-doc
 
 DOCKING STATION DRIVER
 M:      Shaohua Li <shaohua.li@intel.com>
@@ -2018,6 +2026,7 @@ F:	drivers/acpi/dock.c
 DOCUMENTATION
 M:      Randy Dunlap <rdunlap@xenotime.net>
 L:      linux-doc@vger.kernel.org
+T:      quilt oss.oracle.com/~rdunlap/kernel-doc-patches/current/
 S:      Maintained
 F:      Documentation/
 
@@ -2435,9 +2444,12 @@ F:	drivers/net/wan/sdla.c
 FRAMEBUFFER LAYER
 L:      linux-fbdev@vger.kernel.org
 W:      http://linux-fbdev.sourceforge.net/
+Q:      http://patchwork.kernel.org/project/linux-fbdev/list/
+T:      git git://git.kernel.org/pub/scm/linux/kernel/git/lethal/fbdev-2.6.git
 S:      Orphan
 F:      Documentation/fb/
-F:      drivers/video/fb*
+F:      drivers/video/
+F:      include/video/
 F:      include/linux/fb.h
 
 FREESCALE DMA DRIVER
@@ -5705,7 +5717,7 @@ M:	Paul Mundt <lethal@linux-sh.org>
 L:      linux-sh@vger.kernel.org
 W:      http://www.linux-sh.org
 Q:      http://patchwork.kernel.org/project/linux-sh/list/
-T:      git git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6.git
+T:      git git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6.git sh-latest
 S:      Supported
 F:      Documentation/sh/
 F:      arch/sh/
@@ -5827,6 +5839,8 @@ M:	Chris Metcalf <cmetcalf@tilera.com>
 W:      http://www.tilera.com/scm/
 S:      Supported
 F:      arch/tile/
+F:      drivers/char/hvc_tile.c
+F:      drivers/net/tile/
 
 TLAN NETWORK DRIVER
 M:      Samuel Chessman <chessman@tux.org>
diff --git a/Makefile b/Makefile
index 6619720f50dd..b31d21377e4c 100644
--- a/Makefile
+++ b/Makefile
@@ -1,7 +1,7 @@
 VERSION = 2
 PATCHLEVEL = 6
 SUBLEVEL = 37
-EXTRAVERSION = -rc1
+EXTRAVERSION = -rc3
 NAME = Flesh-Eating Bats with Fangs
 
 # *DOCUMENTATION*
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index bd620a481bee..49778bb43782 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -647,7 +647,7 @@ config ARCH_S3C2410
         select ARCH_HAS_CPUFREQ
         select HAVE_CLK
         select ARCH_USES_GETTIMEOFFSET
-        select HAVE_S3C2410_I2C
+        select HAVE_S3C2410_I2C if I2C
         help
           Samsung S3C2410X CPU based systems, such as the Simtec Electronics
           BAST (<http://www.simtec.co.uk/products/EB110ITX/>), the IPAQ 1940 or
@@ -677,8 +677,8 @@ config ARCH_S3C64XX
         select S3C_DEV_NAND
         select USB_ARCH_HAS_OHCI
         select SAMSUNG_GPIOLIB_4BIT
-        select HAVE_S3C2410_I2C
-        select HAVE_S3C2410_WATCHDOG
+        select HAVE_S3C2410_I2C if I2C
+        select HAVE_S3C2410_WATCHDOG if WATCHDOG
         help
           Samsung S3C64XX series based systems
 
@@ -687,10 +687,10 @@ config ARCH_S5P64X0
         select CPU_V6
         select GENERIC_GPIO
         select HAVE_CLK
-        select HAVE_S3C2410_WATCHDOG
+        select HAVE_S3C2410_WATCHDOG if WATCHDOG
         select ARCH_USES_GETTIMEOFFSET
-        select HAVE_S3C2410_I2C
-        select HAVE_S3C_RTC
+        select HAVE_S3C2410_I2C if I2C
+        select HAVE_S3C_RTC if RTC_CLASS
         help
           Samsung S5P64X0 CPU based systems, such as the Samsung SMDK6440,
           SMDK6450.
@@ -701,7 +701,7 @@ config ARCH_S5P6442
         select GENERIC_GPIO
         select HAVE_CLK
         select ARCH_USES_GETTIMEOFFSET
-        select HAVE_S3C2410_WATCHDOG
+        select HAVE_S3C2410_WATCHDOG if WATCHDOG
         help
           Samsung S5P6442 CPU based systems
 
@@ -712,9 +712,9 @@ config ARCH_S5PC100
         select CPU_V7
         select ARM_L1_CACHE_SHIFT_6
         select ARCH_USES_GETTIMEOFFSET
-        select HAVE_S3C2410_I2C
-        select HAVE_S3C_RTC
-        select HAVE_S3C2410_WATCHDOG
+        select HAVE_S3C2410_I2C if I2C
+        select HAVE_S3C_RTC if RTC_CLASS
+        select HAVE_S3C2410_WATCHDOG if WATCHDOG
         help
           Samsung S5PC100 series based systems
 
@@ -727,9 +727,9 @@ config ARCH_S5PV210
         select ARM_L1_CACHE_SHIFT_6
         select ARCH_HAS_CPUFREQ
         select ARCH_USES_GETTIMEOFFSET
-        select HAVE_S3C2410_I2C
-        select HAVE_S3C_RTC
-        select HAVE_S3C2410_WATCHDOG
+        select HAVE_S3C2410_I2C if I2C
+        select HAVE_S3C_RTC if RTC_CLASS
+        select HAVE_S3C2410_WATCHDOG if WATCHDOG
         help
           Samsung S5PV210/S5PC110 series based systems
 
@@ -740,9 +740,9 @@ config ARCH_S5PV310
         select GENERIC_GPIO
         select HAVE_CLK
         select GENERIC_CLOCKEVENTS
-        select HAVE_S3C_RTC
-        select HAVE_S3C2410_I2C
-        select HAVE_S3C2410_WATCHDOG
+        select HAVE_S3C_RTC if RTC_CLASS
+        select HAVE_S3C2410_I2C if I2C
+        select HAVE_S3C2410_WATCHDOG if WATCHDOG
         help
           Samsung S5PV310 series based systems
 
@@ -1206,10 +1206,11 @@ config SMP
         depends on EXPERIMENTAL
         depends on GENERIC_CLOCKEVENTS
         depends on REALVIEW_EB_ARM11MP || REALVIEW_EB_A9MP || \
-                 MACH_REALVIEW_PB11MP || MACH_REALVIEW_PBX || ARCH_OMAP4 ||\
-                 ARCH_S5PV310 || ARCH_TEGRA || ARCH_U8500 || ARCH_VEXPRESS_CA9X4
+                 MACH_REALVIEW_PB11MP || MACH_REALVIEW_PBX || ARCH_OMAP4 || \
+                 ARCH_S5PV310 || ARCH_TEGRA || ARCH_U8500 || ARCH_VEXPRESS_CA9X4 || \
+                 ARCH_MSM_SCORPIONMP
         select USE_GENERIC_SMP_HELPERS
-        select HAVE_ARM_SCU
+        select HAVE_ARM_SCU if !ARCH_MSM_SCORPIONMP
         help
           This enables support for systems with more than one CPU. If you have
           a system with only one CPU, like most personal computers, say N. If
@@ -1284,6 +1285,7 @@ config NR_CPUS
 config HOTPLUG_CPU
         bool "Support for hot-pluggable CPUs (EXPERIMENTAL)"
         depends on SMP && HOTPLUG && EXPERIMENTAL
+        depends on !ARCH_MSM
         help
           Say Y here to experiment with turning CPUs off and on.  CPUs
           can be controlled through /sys/devices/system/cpu.
@@ -1292,7 +1294,7 @@ config LOCAL_TIMERS
         bool "Use local timer interrupts"
         depends on SMP
         default y
-        select HAVE_ARM_TWD
+        select HAVE_ARM_TWD if !ARCH_MSM_SCORPIONMP
         help
           Enable support for local timers on SMP platforms, rather then the
           legacy IPI broadcast method.  Local timers allows the system
diff --git a/arch/arm/boot/compressed/head.S b/arch/arm/boot/compressed/head.S
index 6825c34646d4..9be21ba648cd 100644
--- a/arch/arm/boot/compressed/head.S
+++ b/arch/arm/boot/compressed/head.S
@@ -1084,6 +1084,6 @@ memdump:	mov	r12, r0
 reloc_end:
 
                 .align
-                .section ".stack", "w"
+                .section ".stack", "aw", %nobits
 user_stack:     .space  4096
 user_stack_end:
diff --git a/arch/arm/boot/compressed/vmlinux.lds.in b/arch/arm/boot/compressed/vmlinux.lds.in
index d08168941bd6..366a924019ac 100644
--- a/arch/arm/boot/compressed/vmlinux.lds.in
+++ b/arch/arm/boot/compressed/vmlinux.lds.in
@@ -57,7 +57,7 @@ SECTIONS
   .bss                  : { *(.bss) }
   _end = .;
 
-  .stack (NOLOAD)       : { *(.stack) }
+  .stack                : { *(.stack) }
 
   .stab 0               : { *(.stab) }
   .stabstr 0            : { *(.stabstr) }
diff --git a/arch/arm/include/asm/assembler.h b/arch/arm/include/asm/assembler.h
index 062b58c029ab..749bb6622404 100644
--- a/arch/arm/include/asm/assembler.h
+++ b/arch/arm/include/asm/assembler.h
@@ -238,7 +238,7 @@
         @ Slightly optimised to avoid incrementing the pointer twice
         usraccoff \instr, \reg, \ptr, \inc, 0, \cond, \abort
         .if     \rept == 2
-        usraccoff \instr, \reg, \ptr, \inc, 4, \cond, \abort
+        usraccoff \instr, \reg, \ptr, \inc, \inc, \cond, \abort
         .endif
 
         add\cond \ptr, #\rept * \inc
diff --git a/arch/arm/include/asm/mmu.h b/arch/arm/include/asm/mmu.h
index 68870c776671..b4ffe9d5b526 100644
--- a/arch/arm/include/asm/mmu.h
+++ b/arch/arm/include/asm/mmu.h
@@ -13,6 +13,10 @@ typedef struct {
 
 #ifdef CONFIG_CPU_HAS_ASID
 #define ASID(mm)        ((mm)->context.id & 255)
+
+/* init_mm.context.id_lock should be initialized. */
+#define INIT_MM_CONTEXT(name)                                                 \
+        .context.id_lock    = __SPIN_LOCK_UNLOCKED(name.context.id_lock),
 #else
 #define ASID(mm)        (0)
 #endif
diff --git a/arch/arm/include/asm/pgtable.h b/arch/arm/include/asm/pgtable.h
index b155414192da..53d1d5deb111 100644
--- a/arch/arm/include/asm/pgtable.h
+++ b/arch/arm/include/asm/pgtable.h
@@ -374,6 +374,9 @@ static inline pte_t *pmd_page_vaddr(pmd_t pmd)
 
 #define pmd_page(pmd)           pfn_to_page(__phys_to_pfn(pmd_val(pmd)))
 
+/* we don't need complex calculations here as the pmd is folded into the pgd */
+#define pmd_addr_end(addr,end)  (end)
+
 /*
  * Conversion functions: convert a page and protection to a page entry,
  * and a page entry and page directory to the page they refer to.
diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c
index 07a50357492a..421a4bb88fed 100644
--- a/arch/arm/kernel/perf_event.c
+++ b/arch/arm/kernel/perf_event.c
@@ -4,9 +4,7 @@
  * ARM performance counter support.
  *
  * Copyright (C) 2009 picoChip Designs, Ltd., Jamie Iles
- *
- * ARMv7 support: Jean Pihet <jpihet@mvista.com>
- * 2010 (c) MontaVista Software, LLC.
+ * Copyright (C) 2010 ARM Ltd., Will Deacon <will.deacon@arm.com>
  *
  * This code is based on the sparc64 perf event code, which is in turn based
  * on the x86 code. Callchain code is based on the ARM OProfile backtrace
@@ -69,29 +67,23 @@ struct cpu_hw_events {
 };
 DEFINE_PER_CPU(struct cpu_hw_events, cpu_hw_events);
 
-/* PMU names. */
-static const char *arm_pmu_names[] = {
-        [ARM_PERF_PMU_ID_XSCALE1] = "xscale1",
-        [ARM_PERF_PMU_ID_XSCALE2] = "xscale2",
-        [ARM_PERF_PMU_ID_V6]      = "v6",
-        [ARM_PERF_PMU_ID_V6MP]    = "v6mpcore",
-        [ARM_PERF_PMU_ID_CA8]     = "ARMv7 Cortex-A8",
-        [ARM_PERF_PMU_ID_CA9]     = "ARMv7 Cortex-A9",
-};
-
 struct arm_pmu {
         enum arm_perf_pmu_ids id;
+        const char      *name;
         irqreturn_t     (*handle_irq)(int irq_num, void *dev);
         void            (*enable)(struct hw_perf_event *evt, int idx);
         void            (*disable)(struct hw_perf_event *evt, int idx);
-        int             (*event_map)(int evt);
-        u64             (*raw_event)(u64);
         int             (*get_event_idx)(struct cpu_hw_events *cpuc,
                                          struct hw_perf_event *hwc);
         u32             (*read_counter)(int idx);
         void            (*write_counter)(int idx, u32 val);
         void            (*start)(void);
         void            (*stop)(void);
+        const unsigned  (*cache_map)[PERF_COUNT_HW_CACHE_MAX]
+                                    [PERF_COUNT_HW_CACHE_OP_MAX]
+                                    [PERF_COUNT_HW_CACHE_RESULT_MAX];
+        const unsigned  (*event_map)[PERF_COUNT_HW_MAX];
+        u32             raw_event_mask;
         int             num_events;
         u64             max_period;
 };
@@ -136,10 +128,6 @@ EXPORT_SYMBOL_GPL(perf_num_counters);
 
 #define CACHE_OP_UNSUPPORTED            0xFFFF
 
-static unsigned armpmu_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
-                                     [PERF_COUNT_HW_CACHE_OP_MAX]
-                                     [PERF_COUNT_HW_CACHE_RESULT_MAX];
-
 static int
 armpmu_map_cache_event(u64 config)
 {
@@ -157,7 +145,7 @@ armpmu_map_cache_event(u64 config)
         if (cache_result >= PERF_COUNT_HW_CACHE_RESULT_MAX)
                 return -EINVAL;
 
-        ret = (int)armpmu_perf_cache_map[cache_type][cache_op][cache_result];
+        ret = (int)(*armpmu->cache_map)[cache_type][cache_op][cache_result];
 
         if (ret == CACHE_OP_UNSUPPORTED)
                 return -ENOENT;
@@ -166,6 +154,19 @@ armpmu_map_cache_event(u64 config)
 }
 
 static int
+armpmu_map_event(u64 config)
+{
+        int mapping = (*armpmu->event_map)[config];
+        return mapping == HW_OP_UNSUPPORTED ? -EOPNOTSUPP : mapping;
+}
+
+static int
+armpmu_map_raw_event(u64 config)
+{
+        return (int)(config & armpmu->raw_event_mask);
+}
+
+static int
 armpmu_event_set_period(struct perf_event *event,
                         struct hw_perf_event *hwc,
                         int idx)
@@ -458,11 +459,11 @@ __hw_perf_event_init(struct perf_event *event)
 
         /* Decode the generic type into an ARM event identifier. */
         if (PERF_TYPE_HARDWARE == event->attr.type) {
-                mapping = armpmu->event_map(event->attr.config);
+                mapping = armpmu_map_event(event->attr.config);
         } else if (PERF_TYPE_HW_CACHE == event->attr.type) {
                 mapping = armpmu_map_cache_event(event->attr.config);
         } else if (PERF_TYPE_RAW == event->attr.type) {
-                mapping = armpmu->raw_event(event->attr.config);
+                mapping = armpmu_map_raw_event(event->attr.config);
         } else {
                 pr_debug("event type %x not supported\n", event->attr.type);
                 return -EOPNOTSUPP;
@@ -603,2366 +604,10 @@ static struct pmu pmu = {
         .read           = armpmu_read,
 };
 
-/*
- * ARMv6 Performance counter handling code.
- *
- * ARMv6 has 2 configurable performance counters and a single cycle counter.
- * They all share a single reset bit but can be written to zero so we can use
- * that for a reset.
- *
- * The counters can't be individually enabled or disabled so when we remove
- * one event and replace it with another we could get spurious counts from the
- * wrong event. However, we can take advantage of the fact that the
- * performance counters can export events to the event bus, and the event bus
- * itself can be monitored. This requires that we *don't* export the events to
- * the event bus. The procedure for disabling a configurable counter is:
- *      - change the counter to count the ETMEXTOUT[0] signal (0x20). This
- *        effectively stops the counter from counting.
- *      - disable the counter's interrupt generation (each counter has it's
- *        own interrupt enable bit).
- * Once stopped, the counter value can be written as 0 to reset.
- *
- * To enable a counter:
- *      - enable the counter's interrupt generation.
- *      - set the new event type.
- *
- * Note: the dedicated cycle counter only counts cycles and can't be
- * enabled/disabled independently of the others. When we want to disable the
- * cycle counter, we have to just disable the interrupt reporting and start
- * ignoring that counter. When re-enabling, we have to reset the value and
- * enable the interrupt.
- */
-
-enum armv6_perf_types {
-        ARMV6_PERFCTR_ICACHE_MISS           = 0x0,
-        ARMV6_PERFCTR_IBUF_STALL            = 0x1,
-        ARMV6_PERFCTR_DDEP_STALL            = 0x2,
-        ARMV6_PERFCTR_ITLB_MISS             = 0x3,
-        ARMV6_PERFCTR_DTLB_MISS             = 0x4,
-        ARMV6_PERFCTR_BR_EXEC               = 0x5,
-        ARMV6_PERFCTR_BR_MISPREDICT         = 0x6,
-        ARMV6_PERFCTR_INSTR_EXEC            = 0x7,
-        ARMV6_PERFCTR_DCACHE_HIT            = 0x9,
-        ARMV6_PERFCTR_DCACHE_ACCESS         = 0xA,
-        ARMV6_PERFCTR_DCACHE_MISS           = 0xB,
-        ARMV6_PERFCTR_DCACHE_WBACK          = 0xC,
-        ARMV6_PERFCTR_SW_PC_CHANGE          = 0xD,
-        ARMV6_PERFCTR_MAIN_TLB_MISS         = 0xF,
-        ARMV6_PERFCTR_EXPL_D_ACCESS         = 0x10,
-        ARMV6_PERFCTR_LSU_FULL_STALL        = 0x11,
-        ARMV6_PERFCTR_WBUF_DRAINED          = 0x12,
-        ARMV6_PERFCTR_CPU_CYCLES            = 0xFF,
-        ARMV6_PERFCTR_NOP                   = 0x20,
-};
-
-enum armv6_counters {
-        ARMV6_CYCLE_COUNTER = 1,
-        ARMV6_COUNTER0,
-        ARMV6_COUNTER1,
-};
-
-/*
- * The hardware events that we support. We do support cache operations but
- * we have harvard caches and no way to combine instruction and data
- * accesses/misses in hardware.
- */
-static const unsigned armv6_perf_map[PERF_COUNT_HW_MAX] = {
-        [PERF_COUNT_HW_CPU_CYCLES]          = ARMV6_PERFCTR_CPU_CYCLES,
-        [PERF_COUNT_HW_INSTRUCTIONS]        = ARMV6_PERFCTR_INSTR_EXEC,
-        [PERF_COUNT_HW_CACHE_REFERENCES]    = HW_OP_UNSUPPORTED,
-        [PERF_COUNT_HW_CACHE_MISSES]        = HW_OP_UNSUPPORTED,
-        [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = ARMV6_PERFCTR_BR_EXEC,
-        [PERF_COUNT_HW_BRANCH_MISSES]       = ARMV6_PERFCTR_BR_MISPREDICT,
-        [PERF_COUNT_HW_BUS_CYCLES]          = HW_OP_UNSUPPORTED,
-};
-
-static const unsigned armv6_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
-                                          [PERF_COUNT_HW_CACHE_OP_MAX]
-                                          [PERF_COUNT_HW_CACHE_RESULT_MAX] = {
-        [C(L1D)] = {
-                /*
-                 * The performance counters don't differentiate between read
-                 * and write accesses/misses so this isn't strictly correct,
-                 * but it's the best we can do. Writes and reads get
-                 * combined.
-                 */
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = ARMV6_PERFCTR_DCACHE_ACCESS,
-                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_DCACHE_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = ARMV6_PERFCTR_DCACHE_ACCESS,
-                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_DCACHE_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(L1I)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_ICACHE_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_ICACHE_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(LL)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(DTLB)] = {
-                /*
-                 * The ARM performance counters can count micro DTLB misses,
-                 * micro ITLB misses and main TLB misses. There isn't an event
-                 * for TLB misses, so use the micro misses here and if users
-                 * want the main TLB misses they can use a raw counter.
-                 */
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_DTLB_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_DTLB_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(ITLB)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_ITLB_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_ITLB_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(BPU)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-};
-
-enum armv6mpcore_perf_types {
-        ARMV6MPCORE_PERFCTR_ICACHE_MISS     = 0x0,
-        ARMV6MPCORE_PERFCTR_IBUF_STALL      = 0x1,
-        ARMV6MPCORE_PERFCTR_DDEP_STALL      = 0x2,
-        ARMV6MPCORE_PERFCTR_ITLB_MISS       = 0x3,
-        ARMV6MPCORE_PERFCTR_DTLB_MISS       = 0x4,
-        ARMV6MPCORE_PERFCTR_BR_EXEC         = 0x5,
-        ARMV6MPCORE_PERFCTR_BR_NOTPREDICT   = 0x6,
-        ARMV6MPCORE_PERFCTR_BR_MISPREDICT   = 0x7,
-        ARMV6MPCORE_PERFCTR_INSTR_EXEC      = 0x8,
-        ARMV6MPCORE_PERFCTR_DCACHE_RDACCESS = 0xA,
-        ARMV6MPCORE_PERFCTR_DCACHE_RDMISS   = 0xB,
-        ARMV6MPCORE_PERFCTR_DCACHE_WRACCESS = 0xC,
-        ARMV6MPCORE_PERFCTR_DCACHE_WRMISS   = 0xD,
-        ARMV6MPCORE_PERFCTR_DCACHE_EVICTION = 0xE,
-        ARMV6MPCORE_PERFCTR_SW_PC_CHANGE    = 0xF,
-        ARMV6MPCORE_PERFCTR_MAIN_TLB_MISS   = 0x10,
-        ARMV6MPCORE_PERFCTR_EXPL_MEM_ACCESS = 0x11,
-        ARMV6MPCORE_PERFCTR_LSU_FULL_STALL  = 0x12,
-        ARMV6MPCORE_PERFCTR_WBUF_DRAINED    = 0x13,
-        ARMV6MPCORE_PERFCTR_CPU_CYCLES      = 0xFF,
-};
-
-/*
- * The hardware events that we support. We do support cache operations but
- * we have harvard caches and no way to combine instruction and data
- * accesses/misses in hardware.
- */
-static const unsigned armv6mpcore_perf_map[PERF_COUNT_HW_MAX] = {
-        [PERF_COUNT_HW_CPU_CYCLES]          = ARMV6MPCORE_PERFCTR_CPU_CYCLES,
-        [PERF_COUNT_HW_INSTRUCTIONS]        = ARMV6MPCORE_PERFCTR_INSTR_EXEC,
-        [PERF_COUNT_HW_CACHE_REFERENCES]    = HW_OP_UNSUPPORTED,
-        [PERF_COUNT_HW_CACHE_MISSES]        = HW_OP_UNSUPPORTED,
-        [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = ARMV6MPCORE_PERFCTR_BR_EXEC,
-        [PERF_COUNT_HW_BRANCH_MISSES]       = ARMV6MPCORE_PERFCTR_BR_MISPREDICT,
-        [PERF_COUNT_HW_BUS_CYCLES]          = HW_OP_UNSUPPORTED,
-};
-
-static const unsigned armv6mpcore_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
-                                        [PERF_COUNT_HW_CACHE_OP_MAX]
-                                        [PERF_COUNT_HW_CACHE_RESULT_MAX] = {
-        [C(L1D)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]  =
-                                ARMV6MPCORE_PERFCTR_DCACHE_RDACCESS,
-                        [C(RESULT_MISS)]    =
-                                ARMV6MPCORE_PERFCTR_DCACHE_RDMISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]  =
-                                ARMV6MPCORE_PERFCTR_DCACHE_WRACCESS,
-                        [C(RESULT_MISS)]    =
-                                ARMV6MPCORE_PERFCTR_DCACHE_WRMISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(L1I)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_ICACHE_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_ICACHE_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(LL)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(DTLB)] = {
-                /*
-                 * The ARM performance counters can count micro DTLB misses,
-                 * micro ITLB misses and main TLB misses. There isn't an event
-                 * for TLB misses, so use the micro misses here and if users
-                 * want the main TLB misses they can use a raw counter.
-                 */
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_DTLB_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_DTLB_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(ITLB)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_ITLB_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_ITLB_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(BPU)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
-                },
-        },
-};
-
-static inline unsigned long
-armv6_pmcr_read(void)
-{
-        u32 val;
-        asm volatile("mrc   p15, 0, %0, c15, c12, 0" : "=r"(val));
-        return val;
-}
-
-static inline void
-armv6_pmcr_write(unsigned long val)
-{
-        asm volatile("mcr   p15, 0, %0, c15, c12, 0" : : "r"(val));
-}
-
-#define ARMV6_PMCR_ENABLE               (1 << 0)
-#define ARMV6_PMCR_CTR01_RESET          (1 << 1)
-#define ARMV6_PMCR_CCOUNT_RESET         (1 << 2)
-#define ARMV6_PMCR_CCOUNT_DIV           (1 << 3)
-#define ARMV6_PMCR_COUNT0_IEN           (1 << 4)
-#define ARMV6_PMCR_COUNT1_IEN           (1 << 5)
-#define ARMV6_PMCR_CCOUNT_IEN           (1 << 6)
-#define ARMV6_PMCR_COUNT0_OVERFLOW      (1 << 8)
-#define ARMV6_PMCR_COUNT1_OVERFLOW      (1 << 9)
-#define ARMV6_PMCR_CCOUNT_OVERFLOW      (1 << 10)
-#define ARMV6_PMCR_EVT_COUNT0_SHIFT     20
-#define ARMV6_PMCR_EVT_COUNT0_MASK      (0xFF << ARMV6_PMCR_EVT_COUNT0_SHIFT)
-#define ARMV6_PMCR_EVT_COUNT1_SHIFT     12
-#define ARMV6_PMCR_EVT_COUNT1_MASK      (0xFF << ARMV6_PMCR_EVT_COUNT1_SHIFT)
-
-#define ARMV6_PMCR_OVERFLOWED_MASK \
-        (ARMV6_PMCR_COUNT0_OVERFLOW | ARMV6_PMCR_COUNT1_OVERFLOW | \
-         ARMV6_PMCR_CCOUNT_OVERFLOW)
-
-static inline int
-armv6_pmcr_has_overflowed(unsigned long pmcr)
-{
-        return (pmcr & ARMV6_PMCR_OVERFLOWED_MASK);
-}
-
-static inline int
-armv6_pmcr_counter_has_overflowed(unsigned long pmcr,
-                                  enum armv6_counters counter)
-{
-        int ret = 0;
-
-        if (ARMV6_CYCLE_COUNTER == counter)
-                ret = pmcr & ARMV6_PMCR_CCOUNT_OVERFLOW;
-        else if (ARMV6_COUNTER0 == counter)
-                ret = pmcr & ARMV6_PMCR_COUNT0_OVERFLOW;
-        else if (ARMV6_COUNTER1 == counter)
-                ret = pmcr & ARMV6_PMCR_COUNT1_OVERFLOW;
-        else
-                WARN_ONCE(1, "invalid counter number (%d)\n", counter);
-
-        return ret;
-}
-
-static inline u32
-armv6pmu_read_counter(int counter)
-{
-        unsigned long value = 0;
-
-        if (ARMV6_CYCLE_COUNTER == counter)
-                asm volatile("mrc   p15, 0, %0, c15, c12, 1" : "=r"(value));
-        else if (ARMV6_COUNTER0 == counter)
-                asm volatile("mrc   p15, 0, %0, c15, c12, 2" : "=r"(value));
-        else if (ARMV6_COUNTER1 == counter)
-                asm volatile("mrc   p15, 0, %0, c15, c12, 3" : "=r"(value));
-        else
-                WARN_ONCE(1, "invalid counter number (%d)\n", counter);
-
-        return value;
-}
-
-static inline void
-armv6pmu_write_counter(int counter,
-                       u32 value)
-{
-        if (ARMV6_CYCLE_COUNTER == counter)
-                asm volatile("mcr   p15, 0, %0, c15, c12, 1" : : "r"(value));
-        else if (ARMV6_COUNTER0 == counter)
-                asm volatile("mcr   p15, 0, %0, c15, c12, 2" : : "r"(value));
-        else if (ARMV6_COUNTER1 == counter)
-                asm volatile("mcr   p15, 0, %0, c15, c12, 3" : : "r"(value));
-        else
-                WARN_ONCE(1, "invalid counter number (%d)\n", counter);
-}
-
-void
-armv6pmu_enable_event(struct hw_perf_event *hwc,
-                      int idx)
-{
-        unsigned long val, mask, evt, flags;
-
-        if (ARMV6_CYCLE_COUNTER == idx) {
-                mask    = 0;
-                evt     = ARMV6_PMCR_CCOUNT_IEN;
-        } else if (ARMV6_COUNTER0 == idx) {
-                mask    = ARMV6_PMCR_EVT_COUNT0_MASK;
-                evt     = (hwc->config_base << ARMV6_PMCR_EVT_COUNT0_SHIFT) |
-                          ARMV6_PMCR_COUNT0_IEN;
-        } else if (ARMV6_COUNTER1 == idx) {
-                mask    = ARMV6_PMCR_EVT_COUNT1_MASK;
-                evt     = (hwc->config_base << ARMV6_PMCR_EVT_COUNT1_SHIFT) |
-                          ARMV6_PMCR_COUNT1_IEN;
-        } else {
-                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
-                return;
-        }
-
-        /*
-         * Mask out the current event and set the counter to count the event
-         * that we're interested in.
-         */
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = armv6_pmcr_read();
-        val &= ~mask;
-        val |= evt;
-        armv6_pmcr_write(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static irqreturn_t
-armv6pmu_handle_irq(int irq_num,
-                    void *dev)
-{
-        unsigned long pmcr = armv6_pmcr_read();
-        struct perf_sample_data data;
-        struct cpu_hw_events *cpuc;
-        struct pt_regs *regs;
-        int idx;
-
-        if (!armv6_pmcr_has_overflowed(pmcr))
-                return IRQ_NONE;
-
-        regs = get_irq_regs();
-
-        /*
-         * The interrupts are cleared by writing the overflow flags back to
-         * the control register. All of the other bits don't have any effect
-         * if they are rewritten, so write the whole value back.
-         */
-        armv6_pmcr_write(pmcr);
-
-        perf_sample_data_init(&data, 0);
-
-        cpuc = &__get_cpu_var(cpu_hw_events);
-        for (idx = 0; idx <= armpmu->num_events; ++idx) {
-                struct perf_event *event = cpuc->events[idx];
-                struct hw_perf_event *hwc;
-
-                if (!test_bit(idx, cpuc->active_mask))
-                        continue;
-
-                /*
-                 * We have a single interrupt for all counters. Check that
-                 * each counter has overflowed before we process it.
-                 */
-                if (!armv6_pmcr_counter_has_overflowed(pmcr, idx))
-                        continue;
-
-                hwc = &event->hw;
-                armpmu_event_update(event, hwc, idx);
-                data.period = event->hw.last_period;
-                if (!armpmu_event_set_period(event, hwc, idx))
-                        continue;
-
-                if (perf_event_overflow(event, 0, &data, regs))
-                        armpmu->disable(hwc, idx);
-        }
-
-        /*
-         * Handle the pending perf events.
-         *
-         * Note: this call *must* be run with interrupts disabled. For
-         * platforms that can have the PMU interrupts raised as an NMI, this
-         * will not work.
-         */
-        irq_work_run();
-
-        return IRQ_HANDLED;
-}
-
-static void
-armv6pmu_start(void)
-{
-        unsigned long flags, val;
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = armv6_pmcr_read();
-        val |= ARMV6_PMCR_ENABLE;
-        armv6_pmcr_write(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-void
-armv6pmu_stop(void)
-{
-        unsigned long flags, val;
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = armv6_pmcr_read();
-        val &= ~ARMV6_PMCR_ENABLE;
-        armv6_pmcr_write(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static inline int
-armv6pmu_event_map(int config)
-{
-        int mapping = armv6_perf_map[config];
-        if (HW_OP_UNSUPPORTED == mapping)
-                mapping = -EOPNOTSUPP;
-        return mapping;
-}
-
-static inline int
-armv6mpcore_pmu_event_map(int config)
-{
-        int mapping = armv6mpcore_perf_map[config];
-        if (HW_OP_UNSUPPORTED == mapping)
-                mapping = -EOPNOTSUPP;
-        return mapping;
-}
-
-static u64
-armv6pmu_raw_event(u64 config)
-{
-        return config & 0xff;
-}
-
-static int
-armv6pmu_get_event_idx(struct cpu_hw_events *cpuc,
-                       struct hw_perf_event *event)
-{
-        /* Always place a cycle counter into the cycle counter. */
-        if (ARMV6_PERFCTR_CPU_CYCLES == event->config_base) {
-                if (test_and_set_bit(ARMV6_CYCLE_COUNTER, cpuc->used_mask))
-                        return -EAGAIN;
-
-                return ARMV6_CYCLE_COUNTER;
-        } else {
-                /*
-                 * For anything other than a cycle counter, try and use
-                 * counter0 and counter1.
-                 */
-                if (!test_and_set_bit(ARMV6_COUNTER1, cpuc->used_mask)) {
-                        return ARMV6_COUNTER1;
-                }
-
-                if (!test_and_set_bit(ARMV6_COUNTER0, cpuc->used_mask)) {
-                        return ARMV6_COUNTER0;
-                }
-
-                /* The counters are all in use. */
-                return -EAGAIN;
-        }
-}
-
-static void
-armv6pmu_disable_event(struct hw_perf_event *hwc,
-                       int idx)
-{
-        unsigned long val, mask, evt, flags;
-
-        if (ARMV6_CYCLE_COUNTER == idx) {
-                mask    = ARMV6_PMCR_CCOUNT_IEN;
-                evt     = 0;
-        } else if (ARMV6_COUNTER0 == idx) {
-                mask    = ARMV6_PMCR_COUNT0_IEN | ARMV6_PMCR_EVT_COUNT0_MASK;
-                evt     = ARMV6_PERFCTR_NOP << ARMV6_PMCR_EVT_COUNT0_SHIFT;
-        } else if (ARMV6_COUNTER1 == idx) {
-                mask    = ARMV6_PMCR_COUNT1_IEN | ARMV6_PMCR_EVT_COUNT1_MASK;
-                evt     = ARMV6_PERFCTR_NOP << ARMV6_PMCR_EVT_COUNT1_SHIFT;
-        } else {
-                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
-                return;
-        }
-
-        /*
-         * Mask out the current event and set the counter to count the number
-         * of ETM bus signal assertion cycles. The external reporting should
-         * be disabled and so this should never increment.
-         */
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = armv6_pmcr_read();
-        val &= ~mask;
-        val |= evt;
-        armv6_pmcr_write(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static void
-armv6mpcore_pmu_disable_event(struct hw_perf_event *hwc,
-                              int idx)
-{
-        unsigned long val, mask, flags, evt = 0;
-
-        if (ARMV6_CYCLE_COUNTER == idx) {
-                mask    = ARMV6_PMCR_CCOUNT_IEN;
-        } else if (ARMV6_COUNTER0 == idx) {
-                mask    = ARMV6_PMCR_COUNT0_IEN;
-        } else if (ARMV6_COUNTER1 == idx) {
-                mask    = ARMV6_PMCR_COUNT1_IEN;
-        } else {
-                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
-                return;
-        }
-
-        /*
-         * Unlike UP ARMv6, we don't have a way of stopping the counters. We
-         * simply disable the interrupt reporting.
-         */
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = armv6_pmcr_read();
-        val &= ~mask;
-        val |= evt;
-        armv6_pmcr_write(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static const struct arm_pmu armv6pmu = {
-        .id                     = ARM_PERF_PMU_ID_V6,
-        .handle_irq             = armv6pmu_handle_irq,
-        .enable                 = armv6pmu_enable_event,
-        .disable                = armv6pmu_disable_event,
-        .event_map              = armv6pmu_event_map,
-        .raw_event              = armv6pmu_raw_event,
-        .read_counter           = armv6pmu_read_counter,
-        .write_counter          = armv6pmu_write_counter,
-        .get_event_idx          = armv6pmu_get_event_idx,
-        .start                  = armv6pmu_start,
-        .stop                   = armv6pmu_stop,
-        .num_events             = 3,
-        .max_period             = (1LLU << 32) - 1,
-};
-
-/*
- * ARMv6mpcore is almost identical to single core ARMv6 with the exception
- * that some of the events have different enumerations and that there is no
- * *hack* to stop the programmable counters. To stop the counters we simply
- * disable the interrupt reporting and update the event. When unthrottling we
- * reset the period and enable the interrupt reporting.
- */
-static const struct arm_pmu armv6mpcore_pmu = {
-        .id                     = ARM_PERF_PMU_ID_V6MP,
-        .handle_irq             = armv6pmu_handle_irq,
-        .enable                 = armv6pmu_enable_event,
-        .disable                = armv6mpcore_pmu_disable_event,
-        .event_map              = armv6mpcore_pmu_event_map,
-        .raw_event              = armv6pmu_raw_event,
-        .read_counter           = armv6pmu_read_counter,
-        .write_counter          = armv6pmu_write_counter,
-        .get_event_idx          = armv6pmu_get_event_idx,
-        .start                  = armv6pmu_start,
-        .stop                   = armv6pmu_stop,
-        .num_events             = 3,
-        .max_period             = (1LLU << 32) - 1,
-};
-
-/*
- * ARMv7 Cortex-A8 and Cortex-A9 Performance Events handling code.
- *
- * Copied from ARMv6 code, with the low level code inspired
- *  by the ARMv7 Oprofile code.
- *
- * Cortex-A8 has up to 4 configurable performance counters and
- *  a single cycle counter.
- * Cortex-A9 has up to 31 configurable performance counters and
- *  a single cycle counter.
- *
- * All counters can be enabled/disabled and IRQ masked separately. The cycle
- *  counter and all 4 performance counters together can be reset separately.
- */
-
-/* Common ARMv7 event types */
-enum armv7_perf_types {
-        ARMV7_PERFCTR_PMNC_SW_INCR              = 0x00,
-        ARMV7_PERFCTR_IFETCH_MISS               = 0x01,
-        ARMV7_PERFCTR_ITLB_MISS                 = 0x02,
-        ARMV7_PERFCTR_DCACHE_REFILL             = 0x03,
-        ARMV7_PERFCTR_DCACHE_ACCESS             = 0x04,
-        ARMV7_PERFCTR_DTLB_REFILL               = 0x05,
-        ARMV7_PERFCTR_DREAD                     = 0x06,
-        ARMV7_PERFCTR_DWRITE                    = 0x07,
-
-        ARMV7_PERFCTR_EXC_TAKEN                 = 0x09,
-        ARMV7_PERFCTR_EXC_EXECUTED              = 0x0A,
-        ARMV7_PERFCTR_CID_WRITE                 = 0x0B,
-        /* ARMV7_PERFCTR_PC_WRITE is equivalent to HW_BRANCH_INSTRUCTIONS.
-         * It counts:
-         *  - all branch instructions,
-         *  - instructions that explicitly write the PC,
-         *  - exception generating instructions.
-         */
-        ARMV7_PERFCTR_PC_WRITE                  = 0x0C,
-        ARMV7_PERFCTR_PC_IMM_BRANCH             = 0x0D,
-        ARMV7_PERFCTR_UNALIGNED_ACCESS          = 0x0F,
-        ARMV7_PERFCTR_PC_BRANCH_MIS_PRED        = 0x10,
-        ARMV7_PERFCTR_CLOCK_CYCLES              = 0x11,
-
-        ARMV7_PERFCTR_PC_BRANCH_MIS_USED        = 0x12,
-
-        ARMV7_PERFCTR_CPU_CYCLES                = 0xFF
-};
-
-/* ARMv7 Cortex-A8 specific event types */
-enum armv7_a8_perf_types {
-        ARMV7_PERFCTR_INSTR_EXECUTED            = 0x08,
-
-        ARMV7_PERFCTR_PC_PROC_RETURN            = 0x0E,
-
-        ARMV7_PERFCTR_WRITE_BUFFER_FULL         = 0x40,
-        ARMV7_PERFCTR_L2_STORE_MERGED           = 0x41,
-        ARMV7_PERFCTR_L2_STORE_BUFF             = 0x42,
-        ARMV7_PERFCTR_L2_ACCESS                 = 0x43,
-        ARMV7_PERFCTR_L2_CACH_MISS              = 0x44,
-        ARMV7_PERFCTR_AXI_READ_CYCLES           = 0x45,
-        ARMV7_PERFCTR_AXI_WRITE_CYCLES          = 0x46,
-        ARMV7_PERFCTR_MEMORY_REPLAY             = 0x47,
-        ARMV7_PERFCTR_UNALIGNED_ACCESS_REPLAY   = 0x48,
-        ARMV7_PERFCTR_L1_DATA_MISS              = 0x49,
-        ARMV7_PERFCTR_L1_INST_MISS              = 0x4A,
-        ARMV7_PERFCTR_L1_DATA_COLORING          = 0x4B,
-        ARMV7_PERFCTR_L1_NEON_DATA              = 0x4C,
-        ARMV7_PERFCTR_L1_NEON_CACH_DATA         = 0x4D,
-        ARMV7_PERFCTR_L2_NEON                   = 0x4E,
-        ARMV7_PERFCTR_L2_NEON_HIT               = 0x4F,
-        ARMV7_PERFCTR_L1_INST                   = 0x50,
-        ARMV7_PERFCTR_PC_RETURN_MIS_PRED        = 0x51,
-        ARMV7_PERFCTR_PC_BRANCH_FAILED          = 0x52,
-        ARMV7_PERFCTR_PC_BRANCH_TAKEN           = 0x53,
-        ARMV7_PERFCTR_PC_BRANCH_EXECUTED        = 0x54,
-        ARMV7_PERFCTR_OP_EXECUTED               = 0x55,
-        ARMV7_PERFCTR_CYCLES_INST_STALL         = 0x56,
-        ARMV7_PERFCTR_CYCLES_INST               = 0x57,
-        ARMV7_PERFCTR_CYCLES_NEON_DATA_STALL    = 0x58,
-        ARMV7_PERFCTR_CYCLES_NEON_INST_STALL    = 0x59,
-        ARMV7_PERFCTR_NEON_CYCLES               = 0x5A,
-
-        ARMV7_PERFCTR_PMU0_EVENTS               = 0x70,
-        ARMV7_PERFCTR_PMU1_EVENTS               = 0x71,
-        ARMV7_PERFCTR_PMU_EVENTS                = 0x72,
-};
-
-/* ARMv7 Cortex-A9 specific event types */
-enum armv7_a9_perf_types {
-        ARMV7_PERFCTR_JAVA_HW_BYTECODE_EXEC     = 0x40,
-        ARMV7_PERFCTR_JAVA_SW_BYTECODE_EXEC     = 0x41,
-        ARMV7_PERFCTR_JAZELLE_BRANCH_EXEC       = 0x42,
-
-        ARMV7_PERFCTR_COHERENT_LINE_MISS        = 0x50,
-        ARMV7_PERFCTR_COHERENT_LINE_HIT         = 0x51,
-
-        ARMV7_PERFCTR_ICACHE_DEP_STALL_CYCLES   = 0x60,
-        ARMV7_PERFCTR_DCACHE_DEP_STALL_CYCLES   = 0x61,
-        ARMV7_PERFCTR_TLB_MISS_DEP_STALL_CYCLES = 0x62,
-        ARMV7_PERFCTR_STREX_EXECUTED_PASSED     = 0x63,
-        ARMV7_PERFCTR_STREX_EXECUTED_FAILED     = 0x64,
-        ARMV7_PERFCTR_DATA_EVICTION             = 0x65,
-        ARMV7_PERFCTR_ISSUE_STAGE_NO_INST       = 0x66,
-        ARMV7_PERFCTR_ISSUE_STAGE_EMPTY         = 0x67,
-        ARMV7_PERFCTR_INST_OUT_OF_RENAME_STAGE  = 0x68,
-
-        ARMV7_PERFCTR_PREDICTABLE_FUNCT_RETURNS = 0x6E,
-
-        ARMV7_PERFCTR_MAIN_UNIT_EXECUTED_INST   = 0x70,
-        ARMV7_PERFCTR_SECOND_UNIT_EXECUTED_INST = 0x71,
-        ARMV7_PERFCTR_LD_ST_UNIT_EXECUTED_INST  = 0x72,
-        ARMV7_PERFCTR_FP_EXECUTED_INST          = 0x73,
-        ARMV7_PERFCTR_NEON_EXECUTED_INST        = 0x74,
-
-        ARMV7_PERFCTR_PLD_FULL_DEP_STALL_CYCLES = 0x80,
-        ARMV7_PERFCTR_DATA_WR_DEP_STALL_CYCLES  = 0x81,
-        ARMV7_PERFCTR_ITLB_MISS_DEP_STALL_CYCLES        = 0x82,
-        ARMV7_PERFCTR_DTLB_MISS_DEP_STALL_CYCLES        = 0x83,
-        ARMV7_PERFCTR_MICRO_ITLB_MISS_DEP_STALL_CYCLES  = 0x84,
-        ARMV7_PERFCTR_MICRO_DTLB_MISS_DEP_STALL_CYCLES  = 0x85,
-        ARMV7_PERFCTR_DMB_DEP_STALL_CYCLES      = 0x86,
-
-        ARMV7_PERFCTR_INTGR_CLK_ENABLED_CYCLES  = 0x8A,
-        ARMV7_PERFCTR_DATA_ENGINE_CLK_EN_CYCLES = 0x8B,
-
-        ARMV7_PERFCTR_ISB_INST                  = 0x90,
-        ARMV7_PERFCTR_DSB_INST                  = 0x91,
-        ARMV7_PERFCTR_DMB_INST                  = 0x92,
-        ARMV7_PERFCTR_EXT_INTERRUPTS            = 0x93,
-
-        ARMV7_PERFCTR_PLE_CACHE_LINE_RQST_COMPLETED     = 0xA0,
-        ARMV7_PERFCTR_PLE_CACHE_LINE_RQST_SKIPPED       = 0xA1,
-        ARMV7_PERFCTR_PLE_FIFO_FLUSH            = 0xA2,
-        ARMV7_PERFCTR_PLE_RQST_COMPLETED        = 0xA3,
-        ARMV7_PERFCTR_PLE_FIFO_OVERFLOW         = 0xA4,
-        ARMV7_PERFCTR_PLE_RQST_PROG             = 0xA5
-};
-
-/*
- * Cortex-A8 HW events mapping
- *
- * The hardware events that we support. We do support cache operations but
- * we have harvard caches and no way to combine instruction and data
- * accesses/misses in hardware.
- */
-static const unsigned armv7_a8_perf_map[PERF_COUNT_HW_MAX] = {
-        [PERF_COUNT_HW_CPU_CYCLES]          = ARMV7_PERFCTR_CPU_CYCLES,
-        [PERF_COUNT_HW_INSTRUCTIONS]        = ARMV7_PERFCTR_INSTR_EXECUTED,
-        [PERF_COUNT_HW_CACHE_REFERENCES]    = HW_OP_UNSUPPORTED,
-        [PERF_COUNT_HW_CACHE_MISSES]        = HW_OP_UNSUPPORTED,
-        [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = ARMV7_PERFCTR_PC_WRITE,
-        [PERF_COUNT_HW_BRANCH_MISSES]       = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
-        [PERF_COUNT_HW_BUS_CYCLES]          = ARMV7_PERFCTR_CLOCK_CYCLES,
-};
-
-static const unsigned armv7_a8_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
-                                          [PERF_COUNT_HW_CACHE_OP_MAX]
-                                          [PERF_COUNT_HW_CACHE_RESULT_MAX] = {
-        [C(L1D)] = {
-                /*
-                 * The performance counters don't differentiate between read
-                 * and write accesses/misses so this isn't strictly correct,
-                 * but it's the best we can do. Writes and reads get
-                 * combined.
-                 */
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_DCACHE_ACCESS,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DCACHE_REFILL,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_DCACHE_ACCESS,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DCACHE_REFILL,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(L1I)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_L1_INST,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_L1_INST_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_L1_INST,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_L1_INST_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(LL)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_L2_ACCESS,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_L2_CACH_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_L2_ACCESS,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_L2_CACH_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(DTLB)] = {
-                /*
-                 * Only ITLB misses and DTLB refills are supported.
-                 * If users want the DTLB refills misses a raw counter
-                 * must be used.
-                 */
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DTLB_REFILL,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DTLB_REFILL,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(ITLB)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_ITLB_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_ITLB_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(BPU)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_PC_WRITE,
-                        [C(RESULT_MISS)]
-                                        = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_PC_WRITE,
-                        [C(RESULT_MISS)]
-                                        = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-};
-
-/*
- * Cortex-A9 HW events mapping
- */
-static const unsigned armv7_a9_perf_map[PERF_COUNT_HW_MAX] = {
-        [PERF_COUNT_HW_CPU_CYCLES]          = ARMV7_PERFCTR_CPU_CYCLES,
-        [PERF_COUNT_HW_INSTRUCTIONS]        =
-                                        ARMV7_PERFCTR_INST_OUT_OF_RENAME_STAGE,
-        [PERF_COUNT_HW_CACHE_REFERENCES]    = ARMV7_PERFCTR_COHERENT_LINE_HIT,
-        [PERF_COUNT_HW_CACHE_MISSES]        = ARMV7_PERFCTR_COHERENT_LINE_MISS,
-        [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = ARMV7_PERFCTR_PC_WRITE,
-        [PERF_COUNT_HW_BRANCH_MISSES]       = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
-        [PERF_COUNT_HW_BUS_CYCLES]          = ARMV7_PERFCTR_CLOCK_CYCLES,
-};
-
-static const unsigned armv7_a9_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
-                                          [PERF_COUNT_HW_CACHE_OP_MAX]
-                                          [PERF_COUNT_HW_CACHE_RESULT_MAX] = {
-        [C(L1D)] = {
-                /*
-                 * The performance counters don't differentiate between read
-                 * and write accesses/misses so this isn't strictly correct,
-                 * but it's the best we can do. Writes and reads get
-                 * combined.
-                 */
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_DCACHE_ACCESS,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DCACHE_REFILL,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_DCACHE_ACCESS,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DCACHE_REFILL,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(L1I)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_IFETCH_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_IFETCH_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(LL)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(DTLB)] = {
-                /*
-                 * Only ITLB misses and DTLB refills are supported.
-                 * If users want the DTLB refills misses a raw counter
-                 * must be used.
-                 */
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DTLB_REFILL,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DTLB_REFILL,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(ITLB)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_ITLB_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_ITLB_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(BPU)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_PC_WRITE,
-                        [C(RESULT_MISS)]
-                                        = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_PC_WRITE,
-                        [C(RESULT_MISS)]
-                                        = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-};
-
-/*
- * Perf Events counters
- */
-enum armv7_counters {
-        ARMV7_CYCLE_COUNTER             = 1,    /* Cycle counter */
-        ARMV7_COUNTER0                  = 2,    /* First event counter */
-};
-
-/*
- * The cycle counter is ARMV7_CYCLE_COUNTER.
- * The first event counter is ARMV7_COUNTER0.
- * The last event counter is (ARMV7_COUNTER0 + armpmu->num_events - 1).
- */
-#define ARMV7_COUNTER_LAST      (ARMV7_COUNTER0 + armpmu->num_events - 1)
-
-/*
- * ARMv7 low level PMNC access
- */
-
-/*
- * Per-CPU PMNC: config reg
- */
-#define ARMV7_PMNC_E            (1 << 0) /* Enable all counters */
-#define ARMV7_PMNC_P            (1 << 1) /* Reset all counters */
-#define ARMV7_PMNC_C            (1 << 2) /* Cycle counter reset */
-#define ARMV7_PMNC_D            (1 << 3) /* CCNT counts every 64th cpu cycle */
-#define ARMV7_PMNC_X            (1 << 4) /* Export to ETM */
-#define ARMV7_PMNC_DP           (1 << 5) /* Disable CCNT if non-invasive debug*/
-#define ARMV7_PMNC_N_SHIFT      11       /* Number of counters supported */
-#define ARMV7_PMNC_N_MASK       0x1f
-#define ARMV7_PMNC_MASK         0x3f     /* Mask for writable bits */
-
-/*
- * Available counters
- */
-#define ARMV7_CNT0              0       /* First event counter */
-#define ARMV7_CCNT              31      /* Cycle counter */
-
-/* Perf Event to low level counters mapping */
-#define ARMV7_EVENT_CNT_TO_CNTx (ARMV7_COUNTER0 - ARMV7_CNT0)
-
-/*
- * CNTENS: counters enable reg
- */
-#define ARMV7_CNTENS_P(idx)     (1 << (idx - ARMV7_EVENT_CNT_TO_CNTx))
-#define ARMV7_CNTENS_C          (1 << ARMV7_CCNT)
-
-/*
- * CNTENC: counters disable reg
- */
-#define ARMV7_CNTENC_P(idx)     (1 << (idx - ARMV7_EVENT_CNT_TO_CNTx))
-#define ARMV7_CNTENC_C          (1 << ARMV7_CCNT)
-
-/*
- * INTENS: counters overflow interrupt enable reg
- */
-#define ARMV7_INTENS_P(idx)     (1 << (idx - ARMV7_EVENT_CNT_TO_CNTx))
-#define ARMV7_INTENS_C          (1 << ARMV7_CCNT)
-
-/*
- * INTENC: counters overflow interrupt disable reg
- */
-#define ARMV7_INTENC_P(idx)     (1 << (idx - ARMV7_EVENT_CNT_TO_CNTx))
-#define ARMV7_INTENC_C          (1 << ARMV7_CCNT)
-
-/*
- * EVTSEL: Event selection reg
- */
-#define ARMV7_EVTSEL_MASK       0xff            /* Mask for writable bits */
-
-/*
- * SELECT: Counter selection reg
- */
-#define ARMV7_SELECT_MASK       0x1f            /* Mask for writable bits */
-
-/*
- * FLAG: counters overflow flag status reg
- */
-#define ARMV7_FLAG_P(idx)       (1 << (idx - ARMV7_EVENT_CNT_TO_CNTx))
-#define ARMV7_FLAG_C            (1 << ARMV7_CCNT)
-#define ARMV7_FLAG_MASK         0xffffffff      /* Mask for writable bits */
-#define ARMV7_OVERFLOWED_MASK   ARMV7_FLAG_MASK
-
-static inline unsigned long armv7_pmnc_read(void)
-{
-        u32 val;
-        asm volatile("mrc p15, 0, %0, c9, c12, 0" : "=r"(val));
-        return val;
-}
-
-static inline void armv7_pmnc_write(unsigned long val)
-{
-        val &= ARMV7_PMNC_MASK;
-        asm volatile("mcr p15, 0, %0, c9, c12, 0" : : "r"(val));
-}
-
-static inline int armv7_pmnc_has_overflowed(unsigned long pmnc)
-{
-        return pmnc & ARMV7_OVERFLOWED_MASK;
-}
-
-static inline int armv7_pmnc_counter_has_overflowed(unsigned long pmnc,
-                                        enum armv7_counters counter)
-{
-        int ret = 0;
-
-        if (counter == ARMV7_CYCLE_COUNTER)
-                ret = pmnc & ARMV7_FLAG_C;
-        else if ((counter >= ARMV7_COUNTER0) && (counter <= ARMV7_COUNTER_LAST))
-                ret = pmnc & ARMV7_FLAG_P(counter);
-        else
-                pr_err("CPU%u checking wrong counter %d overflow status\n",
-                        smp_processor_id(), counter);
-
-        return ret;
-}
-
-static inline int armv7_pmnc_select_counter(unsigned int idx)
-{
-        u32 val;
-
-        if ((idx < ARMV7_COUNTER0) || (idx > ARMV7_COUNTER_LAST)) {
-                pr_err("CPU%u selecting wrong PMNC counter"
-                        " %d\n", smp_processor_id(), idx);
-                return -1;
-        }
-
-        val = (idx - ARMV7_EVENT_CNT_TO_CNTx) & ARMV7_SELECT_MASK;
-        asm volatile("mcr p15, 0, %0, c9, c12, 5" : : "r" (val));
-
-        return idx;
-}
-
-static inline u32 armv7pmu_read_counter(int idx)
-{
-        unsigned long value = 0;
-
-        if (idx == ARMV7_CYCLE_COUNTER)
-                asm volatile("mrc p15, 0, %0, c9, c13, 0" : "=r" (value));
-        else if ((idx >= ARMV7_COUNTER0) && (idx <= ARMV7_COUNTER_LAST)) {
-                if (armv7_pmnc_select_counter(idx) == idx)
-                        asm volatile("mrc p15, 0, %0, c9, c13, 2"
-                                     : "=r" (value));
-        } else
-                pr_err("CPU%u reading wrong counter %d\n",
-                        smp_processor_id(), idx);
-
-        return value;
-}
-
-static inline void armv7pmu_write_counter(int idx, u32 value)
-{
-        if (idx == ARMV7_CYCLE_COUNTER)
-                asm volatile("mcr p15, 0, %0, c9, c13, 0" : : "r" (value));
-        else if ((idx >= ARMV7_COUNTER0) && (idx <= ARMV7_COUNTER_LAST)) {
-                if (armv7_pmnc_select_counter(idx) == idx)
-                        asm volatile("mcr p15, 0, %0, c9, c13, 2"
-                                     : : "r" (value));
-        } else
-                pr_err("CPU%u writing wrong counter %d\n",
-                        smp_processor_id(), idx);
-}
-
-static inline void armv7_pmnc_write_evtsel(unsigned int idx, u32 val)
-{
-        if (armv7_pmnc_select_counter(idx) == idx) {
-                val &= ARMV7_EVTSEL_MASK;
-                asm volatile("mcr p15, 0, %0, c9, c13, 1" : : "r" (val));
-        }
-}
-
-static inline u32 armv7_pmnc_enable_counter(unsigned int idx)
-{
-        u32 val;
-
-        if ((idx != ARMV7_CYCLE_COUNTER) &&
-            ((idx < ARMV7_COUNTER0) || (idx > ARMV7_COUNTER_LAST))) {
-                pr_err("CPU%u enabling wrong PMNC counter"
-                        " %d\n", smp_processor_id(), idx);
-                return -1;
-        }
-
-        if (idx == ARMV7_CYCLE_COUNTER)
-                val = ARMV7_CNTENS_C;
-        else
-                val = ARMV7_CNTENS_P(idx);
-
-        asm volatile("mcr p15, 0, %0, c9, c12, 1" : : "r" (val));
-
-        return idx;
-}
-
-static inline u32 armv7_pmnc_disable_counter(unsigned int idx)
-{
-        u32 val;
-
-
-        if ((idx != ARMV7_CYCLE_COUNTER) &&
-            ((idx < ARMV7_COUNTER0) || (idx > ARMV7_COUNTER_LAST))) {
-                pr_err("CPU%u disabling wrong PMNC counter"
-                        " %d\n", smp_processor_id(), idx);
-                return -1;
-        }
-
-        if (idx == ARMV7_CYCLE_COUNTER)
-                val = ARMV7_CNTENC_C;
-        else
-                val = ARMV7_CNTENC_P(idx);
-
-        asm volatile("mcr p15, 0, %0, c9, c12, 2" : : "r" (val));
-
-        return idx;
-}
-
-static inline u32 armv7_pmnc_enable_intens(unsigned int idx)
-{
-        u32 val;
-
-        if ((idx != ARMV7_CYCLE_COUNTER) &&
-            ((idx < ARMV7_COUNTER0) || (idx > ARMV7_COUNTER_LAST))) {
-                pr_err("CPU%u enabling wrong PMNC counter"
-                        " interrupt enable %d\n", smp_processor_id(), idx);
-                return -1;
-        }
-
-        if (idx == ARMV7_CYCLE_COUNTER)
-                val = ARMV7_INTENS_C;
-        else
-                val = ARMV7_INTENS_P(idx);
-
-        asm volatile("mcr p15, 0, %0, c9, c14, 1" : : "r" (val));
-
-        return idx;
-}
-
-static inline u32 armv7_pmnc_disable_intens(unsigned int idx)
-{
-        u32 val;
-
-        if ((idx != ARMV7_CYCLE_COUNTER) &&
-            ((idx < ARMV7_COUNTER0) || (idx > ARMV7_COUNTER_LAST))) {
-                pr_err("CPU%u disabling wrong PMNC counter"
-                        " interrupt enable %d\n", smp_processor_id(), idx);
-                return -1;
-        }
-
-        if (idx == ARMV7_CYCLE_COUNTER)
-                val = ARMV7_INTENC_C;
-        else
-                val = ARMV7_INTENC_P(idx);
-
-        asm volatile("mcr p15, 0, %0, c9, c14, 2" : : "r" (val));
-
-        return idx;
-}
-
-static inline u32 armv7_pmnc_getreset_flags(void)
-{
-        u32 val;
-
-        /* Read */
-        asm volatile("mrc p15, 0, %0, c9, c12, 3" : "=r" (val));
-
-        /* Write to clear flags */
-        val &= ARMV7_FLAG_MASK;
-        asm volatile("mcr p15, 0, %0, c9, c12, 3" : : "r" (val));
-
-        return val;
-}
-
-#ifdef DEBUG
-static void armv7_pmnc_dump_regs(void)
-{
-        u32 val;
-        unsigned int cnt;
-
-        printk(KERN_INFO "PMNC registers dump:\n");
-
-        asm volatile("mrc p15, 0, %0, c9, c12, 0" : "=r" (val));
-        printk(KERN_INFO "PMNC  =0x%08x\n", val);
-
-        asm volatile("mrc p15, 0, %0, c9, c12, 1" : "=r" (val));
-        printk(KERN_INFO "CNTENS=0x%08x\n", val);
-
-        asm volatile("mrc p15, 0, %0, c9, c14, 1" : "=r" (val));
-        printk(KERN_INFO "INTENS=0x%08x\n", val);
-
-        asm volatile("mrc p15, 0, %0, c9, c12, 3" : "=r" (val));
-        printk(KERN_INFO "FLAGS =0x%08x\n", val);
-
-        asm volatile("mrc p15, 0, %0, c9, c12, 5" : "=r" (val));
-        printk(KERN_INFO "SELECT=0x%08x\n", val);
-
-        asm volatile("mrc p15, 0, %0, c9, c13, 0" : "=r" (val));
-        printk(KERN_INFO "CCNT  =0x%08x\n", val);
-
-        for (cnt = ARMV7_COUNTER0; cnt < ARMV7_COUNTER_LAST; cnt++) {
-                armv7_pmnc_select_counter(cnt);
-                asm volatile("mrc p15, 0, %0, c9, c13, 2" : "=r" (val));
-                printk(KERN_INFO "CNT[%d] count =0x%08x\n",
-                        cnt-ARMV7_EVENT_CNT_TO_CNTx, val);
-                asm volatile("mrc p15, 0, %0, c9, c13, 1" : "=r" (val));
-                printk(KERN_INFO "CNT[%d] evtsel=0x%08x\n",
-                        cnt-ARMV7_EVENT_CNT_TO_CNTx, val);
-        }
-}
-#endif
-
-void armv7pmu_enable_event(struct hw_perf_event *hwc, int idx)
-{
-        unsigned long flags;
-
-        /*
-         * Enable counter and interrupt, and set the counter to count
-         * the event that we're interested in.
-         */
-        spin_lock_irqsave(&pmu_lock, flags);
-
-        /*
-         * Disable counter
-         */
-        armv7_pmnc_disable_counter(idx);
-
-        /*
-         * Set event (if destined for PMNx counters)
-         * We don't need to set the event if it's a cycle count
-         */
-        if (idx != ARMV7_CYCLE_COUNTER)
-                armv7_pmnc_write_evtsel(idx, hwc->config_base);
-
-        /*
-         * Enable interrupt for this counter
-         */
-        armv7_pmnc_enable_intens(idx);
-
-        /*
-         * Enable counter
-         */
-        armv7_pmnc_enable_counter(idx);
-
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static void armv7pmu_disable_event(struct hw_perf_event *hwc, int idx)
-{
-        unsigned long flags;
-
-        /*
-         * Disable counter and interrupt
-         */
-        spin_lock_irqsave(&pmu_lock, flags);
-
-        /*
-         * Disable counter
-         */
-        armv7_pmnc_disable_counter(idx);
-
-        /*
-         * Disable interrupt for this counter
-         */
-        armv7_pmnc_disable_intens(idx);
-
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static irqreturn_t armv7pmu_handle_irq(int irq_num, void *dev)
-{
-        unsigned long pmnc;
-        struct perf_sample_data data;
-        struct cpu_hw_events *cpuc;
-        struct pt_regs *regs;
-        int idx;
-
-        /*
-         * Get and reset the IRQ flags
-         */
-        pmnc = armv7_pmnc_getreset_flags();
-
-        /*
-         * Did an overflow occur?
-         */
-        if (!armv7_pmnc_has_overflowed(pmnc))
-                return IRQ_NONE;
-
-        /*
-         * Handle the counter(s) overflow(s)
-         */
-        regs = get_irq_regs();
-
-        perf_sample_data_init(&data, 0);
-
-        cpuc = &__get_cpu_var(cpu_hw_events);
-        for (idx = 0; idx <= armpmu->num_events; ++idx) {
-                struct perf_event *event = cpuc->events[idx];
-                struct hw_perf_event *hwc;
-
-                if (!test_bit(idx, cpuc->active_mask))
-                        continue;
-
-                /*
-                 * We have a single interrupt for all counters. Check that
-                 * each counter has overflowed before we process it.
-                 */
-                if (!armv7_pmnc_counter_has_overflowed(pmnc, idx))
-                        continue;
-
-                hwc = &event->hw;
-                armpmu_event_update(event, hwc, idx);
-                data.period = event->hw.last_period;
-                if (!armpmu_event_set_period(event, hwc, idx))
-                        continue;
-
-                if (perf_event_overflow(event, 0, &data, regs))
-                        armpmu->disable(hwc, idx);
-        }
-
-        /*
-         * Handle the pending perf events.
-         *
-         * Note: this call *must* be run with interrupts disabled. For
-         * platforms that can have the PMU interrupts raised as an NMI, this
-         * will not work.
-         */
-        irq_work_run();
-
-        return IRQ_HANDLED;
-}
-
-static void armv7pmu_start(void)
-{
-        unsigned long flags;
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        /* Enable all counters */
-        armv7_pmnc_write(armv7_pmnc_read() | ARMV7_PMNC_E);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static void armv7pmu_stop(void)
-{
-        unsigned long flags;
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        /* Disable all counters */
-        armv7_pmnc_write(armv7_pmnc_read() & ~ARMV7_PMNC_E);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static inline int armv7_a8_pmu_event_map(int config)
-{
-        int mapping = armv7_a8_perf_map[config];
-        if (HW_OP_UNSUPPORTED == mapping)
-                mapping = -EOPNOTSUPP;
-        return mapping;
-}
-
-static inline int armv7_a9_pmu_event_map(int config)
-{
-        int mapping = armv7_a9_perf_map[config];
-        if (HW_OP_UNSUPPORTED == mapping)
-                mapping = -EOPNOTSUPP;
-        return mapping;
-}
-
-static u64 armv7pmu_raw_event(u64 config)
-{
-        return config & 0xff;
-}
-
-static int armv7pmu_get_event_idx(struct cpu_hw_events *cpuc,
-                                  struct hw_perf_event *event)
-{
-        int idx;
-
-        /* Always place a cycle counter into the cycle counter. */
-        if (event->config_base == ARMV7_PERFCTR_CPU_CYCLES) {
-                if (test_and_set_bit(ARMV7_CYCLE_COUNTER, cpuc->used_mask))
-                        return -EAGAIN;
-
-                return ARMV7_CYCLE_COUNTER;
-        } else {
-                /*
-                 * For anything other than a cycle counter, try and use
-                 * the events counters
-                 */
-                for (idx = ARMV7_COUNTER0; idx <= armpmu->num_events; ++idx) {
-                        if (!test_and_set_bit(idx, cpuc->used_mask))
-                                return idx;
-                }
-
-                /* The counters are all in use. */
-                return -EAGAIN;
-        }
-}
-
-static struct arm_pmu armv7pmu = {
-        .handle_irq             = armv7pmu_handle_irq,
-        .enable                 = armv7pmu_enable_event,
-        .disable                = armv7pmu_disable_event,
-        .raw_event              = armv7pmu_raw_event,
-        .read_counter           = armv7pmu_read_counter,
-        .write_counter          = armv7pmu_write_counter,
-        .get_event_idx          = armv7pmu_get_event_idx,
-        .start                  = armv7pmu_start,
-        .stop                   = armv7pmu_stop,
-        .max_period             = (1LLU << 32) - 1,
-};
-
-static u32 __init armv7_reset_read_pmnc(void)
-{
-        u32 nb_cnt;
-
-        /* Initialize & Reset PMNC: C and P bits */
-        armv7_pmnc_write(ARMV7_PMNC_P | ARMV7_PMNC_C);
-
-        /* Read the nb of CNTx counters supported from PMNC */
-        nb_cnt = (armv7_pmnc_read() >> ARMV7_PMNC_N_SHIFT) & ARMV7_PMNC_N_MASK;
-
-        /* Add the CPU cycles counter and return */
-        return nb_cnt + 1;
-}
-
-/*
- * ARMv5 [xscale] Performance counter handling code.
- *
- * Based on xscale OProfile code.
- *
- * There are two variants of the xscale PMU that we support:
- *      - xscale1pmu: 2 event counters and a cycle counter
- *      - xscale2pmu: 4 event counters and a cycle counter
- * The two variants share event definitions, but have different
- * PMU structures.
- */
-
-enum xscale_perf_types {
-        XSCALE_PERFCTR_ICACHE_MISS              = 0x00,
-        XSCALE_PERFCTR_ICACHE_NO_DELIVER        = 0x01,
-        XSCALE_PERFCTR_DATA_STALL               = 0x02,
-        XSCALE_PERFCTR_ITLB_MISS                = 0x03,
-        XSCALE_PERFCTR_DTLB_MISS                = 0x04,
-        XSCALE_PERFCTR_BRANCH                   = 0x05,
-        XSCALE_PERFCTR_BRANCH_MISS              = 0x06,
-        XSCALE_PERFCTR_INSTRUCTION              = 0x07,
-        XSCALE_PERFCTR_DCACHE_FULL_STALL        = 0x08,
-        XSCALE_PERFCTR_DCACHE_FULL_STALL_CONTIG = 0x09,
-        XSCALE_PERFCTR_DCACHE_ACCESS            = 0x0A,
-        XSCALE_PERFCTR_DCACHE_MISS              = 0x0B,
-        XSCALE_PERFCTR_DCACHE_WRITE_BACK        = 0x0C,
-        XSCALE_PERFCTR_PC_CHANGED               = 0x0D,
-        XSCALE_PERFCTR_BCU_REQUEST              = 0x10,
-        XSCALE_PERFCTR_BCU_FULL                 = 0x11,
-        XSCALE_PERFCTR_BCU_DRAIN                = 0x12,
-        XSCALE_PERFCTR_BCU_ECC_NO_ELOG          = 0x14,
-        XSCALE_PERFCTR_BCU_1_BIT_ERR            = 0x15,
-        XSCALE_PERFCTR_RMW                      = 0x16,
-        /* XSCALE_PERFCTR_CCNT is not hardware defined */
-        XSCALE_PERFCTR_CCNT                     = 0xFE,
-        XSCALE_PERFCTR_UNUSED                   = 0xFF,
-};
-
-enum xscale_counters {
-        XSCALE_CYCLE_COUNTER    = 1,
-        XSCALE_COUNTER0,
-        XSCALE_COUNTER1,
-        XSCALE_COUNTER2,
-        XSCALE_COUNTER3,
-};
-
-static const unsigned xscale_perf_map[PERF_COUNT_HW_MAX] = {
-        [PERF_COUNT_HW_CPU_CYCLES]          = XSCALE_PERFCTR_CCNT,
-        [PERF_COUNT_HW_INSTRUCTIONS]        = XSCALE_PERFCTR_INSTRUCTION,
-        [PERF_COUNT_HW_CACHE_REFERENCES]    = HW_OP_UNSUPPORTED,
-        [PERF_COUNT_HW_CACHE_MISSES]        = HW_OP_UNSUPPORTED,
-        [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = XSCALE_PERFCTR_BRANCH,
-        [PERF_COUNT_HW_BRANCH_MISSES]       = XSCALE_PERFCTR_BRANCH_MISS,
-        [PERF_COUNT_HW_BUS_CYCLES]          = HW_OP_UNSUPPORTED,
-};
-
-static const unsigned xscale_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
-                                           [PERF_COUNT_HW_CACHE_OP_MAX]
-                                           [PERF_COUNT_HW_CACHE_RESULT_MAX] = {
-        [C(L1D)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = XSCALE_PERFCTR_DCACHE_ACCESS,
-                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_DCACHE_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = XSCALE_PERFCTR_DCACHE_ACCESS,
-                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_DCACHE_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(L1I)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_ICACHE_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_ICACHE_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(LL)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(DTLB)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_DTLB_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_DTLB_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(ITLB)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_ITLB_MISS,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_ITLB_MISS,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-        [C(BPU)] = {
-                [C(OP_READ)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_WRITE)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-                [C(OP_PREFETCH)] = {
-                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
-                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
-                },
-        },
-};
-
-#define XSCALE_PMU_ENABLE       0x001
-#define XSCALE_PMN_RESET        0x002
-#define XSCALE_CCNT_RESET       0x004
-#define XSCALE_PMU_RESET        (CCNT_RESET | PMN_RESET)
-#define XSCALE_PMU_CNT64        0x008
-
-static inline int
-xscalepmu_event_map(int config)
-{
-        int mapping = xscale_perf_map[config];
-        if (HW_OP_UNSUPPORTED == mapping)
-                mapping = -EOPNOTSUPP;
-        return mapping;
-}
-
-static u64
-xscalepmu_raw_event(u64 config)
-{
-        return config & 0xff;
-}
-
-#define XSCALE1_OVERFLOWED_MASK 0x700
-#define XSCALE1_CCOUNT_OVERFLOW 0x400
-#define XSCALE1_COUNT0_OVERFLOW 0x100
-#define XSCALE1_COUNT1_OVERFLOW 0x200
-#define XSCALE1_CCOUNT_INT_EN   0x040
-#define XSCALE1_COUNT0_INT_EN   0x010
-#define XSCALE1_COUNT1_INT_EN   0x020
-#define XSCALE1_COUNT0_EVT_SHFT 12
-#define XSCALE1_COUNT0_EVT_MASK (0xff << XSCALE1_COUNT0_EVT_SHFT)
-#define XSCALE1_COUNT1_EVT_SHFT 20
-#define XSCALE1_COUNT1_EVT_MASK (0xff << XSCALE1_COUNT1_EVT_SHFT)
-
-static inline u32
-xscale1pmu_read_pmnc(void)
-{
-        u32 val;
-        asm volatile("mrc p14, 0, %0, c0, c0, 0" : "=r" (val));
-        return val;
-}
-
-static inline void
-xscale1pmu_write_pmnc(u32 val)
-{
-        /* upper 4bits and 7, 11 are write-as-0 */
-        val &= 0xffff77f;
-        asm volatile("mcr p14, 0, %0, c0, c0, 0" : : "r" (val));
-}
-
-static inline int
-xscale1_pmnc_counter_has_overflowed(unsigned long pmnc,
-                                        enum xscale_counters counter)
-{
-        int ret = 0;
-
-        switch (counter) {
-        case XSCALE_CYCLE_COUNTER:
-                ret = pmnc & XSCALE1_CCOUNT_OVERFLOW;
-                break;
-        case XSCALE_COUNTER0:
-                ret = pmnc & XSCALE1_COUNT0_OVERFLOW;
-                break;
-        case XSCALE_COUNTER1:
-                ret = pmnc & XSCALE1_COUNT1_OVERFLOW;
-                break;
-        default:
-                WARN_ONCE(1, "invalid counter number (%d)\n", counter);
-        }
-
-        return ret;
-}
-
-static irqreturn_t
-xscale1pmu_handle_irq(int irq_num, void *dev)
-{
-        unsigned long pmnc;
-        struct perf_sample_data data;
-        struct cpu_hw_events *cpuc;
-        struct pt_regs *regs;
-        int idx;
-
-        /*
-         * NOTE: there's an A stepping erratum that states if an overflow
-         *       bit already exists and another occurs, the previous
-         *       Overflow bit gets cleared. There's no workaround.
-         *       Fixed in B stepping or later.
-         */
-        pmnc = xscale1pmu_read_pmnc();
-
-        /*
-         * Write the value back to clear the overflow flags. Overflow
-         * flags remain in pmnc for use below. We also disable the PMU
-         * while we process the interrupt.
-         */
-        xscale1pmu_write_pmnc(pmnc & ~XSCALE_PMU_ENABLE);
-
-        if (!(pmnc & XSCALE1_OVERFLOWED_MASK))
-                return IRQ_NONE;
-
-        regs = get_irq_regs();
-
-        perf_sample_data_init(&data, 0);
-
-        cpuc = &__get_cpu_var(cpu_hw_events);
-        for (idx = 0; idx <= armpmu->num_events; ++idx) {
-                struct perf_event *event = cpuc->events[idx];
-                struct hw_perf_event *hwc;
-
-                if (!test_bit(idx, cpuc->active_mask))
-                        continue;
-
-                if (!xscale1_pmnc_counter_has_overflowed(pmnc, idx))
-                        continue;
-
-                hwc = &event->hw;
-                armpmu_event_update(event, hwc, idx);
-                data.period = event->hw.last_period;
-                if (!armpmu_event_set_period(event, hwc, idx))
-                        continue;
-
-                if (perf_event_overflow(event, 0, &data, regs))
-                        armpmu->disable(hwc, idx);
-        }
-
-        irq_work_run();
-
-        /*
-         * Re-enable the PMU.
-         */
-        pmnc = xscale1pmu_read_pmnc() | XSCALE_PMU_ENABLE;
-        xscale1pmu_write_pmnc(pmnc);
-
-        return IRQ_HANDLED;
-}
-
-static void
-xscale1pmu_enable_event(struct hw_perf_event *hwc, int idx)
-{
-        unsigned long val, mask, evt, flags;
-
-        switch (idx) {
-        case XSCALE_CYCLE_COUNTER:
-                mask = 0;
-                evt = XSCALE1_CCOUNT_INT_EN;
-                break;
-        case XSCALE_COUNTER0:
-                mask = XSCALE1_COUNT0_EVT_MASK;
-                evt = (hwc->config_base << XSCALE1_COUNT0_EVT_SHFT) |
-                        XSCALE1_COUNT0_INT_EN;
-                break;
-        case XSCALE_COUNTER1:
-                mask = XSCALE1_COUNT1_EVT_MASK;
-                evt = (hwc->config_base << XSCALE1_COUNT1_EVT_SHFT) |
-                        XSCALE1_COUNT1_INT_EN;
-                break;
-        default:
-                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
-                return;
-        }
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = xscale1pmu_read_pmnc();
-        val &= ~mask;
-        val |= evt;
-        xscale1pmu_write_pmnc(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static void
-xscale1pmu_disable_event(struct hw_perf_event *hwc, int idx)
-{
-        unsigned long val, mask, evt, flags;
-
-        switch (idx) {
-        case XSCALE_CYCLE_COUNTER:
-                mask = XSCALE1_CCOUNT_INT_EN;
-                evt = 0;
-                break;
-        case XSCALE_COUNTER0:
-                mask = XSCALE1_COUNT0_INT_EN | XSCALE1_COUNT0_EVT_MASK;
-                evt = XSCALE_PERFCTR_UNUSED << XSCALE1_COUNT0_EVT_SHFT;
-                break;
-        case XSCALE_COUNTER1:
-                mask = XSCALE1_COUNT1_INT_EN | XSCALE1_COUNT1_EVT_MASK;
-                evt = XSCALE_PERFCTR_UNUSED << XSCALE1_COUNT1_EVT_SHFT;
-                break;
-        default:
-                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
-                return;
-        }
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = xscale1pmu_read_pmnc();
-        val &= ~mask;
-        val |= evt;
-        xscale1pmu_write_pmnc(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static int
-xscale1pmu_get_event_idx(struct cpu_hw_events *cpuc,
-                        struct hw_perf_event *event)
-{
-        if (XSCALE_PERFCTR_CCNT == event->config_base) {
-                if (test_and_set_bit(XSCALE_CYCLE_COUNTER, cpuc->used_mask))
-                        return -EAGAIN;
-
-                return XSCALE_CYCLE_COUNTER;
-        } else {
-                if (!test_and_set_bit(XSCALE_COUNTER1, cpuc->used_mask)) {
-                        return XSCALE_COUNTER1;
-                }
-
-                if (!test_and_set_bit(XSCALE_COUNTER0, cpuc->used_mask)) {
-                        return XSCALE_COUNTER0;
-                }
-
-                return -EAGAIN;
-        }
-}
-
-static void
-xscale1pmu_start(void)
-{
-        unsigned long flags, val;
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = xscale1pmu_read_pmnc();
-        val |= XSCALE_PMU_ENABLE;
-        xscale1pmu_write_pmnc(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static void
-xscale1pmu_stop(void)
-{
-        unsigned long flags, val;
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = xscale1pmu_read_pmnc();
-        val &= ~XSCALE_PMU_ENABLE;
-        xscale1pmu_write_pmnc(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static inline u32
-xscale1pmu_read_counter(int counter)
-{
-        u32 val = 0;
-
-        switch (counter) {
-        case XSCALE_CYCLE_COUNTER:
-                asm volatile("mrc p14, 0, %0, c1, c0, 0" : "=r" (val));
-                break;
-        case XSCALE_COUNTER0:
-                asm volatile("mrc p14, 0, %0, c2, c0, 0" : "=r" (val));
-                break;
-        case XSCALE_COUNTER1:
-                asm volatile("mrc p14, 0, %0, c3, c0, 0" : "=r" (val));
-                break;
-        }
-
-        return val;
-}
-
-static inline void
-xscale1pmu_write_counter(int counter, u32 val)
-{
-        switch (counter) {
-        case XSCALE_CYCLE_COUNTER:
-                asm volatile("mcr p14, 0, %0, c1, c0, 0" : : "r" (val));
-                break;
-        case XSCALE_COUNTER0:
-                asm volatile("mcr p14, 0, %0, c2, c0, 0" : : "r" (val));
-                break;
-        case XSCALE_COUNTER1:
-                asm volatile("mcr p14, 0, %0, c3, c0, 0" : : "r" (val));
-                break;
-        }
-}
-
-static const struct arm_pmu xscale1pmu = {
-        .id             = ARM_PERF_PMU_ID_XSCALE1,
-        .handle_irq     = xscale1pmu_handle_irq,
-        .enable         = xscale1pmu_enable_event,
-        .disable        = xscale1pmu_disable_event,
-        .event_map      = xscalepmu_event_map,
-        .raw_event      = xscalepmu_raw_event,
-        .read_counter   = xscale1pmu_read_counter,
-        .write_counter  = xscale1pmu_write_counter,
-        .get_event_idx  = xscale1pmu_get_event_idx,
-        .start          = xscale1pmu_start,
-        .stop           = xscale1pmu_stop,
-        .num_events     = 3,
-        .max_period     = (1LLU << 32) - 1,
-};
-
-#define XSCALE2_OVERFLOWED_MASK 0x01f
-#define XSCALE2_CCOUNT_OVERFLOW 0x001
-#define XSCALE2_COUNT0_OVERFLOW 0x002
-#define XSCALE2_COUNT1_OVERFLOW 0x004
-#define XSCALE2_COUNT2_OVERFLOW 0x008
-#define XSCALE2_COUNT3_OVERFLOW 0x010
-#define XSCALE2_CCOUNT_INT_EN   0x001
-#define XSCALE2_COUNT0_INT_EN   0x002
-#define XSCALE2_COUNT1_INT_EN   0x004
-#define XSCALE2_COUNT2_INT_EN   0x008
-#define XSCALE2_COUNT3_INT_EN   0x010
-#define XSCALE2_COUNT0_EVT_SHFT 0
-#define XSCALE2_COUNT0_EVT_MASK (0xff << XSCALE2_COUNT0_EVT_SHFT)
-#define XSCALE2_COUNT1_EVT_SHFT 8
-#define XSCALE2_COUNT1_EVT_MASK (0xff << XSCALE2_COUNT1_EVT_SHFT)
-#define XSCALE2_COUNT2_EVT_SHFT 16
-#define XSCALE2_COUNT2_EVT_MASK (0xff << XSCALE2_COUNT2_EVT_SHFT)
-#define XSCALE2_COUNT3_EVT_SHFT 24
-#define XSCALE2_COUNT3_EVT_MASK (0xff << XSCALE2_COUNT3_EVT_SHFT)
-
-static inline u32
-xscale2pmu_read_pmnc(void)
-{
-        u32 val;
-        asm volatile("mrc p14, 0, %0, c0, c1, 0" : "=r" (val));
-        /* bits 1-2 and 4-23 are read-unpredictable */
-        return val & 0xff000009;
-}
-
-static inline void
-xscale2pmu_write_pmnc(u32 val)
-{
-        /* bits 4-23 are write-as-0, 24-31 are write ignored */
-        val &= 0xf;
-        asm volatile("mcr p14, 0, %0, c0, c1, 0" : : "r" (val));
-}
-
-static inline u32
-xscale2pmu_read_overflow_flags(void)
-{
-        u32 val;
-        asm volatile("mrc p14, 0, %0, c5, c1, 0" : "=r" (val));
-        return val;
-}
-
-static inline void
-xscale2pmu_write_overflow_flags(u32 val)
-{
-        asm volatile("mcr p14, 0, %0, c5, c1, 0" : : "r" (val));
-}
-
-static inline u32
-xscale2pmu_read_event_select(void)
-{
-        u32 val;
-        asm volatile("mrc p14, 0, %0, c8, c1, 0" : "=r" (val));
-        return val;
-}
-
-static inline void
-xscale2pmu_write_event_select(u32 val)
-{
-        asm volatile("mcr p14, 0, %0, c8, c1, 0" : : "r"(val));
-}
-
-static inline u32
-xscale2pmu_read_int_enable(void)
-{
-        u32 val;
-        asm volatile("mrc p14, 0, %0, c4, c1, 0" : "=r" (val));
-        return val;
-}
-
-static void
-xscale2pmu_write_int_enable(u32 val)
-{
-        asm volatile("mcr p14, 0, %0, c4, c1, 0" : : "r" (val));
-}
-
-static inline int
-xscale2_pmnc_counter_has_overflowed(unsigned long of_flags,
-                                        enum xscale_counters counter)
-{
-        int ret = 0;
-
-        switch (counter) {
-        case XSCALE_CYCLE_COUNTER:
-                ret = of_flags & XSCALE2_CCOUNT_OVERFLOW;
-                break;
-        case XSCALE_COUNTER0:
-                ret = of_flags & XSCALE2_COUNT0_OVERFLOW;
-                break;
-        case XSCALE_COUNTER1:
-                ret = of_flags & XSCALE2_COUNT1_OVERFLOW;
-                break;
-        case XSCALE_COUNTER2:
-                ret = of_flags & XSCALE2_COUNT2_OVERFLOW;
-                break;
-        case XSCALE_COUNTER3:
-                ret = of_flags & XSCALE2_COUNT3_OVERFLOW;
-                break;
-        default:
-                WARN_ONCE(1, "invalid counter number (%d)\n", counter);
-        }
-
-        return ret;
-}
-
-static irqreturn_t
-xscale2pmu_handle_irq(int irq_num, void *dev)
-{
-        unsigned long pmnc, of_flags;
-        struct perf_sample_data data;
-        struct cpu_hw_events *cpuc;
-        struct pt_regs *regs;
-        int idx;
-
-        /* Disable the PMU. */
-        pmnc = xscale2pmu_read_pmnc();
-        xscale2pmu_write_pmnc(pmnc & ~XSCALE_PMU_ENABLE);
-
-        /* Check the overflow flag register. */
-        of_flags = xscale2pmu_read_overflow_flags();
-        if (!(of_flags & XSCALE2_OVERFLOWED_MASK))
-                return IRQ_NONE;
-
-        /* Clear the overflow bits. */
-        xscale2pmu_write_overflow_flags(of_flags);
-
-        regs = get_irq_regs();
-
-        perf_sample_data_init(&data, 0);
-
-        cpuc = &__get_cpu_var(cpu_hw_events);
-        for (idx = 0; idx <= armpmu->num_events; ++idx) {
-                struct perf_event *event = cpuc->events[idx];
-                struct hw_perf_event *hwc;
-
-                if (!test_bit(idx, cpuc->active_mask))
-                        continue;
-
-                if (!xscale2_pmnc_counter_has_overflowed(pmnc, idx))
-                        continue;
-
-                hwc = &event->hw;
-                armpmu_event_update(event, hwc, idx);
-                data.period = event->hw.last_period;
-                if (!armpmu_event_set_period(event, hwc, idx))
-                        continue;
-
-                if (perf_event_overflow(event, 0, &data, regs))
-                        armpmu->disable(hwc, idx);
-        }
-
-        irq_work_run();
-
-        /*
-         * Re-enable the PMU.
-         */
-        pmnc = xscale2pmu_read_pmnc() | XSCALE_PMU_ENABLE;
-        xscale2pmu_write_pmnc(pmnc);
-
-        return IRQ_HANDLED;
-}
-
-static void
-xscale2pmu_enable_event(struct hw_perf_event *hwc, int idx)
-{
-        unsigned long flags, ien, evtsel;
-
-        ien = xscale2pmu_read_int_enable();
-        evtsel = xscale2pmu_read_event_select();
-
-        switch (idx) {
-        case XSCALE_CYCLE_COUNTER:
-                ien |= XSCALE2_CCOUNT_INT_EN;
-                break;
-        case XSCALE_COUNTER0:
-                ien |= XSCALE2_COUNT0_INT_EN;
-                evtsel &= ~XSCALE2_COUNT0_EVT_MASK;
-                evtsel |= hwc->config_base << XSCALE2_COUNT0_EVT_SHFT;
-                break;
-        case XSCALE_COUNTER1:
-                ien |= XSCALE2_COUNT1_INT_EN;
-                evtsel &= ~XSCALE2_COUNT1_EVT_MASK;
-                evtsel |= hwc->config_base << XSCALE2_COUNT1_EVT_SHFT;
-                break;
-        case XSCALE_COUNTER2:
-                ien |= XSCALE2_COUNT2_INT_EN;
-                evtsel &= ~XSCALE2_COUNT2_EVT_MASK;
-                evtsel |= hwc->config_base << XSCALE2_COUNT2_EVT_SHFT;
-                break;
-        case XSCALE_COUNTER3:
-                ien |= XSCALE2_COUNT3_INT_EN;
-                evtsel &= ~XSCALE2_COUNT3_EVT_MASK;
-                evtsel |= hwc->config_base << XSCALE2_COUNT3_EVT_SHFT;
-                break;
-        default:
-                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
-                return;
-        }
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        xscale2pmu_write_event_select(evtsel);
-        xscale2pmu_write_int_enable(ien);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static void
-xscale2pmu_disable_event(struct hw_perf_event *hwc, int idx)
-{
-        unsigned long flags, ien, evtsel;
-
-        ien = xscale2pmu_read_int_enable();
-        evtsel = xscale2pmu_read_event_select();
-
-        switch (idx) {
-        case XSCALE_CYCLE_COUNTER:
-                ien &= ~XSCALE2_CCOUNT_INT_EN;
-                break;
-        case XSCALE_COUNTER0:
-                ien &= ~XSCALE2_COUNT0_INT_EN;
-                evtsel &= ~XSCALE2_COUNT0_EVT_MASK;
-                evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT0_EVT_SHFT;
-                break;
-        case XSCALE_COUNTER1:
-                ien &= ~XSCALE2_COUNT1_INT_EN;
-                evtsel &= ~XSCALE2_COUNT1_EVT_MASK;
-                evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT1_EVT_SHFT;
-                break;
-        case XSCALE_COUNTER2:
-                ien &= ~XSCALE2_COUNT2_INT_EN;
-                evtsel &= ~XSCALE2_COUNT2_EVT_MASK;
-                evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT2_EVT_SHFT;
-                break;
-        case XSCALE_COUNTER3:
-                ien &= ~XSCALE2_COUNT3_INT_EN;
-                evtsel &= ~XSCALE2_COUNT3_EVT_MASK;
-                evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT3_EVT_SHFT;
-                break;
-        default:
-                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
-                return;
-        }
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        xscale2pmu_write_event_select(evtsel);
-        xscale2pmu_write_int_enable(ien);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static int
-xscale2pmu_get_event_idx(struct cpu_hw_events *cpuc,
-                        struct hw_perf_event *event)
-{
-        int idx = xscale1pmu_get_event_idx(cpuc, event);
-        if (idx >= 0)
-                goto out;
-
-        if (!test_and_set_bit(XSCALE_COUNTER3, cpuc->used_mask))
-                idx = XSCALE_COUNTER3;
-        else if (!test_and_set_bit(XSCALE_COUNTER2, cpuc->used_mask))
-                idx = XSCALE_COUNTER2;
-out:
-        return idx;
-}
-
-static void
-xscale2pmu_start(void)
-{
-        unsigned long flags, val;
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = xscale2pmu_read_pmnc() & ~XSCALE_PMU_CNT64;
-        val |= XSCALE_PMU_ENABLE;
-        xscale2pmu_write_pmnc(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static void
-xscale2pmu_stop(void)
-{
-        unsigned long flags, val;
-
-        spin_lock_irqsave(&pmu_lock, flags);
-        val = xscale2pmu_read_pmnc();
-        val &= ~XSCALE_PMU_ENABLE;
-        xscale2pmu_write_pmnc(val);
-        spin_unlock_irqrestore(&pmu_lock, flags);
-}
-
-static inline u32
-xscale2pmu_read_counter(int counter)
-{
-        u32 val = 0;
-
-        switch (counter) {
-        case XSCALE_CYCLE_COUNTER:
-                asm volatile("mrc p14, 0, %0, c1, c1, 0" : "=r" (val));
-                break;
-        case XSCALE_COUNTER0:
-                asm volatile("mrc p14, 0, %0, c0, c2, 0" : "=r" (val));
-                break;
-        case XSCALE_COUNTER1:
-                asm volatile("mrc p14, 0, %0, c1, c2, 0" : "=r" (val));
-                break;
-        case XSCALE_COUNTER2:
-                asm volatile("mrc p14, 0, %0, c2, c2, 0" : "=r" (val));
-                break;
-        case XSCALE_COUNTER3:
-                asm volatile("mrc p14, 0, %0, c3, c2, 0" : "=r" (val));
-                break;
-        }
-
-        return val;
-}
-
-static inline void
-xscale2pmu_write_counter(int counter, u32 val)
-{
-        switch (counter) {
-        case XSCALE_CYCLE_COUNTER:
-                asm volatile("mcr p14, 0, %0, c1, c1, 0" : : "r" (val));
-                break;
-        case XSCALE_COUNTER0:
-                asm volatile("mcr p14, 0, %0, c0, c2, 0" : : "r" (val));
-                break;
-        case XSCALE_COUNTER1:
-                asm volatile("mcr p14, 0, %0, c1, c2, 0" : : "r" (val));
-                break;
-        case XSCALE_COUNTER2:
-                asm volatile("mcr p14, 0, %0, c2, c2, 0" : : "r" (val));
-                break;
-        case XSCALE_COUNTER3:
-                asm volatile("mcr p14, 0, %0, c3, c2, 0" : : "r" (val));
-                break;
-        }
-}
-
-static const struct arm_pmu xscale2pmu = {
-        .id             = ARM_PERF_PMU_ID_XSCALE2,
-        .handle_irq     = xscale2pmu_handle_irq,
-        .enable         = xscale2pmu_enable_event,
-        .disable        = xscale2pmu_disable_event,
-        .event_map      = xscalepmu_event_map,
-        .raw_event      = xscalepmu_raw_event,
-        .read_counter   = xscale2pmu_read_counter,
-        .write_counter  = xscale2pmu_write_counter,
-        .get_event_idx  = xscale2pmu_get_event_idx,
-        .start          = xscale2pmu_start,
-        .stop           = xscale2pmu_stop,
-        .num_events     = 5,
-        .max_period     = (1LLU << 32) - 1,
-};
+/* Include the PMU-specific implementations. */
+#include "perf_event_xscale.c"
+#include "perf_event_v6.c"
+#include "perf_event_v7.c"
 
 static int __init
 init_hw_perf_events(void)
@@ -2977,37 +622,16 @@ init_hw_perf_events(void)
                 case 0xB360:    /* ARM1136 */
                 case 0xB560:    /* ARM1156 */
                 case 0xB760:    /* ARM1176 */
-                        armpmu = &armv6pmu;
-                        memcpy(armpmu_perf_cache_map, armv6_perf_cache_map,
-                                        sizeof(armv6_perf_cache_map));
+                        armpmu = armv6pmu_init();
                         break;
                 case 0xB020:    /* ARM11mpcore */
-                        armpmu = &armv6mpcore_pmu;
-                        memcpy(armpmu_perf_cache_map,
-                               armv6mpcore_perf_cache_map,
-                               sizeof(armv6mpcore_perf_cache_map));
+                        armpmu = armv6mpcore_pmu_init();
                         break;
                 case 0xC080:    /* Cortex-A8 */
-                        armv7pmu.id = ARM_PERF_PMU_ID_CA8;
-                        memcpy(armpmu_perf_cache_map, armv7_a8_perf_cache_map,
-                                sizeof(armv7_a8_perf_cache_map));
-                        armv7pmu.event_map = armv7_a8_pmu_event_map;
-                        armpmu = &armv7pmu;
-
-                        /* Reset PMNC and read the nb of CNTx counters
-                            supported */
-                        armv7pmu.num_events = armv7_reset_read_pmnc();
+                        armpmu = armv7_a8_pmu_init();
                         break;
                 case 0xC090:    /* Cortex-A9 */
-                        armv7pmu.id = ARM_PERF_PMU_ID_CA9;
-                        memcpy(armpmu_perf_cache_map, armv7_a9_perf_cache_map,
-                                sizeof(armv7_a9_perf_cache_map));
-                        armv7pmu.event_map = armv7_a9_pmu_event_map;
-                        armpmu = &armv7pmu;
-
-                        /* Reset PMNC and read the nb of CNTx counters
-                            supported */
-                        armv7pmu.num_events = armv7_reset_read_pmnc();
+                        armpmu = armv7_a9_pmu_init();
                         break;
                 }
         /* Intel CPUs [xscale]. */
@@ -3015,21 +639,17 @@ init_hw_perf_events(void)
                 part_number = (cpuid >> 13) & 0x7;
                 switch (part_number) {
                 case 1:
-                        armpmu = &xscale1pmu;
-                        memcpy(armpmu_perf_cache_map, xscale_perf_cache_map,
-                                        sizeof(xscale_perf_cache_map));
+                        armpmu = xscale1pmu_init();
                         break;
                 case 2:
-                        armpmu = &xscale2pmu;
-                        memcpy(armpmu_perf_cache_map, xscale_perf_cache_map,
-                                        sizeof(xscale_perf_cache_map));
+                        armpmu = xscale2pmu_init();
                         break;
                 }
         }
 
         if (armpmu) {
                 pr_info("enabled with %s PMU driver, %d counters available\n",
-                                arm_pmu_names[armpmu->id], armpmu->num_events);
+                        armpmu->name, armpmu->num_events);
         } else {
                 pr_info("no hardware support available\n");
         }
diff --git a/arch/arm/kernel/perf_event_v6.c b/arch/arm/kernel/perf_event_v6.c
new file mode 100644
index 000000000000..7aeb07da9076
--- /dev/null
+++ b/arch/arm/kernel/perf_event_v6.c
@@ -0,0 +1,672 @@
+/*
+ * ARMv6 Performance counter handling code.
+ *
+ * Copyright (C) 2009 picoChip Designs, Ltd., Jamie Iles
+ *
+ * ARMv6 has 2 configurable performance counters and a single cycle counter.
+ * They all share a single reset bit but can be written to zero so we can use
+ * that for a reset.
+ *
+ * The counters can't be individually enabled or disabled so when we remove
+ * one event and replace it with another we could get spurious counts from the
+ * wrong event. However, we can take advantage of the fact that the
+ * performance counters can export events to the event bus, and the event bus
+ * itself can be monitored. This requires that we *don't* export the events to
+ * the event bus. The procedure for disabling a configurable counter is:
+ *      - change the counter to count the ETMEXTOUT[0] signal (0x20). This
+ *        effectively stops the counter from counting.
+ *      - disable the counter's interrupt generation (each counter has it's
+ *        own interrupt enable bit).
+ * Once stopped, the counter value can be written as 0 to reset.
+ *
+ * To enable a counter:
+ *      - enable the counter's interrupt generation.
+ *      - set the new event type.
+ *
+ * Note: the dedicated cycle counter only counts cycles and can't be
+ * enabled/disabled independently of the others. When we want to disable the
+ * cycle counter, we have to just disable the interrupt reporting and start
+ * ignoring that counter. When re-enabling, we have to reset the value and
+ * enable the interrupt.
+ */
+
+#ifdef CONFIG_CPU_V6
+enum armv6_perf_types {
+        ARMV6_PERFCTR_ICACHE_MISS           = 0x0,
+        ARMV6_PERFCTR_IBUF_STALL            = 0x1,
+        ARMV6_PERFCTR_DDEP_STALL            = 0x2,
+        ARMV6_PERFCTR_ITLB_MISS             = 0x3,
+        ARMV6_PERFCTR_DTLB_MISS             = 0x4,
+        ARMV6_PERFCTR_BR_EXEC               = 0x5,
+        ARMV6_PERFCTR_BR_MISPREDICT         = 0x6,
+        ARMV6_PERFCTR_INSTR_EXEC            = 0x7,
+        ARMV6_PERFCTR_DCACHE_HIT            = 0x9,
+        ARMV6_PERFCTR_DCACHE_ACCESS         = 0xA,
+        ARMV6_PERFCTR_DCACHE_MISS           = 0xB,
+        ARMV6_PERFCTR_DCACHE_WBACK          = 0xC,
+        ARMV6_PERFCTR_SW_PC_CHANGE          = 0xD,
+        ARMV6_PERFCTR_MAIN_TLB_MISS         = 0xF,
+        ARMV6_PERFCTR_EXPL_D_ACCESS         = 0x10,
+        ARMV6_PERFCTR_LSU_FULL_STALL        = 0x11,
+        ARMV6_PERFCTR_WBUF_DRAINED          = 0x12,
+        ARMV6_PERFCTR_CPU_CYCLES            = 0xFF,
+        ARMV6_PERFCTR_NOP                   = 0x20,
+};
+
+enum armv6_counters {
+        ARMV6_CYCLE_COUNTER = 1,
+        ARMV6_COUNTER0,
+        ARMV6_COUNTER1,
+};
+
+/*
+ * The hardware events that we support. We do support cache operations but
+ * we have harvard caches and no way to combine instruction and data
+ * accesses/misses in hardware.
+ */
+static const unsigned armv6_perf_map[PERF_COUNT_HW_MAX] = {
+        [PERF_COUNT_HW_CPU_CYCLES]          = ARMV6_PERFCTR_CPU_CYCLES,
+        [PERF_COUNT_HW_INSTRUCTIONS]        = ARMV6_PERFCTR_INSTR_EXEC,
+        [PERF_COUNT_HW_CACHE_REFERENCES]    = HW_OP_UNSUPPORTED,
+        [PERF_COUNT_HW_CACHE_MISSES]        = HW_OP_UNSUPPORTED,
+        [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = ARMV6_PERFCTR_BR_EXEC,
+        [PERF_COUNT_HW_BRANCH_MISSES]       = ARMV6_PERFCTR_BR_MISPREDICT,
+        [PERF_COUNT_HW_BUS_CYCLES]          = HW_OP_UNSUPPORTED,
+};
+
+static const unsigned armv6_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
+                                          [PERF_COUNT_HW_CACHE_OP_MAX]
+                                          [PERF_COUNT_HW_CACHE_RESULT_MAX] = {
+        [C(L1D)] = {
+                /*
+                 * The performance counters don't differentiate between read
+                 * and write accesses/misses so this isn't strictly correct,
+                 * but it's the best we can do. Writes and reads get
+                 * combined.
+                 */
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = ARMV6_PERFCTR_DCACHE_ACCESS,
+                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_DCACHE_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = ARMV6_PERFCTR_DCACHE_ACCESS,
+                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_DCACHE_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(L1I)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_ICACHE_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_ICACHE_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(LL)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(DTLB)] = {
+                /*
+                 * The ARM performance counters can count micro DTLB misses,
+                 * micro ITLB misses and main TLB misses. There isn't an event
+                 * for TLB misses, so use the micro misses here and if users
+                 * want the main TLB misses they can use a raw counter.
+                 */
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_DTLB_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_DTLB_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(ITLB)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_ITLB_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV6_PERFCTR_ITLB_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(BPU)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+};
+
+enum armv6mpcore_perf_types {
+        ARMV6MPCORE_PERFCTR_ICACHE_MISS     = 0x0,
+        ARMV6MPCORE_PERFCTR_IBUF_STALL      = 0x1,
+        ARMV6MPCORE_PERFCTR_DDEP_STALL      = 0x2,
+        ARMV6MPCORE_PERFCTR_ITLB_MISS       = 0x3,
+        ARMV6MPCORE_PERFCTR_DTLB_MISS       = 0x4,
+        ARMV6MPCORE_PERFCTR_BR_EXEC         = 0x5,
+        ARMV6MPCORE_PERFCTR_BR_NOTPREDICT   = 0x6,
+        ARMV6MPCORE_PERFCTR_BR_MISPREDICT   = 0x7,
+        ARMV6MPCORE_PERFCTR_INSTR_EXEC      = 0x8,
+        ARMV6MPCORE_PERFCTR_DCACHE_RDACCESS = 0xA,
+        ARMV6MPCORE_PERFCTR_DCACHE_RDMISS   = 0xB,
+        ARMV6MPCORE_PERFCTR_DCACHE_WRACCESS = 0xC,
+        ARMV6MPCORE_PERFCTR_DCACHE_WRMISS   = 0xD,
+        ARMV6MPCORE_PERFCTR_DCACHE_EVICTION = 0xE,
+        ARMV6MPCORE_PERFCTR_SW_PC_CHANGE    = 0xF,
+        ARMV6MPCORE_PERFCTR_MAIN_TLB_MISS   = 0x10,
+        ARMV6MPCORE_PERFCTR_EXPL_MEM_ACCESS = 0x11,
+        ARMV6MPCORE_PERFCTR_LSU_FULL_STALL  = 0x12,
+        ARMV6MPCORE_PERFCTR_WBUF_DRAINED    = 0x13,
+        ARMV6MPCORE_PERFCTR_CPU_CYCLES      = 0xFF,
+};
+
+/*
+ * The hardware events that we support. We do support cache operations but
+ * we have harvard caches and no way to combine instruction and data
+ * accesses/misses in hardware.
+ */
+static const unsigned armv6mpcore_perf_map[PERF_COUNT_HW_MAX] = {
+        [PERF_COUNT_HW_CPU_CYCLES]          = ARMV6MPCORE_PERFCTR_CPU_CYCLES,
+        [PERF_COUNT_HW_INSTRUCTIONS]        = ARMV6MPCORE_PERFCTR_INSTR_EXEC,
+        [PERF_COUNT_HW_CACHE_REFERENCES]    = HW_OP_UNSUPPORTED,
+        [PERF_COUNT_HW_CACHE_MISSES]        = HW_OP_UNSUPPORTED,
+        [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = ARMV6MPCORE_PERFCTR_BR_EXEC,
+        [PERF_COUNT_HW_BRANCH_MISSES]       = ARMV6MPCORE_PERFCTR_BR_MISPREDICT,
+        [PERF_COUNT_HW_BUS_CYCLES]          = HW_OP_UNSUPPORTED,
+};
+
+static const unsigned armv6mpcore_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
+                                        [PERF_COUNT_HW_CACHE_OP_MAX]
+                                        [PERF_COUNT_HW_CACHE_RESULT_MAX] = {
+        [C(L1D)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]  =
+                                ARMV6MPCORE_PERFCTR_DCACHE_RDACCESS,
+                        [C(RESULT_MISS)]    =
+                                ARMV6MPCORE_PERFCTR_DCACHE_RDMISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]  =
+                                ARMV6MPCORE_PERFCTR_DCACHE_WRACCESS,
+                        [C(RESULT_MISS)]    =
+                                ARMV6MPCORE_PERFCTR_DCACHE_WRMISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(L1I)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_ICACHE_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_ICACHE_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(LL)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(DTLB)] = {
+                /*
+                 * The ARM performance counters can count micro DTLB misses,
+                 * micro ITLB misses and main TLB misses. There isn't an event
+                 * for TLB misses, so use the micro misses here and if users
+                 * want the main TLB misses they can use a raw counter.
+                 */
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_DTLB_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_DTLB_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(ITLB)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_ITLB_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = ARMV6MPCORE_PERFCTR_ITLB_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(BPU)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]  = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]    = CACHE_OP_UNSUPPORTED,
+                },
+        },
+};
+
+static inline unsigned long
+armv6_pmcr_read(void)
+{
+        u32 val;
+        asm volatile("mrc   p15, 0, %0, c15, c12, 0" : "=r"(val));
+        return val;
+}
+
+static inline void
+armv6_pmcr_write(unsigned long val)
+{
+        asm volatile("mcr   p15, 0, %0, c15, c12, 0" : : "r"(val));
+}
+
+#define ARMV6_PMCR_ENABLE               (1 << 0)
+#define ARMV6_PMCR_CTR01_RESET          (1 << 1)
+#define ARMV6_PMCR_CCOUNT_RESET         (1 << 2)
+#define ARMV6_PMCR_CCOUNT_DIV           (1 << 3)
+#define ARMV6_PMCR_COUNT0_IEN           (1 << 4)
+#define ARMV6_PMCR_COUNT1_IEN           (1 << 5)
+#define ARMV6_PMCR_CCOUNT_IEN           (1 << 6)
+#define ARMV6_PMCR_COUNT0_OVERFLOW      (1 << 8)
+#define ARMV6_PMCR_COUNT1_OVERFLOW      (1 << 9)
+#define ARMV6_PMCR_CCOUNT_OVERFLOW      (1 << 10)
+#define ARMV6_PMCR_EVT_COUNT0_SHIFT     20
+#define ARMV6_PMCR_EVT_COUNT0_MASK      (0xFF << ARMV6_PMCR_EVT_COUNT0_SHIFT)
+#define ARMV6_PMCR_EVT_COUNT1_SHIFT     12
+#define ARMV6_PMCR_EVT_COUNT1_MASK      (0xFF << ARMV6_PMCR_EVT_COUNT1_SHIFT)
+
+#define ARMV6_PMCR_OVERFLOWED_MASK \
+        (ARMV6_PMCR_COUNT0_OVERFLOW | ARMV6_PMCR_COUNT1_OVERFLOW | \
+         ARMV6_PMCR_CCOUNT_OVERFLOW)
+
+static inline int
+armv6_pmcr_has_overflowed(unsigned long pmcr)
+{
+        return pmcr & ARMV6_PMCR_OVERFLOWED_MASK;
+}
+
+static inline int
+armv6_pmcr_counter_has_overflowed(unsigned long pmcr,
+                                  enum armv6_counters counter)
+{
+        int ret = 0;
+
+        if (ARMV6_CYCLE_COUNTER == counter)
+                ret = pmcr & ARMV6_PMCR_CCOUNT_OVERFLOW;
+        else if (ARMV6_COUNTER0 == counter)
+                ret = pmcr & ARMV6_PMCR_COUNT0_OVERFLOW;
+        else if (ARMV6_COUNTER1 == counter)
+                ret = pmcr & ARMV6_PMCR_COUNT1_OVERFLOW;
+        else
+                WARN_ONCE(1, "invalid counter number (%d)\n", counter);
+
+        return ret;
+}
+
+static inline u32
+armv6pmu_read_counter(int counter)
+{
+        unsigned long value = 0;
+
+        if (ARMV6_CYCLE_COUNTER == counter)
+                asm volatile("mrc   p15, 0, %0, c15, c12, 1" : "=r"(value));
+        else if (ARMV6_COUNTER0 == counter)
+                asm volatile("mrc   p15, 0, %0, c15, c12, 2" : "=r"(value));
+        else if (ARMV6_COUNTER1 == counter)
+                asm volatile("mrc   p15, 0, %0, c15, c12, 3" : "=r"(value));
+        else
+                WARN_ONCE(1, "invalid counter number (%d)\n", counter);
+
+        return value;
+}
+
+static inline void
+armv6pmu_write_counter(int counter,
+                       u32 value)
+{
+        if (ARMV6_CYCLE_COUNTER == counter)
+                asm volatile("mcr   p15, 0, %0, c15, c12, 1" : : "r"(value));
+        else if (ARMV6_COUNTER0 == counter)
+                asm volatile("mcr   p15, 0, %0, c15, c12, 2" : : "r"(value));
+        else if (ARMV6_COUNTER1 == counter)
+                asm volatile("mcr   p15, 0, %0, c15, c12, 3" : : "r"(value));
+        else
+                WARN_ONCE(1, "invalid counter number (%d)\n", counter);
+}
+
+void
+armv6pmu_enable_event(struct hw_perf_event *hwc,
+                      int idx)
+{
+        unsigned long val, mask, evt, flags;
+
+        if (ARMV6_CYCLE_COUNTER == idx) {
+                mask    = 0;
+                evt     = ARMV6_PMCR_CCOUNT_IEN;
+        } else if (ARMV6_COUNTER0 == idx) {
+                mask    = ARMV6_PMCR_EVT_COUNT0_MASK;
+                evt     = (hwc->config_base << ARMV6_PMCR_EVT_COUNT0_SHIFT) |
+                          ARMV6_PMCR_COUNT0_IEN;
+        } else if (ARMV6_COUNTER1 == idx) {
+                mask    = ARMV6_PMCR_EVT_COUNT1_MASK;
+                evt     = (hwc->config_base << ARMV6_PMCR_EVT_COUNT1_SHIFT) |
+                          ARMV6_PMCR_COUNT1_IEN;
+        } else {
+                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
+                return;
+        }
+
+        /*
+         * Mask out the current event and set the counter to count the event
+         * that we're interested in.
+         */
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = armv6_pmcr_read();
+        val &= ~mask;
+        val |= evt;
+        armv6_pmcr_write(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static irqreturn_t
+armv6pmu_handle_irq(int irq_num,
+                    void *dev)
+{
+        unsigned long pmcr = armv6_pmcr_read();
+        struct perf_sample_data data;
+        struct cpu_hw_events *cpuc;
+        struct pt_regs *regs;
+        int idx;
+
+        if (!armv6_pmcr_has_overflowed(pmcr))
+                return IRQ_NONE;
+
+        regs = get_irq_regs();
+
+        /*
+         * The interrupts are cleared by writing the overflow flags back to
+         * the control register. All of the other bits don't have any effect
+         * if they are rewritten, so write the whole value back.
+         */
+        armv6_pmcr_write(pmcr);
+
+        perf_sample_data_init(&data, 0);
+
+        cpuc = &__get_cpu_var(cpu_hw_events);
+        for (idx = 0; idx <= armpmu->num_events; ++idx) {
+                struct perf_event *event = cpuc->events[idx];
+                struct hw_perf_event *hwc;
+
+                if (!test_bit(idx, cpuc->active_mask))
+                        continue;
+
+                /*
+                 * We have a single interrupt for all counters. Check that
+                 * each counter has overflowed before we process it.
+                 */
+                if (!armv6_pmcr_counter_has_overflowed(pmcr, idx))
+                        continue;
+
+                hwc = &event->hw;
+                armpmu_event_update(event, hwc, idx);
+                data.period = event->hw.last_period;
+                if (!armpmu_event_set_period(event, hwc, idx))
+                        continue;
+
+                if (perf_event_overflow(event, 0, &data, regs))
+                        armpmu->disable(hwc, idx);
+        }
+
+        /*
+         * Handle the pending perf events.
+         *
+         * Note: this call *must* be run with interrupts disabled. For
+         * platforms that can have the PMU interrupts raised as an NMI, this
+         * will not work.
+         */
+        irq_work_run();
+
+        return IRQ_HANDLED;
+}
+
+static void
+armv6pmu_start(void)
+{
+        unsigned long flags, val;
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = armv6_pmcr_read();
+        val |= ARMV6_PMCR_ENABLE;
+        armv6_pmcr_write(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static void
+armv6pmu_stop(void)
+{
+        unsigned long flags, val;
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = armv6_pmcr_read();
+        val &= ~ARMV6_PMCR_ENABLE;
+        armv6_pmcr_write(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static int
+armv6pmu_get_event_idx(struct cpu_hw_events *cpuc,
+                       struct hw_perf_event *event)
+{
+        /* Always place a cycle counter into the cycle counter. */
+        if (ARMV6_PERFCTR_CPU_CYCLES == event->config_base) {
+                if (test_and_set_bit(ARMV6_CYCLE_COUNTER, cpuc->used_mask))
+                        return -EAGAIN;
+
+                return ARMV6_CYCLE_COUNTER;
+        } else {
+                /*
+                 * For anything other than a cycle counter, try and use
+                 * counter0 and counter1.
+                 */
+                if (!test_and_set_bit(ARMV6_COUNTER1, cpuc->used_mask))
+                        return ARMV6_COUNTER1;
+
+                if (!test_and_set_bit(ARMV6_COUNTER0, cpuc->used_mask))
+                        return ARMV6_COUNTER0;
+
+                /* The counters are all in use. */
+                return -EAGAIN;
+        }
+}
+
+static void
+armv6pmu_disable_event(struct hw_perf_event *hwc,
+                       int idx)
+{
+        unsigned long val, mask, evt, flags;
+
+        if (ARMV6_CYCLE_COUNTER == idx) {
+                mask    = ARMV6_PMCR_CCOUNT_IEN;
+                evt     = 0;
+        } else if (ARMV6_COUNTER0 == idx) {
+                mask    = ARMV6_PMCR_COUNT0_IEN | ARMV6_PMCR_EVT_COUNT0_MASK;
+                evt     = ARMV6_PERFCTR_NOP << ARMV6_PMCR_EVT_COUNT0_SHIFT;
+        } else if (ARMV6_COUNTER1 == idx) {
+                mask    = ARMV6_PMCR_COUNT1_IEN | ARMV6_PMCR_EVT_COUNT1_MASK;
+                evt     = ARMV6_PERFCTR_NOP << ARMV6_PMCR_EVT_COUNT1_SHIFT;
+        } else {
+                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
+                return;
+        }
+
+        /*
+         * Mask out the current event and set the counter to count the number
+         * of ETM bus signal assertion cycles. The external reporting should
+         * be disabled and so this should never increment.
+         */
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = armv6_pmcr_read();
+        val &= ~mask;
+        val |= evt;
+        armv6_pmcr_write(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static void
+armv6mpcore_pmu_disable_event(struct hw_perf_event *hwc,
+                              int idx)
+{
+        unsigned long val, mask, flags, evt = 0;
+
+        if (ARMV6_CYCLE_COUNTER == idx) {
+                mask    = ARMV6_PMCR_CCOUNT_IEN;
+        } else if (ARMV6_COUNTER0 == idx) {
+                mask    = ARMV6_PMCR_COUNT0_IEN;
+        } else if (ARMV6_COUNTER1 == idx) {
+                mask    = ARMV6_PMCR_COUNT1_IEN;
+        } else {
+                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
+                return;
+        }
+
+        /*
+         * Unlike UP ARMv6, we don't have a way of stopping the counters. We
+         * simply disable the interrupt reporting.
+         */
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = armv6_pmcr_read();
+        val &= ~mask;
+        val |= evt;
+        armv6_pmcr_write(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static const struct arm_pmu armv6pmu = {
+        .id                     = ARM_PERF_PMU_ID_V6,
+        .name                   = "v6",
+        .handle_irq             = armv6pmu_handle_irq,
+        .enable                 = armv6pmu_enable_event,
+        .disable                = armv6pmu_disable_event,
+        .read_counter           = armv6pmu_read_counter,
+        .write_counter          = armv6pmu_write_counter,
+        .get_event_idx          = armv6pmu_get_event_idx,
+        .start                  = armv6pmu_start,
+        .stop                   = armv6pmu_stop,
+        .cache_map              = &armv6_perf_cache_map,
+        .event_map              = &armv6_perf_map,
+        .raw_event_mask         = 0xFF,
+        .num_events             = 3,
+        .max_period             = (1LLU << 32) - 1,
+};
+
+const struct arm_pmu *__init armv6pmu_init(void)
+{
+        return &armv6pmu;
+}
+
+/*
+ * ARMv6mpcore is almost identical to single core ARMv6 with the exception
+ * that some of the events have different enumerations and that there is no
+ * *hack* to stop the programmable counters. To stop the counters we simply
+ * disable the interrupt reporting and update the event. When unthrottling we
+ * reset the period and enable the interrupt reporting.
+ */
+static const struct arm_pmu armv6mpcore_pmu = {
+        .id                     = ARM_PERF_PMU_ID_V6MP,
+        .name                   = "v6mpcore",
+        .handle_irq             = armv6pmu_handle_irq,
+        .enable                 = armv6pmu_enable_event,
+        .disable                = armv6mpcore_pmu_disable_event,
+        .read_counter           = armv6pmu_read_counter,
+        .write_counter          = armv6pmu_write_counter,
+        .get_event_idx          = armv6pmu_get_event_idx,
+        .start                  = armv6pmu_start,
+        .stop                   = armv6pmu_stop,
+        .cache_map              = &armv6mpcore_perf_cache_map,
+        .event_map              = &armv6mpcore_perf_map,
+        .raw_event_mask         = 0xFF,
+        .num_events             = 3,
+        .max_period             = (1LLU << 32) - 1,
+};
+
+const struct arm_pmu *__init armv6mpcore_pmu_init(void)
+{
+        return &armv6mpcore_pmu;
+}
+#else
+const struct arm_pmu *__init armv6pmu_init(void)
+{
+        return NULL;
+}
+
+const struct arm_pmu *__init armv6mpcore_pmu_init(void)
+{
+        return NULL;
+}
+#endif  /* CONFIG_CPU_V6 */
diff --git a/arch/arm/kernel/perf_event_v7.c b/arch/arm/kernel/perf_event_v7.c
new file mode 100644
index 000000000000..4d0423969df9
--- /dev/null
+++ b/arch/arm/kernel/perf_event_v7.c
@@ -0,0 +1,906 @@
+/*
+ * ARMv7 Cortex-A8 and Cortex-A9 Performance Events handling code.
+ *
+ * ARMv7 support: Jean Pihet <jpihet@mvista.com>
+ * 2010 (c) MontaVista Software, LLC.
+ *
+ * Copied from ARMv6 code, with the low level code inspired
+ *  by the ARMv7 Oprofile code.
+ *
+ * Cortex-A8 has up to 4 configurable performance counters and
+ *  a single cycle counter.
+ * Cortex-A9 has up to 31 configurable performance counters and
+ *  a single cycle counter.
+ *
+ * All counters can be enabled/disabled and IRQ masked separately. The cycle
+ *  counter and all 4 performance counters together can be reset separately.
+ */
+
+#ifdef CONFIG_CPU_V7
+/* Common ARMv7 event types */
+enum armv7_perf_types {
+        ARMV7_PERFCTR_PMNC_SW_INCR              = 0x00,
+        ARMV7_PERFCTR_IFETCH_MISS               = 0x01,
+        ARMV7_PERFCTR_ITLB_MISS                 = 0x02,
+        ARMV7_PERFCTR_DCACHE_REFILL             = 0x03,
+        ARMV7_PERFCTR_DCACHE_ACCESS             = 0x04,
+        ARMV7_PERFCTR_DTLB_REFILL               = 0x05,
+        ARMV7_PERFCTR_DREAD                     = 0x06,
+        ARMV7_PERFCTR_DWRITE                    = 0x07,
+
+        ARMV7_PERFCTR_EXC_TAKEN                 = 0x09,
+        ARMV7_PERFCTR_EXC_EXECUTED              = 0x0A,
+        ARMV7_PERFCTR_CID_WRITE                 = 0x0B,
+        /* ARMV7_PERFCTR_PC_WRITE is equivalent to HW_BRANCH_INSTRUCTIONS.
+         * It counts:
+         *  - all branch instructions,
+         *  - instructions that explicitly write the PC,
+         *  - exception generating instructions.
+         */
+        ARMV7_PERFCTR_PC_WRITE                  = 0x0C,
+        ARMV7_PERFCTR_PC_IMM_BRANCH             = 0x0D,
+        ARMV7_PERFCTR_UNALIGNED_ACCESS          = 0x0F,
+        ARMV7_PERFCTR_PC_BRANCH_MIS_PRED        = 0x10,
+        ARMV7_PERFCTR_CLOCK_CYCLES              = 0x11,
+
+        ARMV7_PERFCTR_PC_BRANCH_MIS_USED        = 0x12,
+
+        ARMV7_PERFCTR_CPU_CYCLES                = 0xFF
+};
+
+/* ARMv7 Cortex-A8 specific event types */
+enum armv7_a8_perf_types {
+        ARMV7_PERFCTR_INSTR_EXECUTED            = 0x08,
+
+        ARMV7_PERFCTR_PC_PROC_RETURN            = 0x0E,
+
+        ARMV7_PERFCTR_WRITE_BUFFER_FULL         = 0x40,
+        ARMV7_PERFCTR_L2_STORE_MERGED           = 0x41,
+        ARMV7_PERFCTR_L2_STORE_BUFF             = 0x42,
+        ARMV7_PERFCTR_L2_ACCESS                 = 0x43,
+        ARMV7_PERFCTR_L2_CACH_MISS              = 0x44,
+        ARMV7_PERFCTR_AXI_READ_CYCLES           = 0x45,
+        ARMV7_PERFCTR_AXI_WRITE_CYCLES          = 0x46,
+        ARMV7_PERFCTR_MEMORY_REPLAY             = 0x47,
+        ARMV7_PERFCTR_UNALIGNED_ACCESS_REPLAY   = 0x48,
+        ARMV7_PERFCTR_L1_DATA_MISS              = 0x49,
+        ARMV7_PERFCTR_L1_INST_MISS              = 0x4A,
+        ARMV7_PERFCTR_L1_DATA_COLORING          = 0x4B,
+        ARMV7_PERFCTR_L1_NEON_DATA              = 0x4C,
+        ARMV7_PERFCTR_L1_NEON_CACH_DATA         = 0x4D,
+        ARMV7_PERFCTR_L2_NEON                   = 0x4E,
+        ARMV7_PERFCTR_L2_NEON_HIT               = 0x4F,
+        ARMV7_PERFCTR_L1_INST                   = 0x50,
+        ARMV7_PERFCTR_PC_RETURN_MIS_PRED        = 0x51,
+        ARMV7_PERFCTR_PC_BRANCH_FAILED          = 0x52,
+        ARMV7_PERFCTR_PC_BRANCH_TAKEN           = 0x53,
+        ARMV7_PERFCTR_PC_BRANCH_EXECUTED        = 0x54,
+        ARMV7_PERFCTR_OP_EXECUTED               = 0x55,
+        ARMV7_PERFCTR_CYCLES_INST_STALL         = 0x56,
+        ARMV7_PERFCTR_CYCLES_INST               = 0x57,
+        ARMV7_PERFCTR_CYCLES_NEON_DATA_STALL    = 0x58,
+        ARMV7_PERFCTR_CYCLES_NEON_INST_STALL    = 0x59,
+        ARMV7_PERFCTR_NEON_CYCLES               = 0x5A,
+
+        ARMV7_PERFCTR_PMU0_EVENTS               = 0x70,
+        ARMV7_PERFCTR_PMU1_EVENTS               = 0x71,
+        ARMV7_PERFCTR_PMU_EVENTS                = 0x72,
+};
+
+/* ARMv7 Cortex-A9 specific event types */
+enum armv7_a9_perf_types {
+        ARMV7_PERFCTR_JAVA_HW_BYTECODE_EXEC     = 0x40,
+        ARMV7_PERFCTR_JAVA_SW_BYTECODE_EXEC     = 0x41,
+        ARMV7_PERFCTR_JAZELLE_BRANCH_EXEC       = 0x42,
+
+        ARMV7_PERFCTR_COHERENT_LINE_MISS        = 0x50,
+        ARMV7_PERFCTR_COHERENT_LINE_HIT         = 0x51,
+
+        ARMV7_PERFCTR_ICACHE_DEP_STALL_CYCLES   = 0x60,
+        ARMV7_PERFCTR_DCACHE_DEP_STALL_CYCLES   = 0x61,
+        ARMV7_PERFCTR_TLB_MISS_DEP_STALL_CYCLES = 0x62,
+        ARMV7_PERFCTR_STREX_EXECUTED_PASSED     = 0x63,
+        ARMV7_PERFCTR_STREX_EXECUTED_FAILED     = 0x64,
+        ARMV7_PERFCTR_DATA_EVICTION             = 0x65,
+        ARMV7_PERFCTR_ISSUE_STAGE_NO_INST       = 0x66,
+        ARMV7_PERFCTR_ISSUE_STAGE_EMPTY         = 0x67,
+        ARMV7_PERFCTR_INST_OUT_OF_RENAME_STAGE  = 0x68,
+
+        ARMV7_PERFCTR_PREDICTABLE_FUNCT_RETURNS = 0x6E,
+
+        ARMV7_PERFCTR_MAIN_UNIT_EXECUTED_INST   = 0x70,
+        ARMV7_PERFCTR_SECOND_UNIT_EXECUTED_INST = 0x71,
+        ARMV7_PERFCTR_LD_ST_UNIT_EXECUTED_INST  = 0x72,
+        ARMV7_PERFCTR_FP_EXECUTED_INST          = 0x73,
+        ARMV7_PERFCTR_NEON_EXECUTED_INST        = 0x74,
+
+        ARMV7_PERFCTR_PLD_FULL_DEP_STALL_CYCLES = 0x80,
+        ARMV7_PERFCTR_DATA_WR_DEP_STALL_CYCLES  = 0x81,
+        ARMV7_PERFCTR_ITLB_MISS_DEP_STALL_CYCLES        = 0x82,
+        ARMV7_PERFCTR_DTLB_MISS_DEP_STALL_CYCLES        = 0x83,
+        ARMV7_PERFCTR_MICRO_ITLB_MISS_DEP_STALL_CYCLES  = 0x84,
+        ARMV7_PERFCTR_MICRO_DTLB_MISS_DEP_STALL_CYCLES  = 0x85,
+        ARMV7_PERFCTR_DMB_DEP_STALL_CYCLES      = 0x86,
+
+        ARMV7_PERFCTR_INTGR_CLK_ENABLED_CYCLES  = 0x8A,
+        ARMV7_PERFCTR_DATA_ENGINE_CLK_EN_CYCLES = 0x8B,
+
+        ARMV7_PERFCTR_ISB_INST                  = 0x90,
+        ARMV7_PERFCTR_DSB_INST                  = 0x91,
+        ARMV7_PERFCTR_DMB_INST                  = 0x92,
+        ARMV7_PERFCTR_EXT_INTERRUPTS            = 0x93,
+
+        ARMV7_PERFCTR_PLE_CACHE_LINE_RQST_COMPLETED     = 0xA0,
+        ARMV7_PERFCTR_PLE_CACHE_LINE_RQST_SKIPPED       = 0xA1,
+        ARMV7_PERFCTR_PLE_FIFO_FLUSH            = 0xA2,
+        ARMV7_PERFCTR_PLE_RQST_COMPLETED        = 0xA3,
+        ARMV7_PERFCTR_PLE_FIFO_OVERFLOW         = 0xA4,
+        ARMV7_PERFCTR_PLE_RQST_PROG             = 0xA5
+};
+
+/*
+ * Cortex-A8 HW events mapping
+ *
+ * The hardware events that we support. We do support cache operations but
+ * we have harvard caches and no way to combine instruction and data
+ * accesses/misses in hardware.
+ */
+static const unsigned armv7_a8_perf_map[PERF_COUNT_HW_MAX] = {
+        [PERF_COUNT_HW_CPU_CYCLES]          = ARMV7_PERFCTR_CPU_CYCLES,
+        [PERF_COUNT_HW_INSTRUCTIONS]        = ARMV7_PERFCTR_INSTR_EXECUTED,
+        [PERF_COUNT_HW_CACHE_REFERENCES]    = HW_OP_UNSUPPORTED,
+        [PERF_COUNT_HW_CACHE_MISSES]        = HW_OP_UNSUPPORTED,
+        [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = ARMV7_PERFCTR_PC_WRITE,
+        [PERF_COUNT_HW_BRANCH_MISSES]       = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
+        [PERF_COUNT_HW_BUS_CYCLES]          = ARMV7_PERFCTR_CLOCK_CYCLES,
+};
+
+static const unsigned armv7_a8_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
+                                          [PERF_COUNT_HW_CACHE_OP_MAX]
+                                          [PERF_COUNT_HW_CACHE_RESULT_MAX] = {
+        [C(L1D)] = {
+                /*
+                 * The performance counters don't differentiate between read
+                 * and write accesses/misses so this isn't strictly correct,
+                 * but it's the best we can do. Writes and reads get
+                 * combined.
+                 */
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_DCACHE_ACCESS,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DCACHE_REFILL,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_DCACHE_ACCESS,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DCACHE_REFILL,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(L1I)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_L1_INST,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_L1_INST_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_L1_INST,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_L1_INST_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(LL)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_L2_ACCESS,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_L2_CACH_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_L2_ACCESS,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_L2_CACH_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(DTLB)] = {
+                /*
+                 * Only ITLB misses and DTLB refills are supported.
+                 * If users want the DTLB refills misses a raw counter
+                 * must be used.
+                 */
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DTLB_REFILL,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DTLB_REFILL,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(ITLB)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_ITLB_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_ITLB_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(BPU)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_PC_WRITE,
+                        [C(RESULT_MISS)]
+                                        = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_PC_WRITE,
+                        [C(RESULT_MISS)]
+                                        = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+};
+
+/*
+ * Cortex-A9 HW events mapping
+ */
+static const unsigned armv7_a9_perf_map[PERF_COUNT_HW_MAX] = {
+        [PERF_COUNT_HW_CPU_CYCLES]          = ARMV7_PERFCTR_CPU_CYCLES,
+        [PERF_COUNT_HW_INSTRUCTIONS]        =
+                                        ARMV7_PERFCTR_INST_OUT_OF_RENAME_STAGE,
+        [PERF_COUNT_HW_CACHE_REFERENCES]    = ARMV7_PERFCTR_COHERENT_LINE_HIT,
+        [PERF_COUNT_HW_CACHE_MISSES]        = ARMV7_PERFCTR_COHERENT_LINE_MISS,
+        [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = ARMV7_PERFCTR_PC_WRITE,
+        [PERF_COUNT_HW_BRANCH_MISSES]       = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
+        [PERF_COUNT_HW_BUS_CYCLES]          = ARMV7_PERFCTR_CLOCK_CYCLES,
+};
+
+static const unsigned armv7_a9_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
+                                          [PERF_COUNT_HW_CACHE_OP_MAX]
+                                          [PERF_COUNT_HW_CACHE_RESULT_MAX] = {
+        [C(L1D)] = {
+                /*
+                 * The performance counters don't differentiate between read
+                 * and write accesses/misses so this isn't strictly correct,
+                 * but it's the best we can do. Writes and reads get
+                 * combined.
+                 */
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_DCACHE_ACCESS,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DCACHE_REFILL,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_DCACHE_ACCESS,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DCACHE_REFILL,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(L1I)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_IFETCH_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_IFETCH_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(LL)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(DTLB)] = {
+                /*
+                 * Only ITLB misses and DTLB refills are supported.
+                 * If users want the DTLB refills misses a raw counter
+                 * must be used.
+                 */
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DTLB_REFILL,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_DTLB_REFILL,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(ITLB)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_ITLB_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = ARMV7_PERFCTR_ITLB_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(BPU)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_PC_WRITE,
+                        [C(RESULT_MISS)]
+                                        = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = ARMV7_PERFCTR_PC_WRITE,
+                        [C(RESULT_MISS)]
+                                        = ARMV7_PERFCTR_PC_BRANCH_MIS_PRED,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+};
+
+/*
+ * Perf Events counters
+ */
+enum armv7_counters {
+        ARMV7_CYCLE_COUNTER             = 1,    /* Cycle counter */
+        ARMV7_COUNTER0                  = 2,    /* First event counter */
+};
+
+/*
+ * The cycle counter is ARMV7_CYCLE_COUNTER.
+ * The first event counter is ARMV7_COUNTER0.
+ * The last event counter is (ARMV7_COUNTER0 + armpmu->num_events - 1).
+ */
+#define ARMV7_COUNTER_LAST      (ARMV7_COUNTER0 + armpmu->num_events - 1)
+
+/*
+ * ARMv7 low level PMNC access
+ */
+
+/*
+ * Per-CPU PMNC: config reg
+ */
+#define ARMV7_PMNC_E            (1 << 0) /* Enable all counters */
+#define ARMV7_PMNC_P            (1 << 1) /* Reset all counters */
+#define ARMV7_PMNC_C            (1 << 2) /* Cycle counter reset */
+#define ARMV7_PMNC_D            (1 << 3) /* CCNT counts every 64th cpu cycle */
+#define ARMV7_PMNC_X            (1 << 4) /* Export to ETM */
+#define ARMV7_PMNC_DP           (1 << 5) /* Disable CCNT if non-invasive debug*/
+#define ARMV7_PMNC_N_SHIFT      11       /* Number of counters supported */
+#define ARMV7_PMNC_N_MASK       0x1f
+#define ARMV7_PMNC_MASK         0x3f     /* Mask for writable bits */
+
+/*
+ * Available counters
+ */
+#define ARMV7_CNT0              0       /* First event counter */
+#define ARMV7_CCNT              31      /* Cycle counter */
+
+/* Perf Event to low level counters mapping */
+#define ARMV7_EVENT_CNT_TO_CNTx (ARMV7_COUNTER0 - ARMV7_CNT0)
+
+/*
+ * CNTENS: counters enable reg
+ */
+#define ARMV7_CNTENS_P(idx)     (1 << (idx - ARMV7_EVENT_CNT_TO_CNTx))
+#define ARMV7_CNTENS_C          (1 << ARMV7_CCNT)
+
+/*
+ * CNTENC: counters disable reg
+ */
+#define ARMV7_CNTENC_P(idx)     (1 << (idx - ARMV7_EVENT_CNT_TO_CNTx))
+#define ARMV7_CNTENC_C          (1 << ARMV7_CCNT)
+
+/*
+ * INTENS: counters overflow interrupt enable reg
+ */
+#define ARMV7_INTENS_P(idx)     (1 << (idx - ARMV7_EVENT_CNT_TO_CNTx))
+#define ARMV7_INTENS_C          (1 << ARMV7_CCNT)
+
+/*
+ * INTENC: counters overflow interrupt disable reg
+ */
+#define ARMV7_INTENC_P(idx)     (1 << (idx - ARMV7_EVENT_CNT_TO_CNTx))
+#define ARMV7_INTENC_C          (1 << ARMV7_CCNT)
+
+/*
+ * EVTSEL: Event selection reg
+ */
+#define ARMV7_EVTSEL_MASK       0xff            /* Mask for writable bits */
+
+/*
+ * SELECT: Counter selection reg
+ */
+#define ARMV7_SELECT_MASK       0x1f            /* Mask for writable bits */
+
+/*
+ * FLAG: counters overflow flag status reg
+ */
+#define ARMV7_FLAG_P(idx)       (1 << (idx - ARMV7_EVENT_CNT_TO_CNTx))
+#define ARMV7_FLAG_C            (1 << ARMV7_CCNT)
+#define ARMV7_FLAG_MASK         0xffffffff      /* Mask for writable bits */
+#define ARMV7_OVERFLOWED_MASK   ARMV7_FLAG_MASK
+
+static inline unsigned long armv7_pmnc_read(void)
+{
+        u32 val;
+        asm volatile("mrc p15, 0, %0, c9, c12, 0" : "=r"(val));
+        return val;
+}
+
+static inline void armv7_pmnc_write(unsigned long val)
+{
+        val &= ARMV7_PMNC_MASK;
+        asm volatile("mcr p15, 0, %0, c9, c12, 0" : : "r"(val));
+}
+
+static inline int armv7_pmnc_has_overflowed(unsigned long pmnc)
+{
+        return pmnc & ARMV7_OVERFLOWED_MASK;
+}
+
+static inline int armv7_pmnc_counter_has_overflowed(unsigned long pmnc,
+                                        enum armv7_counters counter)
+{
+        int ret = 0;
+
+        if (counter == ARMV7_CYCLE_COUNTER)
+                ret = pmnc & ARMV7_FLAG_C;
+        else if ((counter >= ARMV7_COUNTER0) && (counter <= ARMV7_COUNTER_LAST))
+                ret = pmnc & ARMV7_FLAG_P(counter);
+        else
+                pr_err("CPU%u checking wrong counter %d overflow status\n",
+                        smp_processor_id(), counter);
+
+        return ret;
+}
+
+static inline int armv7_pmnc_select_counter(unsigned int idx)
+{
+        u32 val;
+
+        if ((idx < ARMV7_COUNTER0) || (idx > ARMV7_COUNTER_LAST)) {
+                pr_err("CPU%u selecting wrong PMNC counter"
+                        " %d\n", smp_processor_id(), idx);
+                return -1;
+        }
+
+        val = (idx - ARMV7_EVENT_CNT_TO_CNTx) & ARMV7_SELECT_MASK;
+        asm volatile("mcr p15, 0, %0, c9, c12, 5" : : "r" (val));
+
+        return idx;
+}
+
+static inline u32 armv7pmu_read_counter(int idx)
+{
+        unsigned long value = 0;
+
+        if (idx == ARMV7_CYCLE_COUNTER)
+                asm volatile("mrc p15, 0, %0, c9, c13, 0" : "=r" (value));
+        else if ((idx >= ARMV7_COUNTER0) && (idx <= ARMV7_COUNTER_LAST)) {
+                if (armv7_pmnc_select_counter(idx) == idx)
+                        asm volatile("mrc p15, 0, %0, c9, c13, 2"
+                                     : "=r" (value));
+        } else
+                pr_err("CPU%u reading wrong counter %d\n",
+                        smp_processor_id(), idx);
+
+        return value;
+}
+
+static inline void armv7pmu_write_counter(int idx, u32 value)
+{
+        if (idx == ARMV7_CYCLE_COUNTER)
+                asm volatile("mcr p15, 0, %0, c9, c13, 0" : : "r" (value));
+        else if ((idx >= ARMV7_COUNTER0) && (idx <= ARMV7_COUNTER_LAST)) {
+                if (armv7_pmnc_select_counter(idx) == idx)
+                        asm volatile("mcr p15, 0, %0, c9, c13, 2"
+                                     : : "r" (value));
+        } else
+                pr_err("CPU%u writing wrong counter %d\n",
+                        smp_processor_id(), idx);
+}
+
+static inline void armv7_pmnc_write_evtsel(unsigned int idx, u32 val)
+{
+        if (armv7_pmnc_select_counter(idx) == idx) {
+                val &= ARMV7_EVTSEL_MASK;
+                asm volatile("mcr p15, 0, %0, c9, c13, 1" : : "r" (val));
+        }
+}
+
+static inline u32 armv7_pmnc_enable_counter(unsigned int idx)
+{
+        u32 val;
+
+        if ((idx != ARMV7_CYCLE_COUNTER) &&
+            ((idx < ARMV7_COUNTER0) || (idx > ARMV7_COUNTER_LAST))) {
+                pr_err("CPU%u enabling wrong PMNC counter"
+                        " %d\n", smp_processor_id(), idx);
+                return -1;
+        }
+
+        if (idx == ARMV7_CYCLE_COUNTER)
+                val = ARMV7_CNTENS_C;
+        else
+                val = ARMV7_CNTENS_P(idx);
+
+        asm volatile("mcr p15, 0, %0, c9, c12, 1" : : "r" (val));
+
+        return idx;
+}
+
+static inline u32 armv7_pmnc_disable_counter(unsigned int idx)
+{
+        u32 val;
+
+
+        if ((idx != ARMV7_CYCLE_COUNTER) &&
+            ((idx < ARMV7_COUNTER0) || (idx > ARMV7_COUNTER_LAST))) {
+                pr_err("CPU%u disabling wrong PMNC counter"
+                        " %d\n", smp_processor_id(), idx);
+                return -1;
+        }
+
+        if (idx == ARMV7_CYCLE_COUNTER)
+                val = ARMV7_CNTENC_C;
+        else
+                val = ARMV7_CNTENC_P(idx);
+
+        asm volatile("mcr p15, 0, %0, c9, c12, 2" : : "r" (val));
+
+        return idx;
+}
+
+static inline u32 armv7_pmnc_enable_intens(unsigned int idx)
+{
+        u32 val;
+
+        if ((idx != ARMV7_CYCLE_COUNTER) &&
+            ((idx < ARMV7_COUNTER0) || (idx > ARMV7_COUNTER_LAST))) {
+                pr_err("CPU%u enabling wrong PMNC counter"
+                        " interrupt enable %d\n", smp_processor_id(), idx);
+                return -1;
+        }
+
+        if (idx == ARMV7_CYCLE_COUNTER)
+                val = ARMV7_INTENS_C;
+        else
+                val = ARMV7_INTENS_P(idx);
+
+        asm volatile("mcr p15, 0, %0, c9, c14, 1" : : "r" (val));
+
+        return idx;
+}
+
+static inline u32 armv7_pmnc_disable_intens(unsigned int idx)
+{
+        u32 val;
+
+        if ((idx != ARMV7_CYCLE_COUNTER) &&
+            ((idx < ARMV7_COUNTER0) || (idx > ARMV7_COUNTER_LAST))) {
+                pr_err("CPU%u disabling wrong PMNC counter"
+                        " interrupt enable %d\n", smp_processor_id(), idx);
+                return -1;
+        }
+
+        if (idx == ARMV7_CYCLE_COUNTER)
+                val = ARMV7_INTENC_C;
+        else
+                val = ARMV7_INTENC_P(idx);
+
+        asm volatile("mcr p15, 0, %0, c9, c14, 2" : : "r" (val));
+
+        return idx;
+}
+
+static inline u32 armv7_pmnc_getreset_flags(void)
+{
+        u32 val;
+
+        /* Read */
+        asm volatile("mrc p15, 0, %0, c9, c12, 3" : "=r" (val));
+
+        /* Write to clear flags */
+        val &= ARMV7_FLAG_MASK;
+        asm volatile("mcr p15, 0, %0, c9, c12, 3" : : "r" (val));
+
+        return val;
+}
+
+#ifdef DEBUG
+static void armv7_pmnc_dump_regs(void)
+{
+        u32 val;
+        unsigned int cnt;
+
+        printk(KERN_INFO "PMNC registers dump:\n");
+
+        asm volatile("mrc p15, 0, %0, c9, c12, 0" : "=r" (val));
+        printk(KERN_INFO "PMNC  =0x%08x\n", val);
+
+        asm volatile("mrc p15, 0, %0, c9, c12, 1" : "=r" (val));
+        printk(KERN_INFO "CNTENS=0x%08x\n", val);
+
+        asm volatile("mrc p15, 0, %0, c9, c14, 1" : "=r" (val));
+        printk(KERN_INFO "INTENS=0x%08x\n", val);
+
+        asm volatile("mrc p15, 0, %0, c9, c12, 3" : "=r" (val));
+        printk(KERN_INFO "FLAGS =0x%08x\n", val);
+
+        asm volatile("mrc p15, 0, %0, c9, c12, 5" : "=r" (val));
+        printk(KERN_INFO "SELECT=0x%08x\n", val);
+
+        asm volatile("mrc p15, 0, %0, c9, c13, 0" : "=r" (val));
+        printk(KERN_INFO "CCNT  =0x%08x\n", val);
+
+        for (cnt = ARMV7_COUNTER0; cnt < ARMV7_COUNTER_LAST; cnt++) {
+                armv7_pmnc_select_counter(cnt);
+                asm volatile("mrc p15, 0, %0, c9, c13, 2" : "=r" (val));
+                printk(KERN_INFO "CNT[%d] count =0x%08x\n",
+                        cnt-ARMV7_EVENT_CNT_TO_CNTx, val);
+                asm volatile("mrc p15, 0, %0, c9, c13, 1" : "=r" (val));
+                printk(KERN_INFO "CNT[%d] evtsel=0x%08x\n",
+                        cnt-ARMV7_EVENT_CNT_TO_CNTx, val);
+        }
+}
+#endif
+
+void armv7pmu_enable_event(struct hw_perf_event *hwc, int idx)
+{
+        unsigned long flags;
+
+        /*
+         * Enable counter and interrupt, and set the counter to count
+         * the event that we're interested in.
+         */
+        spin_lock_irqsave(&pmu_lock, flags);
+
+        /*
+         * Disable counter
+         */
+        armv7_pmnc_disable_counter(idx);
+
+        /*
+         * Set event (if destined for PMNx counters)
+         * We don't need to set the event if it's a cycle count
+         */
+        if (idx != ARMV7_CYCLE_COUNTER)
+                armv7_pmnc_write_evtsel(idx, hwc->config_base);
+
+        /*
+         * Enable interrupt for this counter
+         */
+        armv7_pmnc_enable_intens(idx);
+
+        /*
+         * Enable counter
+         */
+        armv7_pmnc_enable_counter(idx);
+
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static void armv7pmu_disable_event(struct hw_perf_event *hwc, int idx)
+{
+        unsigned long flags;
+
+        /*
+         * Disable counter and interrupt
+         */
+        spin_lock_irqsave(&pmu_lock, flags);
+
+        /*
+         * Disable counter
+         */
+        armv7_pmnc_disable_counter(idx);
+
+        /*
+         * Disable interrupt for this counter
+         */
+        armv7_pmnc_disable_intens(idx);
+
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static irqreturn_t armv7pmu_handle_irq(int irq_num, void *dev)
+{
+        unsigned long pmnc;
+        struct perf_sample_data data;
+        struct cpu_hw_events *cpuc;
+        struct pt_regs *regs;
+        int idx;
+
+        /*
+         * Get and reset the IRQ flags
+         */
+        pmnc = armv7_pmnc_getreset_flags();
+
+        /*
+         * Did an overflow occur?
+         */
+        if (!armv7_pmnc_has_overflowed(pmnc))
+                return IRQ_NONE;
+
+        /*
+         * Handle the counter(s) overflow(s)
+         */
+        regs = get_irq_regs();
+
+        perf_sample_data_init(&data, 0);
+
+        cpuc = &__get_cpu_var(cpu_hw_events);
+        for (idx = 0; idx <= armpmu->num_events; ++idx) {
+                struct perf_event *event = cpuc->events[idx];
+                struct hw_perf_event *hwc;
+
+                if (!test_bit(idx, cpuc->active_mask))
+                        continue;
+
+                /*
+                 * We have a single interrupt for all counters. Check that
+                 * each counter has overflowed before we process it.
+                 */
+                if (!armv7_pmnc_counter_has_overflowed(pmnc, idx))
+                        continue;
+
+                hwc = &event->hw;
+                armpmu_event_update(event, hwc, idx);
+                data.period = event->hw.last_period;
+                if (!armpmu_event_set_period(event, hwc, idx))
+                        continue;
+
+                if (perf_event_overflow(event, 0, &data, regs))
+                        armpmu->disable(hwc, idx);
+        }
+
+        /*
+         * Handle the pending perf events.
+         *
+         * Note: this call *must* be run with interrupts disabled. For
+         * platforms that can have the PMU interrupts raised as an NMI, this
+         * will not work.
+         */
+        irq_work_run();
+
+        return IRQ_HANDLED;
+}
+
+static void armv7pmu_start(void)
+{
+        unsigned long flags;
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        /* Enable all counters */
+        armv7_pmnc_write(armv7_pmnc_read() | ARMV7_PMNC_E);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static void armv7pmu_stop(void)
+{
+        unsigned long flags;
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        /* Disable all counters */
+        armv7_pmnc_write(armv7_pmnc_read() & ~ARMV7_PMNC_E);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static int armv7pmu_get_event_idx(struct cpu_hw_events *cpuc,
+                                  struct hw_perf_event *event)
+{
+        int idx;
+
+        /* Always place a cycle counter into the cycle counter. */
+        if (event->config_base == ARMV7_PERFCTR_CPU_CYCLES) {
+                if (test_and_set_bit(ARMV7_CYCLE_COUNTER, cpuc->used_mask))
+                        return -EAGAIN;
+
+                return ARMV7_CYCLE_COUNTER;
+        } else {
+                /*
+                 * For anything other than a cycle counter, try and use
+                 * the events counters
+                 */
+                for (idx = ARMV7_COUNTER0; idx <= armpmu->num_events; ++idx) {
+                        if (!test_and_set_bit(idx, cpuc->used_mask))
+                                return idx;
+                }
+
+                /* The counters are all in use. */
+                return -EAGAIN;
+        }
+}
+
+static struct arm_pmu armv7pmu = {
+        .handle_irq             = armv7pmu_handle_irq,
+        .enable                 = armv7pmu_enable_event,
+        .disable                = armv7pmu_disable_event,
+        .read_counter           = armv7pmu_read_counter,
+        .write_counter          = armv7pmu_write_counter,
+        .get_event_idx          = armv7pmu_get_event_idx,
+        .start                  = armv7pmu_start,
+        .stop                   = armv7pmu_stop,
+        .raw_event_mask         = 0xFF,
+        .max_period             = (1LLU << 32) - 1,
+};
+
+static u32 __init armv7_reset_read_pmnc(void)
+{
+        u32 nb_cnt;
+
+        /* Initialize & Reset PMNC: C and P bits */
+        armv7_pmnc_write(ARMV7_PMNC_P | ARMV7_PMNC_C);
+
+        /* Read the nb of CNTx counters supported from PMNC */
+        nb_cnt = (armv7_pmnc_read() >> ARMV7_PMNC_N_SHIFT) & ARMV7_PMNC_N_MASK;
+
+        /* Add the CPU cycles counter and return */
+        return nb_cnt + 1;
+}
+
+const struct arm_pmu *__init armv7_a8_pmu_init(void)
+{
+        armv7pmu.id             = ARM_PERF_PMU_ID_CA8;
+        armv7pmu.name           = "ARMv7 Cortex-A8";
+        armv7pmu.cache_map      = &armv7_a8_perf_cache_map;
+        armv7pmu.event_map      = &armv7_a8_perf_map;
+        armv7pmu.num_events     = armv7_reset_read_pmnc();
+        return &armv7pmu;
+}
+
+const struct arm_pmu *__init armv7_a9_pmu_init(void)
+{
+        armv7pmu.id             = ARM_PERF_PMU_ID_CA9;
+        armv7pmu.name           = "ARMv7 Cortex-A9";
+        armv7pmu.cache_map      = &armv7_a9_perf_cache_map;
+        armv7pmu.event_map      = &armv7_a9_perf_map;
+        armv7pmu.num_events     = armv7_reset_read_pmnc();
+        return &armv7pmu;
+}
+#else
+const struct arm_pmu *__init armv7_a8_pmu_init(void)
+{
+        return NULL;
+}
+
+const struct arm_pmu *__init armv7_a9_pmu_init(void)
+{
+        return NULL;
+}
+#endif  /* CONFIG_CPU_V7 */
diff --git a/arch/arm/kernel/perf_event_xscale.c b/arch/arm/kernel/perf_event_xscale.c
new file mode 100644
index 000000000000..4e9592789d40
--- /dev/null
+++ b/arch/arm/kernel/perf_event_xscale.c
@@ -0,0 +1,807 @@
+/*
+ * ARMv5 [xscale] Performance counter handling code.
+ *
+ * Copyright (C) 2010, ARM Ltd., Will Deacon <will.deacon@arm.com>
+ *
+ * Based on the previous xscale OProfile code.
+ *
+ * There are two variants of the xscale PMU that we support:
+ *      - xscale1pmu: 2 event counters and a cycle counter
+ *      - xscale2pmu: 4 event counters and a cycle counter
+ * The two variants share event definitions, but have different
+ * PMU structures.
+ */
+
+#ifdef CONFIG_CPU_XSCALE
+enum xscale_perf_types {
+        XSCALE_PERFCTR_ICACHE_MISS              = 0x00,
+        XSCALE_PERFCTR_ICACHE_NO_DELIVER        = 0x01,
+        XSCALE_PERFCTR_DATA_STALL               = 0x02,
+        XSCALE_PERFCTR_ITLB_MISS                = 0x03,
+        XSCALE_PERFCTR_DTLB_MISS                = 0x04,
+        XSCALE_PERFCTR_BRANCH                   = 0x05,
+        XSCALE_PERFCTR_BRANCH_MISS              = 0x06,
+        XSCALE_PERFCTR_INSTRUCTION              = 0x07,
+        XSCALE_PERFCTR_DCACHE_FULL_STALL        = 0x08,
+        XSCALE_PERFCTR_DCACHE_FULL_STALL_CONTIG = 0x09,
+        XSCALE_PERFCTR_DCACHE_ACCESS            = 0x0A,
+        XSCALE_PERFCTR_DCACHE_MISS              = 0x0B,
+        XSCALE_PERFCTR_DCACHE_WRITE_BACK        = 0x0C,
+        XSCALE_PERFCTR_PC_CHANGED               = 0x0D,
+        XSCALE_PERFCTR_BCU_REQUEST              = 0x10,
+        XSCALE_PERFCTR_BCU_FULL                 = 0x11,
+        XSCALE_PERFCTR_BCU_DRAIN                = 0x12,
+        XSCALE_PERFCTR_BCU_ECC_NO_ELOG          = 0x14,
+        XSCALE_PERFCTR_BCU_1_BIT_ERR            = 0x15,
+        XSCALE_PERFCTR_RMW                      = 0x16,
+        /* XSCALE_PERFCTR_CCNT is not hardware defined */
+        XSCALE_PERFCTR_CCNT                     = 0xFE,
+        XSCALE_PERFCTR_UNUSED                   = 0xFF,
+};
+
+enum xscale_counters {
+        XSCALE_CYCLE_COUNTER    = 1,
+        XSCALE_COUNTER0,
+        XSCALE_COUNTER1,
+        XSCALE_COUNTER2,
+        XSCALE_COUNTER3,
+};
+
+static const unsigned xscale_perf_map[PERF_COUNT_HW_MAX] = {
+        [PERF_COUNT_HW_CPU_CYCLES]          = XSCALE_PERFCTR_CCNT,
+        [PERF_COUNT_HW_INSTRUCTIONS]        = XSCALE_PERFCTR_INSTRUCTION,
+        [PERF_COUNT_HW_CACHE_REFERENCES]    = HW_OP_UNSUPPORTED,
+        [PERF_COUNT_HW_CACHE_MISSES]        = HW_OP_UNSUPPORTED,
+        [PERF_COUNT_HW_BRANCH_INSTRUCTIONS] = XSCALE_PERFCTR_BRANCH,
+        [PERF_COUNT_HW_BRANCH_MISSES]       = XSCALE_PERFCTR_BRANCH_MISS,
+        [PERF_COUNT_HW_BUS_CYCLES]          = HW_OP_UNSUPPORTED,
+};
+
+static const unsigned xscale_perf_cache_map[PERF_COUNT_HW_CACHE_MAX]
+                                           [PERF_COUNT_HW_CACHE_OP_MAX]
+                                           [PERF_COUNT_HW_CACHE_RESULT_MAX] = {
+        [C(L1D)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = XSCALE_PERFCTR_DCACHE_ACCESS,
+                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_DCACHE_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = XSCALE_PERFCTR_DCACHE_ACCESS,
+                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_DCACHE_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(L1I)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_ICACHE_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_ICACHE_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(LL)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(DTLB)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_DTLB_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_DTLB_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(ITLB)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_ITLB_MISS,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = XSCALE_PERFCTR_ITLB_MISS,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+        [C(BPU)] = {
+                [C(OP_READ)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_WRITE)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+                [C(OP_PREFETCH)] = {
+                        [C(RESULT_ACCESS)]      = CACHE_OP_UNSUPPORTED,
+                        [C(RESULT_MISS)]        = CACHE_OP_UNSUPPORTED,
+                },
+        },
+};
+
+#define XSCALE_PMU_ENABLE       0x001
+#define XSCALE_PMN_RESET        0x002
+#define XSCALE_CCNT_RESET       0x004
+#define XSCALE_PMU_RESET        (CCNT_RESET | PMN_RESET)
+#define XSCALE_PMU_CNT64        0x008
+
+#define XSCALE1_OVERFLOWED_MASK 0x700
+#define XSCALE1_CCOUNT_OVERFLOW 0x400
+#define XSCALE1_COUNT0_OVERFLOW 0x100
+#define XSCALE1_COUNT1_OVERFLOW 0x200
+#define XSCALE1_CCOUNT_INT_EN   0x040
+#define XSCALE1_COUNT0_INT_EN   0x010
+#define XSCALE1_COUNT1_INT_EN   0x020
+#define XSCALE1_COUNT0_EVT_SHFT 12
+#define XSCALE1_COUNT0_EVT_MASK (0xff << XSCALE1_COUNT0_EVT_SHFT)
+#define XSCALE1_COUNT1_EVT_SHFT 20
+#define XSCALE1_COUNT1_EVT_MASK (0xff << XSCALE1_COUNT1_EVT_SHFT)
+
+static inline u32
+xscale1pmu_read_pmnc(void)
+{
+        u32 val;
+        asm volatile("mrc p14, 0, %0, c0, c0, 0" : "=r" (val));
+        return val;
+}
+
+static inline void
+xscale1pmu_write_pmnc(u32 val)
+{
+        /* upper 4bits and 7, 11 are write-as-0 */
+        val &= 0xffff77f;
+        asm volatile("mcr p14, 0, %0, c0, c0, 0" : : "r" (val));
+}
+
+static inline int
+xscale1_pmnc_counter_has_overflowed(unsigned long pmnc,
+                                        enum xscale_counters counter)
+{
+        int ret = 0;
+
+        switch (counter) {
+        case XSCALE_CYCLE_COUNTER:
+                ret = pmnc & XSCALE1_CCOUNT_OVERFLOW;
+                break;
+        case XSCALE_COUNTER0:
+                ret = pmnc & XSCALE1_COUNT0_OVERFLOW;
+                break;
+        case XSCALE_COUNTER1:
+                ret = pmnc & XSCALE1_COUNT1_OVERFLOW;
+                break;
+        default:
+                WARN_ONCE(1, "invalid counter number (%d)\n", counter);
+        }
+
+        return ret;
+}
+
+static irqreturn_t
+xscale1pmu_handle_irq(int irq_num, void *dev)
+{
+        unsigned long pmnc;
+        struct perf_sample_data data;
+        struct cpu_hw_events *cpuc;
+        struct pt_regs *regs;
+        int idx;
+
+        /*
+         * NOTE: there's an A stepping erratum that states if an overflow
+         *       bit already exists and another occurs, the previous
+         *       Overflow bit gets cleared. There's no workaround.
+         *       Fixed in B stepping or later.
+         */
+        pmnc = xscale1pmu_read_pmnc();
+
+        /*
+         * Write the value back to clear the overflow flags. Overflow
+         * flags remain in pmnc for use below. We also disable the PMU
+         * while we process the interrupt.
+         */
+        xscale1pmu_write_pmnc(pmnc & ~XSCALE_PMU_ENABLE);
+
+        if (!(pmnc & XSCALE1_OVERFLOWED_MASK))
+                return IRQ_NONE;
+
+        regs = get_irq_regs();
+
+        perf_sample_data_init(&data, 0);
+
+        cpuc = &__get_cpu_var(cpu_hw_events);
+        for (idx = 0; idx <= armpmu->num_events; ++idx) {
+                struct perf_event *event = cpuc->events[idx];
+                struct hw_perf_event *hwc;
+
+                if (!test_bit(idx, cpuc->active_mask))
+                        continue;
+
+                if (!xscale1_pmnc_counter_has_overflowed(pmnc, idx))
+                        continue;
+
+                hwc = &event->hw;
+                armpmu_event_update(event, hwc, idx);
+                data.period = event->hw.last_period;
+                if (!armpmu_event_set_period(event, hwc, idx))
+                        continue;
+
+                if (perf_event_overflow(event, 0, &data, regs))
+                        armpmu->disable(hwc, idx);
+        }
+
+        irq_work_run();
+
+        /*
+         * Re-enable the PMU.
+         */
+        pmnc = xscale1pmu_read_pmnc() | XSCALE_PMU_ENABLE;
+        xscale1pmu_write_pmnc(pmnc);
+
+        return IRQ_HANDLED;
+}
+
+static void
+xscale1pmu_enable_event(struct hw_perf_event *hwc, int idx)
+{
+        unsigned long val, mask, evt, flags;
+
+        switch (idx) {
+        case XSCALE_CYCLE_COUNTER:
+                mask = 0;
+                evt = XSCALE1_CCOUNT_INT_EN;
+                break;
+        case XSCALE_COUNTER0:
+                mask = XSCALE1_COUNT0_EVT_MASK;
+                evt = (hwc->config_base << XSCALE1_COUNT0_EVT_SHFT) |
+                        XSCALE1_COUNT0_INT_EN;
+                break;
+        case XSCALE_COUNTER1:
+                mask = XSCALE1_COUNT1_EVT_MASK;
+                evt = (hwc->config_base << XSCALE1_COUNT1_EVT_SHFT) |
+                        XSCALE1_COUNT1_INT_EN;
+                break;
+        default:
+                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
+                return;
+        }
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = xscale1pmu_read_pmnc();
+        val &= ~mask;
+        val |= evt;
+        xscale1pmu_write_pmnc(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static void
+xscale1pmu_disable_event(struct hw_perf_event *hwc, int idx)
+{
+        unsigned long val, mask, evt, flags;
+
+        switch (idx) {
+        case XSCALE_CYCLE_COUNTER:
+                mask = XSCALE1_CCOUNT_INT_EN;
+                evt = 0;
+                break;
+        case XSCALE_COUNTER0:
+                mask = XSCALE1_COUNT0_INT_EN | XSCALE1_COUNT0_EVT_MASK;
+                evt = XSCALE_PERFCTR_UNUSED << XSCALE1_COUNT0_EVT_SHFT;
+                break;
+        case XSCALE_COUNTER1:
+                mask = XSCALE1_COUNT1_INT_EN | XSCALE1_COUNT1_EVT_MASK;
+                evt = XSCALE_PERFCTR_UNUSED << XSCALE1_COUNT1_EVT_SHFT;
+                break;
+        default:
+                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
+                return;
+        }
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = xscale1pmu_read_pmnc();
+        val &= ~mask;
+        val |= evt;
+        xscale1pmu_write_pmnc(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static int
+xscale1pmu_get_event_idx(struct cpu_hw_events *cpuc,
+                        struct hw_perf_event *event)
+{
+        if (XSCALE_PERFCTR_CCNT == event->config_base) {
+                if (test_and_set_bit(XSCALE_CYCLE_COUNTER, cpuc->used_mask))
+                        return -EAGAIN;
+
+                return XSCALE_CYCLE_COUNTER;
+        } else {
+                if (!test_and_set_bit(XSCALE_COUNTER1, cpuc->used_mask))
+                        return XSCALE_COUNTER1;
+
+                if (!test_and_set_bit(XSCALE_COUNTER0, cpuc->used_mask))
+                        return XSCALE_COUNTER0;
+
+                return -EAGAIN;
+        }
+}
+
+static void
+xscale1pmu_start(void)
+{
+        unsigned long flags, val;
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = xscale1pmu_read_pmnc();
+        val |= XSCALE_PMU_ENABLE;
+        xscale1pmu_write_pmnc(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static void
+xscale1pmu_stop(void)
+{
+        unsigned long flags, val;
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = xscale1pmu_read_pmnc();
+        val &= ~XSCALE_PMU_ENABLE;
+        xscale1pmu_write_pmnc(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static inline u32
+xscale1pmu_read_counter(int counter)
+{
+        u32 val = 0;
+
+        switch (counter) {
+        case XSCALE_CYCLE_COUNTER:
+                asm volatile("mrc p14, 0, %0, c1, c0, 0" : "=r" (val));
+                break;
+        case XSCALE_COUNTER0:
+                asm volatile("mrc p14, 0, %0, c2, c0, 0" : "=r" (val));
+                break;
+        case XSCALE_COUNTER1:
+                asm volatile("mrc p14, 0, %0, c3, c0, 0" : "=r" (val));
+                break;
+        }
+
+        return val;
+}
+
+static inline void
+xscale1pmu_write_counter(int counter, u32 val)
+{
+        switch (counter) {
+        case XSCALE_CYCLE_COUNTER:
+                asm volatile("mcr p14, 0, %0, c1, c0, 0" : : "r" (val));
+                break;
+        case XSCALE_COUNTER0:
+                asm volatile("mcr p14, 0, %0, c2, c0, 0" : : "r" (val));
+                break;
+        case XSCALE_COUNTER1:
+                asm volatile("mcr p14, 0, %0, c3, c0, 0" : : "r" (val));
+                break;
+        }
+}
+
+static const struct arm_pmu xscale1pmu = {
+        .id             = ARM_PERF_PMU_ID_XSCALE1,
+        .name           = "xscale1",
+        .handle_irq     = xscale1pmu_handle_irq,
+        .enable         = xscale1pmu_enable_event,
+        .disable        = xscale1pmu_disable_event,
+        .read_counter   = xscale1pmu_read_counter,
+        .write_counter  = xscale1pmu_write_counter,
+        .get_event_idx  = xscale1pmu_get_event_idx,
+        .start          = xscale1pmu_start,
+        .stop           = xscale1pmu_stop,
+        .cache_map      = &xscale_perf_cache_map,
+        .event_map      = &xscale_perf_map,
+        .raw_event_mask = 0xFF,
+        .num_events     = 3,
+        .max_period     = (1LLU << 32) - 1,
+};
+
+const struct arm_pmu *__init xscale1pmu_init(void)
+{
+        return &xscale1pmu;
+}
+
+#define XSCALE2_OVERFLOWED_MASK 0x01f
+#define XSCALE2_CCOUNT_OVERFLOW 0x001
+#define XSCALE2_COUNT0_OVERFLOW 0x002
+#define XSCALE2_COUNT1_OVERFLOW 0x004
+#define XSCALE2_COUNT2_OVERFLOW 0x008
+#define XSCALE2_COUNT3_OVERFLOW 0x010
+#define XSCALE2_CCOUNT_INT_EN   0x001
+#define XSCALE2_COUNT0_INT_EN   0x002
+#define XSCALE2_COUNT1_INT_EN   0x004
+#define XSCALE2_COUNT2_INT_EN   0x008
+#define XSCALE2_COUNT3_INT_EN   0x010
+#define XSCALE2_COUNT0_EVT_SHFT 0
+#define XSCALE2_COUNT0_EVT_MASK (0xff << XSCALE2_COUNT0_EVT_SHFT)
+#define XSCALE2_COUNT1_EVT_SHFT 8
+#define XSCALE2_COUNT1_EVT_MASK (0xff << XSCALE2_COUNT1_EVT_SHFT)
+#define XSCALE2_COUNT2_EVT_SHFT 16
+#define XSCALE2_COUNT2_EVT_MASK (0xff << XSCALE2_COUNT2_EVT_SHFT)
+#define XSCALE2_COUNT3_EVT_SHFT 24
+#define XSCALE2_COUNT3_EVT_MASK (0xff << XSCALE2_COUNT3_EVT_SHFT)
+
+static inline u32
+xscale2pmu_read_pmnc(void)
+{
+        u32 val;
+        asm volatile("mrc p14, 0, %0, c0, c1, 0" : "=r" (val));
+        /* bits 1-2 and 4-23 are read-unpredictable */
+        return val & 0xff000009;
+}
+
+static inline void
+xscale2pmu_write_pmnc(u32 val)
+{
+        /* bits 4-23 are write-as-0, 24-31 are write ignored */
+        val &= 0xf;
+        asm volatile("mcr p14, 0, %0, c0, c1, 0" : : "r" (val));
+}
+
+static inline u32
+xscale2pmu_read_overflow_flags(void)
+{
+        u32 val;
+        asm volatile("mrc p14, 0, %0, c5, c1, 0" : "=r" (val));
+        return val;
+}
+
+static inline void
+xscale2pmu_write_overflow_flags(u32 val)
+{
+        asm volatile("mcr p14, 0, %0, c5, c1, 0" : : "r" (val));
+}
+
+static inline u32
+xscale2pmu_read_event_select(void)
+{
+        u32 val;
+        asm volatile("mrc p14, 0, %0, c8, c1, 0" : "=r" (val));
+        return val;
+}
+
+static inline void
+xscale2pmu_write_event_select(u32 val)
+{
+        asm volatile("mcr p14, 0, %0, c8, c1, 0" : : "r"(val));
+}
+
+static inline u32
+xscale2pmu_read_int_enable(void)
+{
+        u32 val;
+        asm volatile("mrc p14, 0, %0, c4, c1, 0" : "=r" (val));
+        return val;
+}
+
+static void
+xscale2pmu_write_int_enable(u32 val)
+{
+        asm volatile("mcr p14, 0, %0, c4, c1, 0" : : "r" (val));
+}
+
+static inline int
+xscale2_pmnc_counter_has_overflowed(unsigned long of_flags,
+                                        enum xscale_counters counter)
+{
+        int ret = 0;
+
+        switch (counter) {
+        case XSCALE_CYCLE_COUNTER:
+                ret = of_flags & XSCALE2_CCOUNT_OVERFLOW;
+                break;
+        case XSCALE_COUNTER0:
+                ret = of_flags & XSCALE2_COUNT0_OVERFLOW;
+                break;
+        case XSCALE_COUNTER1:
+                ret = of_flags & XSCALE2_COUNT1_OVERFLOW;
+                break;
+        case XSCALE_COUNTER2:
+                ret = of_flags & XSCALE2_COUNT2_OVERFLOW;
+                break;
+        case XSCALE_COUNTER3:
+                ret = of_flags & XSCALE2_COUNT3_OVERFLOW;
+                break;
+        default:
+                WARN_ONCE(1, "invalid counter number (%d)\n", counter);
+        }
+
+        return ret;
+}
+
+static irqreturn_t
+xscale2pmu_handle_irq(int irq_num, void *dev)
+{
+        unsigned long pmnc, of_flags;
+        struct perf_sample_data data;
+        struct cpu_hw_events *cpuc;
+        struct pt_regs *regs;
+        int idx;
+
+        /* Disable the PMU. */
+        pmnc = xscale2pmu_read_pmnc();
+        xscale2pmu_write_pmnc(pmnc & ~XSCALE_PMU_ENABLE);
+
+        /* Check the overflow flag register. */
+        of_flags = xscale2pmu_read_overflow_flags();
+        if (!(of_flags & XSCALE2_OVERFLOWED_MASK))
+                return IRQ_NONE;
+
+        /* Clear the overflow bits. */
+        xscale2pmu_write_overflow_flags(of_flags);
+
+        regs = get_irq_regs();
+
+        perf_sample_data_init(&data, 0);
+
+        cpuc = &__get_cpu_var(cpu_hw_events);
+        for (idx = 0; idx <= armpmu->num_events; ++idx) {
+                struct perf_event *event = cpuc->events[idx];
+                struct hw_perf_event *hwc;
+
+                if (!test_bit(idx, cpuc->active_mask))
+                        continue;
+
+                if (!xscale2_pmnc_counter_has_overflowed(pmnc, idx))
+                        continue;
+
+                hwc = &event->hw;
+                armpmu_event_update(event, hwc, idx);
+                data.period = event->hw.last_period;
+                if (!armpmu_event_set_period(event, hwc, idx))
+                        continue;
+
+                if (perf_event_overflow(event, 0, &data, regs))
+                        armpmu->disable(hwc, idx);
+        }
+
+        irq_work_run();
+
+        /*
+         * Re-enable the PMU.
+         */
+        pmnc = xscale2pmu_read_pmnc() | XSCALE_PMU_ENABLE;
+        xscale2pmu_write_pmnc(pmnc);
+
+        return IRQ_HANDLED;
+}
+
+static void
+xscale2pmu_enable_event(struct hw_perf_event *hwc, int idx)
+{
+        unsigned long flags, ien, evtsel;
+
+        ien = xscale2pmu_read_int_enable();
+        evtsel = xscale2pmu_read_event_select();
+
+        switch (idx) {
+        case XSCALE_CYCLE_COUNTER:
+                ien |= XSCALE2_CCOUNT_INT_EN;
+                break;
+        case XSCALE_COUNTER0:
+                ien |= XSCALE2_COUNT0_INT_EN;
+                evtsel &= ~XSCALE2_COUNT0_EVT_MASK;
+                evtsel |= hwc->config_base << XSCALE2_COUNT0_EVT_SHFT;
+                break;
+        case XSCALE_COUNTER1:
+                ien |= XSCALE2_COUNT1_INT_EN;
+                evtsel &= ~XSCALE2_COUNT1_EVT_MASK;
+                evtsel |= hwc->config_base << XSCALE2_COUNT1_EVT_SHFT;
+                break;
+        case XSCALE_COUNTER2:
+                ien |= XSCALE2_COUNT2_INT_EN;
+                evtsel &= ~XSCALE2_COUNT2_EVT_MASK;
+                evtsel |= hwc->config_base << XSCALE2_COUNT2_EVT_SHFT;
+                break;
+        case XSCALE_COUNTER3:
+                ien |= XSCALE2_COUNT3_INT_EN;
+                evtsel &= ~XSCALE2_COUNT3_EVT_MASK;
+                evtsel |= hwc->config_base << XSCALE2_COUNT3_EVT_SHFT;
+                break;
+        default:
+                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
+                return;
+        }
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        xscale2pmu_write_event_select(evtsel);
+        xscale2pmu_write_int_enable(ien);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static void
+xscale2pmu_disable_event(struct hw_perf_event *hwc, int idx)
+{
+        unsigned long flags, ien, evtsel;
+
+        ien = xscale2pmu_read_int_enable();
+        evtsel = xscale2pmu_read_event_select();
+
+        switch (idx) {
+        case XSCALE_CYCLE_COUNTER:
+                ien &= ~XSCALE2_CCOUNT_INT_EN;
+                break;
+        case XSCALE_COUNTER0:
+                ien &= ~XSCALE2_COUNT0_INT_EN;
+                evtsel &= ~XSCALE2_COUNT0_EVT_MASK;
+                evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT0_EVT_SHFT;
+                break;
+        case XSCALE_COUNTER1:
+                ien &= ~XSCALE2_COUNT1_INT_EN;
+                evtsel &= ~XSCALE2_COUNT1_EVT_MASK;
+                evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT1_EVT_SHFT;
+                break;
+        case XSCALE_COUNTER2:
+                ien &= ~XSCALE2_COUNT2_INT_EN;
+                evtsel &= ~XSCALE2_COUNT2_EVT_MASK;
+                evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT2_EVT_SHFT;
+                break;
+        case XSCALE_COUNTER3:
+                ien &= ~XSCALE2_COUNT3_INT_EN;
+                evtsel &= ~XSCALE2_COUNT3_EVT_MASK;
+                evtsel |= XSCALE_PERFCTR_UNUSED << XSCALE2_COUNT3_EVT_SHFT;
+                break;
+        default:
+                WARN_ONCE(1, "invalid counter number (%d)\n", idx);
+                return;
+        }
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        xscale2pmu_write_event_select(evtsel);
+        xscale2pmu_write_int_enable(ien);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static int
+xscale2pmu_get_event_idx(struct cpu_hw_events *cpuc,
+                        struct hw_perf_event *event)
+{
+        int idx = xscale1pmu_get_event_idx(cpuc, event);
+        if (idx >= 0)
+                goto out;
+
+        if (!test_and_set_bit(XSCALE_COUNTER3, cpuc->used_mask))
+                idx = XSCALE_COUNTER3;
+        else if (!test_and_set_bit(XSCALE_COUNTER2, cpuc->used_mask))
+                idx = XSCALE_COUNTER2;
+out:
+        return idx;
+}
+
+static void
+xscale2pmu_start(void)
+{
+        unsigned long flags, val;
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = xscale2pmu_read_pmnc() & ~XSCALE_PMU_CNT64;
+        val |= XSCALE_PMU_ENABLE;
+        xscale2pmu_write_pmnc(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static void
+xscale2pmu_stop(void)
+{
+        unsigned long flags, val;
+
+        spin_lock_irqsave(&pmu_lock, flags);
+        val = xscale2pmu_read_pmnc();
+        val &= ~XSCALE_PMU_ENABLE;
+        xscale2pmu_write_pmnc(val);
+        spin_unlock_irqrestore(&pmu_lock, flags);
+}
+
+static inline u32
+xscale2pmu_read_counter(int counter)
+{
+        u32 val = 0;
+
+        switch (counter) {
+        case XSCALE_CYCLE_COUNTER:
+                asm volatile("mrc p14, 0, %0, c1, c1, 0" : "=r" (val));
+                break;
+        case XSCALE_COUNTER0:
+                asm volatile("mrc p14, 0, %0, c0, c2, 0" : "=r" (val));
+                break;
+        case XSCALE_COUNTER1:
+                asm volatile("mrc p14, 0, %0, c1, c2, 0" : "=r" (val));
+                break;
+        case XSCALE_COUNTER2:
+                asm volatile("mrc p14, 0, %0, c2, c2, 0" : "=r" (val));
+                break;
+        case XSCALE_COUNTER3:
+                asm volatile("mrc p14, 0, %0, c3, c2, 0" : "=r" (val));
+                break;
+        }
+
+        return val;
+}
+
+static inline void
+xscale2pmu_write_counter(int counter, u32 val)
+{
+        switch (counter) {
+        case XSCALE_CYCLE_COUNTER:
+                asm volatile("mcr p14, 0, %0, c1, c1, 0" : : "r" (val));
+                break;
+        case XSCALE_COUNTER0:
+                asm volatile("mcr p14, 0, %0, c0, c2, 0" : : "r" (val));
+                break;
+        case XSCALE_COUNTER1:
+                asm volatile("mcr p14, 0, %0, c1, c2, 0" : : "r" (val));
+                break;
+        case XSCALE_COUNTER2:
+                asm volatile("mcr p14, 0, %0, c2, c2, 0" : : "r" (val));
+                break;
+        case XSCALE_COUNTER3:
+                asm volatile("mcr p14, 0, %0, c3, c2, 0" : : "r" (val));
+                break;
+        }
+}
+
+static const struct arm_pmu xscale2pmu = {
+        .id             = ARM_PERF_PMU_ID_XSCALE2,
+        .name           = "xscale2",
+        .handle_irq     = xscale2pmu_handle_irq,
+        .enable         = xscale2pmu_enable_event,
+        .disable        = xscale2pmu_disable_event,
+        .read_counter   = xscale2pmu_read_counter,
+        .write_counter  = xscale2pmu_write_counter,
+        .get_event_idx  = xscale2pmu_get_event_idx,
+        .start          = xscale2pmu_start,
+        .stop           = xscale2pmu_stop,
+        .cache_map      = &xscale_perf_cache_map,
+        .event_map      = &xscale_perf_map,
+        .raw_event_mask = 0xFF,
+        .num_events     = 5,
+        .max_period     = (1LLU << 32) - 1,
+};
+
+const struct arm_pmu *__init xscale2pmu_init(void)
+{
+        return &xscale2pmu;
+}
+#else
+const struct arm_pmu *__init xscale1pmu_init(void)
+{
+        return NULL;
+}
+
+const struct arm_pmu *__init xscale2pmu_init(void)
+{
+        return NULL;
+}
+#endif  /* CONFIG_CPU_XSCALE */
diff --git a/arch/arm/lib/findbit.S b/arch/arm/lib/findbit.S
index 1e4cbd4e7be9..64f6bc1a9132 100644
--- a/arch/arm/lib/findbit.S
+++ b/arch/arm/lib/findbit.S
@@ -174,8 +174,8 @@ ENDPROC(_find_next_bit_be)
  */
 .L_found:
 #if __LINUX_ARM_ARCH__ >= 5
-                rsb     r1, r3, #0
-                and     r3, r3, r1
+                rsb     r0, r3, #0
+                and     r3, r3, r0
                 clz     r3, r3
                 rsb     r3, r3, #31
                 add     r0, r2, r3
@@ -190,5 +190,7 @@ ENDPROC(_find_next_bit_be)
                 addeq   r2, r2, #1
                 mov     r0, r2
 #endif
+                cmp     r1, r0                  @ Clamp to maxbit
+                movlo   r0, r1
                 mov     pc, lr
 
diff --git a/arch/arm/mach-aaec2000/include/mach/vmalloc.h b/arch/arm/mach-aaec2000/include/mach/vmalloc.h
index cff4e0a996ce..a6299e8321bd 100644
--- a/arch/arm/mach-aaec2000/include/mach/vmalloc.h
+++ b/arch/arm/mach-aaec2000/include/mach/vmalloc.h
@@ -11,6 +11,6 @@
 #ifndef __ASM_ARCH_VMALLOC_H
 #define __ASM_ARCH_VMALLOC_H
 
-#define VMALLOC_END             0xd0000000
+#define VMALLOC_END             0xd0000000UL
 
 #endif /* __ASM_ARCH_VMALLOC_H */
diff --git a/arch/arm/mach-bcmring/include/mach/vmalloc.h b/arch/arm/mach-bcmring/include/mach/vmalloc.h
index 3db3a09fd398..7397bd7817d9 100644
--- a/arch/arm/mach-bcmring/include/mach/vmalloc.h
+++ b/arch/arm/mach-bcmring/include/mach/vmalloc.h
@@ -22,4 +22,4 @@
  * 0xe0000000 to 0xefffffff. This gives us 256 MB of vm space and handles
  * larger physical memory designs better.
  */
-#define VMALLOC_END       0xf0000000
+#define VMALLOC_END       0xf0000000UL
diff --git a/arch/arm/mach-clps711x/include/mach/vmalloc.h b/arch/arm/mach-clps711x/include/mach/vmalloc.h
index 30b3a287ed88..467b96137e47 100644
--- a/arch/arm/mach-clps711x/include/mach/vmalloc.h
+++ b/arch/arm/mach-clps711x/include/mach/vmalloc.h
@@ -17,4 +17,4 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
-#define VMALLOC_END       0xd0000000
+#define VMALLOC_END       0xd0000000UL
diff --git a/arch/arm/mach-davinci/dm355.c b/arch/arm/mach-davinci/dm355.c
index 9be261beae7d..2652af124acd 100644
--- a/arch/arm/mach-davinci/dm355.c
+++ b/arch/arm/mach-davinci/dm355.c
@@ -359,8 +359,8 @@ static struct clk_lookup dm355_clks[] = {
         CLK(NULL, "uart1", &uart1_clk),
         CLK(NULL, "uart2", &uart2_clk),
         CLK("i2c_davinci.1", NULL, &i2c_clk),
-        CLK("davinci-asp.0", NULL, &asp0_clk),
-        CLK("davinci-asp.1", NULL, &asp1_clk),
+        CLK("davinci-mcbsp.0", NULL, &asp0_clk),
+        CLK("davinci-mcbsp.1", NULL, &asp1_clk),
         CLK("davinci_mmc.0", NULL, &mmcsd0_clk),
         CLK("davinci_mmc.1", NULL, &mmcsd1_clk),
         CLK("spi_davinci.0", NULL, &spi0_clk),
@@ -664,7 +664,7 @@ static struct resource dm355_asp1_resources[] = {
 };
 
 static struct platform_device dm355_asp1_device = {
-        .name           = "davinci-asp",
+        .name           = "davinci-mcbsp",
         .id             = 1,
         .num_resources  = ARRAY_SIZE(dm355_asp1_resources),
         .resource       = dm355_asp1_resources,
diff --git a/arch/arm/mach-davinci/dm365.c b/arch/arm/mach-davinci/dm365.c
index a12065e87266..c466d710d3c1 100644
--- a/arch/arm/mach-davinci/dm365.c
+++ b/arch/arm/mach-davinci/dm365.c
@@ -459,7 +459,7 @@ static struct clk_lookup dm365_clks[] = {
         CLK(NULL, "usb", &usb_clk),
         CLK("davinci_emac.1", NULL, &emac_clk),
         CLK("davinci_voicecodec", NULL, &voicecodec_clk),
-        CLK("davinci-asp.0", NULL, &asp0_clk),
+        CLK("davinci-mcbsp", NULL, &asp0_clk),
         CLK(NULL, "rto", &rto_clk),
         CLK(NULL, "mjcp", &mjcp_clk),
         CLK(NULL, NULL, NULL),
@@ -922,8 +922,8 @@ static struct resource dm365_asp_resources[] = {
 };
 
 static struct platform_device dm365_asp_device = {
-        .name           = "davinci-asp",
-        .id             = 0,
+        .name           = "davinci-mcbsp",
+        .id             = -1,
         .num_resources  = ARRAY_SIZE(dm365_asp_resources),
         .resource       = dm365_asp_resources,
 };
diff --git a/arch/arm/mach-davinci/dm644x.c b/arch/arm/mach-davinci/dm644x.c
index 0608dd776a16..9a2376b3137c 100644
--- a/arch/arm/mach-davinci/dm644x.c
+++ b/arch/arm/mach-davinci/dm644x.c
@@ -302,7 +302,7 @@ static struct clk_lookup dm644x_clks[] = {
         CLK("davinci_emac.1", NULL, &emac_clk),
         CLK("i2c_davinci.1", NULL, &i2c_clk),
         CLK("palm_bk3710", NULL, &ide_clk),
-        CLK("davinci-asp", NULL, &asp_clk),
+        CLK("davinci-mcbsp", NULL, &asp_clk),
         CLK("davinci_mmc.0", NULL, &mmcsd_clk),
         CLK(NULL, "spi", &spi_clk),
         CLK(NULL, "gpio", &gpio_clk),
@@ -580,7 +580,7 @@ static struct resource dm644x_asp_resources[] = {
 };
 
 static struct platform_device dm644x_asp_device = {
-        .name           = "davinci-asp",
+        .name           = "davinci-mcbsp",
         .id             = -1,
         .num_resources  = ARRAY_SIZE(dm644x_asp_resources),
         .resource       = dm644x_asp_resources,
diff --git a/arch/arm/mach-ebsa110/include/mach/vmalloc.h b/arch/arm/mach-ebsa110/include/mach/vmalloc.h
index 60bde56fba4c..ea141b7a3e03 100644
--- a/arch/arm/mach-ebsa110/include/mach/vmalloc.h
+++ b/arch/arm/mach-ebsa110/include/mach/vmalloc.h
@@ -7,4 +7,4 @@
  * it under the terms of the GNU General Public License version 2 as
  * published by the Free Software Foundation.
  */
-#define VMALLOC_END       0xdf000000
+#define VMALLOC_END       0xdf000000UL
diff --git a/arch/arm/mach-footbridge/include/mach/vmalloc.h b/arch/arm/mach-footbridge/include/mach/vmalloc.h
index 0ffbb7c85e59..40ba78e5782b 100644
--- a/arch/arm/mach-footbridge/include/mach/vmalloc.h
+++ b/arch/arm/mach-footbridge/include/mach/vmalloc.h
@@ -7,4 +7,4 @@
  */
 
 
-#define VMALLOC_END       0xf0000000
+#define VMALLOC_END       0xf0000000UL
diff --git a/arch/arm/mach-h720x/include/mach/vmalloc.h b/arch/arm/mach-h720x/include/mach/vmalloc.h
index a45915b88756..8520b4a4d4e6 100644
--- a/arch/arm/mach-h720x/include/mach/vmalloc.h
+++ b/arch/arm/mach-h720x/include/mach/vmalloc.h
@@ -5,6 +5,6 @@
 #ifndef __ARCH_ARM_VMALLOC_H
 #define __ARCH_ARM_VMALLOC_H
 
-#define VMALLOC_END       0xd0000000
+#define VMALLOC_END       0xd0000000UL
 
 #endif
diff --git a/arch/arm/mach-imx/eukrea_mbimx27-baseboard.c b/arch/arm/mach-imx/eukrea_mbimx27-baseboard.c
index 026263c665ca..7e1e9dc2c8fc 100644
--- a/arch/arm/mach-imx/eukrea_mbimx27-baseboard.c
+++ b/arch/arm/mach-imx/eukrea_mbimx27-baseboard.c
@@ -250,9 +250,6 @@ static const struct imxuart_platform_data uart_pdata __initconst = {
         .flags = IMXUART_HAVE_RTSCTS,
 };
 
-#if defined(CONFIG_TOUCHSCREEN_ADS7846) \
-        || defined(CONFIG_TOUCHSCREEN_ADS7846_MODULE)
-
 #define ADS7846_PENDOWN (GPIO_PORTD | 25)
 
 static void ads7846_dev_init(void)
@@ -273,9 +270,7 @@ static struct ads7846_platform_data ads7846_config __initdata = {
         .get_pendown_state      = ads7846_get_pendown_state,
         .keep_vref_on           = 1,
 };
-#endif
 
-#if defined(CONFIG_SPI_IMX) || defined(CONFIG_SPI_IMX_MODULE)
 static struct spi_board_info eukrea_mbimx27_spi_board_info[] __initdata = {
         [0] = {
                 .modalias       = "ads7846",
@@ -294,7 +289,6 @@ static const struct spi_imx_master eukrea_mbimx27_spi0_data __initconst = {
         .chipselect     = eukrea_mbimx27_spi_cs,
         .num_chipselect = ARRAY_SIZE(eukrea_mbimx27_spi_cs),
 };
-#endif
 
 static struct i2c_board_info eukrea_mbimx27_i2c_devices[] = {
         {
diff --git a/arch/arm/mach-integrator/include/mach/vmalloc.h b/arch/arm/mach-integrator/include/mach/vmalloc.h
index e056e7cf5645..2f5a2bafb11f 100644
--- a/arch/arm/mach-integrator/include/mach/vmalloc.h
+++ b/arch/arm/mach-integrator/include/mach/vmalloc.h
@@ -17,4 +17,4 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
-#define VMALLOC_END       0xd0000000
+#define VMALLOC_END       0xd0000000UL
diff --git a/arch/arm/mach-msm/Kconfig b/arch/arm/mach-msm/Kconfig
index dbbcfeb919db..31e5fd63ec9a 100644
--- a/arch/arm/mach-msm/Kconfig
+++ b/arch/arm/mach-msm/Kconfig
@@ -49,6 +49,8 @@ endchoice
 
 config MSM_SOC_REV_A
         bool
+config  ARCH_MSM_SCORPIONMP
+        bool
 
 config  ARCH_MSM_ARM11
         bool
diff --git a/arch/arm/mach-msm/include/mach/vmalloc.h b/arch/arm/mach-msm/include/mach/vmalloc.h
index 31a32ad062dc..d138448eff16 100644
--- a/arch/arm/mach-msm/include/mach/vmalloc.h
+++ b/arch/arm/mach-msm/include/mach/vmalloc.h
@@ -16,7 +16,7 @@
 #ifndef __ASM_ARCH_MSM_VMALLOC_H
 #define __ASM_ARCH_MSM_VMALLOC_H
 
-#define VMALLOC_END       0xd0000000
+#define VMALLOC_END       0xd0000000UL
 
 #endif
 
diff --git a/arch/arm/mach-mx25/devices-imx25.h b/arch/arm/mach-mx25/devices-imx25.h
index 93afa10b13cf..d94d282fa676 100644
--- a/arch/arm/mach-mx25/devices-imx25.h
+++ b/arch/arm/mach-mx25/devices-imx25.h
@@ -42,9 +42,9 @@ extern const struct imx_mxc_nand_data imx25_mxc_nand_data __initconst;
 #define imx25_add_mxc_nand(pdata)       \
         imx_add_mxc_nand(&imx25_mxc_nand_data, pdata)
 
-extern const struct imx_spi_imx_data imx25_spi_imx_data[] __initconst;
+extern const struct imx_spi_imx_data imx25_cspi_data[] __initconst;
 #define imx25_add_spi_imx(id, pdata)    \
-        imx_add_spi_imx(&imx25_spi_imx_data[id], pdata)
+        imx_add_spi_imx(&imx25_cspi_data[id], pdata)
 #define imx25_add_spi_imx0(pdata)       imx25_add_spi_imx(0, pdata)
 #define imx25_add_spi_imx1(pdata)       imx25_add_spi_imx(1, pdata)
 #define imx25_add_spi_imx2(pdata)       imx25_add_spi_imx(2, pdata)
diff --git a/arch/arm/mach-mx3/mach-pcm037_eet.c b/arch/arm/mach-mx3/mach-pcm037_eet.c
index 99e0894e07db..fda56545d2fd 100644
--- a/arch/arm/mach-mx3/mach-pcm037_eet.c
+++ b/arch/arm/mach-mx3/mach-pcm037_eet.c
@@ -14,6 +14,7 @@
 
 #include <mach/common.h>
 #include <mach/iomux-mx3.h>
+#include <mach/spi.h>
 
 #include <asm/mach-types.h>
 
@@ -59,14 +60,12 @@ static struct spi_board_info pcm037_spi_dev[] = {
 };
 
 /* Platform Data for MXC CSPI */
-#if defined(CONFIG_SPI_IMX) || defined(CONFIG_SPI_IMX_MODULE)
 static int pcm037_spi1_cs[] = {MXC_SPI_CS(1), IOMUX_TO_GPIO(MX31_PIN_KEY_COL7)};
 
 static const struct spi_imx_master pcm037_spi1_pdata __initconst = {
         .chipselect = pcm037_spi1_cs,
         .num_chipselect = ARRAY_SIZE(pcm037_spi1_cs),
 };
-#endif
 
 /* GPIO-keys input device */
 static struct gpio_keys_button pcm037_gpio_keys[] = {
@@ -171,7 +170,7 @@ static struct platform_device pcm037_gpio_keys_device = {
         },
 };
 
-static int eet_init_devices(void)
+static int __init eet_init_devices(void)
 {
         if (!machine_is_pcm037() || pcm037_variant() != PCM037_EET)
                 return 0;
diff --git a/arch/arm/mach-netx/include/mach/vmalloc.h b/arch/arm/mach-netx/include/mach/vmalloc.h
index 7cca3574308f..871f1ef7bff5 100644
--- a/arch/arm/mach-netx/include/mach/vmalloc.h
+++ b/arch/arm/mach-netx/include/mach/vmalloc.h
@@ -16,4 +16,4 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
-#define VMALLOC_END       0xd0000000
+#define VMALLOC_END       0xd0000000UL
diff --git a/arch/arm/mach-omap1/devices.c b/arch/arm/mach-omap1/devices.c
index ea0d80a89da7..e7f9ee63dce5 100644
--- a/arch/arm/mach-omap1/devices.c
+++ b/arch/arm/mach-omap1/devices.c
@@ -321,10 +321,9 @@ static struct platform_device omap_wdt_device = {
 static int __init omap_init_wdt(void)
 {
         if (!cpu_is_omap16xx())
-                return;
+                return -ENODEV;
 
-        platform_device_register(&omap_wdt_device);
-        return 0;
+        return platform_device_register(&omap_wdt_device);
 }
 subsys_initcall(omap_init_wdt);
 #endif
diff --git a/arch/arm/mach-omap1/include/mach/camera.h b/arch/arm/mach-omap1/include/mach/camera.h
index fd54b452eb22..847d00f0bb0a 100644
--- a/arch/arm/mach-omap1/include/mach/camera.h
+++ b/arch/arm/mach-omap1/include/mach/camera.h
@@ -1,6 +1,8 @@
 #ifndef __ASM_ARCH_CAMERA_H_
 #define __ASM_ARCH_CAMERA_H_
 
+#include <media/omap1_camera.h>
+
 void omap1_camera_init(void *);
 
 static inline void omap1_set_camera_info(struct omap1_cam_platform_data *info)
diff --git a/arch/arm/mach-omap1/include/mach/vmalloc.h b/arch/arm/mach-omap1/include/mach/vmalloc.h
index b001f67d695b..22ec4a479577 100644
--- a/arch/arm/mach-omap1/include/mach/vmalloc.h
+++ b/arch/arm/mach-omap1/include/mach/vmalloc.h
@@ -17,4 +17,4 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
  */
-#define VMALLOC_END     0xd8000000
+#define VMALLOC_END     0xd8000000UL
diff --git a/arch/arm/mach-omap2/board-devkit8000.c b/arch/arm/mach-omap2/board-devkit8000.c
index 067f4379c87f..53ac762518bd 100644
--- a/arch/arm/mach-omap2/board-devkit8000.c
+++ b/arch/arm/mach-omap2/board-devkit8000.c
@@ -242,9 +242,6 @@ static int devkit8000_twl_gpio_setup(struct device *dev,
         mmc[0].gpio_cd = gpio + 0;
         omap2_hsmmc_init(mmc);
 
-        /* link regulators to MMC adapters */
-        devkit8000_vmmc1_supply.dev = mmc[0].dev;
-
         /* TWL4030_GPIO_MAX + 1 == ledB, PMU_STAT (out, active low LED) */
         gpio_leds[2].gpio = gpio + TWL4030_GPIO_MAX + 1;
 
diff --git a/arch/arm/mach-omap2/include/mach/vmalloc.h b/arch/arm/mach-omap2/include/mach/vmalloc.h
index 4da31e997efe..866319947760 100644
--- a/arch/arm/mach-omap2/include/mach/vmalloc.h
+++ b/arch/arm/mach-omap2/include/mach/vmalloc.h
@@ -17,4 +17,4 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
  */
-#define VMALLOC_END       0xf8000000
+#define VMALLOC_END       0xf8000000UL
diff --git a/arch/arm/mach-pnx4008/include/mach/vmalloc.h b/arch/arm/mach-pnx4008/include/mach/vmalloc.h
index 31b65ee07b0b..184913c71141 100644
--- a/arch/arm/mach-pnx4008/include/mach/vmalloc.h
+++ b/arch/arm/mach-pnx4008/include/mach/vmalloc.h
@@ -17,4 +17,4 @@
  * The vmalloc() routines leaves a hole of 4kB between each vmalloced
  * area for the same reason. ;)
  */
-#define VMALLOC_END       0xd0000000
+#define VMALLOC_END       0xd0000000UL
diff --git a/arch/arm/mach-rpc/include/mach/vmalloc.h b/arch/arm/mach-rpc/include/mach/vmalloc.h
index 3bcd86fadb81..fb700228637a 100644
--- a/arch/arm/mach-rpc/include/mach/vmalloc.h
+++ b/arch/arm/mach-rpc/include/mach/vmalloc.h
@@ -7,4 +7,4 @@
  * it under the terms of the GNU General Public License version 2 as
  * published by the Free Software Foundation.
  */
-#define VMALLOC_END       0xdc000000
+#define VMALLOC_END       0xdc000000UL
diff --git a/arch/arm/mach-s3c64xx/Kconfig b/arch/arm/mach-s3c64xx/Kconfig
index 1ca7bdc6485c..579d2f0f4dd0 100644
--- a/arch/arm/mach-s3c64xx/Kconfig
+++ b/arch/arm/mach-s3c64xx/Kconfig
@@ -143,7 +143,7 @@ config MACH_SMDK6410
         select S3C_DEV_USB_HSOTG
         select S3C_DEV_WDT
         select SAMSUNG_DEV_KEYPAD
-        select HAVE_S3C2410_WATCHDOG
+        select HAVE_S3C2410_WATCHDOG if WATCHDOG
         select S3C64XX_SETUP_SDHCI
         select S3C64XX_SETUP_I2C1
         select S3C64XX_SETUP_IDE
diff --git a/arch/arm/mach-shark/include/mach/vmalloc.h b/arch/arm/mach-shark/include/mach/vmalloc.h
index 8e845b6a7cb5..b10df988526d 100644
--- a/arch/arm/mach-shark/include/mach/vmalloc.h
+++ b/arch/arm/mach-shark/include/mach/vmalloc.h
@@ -1,4 +1,4 @@
 /*
  * arch/arm/mach-shark/include/mach/vmalloc.h
  */
-#define VMALLOC_END       0xd0000000
+#define VMALLOC_END       0xd0000000UL
diff --git a/arch/arm/mach-shmobile/board-ap4evb.c b/arch/arm/mach-shmobile/board-ap4evb.c
index 32d9e2816e56..d3260542b943 100644
--- a/arch/arm/mach-shmobile/board-ap4evb.c
+++ b/arch/arm/mach-shmobile/board-ap4evb.c
@@ -163,11 +163,13 @@ static struct mtd_partition nor_flash_partitions[] = {
                 .name           = "loader",
                 .offset         = 0x00000000,
                 .size           = 512 * 1024,
+                .mask_flags     = MTD_WRITEABLE,
         },
         {
                 .name           = "bootenv",
                 .offset         = MTDPART_OFS_APPEND,
                 .size           = 512 * 1024,
+                .mask_flags     = MTD_WRITEABLE,
         },
         {
                 .name           = "kernel_ro",
@@ -581,6 +583,10 @@ static int fsi_set_rate(int is_porta, int rate)
                 return -EINVAL;
 
         switch (rate) {
+        case 44100:
+                clk_set_rate(fsib_clk, clk_round_rate(fsib_clk, 11283000));
+                ret = SH_FSI_ACKMD_256 | SH_FSI_BPFMD_64;
+                break;
         case 48000:
                 clk_set_rate(fsib_clk, clk_round_rate(fsib_clk, 85428000));
                 clk_set_rate(fdiv_clk, clk_round_rate(fdiv_clk, 12204000));
diff --git a/arch/arm/mach-shmobile/clock-sh7372.c b/arch/arm/mach-shmobile/clock-sh7372.c
index 7db31e6c6bf2..b25ce90a346e 100644
--- a/arch/arm/mach-shmobile/clock-sh7372.c
+++ b/arch/arm/mach-shmobile/clock-sh7372.c
@@ -220,8 +220,7 @@ static void pllc2_disable(struct clk *clk)
         __raw_writel(__raw_readl(PLLC2CR) & ~0x80000000, PLLC2CR);
 }
 
-static int pllc2_set_rate(struct clk *clk,
-                          unsigned long rate, int algo_id)
+static int pllc2_set_rate(struct clk *clk, unsigned long rate)
 {
         unsigned long value;
         int idx;
@@ -463,8 +462,7 @@ static int fsidiv_enable(struct clk *clk)
         return 0;
 }
 
-static int fsidiv_set_rate(struct clk *clk,
-                           unsigned long rate, int algo_id)
+static int fsidiv_set_rate(struct clk *clk, unsigned long rate)
 {
         int idx;
 
diff --git a/arch/arm/mach-shmobile/intc-sh7372.c b/arch/arm/mach-shmobile/intc-sh7372.c
index 4cd3cae38e72..30b2f400666a 100644
--- a/arch/arm/mach-shmobile/intc-sh7372.c
+++ b/arch/arm/mach-shmobile/intc-sh7372.c
@@ -98,7 +98,7 @@ static struct intc_vect intca_vectors[] __initdata = {
         INTC_VECT(IRQ14A, 0x03c0), INTC_VECT(IRQ15A, 0x03e0),
         INTC_VECT(IRQ16A, 0x3200), INTC_VECT(IRQ17A, 0x3220),
         INTC_VECT(IRQ18A, 0x3240), INTC_VECT(IRQ19A, 0x3260),
-        INTC_VECT(IRQ20A, 0x3280), INTC_VECT(IRQ31A, 0x32a0),
+        INTC_VECT(IRQ20A, 0x3280), INTC_VECT(IRQ21A, 0x32a0),
         INTC_VECT(IRQ22A, 0x32c0), INTC_VECT(IRQ23A, 0x32e0),
         INTC_VECT(IRQ24A, 0x3300), INTC_VECT(IRQ25A, 0x3320),
         INTC_VECT(IRQ26A, 0x3340), INTC_VECT(IRQ27A, 0x3360),
diff --git a/arch/arm/mach-ux500/cpu.c b/arch/arm/mach-ux500/cpu.c
index 73fb1a551ec6..608a1372b172 100644
--- a/arch/arm/mach-ux500/cpu.c
+++ b/arch/arm/mach-ux500/cpu.c
@@ -75,14 +75,14 @@ void __init ux500_init_irq(void)
 static inline void ux500_cache_wait(void __iomem *reg, unsigned long mask)
 {
         /* wait for the operation to complete */
-        while (readl(reg) & mask)
+        while (readl_relaxed(reg) & mask)
                 ;
 }
 
 static inline void ux500_cache_sync(void)
 {
         void __iomem *base = __io_address(UX500_L2CC_BASE);
-        writel(0, base + L2X0_CACHE_SYNC);
+        writel_relaxed(0, base + L2X0_CACHE_SYNC);
         ux500_cache_wait(base + L2X0_CACHE_SYNC, 1);
 }
 
@@ -107,7 +107,7 @@ static void ux500_l2x0_inv_all(void)
         uint32_t l2x0_way_mask = (1<<16) - 1;   /* Bitmask of active ways */
 
         /* invalidate all ways */
-        writel(l2x0_way_mask, l2x0_base + L2X0_INV_WAY);
+        writel_relaxed(l2x0_way_mask, l2x0_base + L2X0_INV_WAY);
         ux500_cache_wait(l2x0_base + L2X0_INV_WAY, l2x0_way_mask);
         ux500_cache_sync();
 }
diff --git a/arch/arm/mach-versatile/include/mach/vmalloc.h b/arch/arm/mach-versatile/include/mach/vmalloc.h
index ebd8a2543d3b..7d8e069ad51b 100644
--- a/arch/arm/mach-versatile/include/mach/vmalloc.h
+++ b/arch/arm/mach-versatile/include/mach/vmalloc.h
@@ -18,4 +18,4 @@
  * along with this program; if not, write to the Free Software
  * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
  */
-#define VMALLOC_END             0xd8000000
+#define VMALLOC_END             0xd8000000UL
diff --git a/arch/arm/mm/ioremap.c b/arch/arm/mm/ioremap.c
index 17e7b0b57e49..55c17a6fb22f 100644
--- a/arch/arm/mm/ioremap.c
+++ b/arch/arm/mm/ioremap.c
@@ -206,8 +206,8 @@ void __iomem * __arm_ioremap_pfn_caller(unsigned long pfn,
          */
         if (pfn_valid(pfn)) {
                 printk(KERN_WARNING "BUG: Your driver calls ioremap() on system memory.  This leads\n"
-                       KERN_WARNING "to architecturally unpredictable behaviour on ARMv6+, and ioremap()\n"
-                       KERN_WARNING "will fail in the next kernel release.  Please fix your driver.\n");
+                       "to architecturally unpredictable behaviour on ARMv6+, and ioremap()\n"
+                       "will fail in the next kernel release.  Please fix your driver.\n");
                 WARN_ON(1);
         }
 
diff --git a/arch/arm/plat-mxc/devices/platform-imx-dma.c b/arch/arm/plat-mxc/devices/platform-imx-dma.c
index 02d989018059..3a705c7877dd 100644
--- a/arch/arm/plat-mxc/devices/platform-imx-dma.c
+++ b/arch/arm/plat-mxc/devices/platform-imx-dma.c
@@ -12,15 +12,7 @@
 
 #include <mach/hardware.h>
 #include <mach/devices-common.h>
-#ifdef SDMA_IS_MERGED
 #include <mach/sdma.h>
-#else
-struct sdma_platform_data {
-        int sdma_version;
-        char *cpu_name;
-        int to_version;
-};
-#endif
 
 struct imx_imx_sdma_data {
         resource_size_t iobase;
diff --git a/arch/arm/plat-mxc/devices/platform-spi_imx.c b/arch/arm/plat-mxc/devices/platform-spi_imx.c
index e48340ec331e..17f724c9452d 100644
--- a/arch/arm/plat-mxc/devices/platform-spi_imx.c
+++ b/arch/arm/plat-mxc/devices/platform-spi_imx.c
@@ -27,6 +27,7 @@ const struct imx_spi_imx_data imx21_cspi_data[] __initconst = {
         imx_spi_imx_data_entry(MX21, CSPI, "imx21-cspi", _id, _hwid, SZ_4K)
         imx21_cspi_data_entry(0, 1),
         imx21_cspi_data_entry(1, 2),
+};
 #endif
 
 #ifdef CONFIG_ARCH_MX25
diff --git a/arch/arm/plat-nomadik/timer.c b/arch/arm/plat-nomadik/timer.c
index aedf9c1d645e..63cdc6025bd7 100644
--- a/arch/arm/plat-nomadik/timer.c
+++ b/arch/arm/plat-nomadik/timer.c
@@ -3,6 +3,7 @@
  *
  * Copyright (C) 2008 STMicroelectronics
  * Copyright (C) 2010 Alessandro Rubini
+ * Copyright (C) 2010 Linus Walleij for ST-Ericsson
  *
  * This program is free software; you can redistribute it and/or modify
  * it under the terms of the GNU General Public License version 2, as
@@ -16,11 +17,13 @@
 #include <linux/clk.h>
 #include <linux/jiffies.h>
 #include <linux/err.h>
+#include <linux/cnt32_to_63.h>
+#include <linux/timer.h>
 #include <asm/mach/time.h>
 
 #include <plat/mtu.h>
 
-void __iomem *mtu_base; /* ssigned by machine code */
+void __iomem *mtu_base; /* Assigned by machine code */
 
 /*
  * Kernel assumes that sched_clock can be called early
@@ -48,16 +51,82 @@ static struct clocksource nmdk_clksrc = {
 /*
  * Override the global weak sched_clock symbol with this
  * local implementation which uses the clocksource to get some
- * better resolution when scheduling the kernel. We accept that
- * this wraps around for now, since it is just a relative time
- * stamp. (Inspired by OMAP implementation.)
+ * better resolution when scheduling the kernel.
+ *
+ * Because the hardware timer period may be quite short
+ * (32.3 secs on the 133 MHz MTU timer selection on ux500)
+ * and because cnt32_to_63() needs to be called at least once per
+ * half period to work properly, a kernel keepwarm() timer is set up
+ * to ensure this requirement is always met.
+ *
+ * Also the sched_clock timer will wrap around at some point,
+ * here we set it to run continously for a year.
  */
+#define SCHED_CLOCK_MIN_WRAP 3600*24*365
+static struct timer_list cnt32_to_63_keepwarm_timer;
+static u32 sched_mult;
+static u32 sched_shift;
+
 unsigned long long notrace sched_clock(void)
 {
-        return clocksource_cyc2ns(nmdk_clksrc.read(
-                                  &nmdk_clksrc),
-                                  nmdk_clksrc.mult,
-                                  nmdk_clksrc.shift);
+        u64 cycles;
+
+        if (unlikely(!mtu_base))
+                return 0;
+
+        cycles = cnt32_to_63(-readl(mtu_base + MTU_VAL(0)));
+        /*
+         * sched_mult is guaranteed to be even so will
+         * shift out bit 63
+         */
+        return (cycles * sched_mult) >> sched_shift;
+}
+
+/* Just kick sched_clock every so often */
+static void cnt32_to_63_keepwarm(unsigned long data)
+{
+        mod_timer(&cnt32_to_63_keepwarm_timer, round_jiffies(jiffies + data));
+        (void) sched_clock();
+}
+
+/*
+ * Set up a timer to keep sched_clock():s 32_to_63 algorithm warm
+ * once in half a 32bit timer wrap interval.
+ */
+static void __init nmdk_sched_clock_init(unsigned long rate)
+{
+        u32 v;
+        unsigned long delta;
+        u64 days;
+
+        /* Find the apropriate mult and shift factors */
+        clocks_calc_mult_shift(&sched_mult, &sched_shift,
+                               rate, NSEC_PER_SEC, SCHED_CLOCK_MIN_WRAP);
+        /* We need to multiply by an even number to get rid of bit 63 */
+        if (sched_mult & 1)
+                sched_mult++;
+
+        /* Let's see what we get, take max counter and scale it */
+        days = (0xFFFFFFFFFFFFFFFFLLU * sched_mult) >> sched_shift;
+        do_div(days, NSEC_PER_SEC);
+        do_div(days, (3600*24));
+
+        pr_info("sched_clock: using %d bits @ %lu Hz wrap in %lu days\n",
+                (64 - sched_shift), rate, (unsigned long) days);
+
+        /*
+         * Program a timer to kick us at half 32bit wraparound
+         * Formula: seconds per wrap = (2^32) / f
+         */
+        v = 0xFFFFFFFFUL / rate;
+        /* We want half of the wrap time to keep cnt32_to_63 warm */
+        v /= 2;
+        pr_debug("sched_clock: prescaled timer rate: %lu Hz, "
+                 "initialize keepwarm timer every %d seconds\n", rate, v);
+        /* Convert seconds to jiffies */
+        delta = msecs_to_jiffies(v*1000);
+        setup_timer(&cnt32_to_63_keepwarm_timer, cnt32_to_63_keepwarm, delta);
+        mod_timer(&cnt32_to_63_keepwarm_timer, round_jiffies(jiffies + delta));
 }
 
 /* Clockevent device: use one-shot mode */
@@ -161,13 +230,15 @@ void __init nmdk_timer_init(void)
         writel(0, mtu_base + MTU_BGLR(0));
         writel(cr | MTU_CRn_ENA, mtu_base + MTU_CR(0));
 
-        /* Now the scheduling clock is ready */
+        /* Now the clock source is ready */
         nmdk_clksrc.read = nmdk_read_timer;
 
         if (clocksource_register(&nmdk_clksrc))
                 pr_err("timer: failed to initialize clock source %s\n",
                        nmdk_clksrc.name);
 
+        nmdk_sched_clock_init(rate);
+
         /* Timer 1 is used for events */
 
         clockevents_calc_mult_shift(&nmdk_clkevt, rate, MTU_MIN_RANGE);
diff --git a/arch/arm/plat-omap/dma.c b/arch/arm/plat-omap/dma.c
index f5c5b8da9a87..2c2826571d45 100644
--- a/arch/arm/plat-omap/dma.c
+++ b/arch/arm/plat-omap/dma.c
@@ -1983,6 +1983,8 @@ static int omap2_dma_handle_ch(int ch)
 
         dma_write(OMAP2_DMA_CSR_CLEAR_MASK, CSR(ch));
         dma_write(1 << ch, IRQSTATUS_L0);
+        /* read back the register to flush the write */
+        dma_read(IRQSTATUS_L0);
 
         /* If the ch is not chained then chain_id will be -1 */
         if (dma_chan[ch].chain_id != -1) {
diff --git a/arch/blackfin/kernel/process.c b/arch/blackfin/kernel/process.c
index cd0c090ebc54..b407bc8ad918 100644
--- a/arch/blackfin/kernel/process.c
+++ b/arch/blackfin/kernel/process.c
@@ -7,7 +7,6 @@
  */
 
 #include <linux/module.h>
-#include <linux/smp_lock.h>
 #include <linux/unistd.h>
 #include <linux/user.h>
 #include <linux/uaccess.h>
diff --git a/arch/frv/kernel/process.c b/arch/frv/kernel/process.c
index 2b63b0191f52..efad12071c2e 100644
--- a/arch/frv/kernel/process.c
+++ b/arch/frv/kernel/process.c
@@ -16,7 +16,6 @@
 #include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/stddef.h>
 #include <linux/unistd.h>
 #include <linux/ptrace.h>
diff --git a/arch/h8300/kernel/process.c b/arch/h8300/kernel/process.c
index 97478138e361..933bd388efb2 100644
--- a/arch/h8300/kernel/process.c
+++ b/arch/h8300/kernel/process.c
@@ -28,7 +28,6 @@
 #include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/stddef.h>
 #include <linux/unistd.h>
 #include <linux/ptrace.h>
diff --git a/arch/ia64/hp/sim/simscsi.c b/arch/ia64/hp/sim/simscsi.c
index 3a078ad3aa44..331de723c676 100644
--- a/arch/ia64/hp/sim/simscsi.c
+++ b/arch/ia64/hp/sim/simscsi.c
@@ -202,7 +202,7 @@ simscsi_readwrite10 (struct scsi_cmnd *sc, int mode)
 }
 
 static int
-simscsi_queuecommand (struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *))
+simscsi_queuecommand_lck (struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *))
 {
         unsigned int target_id = sc->device->id;
         char fname[MAX_ROOT_LEN+16];
@@ -326,6 +326,8 @@ simscsi_queuecommand (struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(simscsi_queuecommand)
+
 static int
 simscsi_host_reset (struct scsi_cmnd *sc)
 {
diff --git a/arch/m68k/kernel/process.c b/arch/m68k/kernel/process.c
index 18732ab23292..c2a1fc23dd75 100644
--- a/arch/m68k/kernel/process.c
+++ b/arch/m68k/kernel/process.c
@@ -18,7 +18,6 @@
 #include <linux/slab.h>
 #include <linux/fs.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/stddef.h>
 #include <linux/unistd.h>
 #include <linux/ptrace.h>
diff --git a/arch/m68knommu/kernel/process.c b/arch/m68knommu/kernel/process.c
index 6d3390590e5b..e2a63af5d517 100644
--- a/arch/m68knommu/kernel/process.c
+++ b/arch/m68knommu/kernel/process.c
@@ -19,7 +19,6 @@
 #include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/stddef.h>
 #include <linux/unistd.h>
 #include <linux/ptrace.h>
diff --git a/arch/mn10300/kernel/process.c b/arch/mn10300/kernel/process.c
index 0d0f8049a17b..e1b14a6ed544 100644
--- a/arch/mn10300/kernel/process.c
+++ b/arch/mn10300/kernel/process.c
@@ -14,7 +14,6 @@
 #include <linux/kernel.h>
 #include <linux/mm.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/stddef.h>
 #include <linux/unistd.h>
 #include <linux/ptrace.h>
diff --git a/arch/parisc/hpux/sys_hpux.c b/arch/parisc/hpux/sys_hpux.c
index ba430a03bc7a..30394081d9b6 100644
--- a/arch/parisc/hpux/sys_hpux.c
+++ b/arch/parisc/hpux/sys_hpux.c
@@ -28,7 +28,6 @@
 #include <linux/namei.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/syscalls.h>
 #include <linux/utsname.h>
 #include <linux/vfs.h>
diff --git a/arch/parisc/kernel/sys_parisc32.c b/arch/parisc/kernel/sys_parisc32.c
index 9779ece2b070..88a0ad14a9c9 100644
--- a/arch/parisc/kernel/sys_parisc32.c
+++ b/arch/parisc/kernel/sys_parisc32.c
@@ -20,7 +20,6 @@
 #include <linux/times.h>
 #include <linux/time.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/sem.h>
 #include <linux/msg.h>
 #include <linux/shm.h>
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index b6447190e1a2..e625e9e034ae 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -4,6 +4,10 @@ config PPC32
         bool
         default y if !PPC64
 
+config 32BIT
+        bool
+        default y if PPC32
+
 config 64BIT
         bool
         default y if PPC64
diff --git a/arch/powerpc/boot/div64.S b/arch/powerpc/boot/div64.S
index 722f360a32a9..d271ab542673 100644
--- a/arch/powerpc/boot/div64.S
+++ b/arch/powerpc/boot/div64.S
@@ -33,9 +33,10 @@ __div64_32:
         cntlzw  r0,r5           # we are shifting the dividend right
         li      r10,-1          # to make it < 2^32, and shifting
         srw     r10,r10,r0      # the divisor right the same amount,
-        add     r9,r4,r10       # rounding up (so the estimate cannot
+        addc    r9,r4,r10       # rounding up (so the estimate cannot
         andc    r11,r6,r10      # ever be too large, only too small)
         andc    r9,r9,r10
+        addze   r9,r9
         or      r11,r5,r11
         rotlw   r9,r9,r0
         rotlw   r11,r11,r0
diff --git a/arch/powerpc/kernel/kgdb.c b/arch/powerpc/kernel/kgdb.c
index 7a9db64f3f04..42850ee00ada 100644
--- a/arch/powerpc/kernel/kgdb.c
+++ b/arch/powerpc/kernel/kgdb.c
@@ -337,7 +337,7 @@ char *dbg_get_reg(int regno, void *mem, struct pt_regs *regs)
                 /* FP registers 32 -> 63 */
 #if defined(CONFIG_FSL_BOOKE) && defined(CONFIG_SPE)
                 if (current)
-                        memcpy(mem, current->thread.evr[regno-32],
+                        memcpy(mem, &current->thread.evr[regno-32],
                                         dbg_reg_def[regno].size);
 #else
                 /* fp registers not used by kernel, leave zero */
@@ -362,7 +362,7 @@ int dbg_set_reg(int regno, void *mem, struct pt_regs *regs)
         if (regno >= 32 && regno < 64) {
                 /* FP registers 32 -> 63 */
 #if defined(CONFIG_FSL_BOOKE) && defined(CONFIG_SPE)
-                memcpy(current->thread.evr[regno-32], mem,
+                memcpy(&current->thread.evr[regno-32], mem,
                                 dbg_reg_def[regno].size);
 #else
                 /* fp registers not used by kernel, leave zero */
diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c
index 2a178b0ebcdf..ce6f61c6f871 100644
--- a/arch/powerpc/kernel/setup_64.c
+++ b/arch/powerpc/kernel/setup_64.c
@@ -497,9 +497,8 @@ static void __init emergency_stack_init(void)
 }
 
 /*
- * Called into from start_kernel, after lock_kernel has been called.
- * Initializes bootmem, which is unsed to manage page allocation until
- * mem_init is called.
+ * Called into from start_kernel this initializes bootmem, which is used
+ * to manage page allocation until mem_init is called.
  */
 void __init setup_arch(char **cmdline_p)
 {
diff --git a/arch/powerpc/kernel/sys_ppc32.c b/arch/powerpc/kernel/sys_ppc32.c
index b1b6043a56c4..4e5bf1edc0f2 100644
--- a/arch/powerpc/kernel/sys_ppc32.c
+++ b/arch/powerpc/kernel/sys_ppc32.c
@@ -23,7 +23,6 @@
 #include <linux/resource.h>
 #include <linux/times.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/sem.h>
 #include <linux/msg.h>
 #include <linux/shm.h>
diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c
index 83f534d862db..5e9584405c45 100644
--- a/arch/powerpc/mm/hash_utils_64.c
+++ b/arch/powerpc/mm/hash_utils_64.c
@@ -1123,7 +1123,7 @@ void hash_preload(struct mm_struct *mm, unsigned long ea,
         else
 #endif /* CONFIG_PPC_HAS_HASH_64K */
                 rc = __hash_page_4K(ea, access, vsid, ptep, trap, local, ssize,
-                                    subpage_protection(pgdir, ea));
+                                    subpage_protection(mm, ea));
 
         /* Dump some info in case of hash insertion failure, they should
          * never happen so it is really useful to know if/when they do
diff --git a/arch/powerpc/mm/tlb_low_64e.S b/arch/powerpc/mm/tlb_low_64e.S
index 8b04c54e596f..8526bd9d2aa3 100644
--- a/arch/powerpc/mm/tlb_low_64e.S
+++ b/arch/powerpc/mm/tlb_low_64e.S
@@ -138,8 +138,11 @@
         cmpldi  cr0,r15,0                       /* Check for user region */
         std     r14,EX_TLB_ESR(r12)             /* write crazy -1 to frame */
         beq     normal_tlb_miss
+
+        li      r11,_PAGE_PRESENT|_PAGE_BAP_SX  /* Base perm */
+        oris    r11,r11,_PAGE_ACCESSED@h
         /* XXX replace the RMW cycles with immediate loads + writes */
-1:      mfspr   r10,SPRN_MAS1
+        mfspr   r10,SPRN_MAS1
         cmpldi  cr0,r15,8                       /* Check for vmalloc region */
         rlwinm  r10,r10,0,16,1                  /* Clear TID */
         mtspr   SPRN_MAS1,r10
diff --git a/arch/powerpc/mm/tlb_nohash.c b/arch/powerpc/mm/tlb_nohash.c
index 36c0c449a899..2a030d89bbc6 100644
--- a/arch/powerpc/mm/tlb_nohash.c
+++ b/arch/powerpc/mm/tlb_nohash.c
@@ -585,6 +585,6 @@ void setup_initial_memory_limit(phys_addr_t first_memblock_base,
         ppc64_rma_size = min_t(u64, first_memblock_size, 0x40000000);
 
         /* Finally limit subsequent allocations */
-        memblock_set_current_limit(ppc64_memblock_base + ppc64_rma_size);
+        memblock_set_current_limit(first_memblock_base + ppc64_rma_size);
 }
 #endif /* CONFIG_PPC64 */
diff --git a/arch/powerpc/platforms/pseries/Kconfig b/arch/powerpc/platforms/pseries/Kconfig
index c667f0f02c34..3139814f6439 100644
--- a/arch/powerpc/platforms/pseries/Kconfig
+++ b/arch/powerpc/platforms/pseries/Kconfig
@@ -47,6 +47,12 @@ config LPARCFG
 config PPC_PSERIES_DEBUG
         depends on PPC_PSERIES && PPC_EARLY_DEBUG
         bool "Enable extra debug logging in platforms/pseries"
+        help
+          Say Y here if you want the pseries core to produce a bunch of
+          debug messages to the system log. Select this if you are having a
+          problem with the pseries core and want to see more of what is
+          going on. This does not enable debugging in lpar.c, which must
+          be manually done due to its verbosity.
         default y
 
 config PPC_SMLPAR
diff --git a/arch/powerpc/platforms/pseries/eeh.c b/arch/powerpc/platforms/pseries/eeh.c
index 34b7dc12e731..17a11c82e6f8 100644
--- a/arch/powerpc/platforms/pseries/eeh.c
+++ b/arch/powerpc/platforms/pseries/eeh.c
@@ -21,8 +21,6 @@
  * Please address comments and feedback to Linas Vepstas <linas@austin.ibm.com>
  */
 
-#undef DEBUG
-
 #include <linux/delay.h>
 #include <linux/init.h>
 #include <linux/list.h>
diff --git a/arch/powerpc/platforms/pseries/pci_dlpar.c b/arch/powerpc/platforms/pseries/pci_dlpar.c
index 4b7a062dee15..5fcc92a12d3e 100644
--- a/arch/powerpc/platforms/pseries/pci_dlpar.c
+++ b/arch/powerpc/platforms/pseries/pci_dlpar.c
@@ -25,8 +25,6 @@
  * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  */
 
-#undef DEBUG
-
 #include <linux/pci.h>
 #include <asm/pci-bridge.h>
 #include <asm/ppc-pci.h>
diff --git a/arch/s390/Kconfig.debug b/arch/s390/Kconfig.debug
index 45e0c6199f36..05221b13ffb1 100644
--- a/arch/s390/Kconfig.debug
+++ b/arch/s390/Kconfig.debug
@@ -6,6 +6,18 @@ config TRACE_IRQFLAGS_SUPPORT
 
 source "lib/Kconfig.debug"
 
+config STRICT_DEVMEM
+        def_bool y
+        prompt "Filter access to /dev/mem"
+        ---help---
+          This option restricts access to /dev/mem.  If this option is
+          disabled, you allow userspace access to all memory, including
+          kernel and userspace memory. Accidental memory access is likely
+          to be disastrous.
+          Memory access is required for experts who want to debug the kernel.
+
+          If you are unsure, say Y.
+
 config DEBUG_STRICT_USER_COPY_CHECKS
         bool "Strict user copy size checks"
         ---help---
diff --git a/arch/s390/include/asm/page.h b/arch/s390/include/asm/page.h
index a8729ea7e9ac..3c987e9ec8d6 100644
--- a/arch/s390/include/asm/page.h
+++ b/arch/s390/include/asm/page.h
@@ -130,6 +130,11 @@ struct page;
 void arch_free_page(struct page *page, int order);
 void arch_alloc_page(struct page *page, int order);
 
+static inline int devmem_is_allowed(unsigned long pfn)
+{
+        return 0;
+}
+
 #define HAVE_ARCH_FREE_PAGE
 #define HAVE_ARCH_ALLOC_PAGE
 
diff --git a/arch/s390/kernel/compat_linux.c b/arch/s390/kernel/compat_linux.c
index 1e6449c79ab6..53acaa86dd94 100644
--- a/arch/s390/kernel/compat_linux.c
+++ b/arch/s390/kernel/compat_linux.c
@@ -25,7 +25,6 @@
 #include <linux/resource.h>
 #include <linux/times.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/sem.h>
 #include <linux/msg.h>
 #include <linux/shm.h>
diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c
index d60fc4398516..2564793ec2b6 100644
--- a/arch/s390/kernel/kprobes.c
+++ b/arch/s390/kernel/kprobes.c
@@ -30,6 +30,7 @@
 #include <asm/sections.h>
 #include <linux/module.h>
 #include <linux/slab.h>
+#include <linux/hardirq.h>
 
 DEFINE_PER_CPU(struct kprobe *, current_kprobe) = NULL;
 DEFINE_PER_CPU(struct kprobe_ctlblk, kprobe_ctlblk);
@@ -212,7 +213,7 @@ static void __kprobes prepare_singlestep(struct kprobe *p, struct pt_regs *regs)
         /* Set the PER control regs, turns on single step for this address */
         __ctl_load(kprobe_per_regs, 9, 11);
         regs->psw.mask |= PSW_MASK_PER;
-        regs->psw.mask &= ~(PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK);
+        regs->psw.mask &= ~(PSW_MASK_IO | PSW_MASK_EXT);
 }
 
 static void __kprobes save_previous_kprobe(struct kprobe_ctlblk *kcb)
@@ -239,7 +240,7 @@ static void __kprobes set_current_kprobe(struct kprobe *p, struct pt_regs *regs,
         __get_cpu_var(current_kprobe) = p;
         /* Save the interrupt and per flags */
         kcb->kprobe_saved_imask = regs->psw.mask &
-            (PSW_MASK_PER | PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK);
+                (PSW_MASK_PER | PSW_MASK_IO | PSW_MASK_EXT);
         /* Save the control regs that govern PER */
         __ctl_store(kcb->kprobe_saved_ctl, 9, 11);
 }
@@ -316,8 +317,6 @@ static int __kprobes kprobe_handler(struct pt_regs *regs)
                 return 1;
 
 ss_probe:
-        if (regs->psw.mask & (PSW_MASK_PER | PSW_MASK_IO))
-                local_irq_disable();
         prepare_singlestep(p, regs);
         kcb->kprobe_status = KPROBE_HIT_SS;
         return 1;
@@ -350,6 +349,7 @@ static int __kprobes trampoline_probe_handler(struct kprobe *p,
         struct hlist_node *node, *tmp;
         unsigned long flags, orig_ret_address = 0;
         unsigned long trampoline_address = (unsigned long)&kretprobe_trampoline;
+        kprobe_opcode_t *correct_ret_addr = NULL;
 
         INIT_HLIST_HEAD(&empty_rp);
         kretprobe_hash_lock(current, &head, &flags);
@@ -372,10 +372,32 @@ static int __kprobes trampoline_probe_handler(struct kprobe *p,
                         /* another task is sharing our hash bucket */
                         continue;
 
-                if (ri->rp && ri->rp->handler)
-                        ri->rp->handler(ri, regs);
+                orig_ret_address = (unsigned long)ri->ret_addr;
+
+                if (orig_ret_address != trampoline_address)
+                        /*
+                         * This is the real return address. Any other
+                         * instances associated with this task are for
+                         * other calls deeper on the call stack
+                         */
+                        break;
+        }
+
+        kretprobe_assert(ri, orig_ret_address, trampoline_address);
+
+        correct_ret_addr = ri->ret_addr;
+        hlist_for_each_entry_safe(ri, node, tmp, head, hlist) {
+                if (ri->task != current)
+                        /* another task is sharing our hash bucket */
+                        continue;
 
                 orig_ret_address = (unsigned long)ri->ret_addr;
+
+                if (ri->rp && ri->rp->handler) {
+                        ri->ret_addr = correct_ret_addr;
+                        ri->rp->handler(ri, regs);
+                }
+
                 recycle_rp_inst(ri, &empty_rp);
 
                 if (orig_ret_address != trampoline_address) {
@@ -387,7 +409,7 @@ static int __kprobes trampoline_probe_handler(struct kprobe *p,
                         break;
                 }
         }
-        kretprobe_assert(ri, orig_ret_address, trampoline_address);
+
         regs->psw.addr = orig_ret_address | PSW_ADDR_AMODE;
 
         reset_current_kprobe();
@@ -465,8 +487,6 @@ static int __kprobes post_kprobe_handler(struct pt_regs *regs)
                 goto out;
         }
         reset_current_kprobe();
-        if (regs->psw.mask & (PSW_MASK_PER | PSW_MASK_IO))
-                local_irq_enable();
 out:
         preempt_enable_no_resched();
 
@@ -482,7 +502,7 @@ out:
         return 1;
 }
 
-int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
+static int __kprobes kprobe_trap_handler(struct pt_regs *regs, int trapnr)
 {
         struct kprobe *cur = kprobe_running();
         struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
@@ -508,8 +528,6 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
                         restore_previous_kprobe(kcb);
                 else {
                         reset_current_kprobe();
-                        if (regs->psw.mask & (PSW_MASK_PER | PSW_MASK_IO))
-                                local_irq_enable();
                 }
                 preempt_enable_no_resched();
                 break;
@@ -553,6 +571,18 @@ int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
         return 0;
 }
 
+int __kprobes kprobe_fault_handler(struct pt_regs *regs, int trapnr)
+{
+        int ret;
+
+        if (regs->psw.mask & (PSW_MASK_IO | PSW_MASK_EXT))
+                local_irq_disable();
+        ret = kprobe_trap_handler(regs, trapnr);
+        if (regs->psw.mask & (PSW_MASK_IO | PSW_MASK_EXT))
+                local_irq_restore(regs->psw.mask & ~PSW_MASK_PER);
+        return ret;
+}
+
 /*
  * Wrapper routine to for handling exceptions.
  */
@@ -560,8 +590,12 @@ int __kprobes kprobe_exceptions_notify(struct notifier_block *self,
                                        unsigned long val, void *data)
 {
         struct die_args *args = (struct die_args *)data;
+        struct pt_regs *regs = args->regs;
         int ret = NOTIFY_DONE;
 
+        if (regs->psw.mask & (PSW_MASK_IO | PSW_MASK_EXT))
+                local_irq_disable();
+
         switch (val) {
         case DIE_BPT:
                 if (kprobe_handler(args->regs))
@@ -572,16 +606,17 @@ int __kprobes kprobe_exceptions_notify(struct notifier_block *self,
                         ret = NOTIFY_STOP;
                 break;
         case DIE_TRAP:
-                /* kprobe_running() needs smp_processor_id() */
-                preempt_disable();
-                if (kprobe_running() &&
-                    kprobe_fault_handler(args->regs, args->trapnr))
+                if (!preemptible() && kprobe_running() &&
+                    kprobe_trap_handler(args->regs, args->trapnr))
                         ret = NOTIFY_STOP;
-                preempt_enable();
                 break;
         default:
                 break;
         }
+
+        if (regs->psw.mask & (PSW_MASK_IO | PSW_MASK_EXT))
+                local_irq_restore(regs->psw.mask & ~PSW_MASK_PER);
+
         return ret;
 }
 
@@ -595,6 +630,7 @@ int __kprobes setjmp_pre_handler(struct kprobe *p, struct pt_regs *regs)
 
         /* setup return addr to the jprobe handler routine */
         regs->psw.addr = (unsigned long)(jp->entry) | PSW_ADDR_AMODE;
+        regs->psw.mask &= ~(PSW_MASK_IO | PSW_MASK_EXT);
 
         /* r14 is the function return address */
         kcb->jprobe_saved_r14 = (unsigned long)regs->gprs[14];
diff --git a/arch/s390/mm/gup.c b/arch/s390/mm/gup.c
index 38e641cdd977..45b405ca2567 100644
--- a/arch/s390/mm/gup.c
+++ b/arch/s390/mm/gup.c
@@ -20,18 +20,17 @@
 static inline int gup_pte_range(pmd_t *pmdp, pmd_t pmd, unsigned long addr,
                 unsigned long end, int write, struct page **pages, int *nr)
 {
-        unsigned long mask, result;
+        unsigned long mask;
         pte_t *ptep, pte;
         struct page *page;
 
-        result = write ? 0 : _PAGE_RO;
-        mask = result | _PAGE_INVALID | _PAGE_SPECIAL;
+        mask = (write ? _PAGE_RO : 0) | _PAGE_INVALID | _PAGE_SPECIAL;
 
         ptep = ((pte_t *) pmd_deref(pmd)) + pte_index(addr);
         do {
                 pte = *ptep;
                 barrier();
-                if ((pte_val(pte) & mask) != result)
+                if ((pte_val(pte) & mask) != 0)
                         return 0;
                 VM_BUG_ON(!pfn_valid(pte_pfn(pte)));
                 page = pte_page(pte);
diff --git a/arch/sh/include/asm/processor_32.h b/arch/sh/include/asm/processor_32.h
index 46d5179c9f49..e3c73cdd8c90 100644
--- a/arch/sh/include/asm/processor_32.h
+++ b/arch/sh/include/asm/processor_32.h
@@ -199,10 +199,13 @@ extern unsigned long get_wchan(struct task_struct *p);
 #define ARCH_HAS_PREFETCHW
 static inline void prefetch(void *x)
 {
-        __asm__ __volatile__ ("pref @%0\n\t" : : "r" (x) : "memory");
+        __builtin_prefetch(x, 0, 3);
 }
 
-#define prefetchw(x)    prefetch(x)
+static inline void prefetchw(void *x)
+{
+        __builtin_prefetch(x, 1, 3);
+}
 #endif
 
 #endif /* __KERNEL__ */
diff --git a/arch/sh/kernel/cpu/sh4/clock-sh4-202.c b/arch/sh/kernel/cpu/sh4/clock-sh4-202.c
index 4eabc68cd753..b601fa3978d1 100644
--- a/arch/sh/kernel/cpu/sh4/clock-sh4-202.c
+++ b/arch/sh/kernel/cpu/sh4/clock-sh4-202.c
@@ -110,7 +110,7 @@ static int shoc_clk_verify_rate(struct clk *clk, unsigned long rate)
         return 0;
 }
 
-static int shoc_clk_set_rate(struct clk *clk, unsigned long rate, int algo_id)
+static int shoc_clk_set_rate(struct clk *clk, unsigned long rate)
 {
         unsigned long frqcr3;
         unsigned int tmp;
diff --git a/arch/sh/kernel/sys_sh.c b/arch/sh/kernel/sys_sh.c
index 81f58371613d..8c6a350df751 100644
--- a/arch/sh/kernel/sys_sh.c
+++ b/arch/sh/kernel/sys_sh.c
@@ -88,7 +88,7 @@ asmlinkage int sys_cacheflush(unsigned long addr, unsigned long len, int op)
         }
 
         if (op & CACHEFLUSH_I)
-                flush_cache_all();
+                flush_icache_range(addr, addr+len);
 
         up_read(&current->mm->mmap_sem);
         return 0;
diff --git a/arch/sh/kernel/vsyscall/vsyscall-trapa.S b/arch/sh/kernel/vsyscall/vsyscall-trapa.S
index 3b6eb34c43fa..3e70f851cdc6 100644
--- a/arch/sh/kernel/vsyscall/vsyscall-trapa.S
+++ b/arch/sh/kernel/vsyscall/vsyscall-trapa.S
@@ -8,9 +8,9 @@ __kernel_vsyscall:
          * fill out .eh_frame -- PFM. */
 .LEND_vsyscall:
         .size __kernel_vsyscall,.-.LSTART_vsyscall
-        .previous
 
         .section .eh_frame,"a",@progbits
+        .previous
 .LCIE:
         .ualong .LCIE_end - .LCIE_start
 .LCIE_start:
diff --git a/arch/sparc/kernel/leon_smp.c b/arch/sparc/kernel/leon_smp.c
index 7524689b03d2..16582d85368a 100644
--- a/arch/sparc/kernel/leon_smp.c
+++ b/arch/sparc/kernel/leon_smp.c
@@ -12,7 +12,6 @@
 #include <linux/sched.h>
 #include <linux/threads.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/interrupt.h>
 #include <linux/kernel_stat.h>
 #include <linux/init.h>
diff --git a/arch/sparc/kernel/sys_sparc32.c b/arch/sparc/kernel/sys_sparc32.c
index e6375a750d9a..6db18c6927fb 100644
--- a/arch/sparc/kernel/sys_sparc32.c
+++ b/arch/sparc/kernel/sys_sparc32.c
@@ -17,7 +17,6 @@
 #include <linux/resource.h>
 #include <linux/times.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/sem.h>
 #include <linux/msg.h>
 #include <linux/shm.h>
diff --git a/arch/sparc/kernel/sys_sparc_32.c b/arch/sparc/kernel/sys_sparc_32.c
index 675c9e11ada5..42b282fa6112 100644
--- a/arch/sparc/kernel/sys_sparc_32.c
+++ b/arch/sparc/kernel/sys_sparc_32.c
@@ -19,7 +19,6 @@
 #include <linux/mman.h>
 #include <linux/utsname.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/ipc.h>
 
 #include <asm/uaccess.h>
diff --git a/arch/sparc/kernel/unaligned_32.c b/arch/sparc/kernel/unaligned_32.c
index 12b9f352595f..4491f4cb2695 100644
--- a/arch/sparc/kernel/unaligned_32.c
+++ b/arch/sparc/kernel/unaligned_32.c
@@ -16,7 +16,6 @@
 #include <asm/system.h>
 #include <asm/uaccess.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/perf_event.h>
 
 enum direction {
diff --git a/arch/sparc/kernel/windows.c b/arch/sparc/kernel/windows.c
index b351770cbdd6..3107381e576d 100644
--- a/arch/sparc/kernel/windows.c
+++ b/arch/sparc/kernel/windows.c
@@ -9,7 +9,6 @@
 #include <linux/string.h>
 #include <linux/mm.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 
 #include <asm/uaccess.h>
 
diff --git a/arch/tile/Kconfig b/arch/tile/Kconfig
index 07ec8a865c1d..e11b5fcb70eb 100644
--- a/arch/tile/Kconfig
+++ b/arch/tile/Kconfig
@@ -329,6 +329,18 @@ endmenu  # Tilera-specific configuration
 
 menu "Bus options"
 
+config PCI
+        bool "PCI support"
+        default y
+        select PCI_DOMAINS
+        ---help---
+          Enable PCI root complex support, so PCIe endpoint devices can
+          be attached to the Tile chip.  Many, but not all, PCI devices
+          are supported under Tilera's root complex driver.
+
+config PCI_DOMAINS
+        bool
+
 config NO_IOMEM
         def_bool !PCI
 
diff --git a/arch/tile/include/asm/cacheflush.h b/arch/tile/include/asm/cacheflush.h
index c5741da4eeac..14a3f8556ace 100644
--- a/arch/tile/include/asm/cacheflush.h
+++ b/arch/tile/include/asm/cacheflush.h
@@ -137,4 +137,56 @@ static inline void finv_buffer(void *buffer, size_t size)
         mb_incoherent();
 }
 
+/*
+ * Flush & invalidate a VA range that is homed remotely on a single core,
+ * waiting until the memory controller holds the flushed values.
+ */
+static inline void finv_buffer_remote(void *buffer, size_t size)
+{
+        char *p;
+        int i;
+
+        /*
+         * Flush and invalidate the buffer out of the local L1/L2
+         * and request the home cache to flush and invalidate as well.
+         */
+        __finv_buffer(buffer, size);
+
+        /*
+         * Wait for the home cache to acknowledge that it has processed
+         * all the flush-and-invalidate requests.  This does not mean
+         * that the flushed data has reached the memory controller yet,
+         * but it does mean the home cache is processing the flushes.
+         */
+        __insn_mf();
+
+        /*
+         * Issue a load to the last cache line, which can't complete
+         * until all the previously-issued flushes to the same memory
+         * controller have also completed.  If we weren't striping
+         * memory, that one load would be sufficient, but since we may
+         * be, we also need to back up to the last load issued to
+         * another memory controller, which would be the point where
+         * we crossed an 8KB boundary (the granularity of striping
+         * across memory controllers).  Keep backing up and doing this
+         * until we are before the beginning of the buffer, or have
+         * hit all the controllers.
+         */
+        for (i = 0, p = (char *)buffer + size - 1;
+             i < (1 << CHIP_LOG_NUM_MSHIMS()) && p >= (char *)buffer;
+             ++i) {
+                const unsigned long STRIPE_WIDTH = 8192;
+
+                /* Force a load instruction to issue. */
+                *(volatile char *)p;
+
+                /* Jump to end of previous stripe. */
+                p -= STRIPE_WIDTH;
+                p = (char *)((unsigned long)p | (STRIPE_WIDTH - 1));
+        }
+
+        /* Wait for the loads (and thus flushes) to have completed. */
+        __insn_mf();
+}
+
 #endif /* _ASM_TILE_CACHEFLUSH_H */
diff --git a/arch/tile/include/asm/io.h b/arch/tile/include/asm/io.h
index ee43328713ab..d3cbb9b14cbe 100644
--- a/arch/tile/include/asm/io.h
+++ b/arch/tile/include/asm/io.h
@@ -55,9 +55,6 @@ extern void iounmap(volatile void __iomem *addr);
 #define ioremap_writethrough(physaddr, size)    ioremap(physaddr, size)
 #define ioremap_fullcache(physaddr, size)       ioremap(physaddr, size)
 
-void __iomem *ioport_map(unsigned long port, unsigned int len);
-extern inline void ioport_unmap(void __iomem *addr) {}
-
 #define mmiowb()
 
 /* Conversion between virtual and physical mappings.  */
@@ -189,12 +186,22 @@ static inline void memcpy_toio(volatile void __iomem *dst, const void *src,
  * we never run, uses them unconditionally.
  */
 
-static inline int ioport_panic(void)
+static inline long ioport_panic(void)
 {
         panic("inb/outb and friends do not exist on tile");
         return 0;
 }
 
+static inline void __iomem *ioport_map(unsigned long port, unsigned int len)
+{
+        return (void __iomem *) ioport_panic();
+}
+
+static inline void ioport_unmap(void __iomem *addr)
+{
+        ioport_panic();
+}
+
 static inline u8 inb(unsigned long addr)
 {
         return ioport_panic();
diff --git a/arch/tile/include/asm/pci-bridge.h b/arch/tile/include/asm/pci-bridge.h
deleted file mode 100644
index e853b0e2793b..000000000000
--- a/arch/tile/include/asm/pci-bridge.h
+++ /dev/null
@@ -1,117 +0,0 @@
-/*
- * Copyright 2010 Tilera Corporation. All Rights Reserved.
- *
- *   This program is free software; you can redistribute it and/or
- *   modify it under the terms of the GNU General Public License
- *   as published by the Free Software Foundation, version 2.
- *
- *   This program is distributed in the hope that it will be useful, but
- *   WITHOUT ANY WARRANTY; without even the implied warranty of
- *   MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
- *   NON INFRINGEMENT.  See the GNU General Public License for
- *   more details.
- */
-
-#ifndef _ASM_TILE_PCI_BRIDGE_H
-#define _ASM_TILE_PCI_BRIDGE_H
-
-#include <linux/ioport.h>
-#include <linux/pci.h>
-
-struct device_node;
-struct pci_controller;
-
-/*
- * pci_io_base returns the memory address at which you can access
- * the I/O space for PCI bus number `bus' (or NULL on error).
- */
-extern void __iomem *pci_bus_io_base(unsigned int bus);
-extern unsigned long pci_bus_io_base_phys(unsigned int bus);
-extern unsigned long pci_bus_mem_base_phys(unsigned int bus);
-
-/* Allocate a new PCI host bridge structure */
-extern struct pci_controller *pcibios_alloc_controller(void);
-
-/* Helper function for setting up resources */
-extern void pci_init_resource(struct resource *res, unsigned long start,
-                              unsigned long end, int flags, char *name);
-
-/* Get the PCI host controller for a bus */
-extern struct pci_controller *pci_bus_to_hose(int bus);
-
-/*
- * Structure of a PCI controller (host bridge)
- */
-struct pci_controller {
-        int index;              /* PCI domain number */
-        struct pci_bus *root_bus;
-
-        int first_busno;
-        int last_busno;
-
-        int hv_cfg_fd[2];       /* config{0,1} fds for this PCIe controller */
-        int hv_mem_fd;          /* fd to Hypervisor for MMIO operations */
-
-        struct pci_ops *ops;
-
-        int irq_base;           /* Base IRQ from the Hypervisor */
-        int plx_gen1;           /* flag for PLX Gen 1 configuration */
-
-        /* Address ranges that are routed to this controller/bridge. */
-        struct resource mem_resources[3];
-};
-
-static inline struct pci_controller *pci_bus_to_host(struct pci_bus *bus)
-{
-        return bus->sysdata;
-}
-
-extern void setup_indirect_pci_nomap(struct pci_controller *hose,
-                               void __iomem *cfg_addr, void __iomem *cfg_data);
-extern void setup_indirect_pci(struct pci_controller *hose,
-                               u32 cfg_addr, u32 cfg_data);
-extern void setup_grackle(struct pci_controller *hose);
-
-extern unsigned char common_swizzle(struct pci_dev *, unsigned char *);
-
-/*
- *   The following code swizzles for exactly one bridge.  The routine
- *   common_swizzle below handles multiple bridges.  But there are a
- *   some boards that don't follow the PCI spec's suggestion so we
- *   break this piece out separately.
- */
-static inline unsigned char bridge_swizzle(unsigned char pin,
-                unsigned char idsel)
-{
-        return (((pin-1) + idsel) % 4) + 1;
-}
-
-/*
- * The following macro is used to lookup irqs in a standard table
- * format for those PPC systems that do not already have PCI
- * interrupts properly routed.
- */
-/* FIXME - double check this */
-#define PCI_IRQ_TABLE_LOOKUP ({ \
-        long _ctl_ = -1; \
-        if (idsel >= min_idsel && idsel <= max_idsel && pin <= irqs_per_slot) \
-                _ctl_ = pci_irq_table[idsel - min_idsel][pin-1]; \
-        _ctl_; \
-})
-
-/*
- * Scan the buses below a given PCI host bridge and assign suitable
- * resources to all devices found.
- */
-extern int pciauto_bus_scan(struct pci_controller *, int);
-
-#ifdef CONFIG_PCI
-extern unsigned long pci_address_to_pio(phys_addr_t address);
-#else
-static inline unsigned long pci_address_to_pio(phys_addr_t address)
-{
-        return (unsigned long)-1;
-}
-#endif
-
-#endif /* _ASM_TILE_PCI_BRIDGE_H */
diff --git a/arch/tile/include/asm/pci.h b/arch/tile/include/asm/pci.h
index b0c15da2d5d5..c3fc458a0d32 100644
--- a/arch/tile/include/asm/pci.h
+++ b/arch/tile/include/asm/pci.h
@@ -15,7 +15,29 @@
 #ifndef _ASM_TILE_PCI_H
 #define _ASM_TILE_PCI_H
 
-#include <asm/pci-bridge.h>
+#include <linux/pci.h>
+
+/*
+ * Structure of a PCI controller (host bridge)
+ */
+struct pci_controller {
+        int index;              /* PCI domain number */
+        struct pci_bus *root_bus;
+
+        int first_busno;
+        int last_busno;
+
+        int hv_cfg_fd[2];       /* config{0,1} fds for this PCIe controller */
+        int hv_mem_fd;          /* fd to Hypervisor for MMIO operations */
+
+        struct pci_ops *ops;
+
+        int irq_base;           /* Base IRQ from the Hypervisor */
+        int plx_gen1;           /* flag for PLX Gen 1 configuration */
+
+        /* Address ranges that are routed to this controller/bridge. */
+        struct resource mem_resources[3];
+};
 
 /*
  * The hypervisor maps the entirety of CPA-space as bus addresses, so
@@ -24,56 +46,12 @@
  */
 #define PCI_DMA_BUS_IS_PHYS     1
 
-struct pci_controller *pci_bus_to_hose(int bus);
-unsigned char __init common_swizzle(struct pci_dev *dev, unsigned char *pinp);
 int __init tile_pci_init(void);
-void pci_iounmap(struct pci_dev *dev, void __iomem *addr);
-void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max);
-void __devinit pcibios_fixup_bus(struct pci_bus *bus);
 
-int __devinit _tile_cfg_read(struct pci_controller *hose,
-                                    int bus,
-                                    int slot,
-                                    int function,
-                                    int offset,
-                                    int size,
-                                    u32 *val);
-int __devinit _tile_cfg_write(struct pci_controller *hose,
-                                     int bus,
-                                     int slot,
-                                     int function,
-                                     int offset,
-                                     int size,
-                                     u32 val);
+void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max);
+static inline void pci_iounmap(struct pci_dev *dev, void __iomem *addr) {}
 
-/*
- * These are used to to config reads and writes in the early stages of
- * setup before the driver infrastructure has been set up enough to be
- * able to do config reads and writes.
- */
-#define early_cfg_read(where, size, value) \
-        _tile_cfg_read(controller, \
-                       current_bus, \
-                       pci_slot, \
-                       pci_fn, \
-                       where, \
-                       size, \
-                       value)
-
-#define early_cfg_write(where, size, value) \
-        _tile_cfg_write(controller, \
-                       current_bus, \
-                       pci_slot, \
-                       pci_fn, \
-                       where, \
-                       size, \
-                       value)
-
-
-
-#define PCICFG_BYTE     1
-#define PCICFG_WORD     2
-#define PCICFG_DWORD    4
+void __devinit pcibios_fixup_bus(struct pci_bus *bus);
 
 #define TILE_NUM_PCIE   2
 
@@ -88,33 +66,33 @@ static inline int pci_proc_domain(struct pci_bus *bus)
 }
 
 /*
- * I/O space is currently not supported.
+ * pcibios_assign_all_busses() tells whether or not the bus numbers
+ * should be reassigned, in case the BIOS didn't do it correctly, or
+ * in case we don't have a BIOS and we want to let Linux do it.
  */
+static inline int pcibios_assign_all_busses(void)
+{
+        return 1;
+}
 
-#define TILE_PCIE_LOWER_IO              0x0
-#define TILE_PCIE_UPPER_IO              0x10000
-#define TILE_PCIE_PCIE_IO_SIZE          0x0000FFFF
-
-#define _PAGE_NO_CACHE          0
-#define _PAGE_GUARDED           0
-
-
-#define pcibios_assign_all_busses()    pci_assign_all_buses
-extern int pci_assign_all_buses;
-
+/*
+ * No special bus mastering setup handling.
+ */
 static inline void pcibios_set_master(struct pci_dev *dev)
 {
-        /* No special bus mastering setup handling */
 }
 
 #define PCIBIOS_MIN_MEM         0
-#define PCIBIOS_MIN_IO          TILE_PCIE_LOWER_IO
+#define PCIBIOS_MIN_IO          0
 
 /*
  * This flag tells if the platform is TILEmpower that needs
  * special configuration for the PLX switch chip.
  */
-extern int blade_pci;
+extern int tile_plx_gen1;
+
+/* Use any cpu for PCI. */
+#define cpumask_of_pcibus(bus) cpu_online_mask
 
 /* implement the pci_ DMA API in terms of the generic device dma_ one */
 #include <asm-generic/pci-dma-compat.h>
@@ -122,7 +100,4 @@ extern int blade_pci;
 /* generic pci stuff */
 #include <asm-generic/pci.h>
 
-/* Use any cpu for PCI. */
-#define cpumask_of_pcibus(bus) cpu_online_mask
-
 #endif /* _ASM_TILE_PCI_H */
diff --git a/arch/tile/include/asm/processor.h b/arch/tile/include/asm/processor.h
index 1747ff3946b2..a9e7c8760334 100644
--- a/arch/tile/include/asm/processor.h
+++ b/arch/tile/include/asm/processor.h
@@ -292,8 +292,18 @@ extern int kstack_hash;
 /* Are we using huge pages in the TLB for kernel data? */
 extern int kdata_huge;
 
+/* Support standard Linux prefetching. */
+#define ARCH_HAS_PREFETCH
+#define prefetch(x) __builtin_prefetch(x)
 #define PREFETCH_STRIDE CHIP_L2_LINE_SIZE()
 
+/* Bring a value into the L1D, faulting the TLB if necessary. */
+#ifdef __tilegx__
+#define prefetch_L1(x) __insn_prefetch_l1_fault((void *)(x))
+#else
+#define prefetch_L1(x) __insn_prefetch_L1((void *)(x))
+#endif
+
 #else /* __ASSEMBLY__ */
 
 /* Do some slow action (e.g. read a slow SPR). */
diff --git a/arch/tile/include/hv/drv_xgbe_impl.h b/arch/tile/include/hv/drv_xgbe_impl.h
new file mode 100644
index 000000000000..3a73b2b44913
--- /dev/null
+++ b/arch/tile/include/hv/drv_xgbe_impl.h
@@ -0,0 +1,300 @@
+/*
+ * Copyright 2010 Tilera Corporation. All Rights Reserved.
+ *
+ *   This program is free software; you can redistribute it and/or
+ *   modify it under the terms of the GNU General Public License
+ *   as published by the Free Software Foundation, version 2.
+ *
+ *   This program is distributed in the hope that it will be useful, but
+ *   WITHOUT ANY WARRANTY; without even the implied warranty of
+ *   MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
+ *   NON INFRINGEMENT.  See the GNU General Public License for
+ *   more details.
+ */
+
+/**
+ * @file drivers/xgbe/impl.h
+ * Implementation details for the NetIO library.
+ */
+
+#ifndef __DRV_XGBE_IMPL_H__
+#define __DRV_XGBE_IMPL_H__
+
+#include <hv/netio_errors.h>
+#include <hv/netio_intf.h>
+#include <hv/drv_xgbe_intf.h>
+
+
+/** How many groups we have (log2). */
+#define LOG2_NUM_GROUPS (12)
+/** How many groups we have. */
+#define NUM_GROUPS (1 << LOG2_NUM_GROUPS)
+
+/** Number of output requests we'll buffer per tile. */
+#define EPP_REQS_PER_TILE (32)
+
+/** Words used in an eDMA command without checksum acceleration. */
+#define EDMA_WDS_NO_CSUM      8
+/** Words used in an eDMA command with checksum acceleration. */
+#define EDMA_WDS_CSUM        10
+/** Total available words in the eDMA command FIFO. */
+#define EDMA_WDS_TOTAL      128
+
+
+/*
+ * FIXME: These definitions are internal and should have underscores!
+ * NOTE: The actual numeric values here are intentional and allow us to
+ * optimize the concept "if small ... else if large ... else ...", by
+ * checking for the low bit being set, and then for non-zero.
+ * These are used as array indices, so they must have the values (0, 1, 2)
+ * in some order.
+ */
+#define SIZE_SMALL (1)       /**< Small packet queue. */
+#define SIZE_LARGE (2)       /**< Large packet queue. */
+#define SIZE_JUMBO (0)       /**< Jumbo packet queue. */
+
+/** The number of "SIZE_xxx" values. */
+#define NETIO_NUM_SIZES 3
+
+
+/*
+ * Default numbers of packets for IPP drivers.  These values are chosen
+ * such that CIPP1 will not overflow its L2 cache.
+ */
+
+/** The default number of small packets. */
+#define NETIO_DEFAULT_SMALL_PACKETS 2750
+/** The default number of large packets. */
+#define NETIO_DEFAULT_LARGE_PACKETS 2500
+/** The default number of jumbo packets. */
+#define NETIO_DEFAULT_JUMBO_PACKETS 250
+
+
+/** Log2 of the size of a memory arena. */
+#define NETIO_ARENA_SHIFT      24      /* 16 MB */
+/** Size of a memory arena. */
+#define NETIO_ARENA_SIZE       (1 << NETIO_ARENA_SHIFT)
+
+
+/** A queue of packets.
+ *
+ * This structure partially defines a queue of packets waiting to be
+ * processed.  The queue as a whole is written to by an interrupt handler and
+ * read by non-interrupt code; this data structure is what's touched by the
+ * interrupt handler.  The other part of the queue state, the read offset, is
+ * kept in user space, not in hypervisor space, so it is in a separate data
+ * structure.
+ *
+ * The read offset (__packet_receive_read in the user part of the queue
+ * structure) points to the next packet to be read. When the read offset is
+ * equal to the write offset, the queue is empty; therefore the queue must
+ * contain one more slot than the required maximum queue size.
+ *
+ * Here's an example of all 3 state variables and what they mean.  All
+ * pointers move left to right.
+ *
+ * @code
+ *   I   I   V   V   V   V   I   I   I   I
+ *   0   1   2   3   4   5   6   7   8   9  10
+ *           ^       ^       ^               ^
+ *           |               |               |
+ *           |               |               __last_packet_plus_one
+ *           |               __buffer_write
+ *           __packet_receive_read
+ * @endcode
+ *
+ * This queue has 10 slots, and thus can hold 9 packets (_last_packet_plus_one
+ * = 10).  The read pointer is at 2, and the write pointer is at 6; thus,
+ * there are valid, unread packets in slots 2, 3, 4, and 5.  The remaining
+ * slots are invalid (do not contain a packet).
+ */
+typedef struct {
+  /** Byte offset of the next notify packet to be written: zero for the first
+   *  packet on the queue, sizeof (netio_pkt_t) for the second packet on the
+   *  queue, etc. */
+  volatile uint32_t __packet_write;
+
+  /** Offset of the packet after the last valid packet (i.e., when any
+   *  pointer is incremented to this value, it wraps back to zero). */
+  uint32_t __last_packet_plus_one;
+}
+__netio_packet_queue_t;
+
+
+/** A queue of buffers.
+ *
+ * This structure partially defines a queue of empty buffers which have been
+ * obtained via requests to the IPP.  (The elements of the queue are packet
+ * handles, which are transformed into a full netio_pkt_t when the buffer is
+ * retrieved.)  The queue as a whole is written to by an interrupt handler and
+ * read by non-interrupt code; this data structure is what's touched by the
+ * interrupt handler.  The other parts of the queue state, the read offset and
+ * requested write offset, are kept in user space, not in hypervisor space, so
+ * they are in a separate data structure.
+ *
+ * The read offset (__buffer_read in the user part of the queue structure)
+ * points to the next buffer to be read. When the read offset is equal to the
+ * write offset, the queue is empty; therefore the queue must contain one more
+ * slot than the required maximum queue size.
+ *
+ * The requested write offset (__buffer_requested_write in the user part of
+ * the queue structure) points to the slot which will hold the next buffer we
+ * request from the IPP, once we get around to sending such a request.  When
+ * the requested write offset is equal to the write offset, no requests for
+ * new buffers are outstanding; when the requested write offset is one greater
+ * than the read offset, no more requests may be sent.
+ *
+ * Note that, unlike the packet_queue, the buffer_queue places incoming
+ * buffers at decreasing addresses.  This makes the check for "is it time to
+ * wrap the buffer pointer" cheaper in the assembly code which receives new
+ * buffers, and means that the value which defines the queue size,
+ * __last_buffer, is different than in the packet queue.  Also, the offset
+ * used in the packet_queue is already scaled by the size of a packet; here we
+ * use unscaled slot indices for the offsets.  (These differences are
+ * historical, and in the future it's possible that the packet_queue will look
+ * more like this queue.)
+ *
+ * @code
+ * Here's an example of all 4 state variables and what they mean.  Remember:
+ * all pointers move right to left.
+ *
+ *   V   V   V   I   I   R   R   V   V   V
+ *   0   1   2   3   4   5   6   7   8   9
+ *           ^       ^       ^           ^
+ *           |       |       |           |
+ *           |       |       |           __last_buffer
+ *           |       |       __buffer_write
+ *           |       __buffer_requested_write
+ *           __buffer_read
+ * @endcode
+ *
+ * This queue has 10 slots, and thus can hold 9 buffers (_last_buffer = 9).
+ * The read pointer is at 2, and the write pointer is at 6; thus, there are
+ * valid, unread buffers in slots 2, 1, 0, 9, 8, and 7.  The requested write
+ * pointer is at 4; thus, requests have been made to the IPP for buffers which
+ * will be placed in slots 6 and 5 when they arrive.  Finally, the remaining
+ * slots are invalid (do not contain a buffer).
+ */
+typedef struct
+{
+  /** Ordinal number of the next buffer to be written: 0 for the first slot in
+   *  the queue, 1 for the second slot in the queue, etc. */
+  volatile uint32_t __buffer_write;
+
+  /** Ordinal number of the last buffer (i.e., when any pointer is decremented
+   *  below zero, it is reloaded with this value). */
+  uint32_t __last_buffer;
+}
+__netio_buffer_queue_t;
+
+
+/**
+ * An object for providing Ethernet packets to a process.
+ */
+typedef struct __netio_queue_impl_t
+{
+  /** The queue of packets waiting to be received. */
+  __netio_packet_queue_t __packet_receive_queue;
+  /** The intr bit mask that IDs this device. */
+  unsigned int __intr_id;
+  /** Offset to queues of empty buffers, one per size. */
+  uint32_t __buffer_queue[NETIO_NUM_SIZES];
+  /** The address of the first EPP tile, or -1 if no EPP. */
+  /* ISSUE: Actually this is always "0" or "~0". */
+  uint32_t __epp_location;
+  /** The queue ID that this queue represents. */
+  unsigned int __queue_id;
+  /** Number of acknowledgements received. */
+  volatile uint32_t __acks_received;
+  /** Last completion number received for packet_sendv. */
+  volatile uint32_t __last_completion_rcv;
+  /** Number of packets allowed to be outstanding. */
+  uint32_t __max_outstanding;
+  /** First VA available for packets. */
+  void* __va_0;
+  /** First VA in second range available for packets. */
+  void* __va_1;
+  /** Padding to align the "__packets" field to the size of a netio_pkt_t. */
+  uint32_t __padding[3];
+  /** The packets themselves. */
+  netio_pkt_t __packets[0];
+}
+netio_queue_impl_t;
+
+
+/**
+ * An object for managing the user end of a NetIO queue.
+ */
+typedef struct __netio_queue_user_impl_t
+{
+  /** The next incoming packet to be read. */
+  uint32_t __packet_receive_read;
+  /** The next empty buffers to be read, one index per size. */
+  uint8_t __buffer_read[NETIO_NUM_SIZES];
+  /** Where the empty buffer we next request from the IPP will go, one index
+   * per size. */
+  uint8_t __buffer_requested_write[NETIO_NUM_SIZES];
+  /** PCIe interface flag. */
+  uint8_t __pcie;
+  /** Number of packets left to be received before we send a credit update. */
+  uint32_t __receive_credit_remaining;
+  /** Value placed in __receive_credit_remaining when it reaches zero. */
+  uint32_t __receive_credit_interval;
+  /** First fast I/O routine index. */
+  uint32_t __fastio_index;
+  /** Number of acknowledgements expected. */
+  uint32_t __acks_outstanding;
+  /** Last completion number requested. */
+  uint32_t __last_completion_req;
+  /** File descriptor for driver. */
+  int __fd;
+}
+netio_queue_user_impl_t;
+
+
+#define NETIO_GROUP_CHUNK_SIZE   64   /**< Max # groups in one IPP request */
+#define NETIO_BUCKET_CHUNK_SIZE  64   /**< Max # buckets in one IPP request */
+
+
+/** Internal structure used to convey packet send information to the
+ * hypervisor.  FIXME: Actually, it's not used for that anymore, but
+ * netio_packet_send() still uses it internally.
+ */
+typedef struct
+{
+  uint16_t flags;              /**< Packet flags (__NETIO_SEND_FLG_xxx) */
+  uint16_t transfer_size;      /**< Size of packet */
+  uint32_t va;                 /**< VA of start of packet */
+  __netio_pkt_handle_t handle; /**< Packet handle */
+  uint32_t csum0;              /**< First checksum word */
+  uint32_t csum1;              /**< Second checksum word */
+}
+__netio_send_cmd_t;
+
+
+/** Flags used in two contexts:
+ *  - As the "flags" member in the __netio_send_cmd_t, above; used only
+ *    for netio_pkt_send_{prepare,commit}.
+ *  - As part of the flags passed to the various send packet fast I/O calls.
+ */
+
+/** Need acknowledgement on this packet.  Note that some code in the
+ *  normal send_pkt fast I/O handler assumes that this is equal to 1. */
+#define __NETIO_SEND_FLG_ACK    0x1
+
+/** Do checksum on this packet.  (Only used with the __netio_send_cmd_t;
+ *  normal packet sends use a special fast I/O index to denote checksumming,
+ *  and multi-segment sends test the checksum descriptor.) */
+#define __NETIO_SEND_FLG_CSUM   0x2
+
+/** Get a completion on this packet.  Only used with multi-segment sends.  */
+#define __NETIO_SEND_FLG_COMPLETION 0x4
+
+/** Position of the number-of-extra-segments value in the flags word.
+    Only used with multi-segment sends. */
+#define __NETIO_SEND_FLG_XSEG_SHIFT 3
+
+/** Width of the number-of-extra-segments value in the flags word. */
+#define __NETIO_SEND_FLG_XSEG_WIDTH 2
+
+#endif /* __DRV_XGBE_IMPL_H__ */
diff --git a/arch/tile/include/hv/drv_xgbe_intf.h b/arch/tile/include/hv/drv_xgbe_intf.h
new file mode 100644
index 000000000000..146e47d5334b
--- /dev/null
+++ b/arch/tile/include/hv/drv_xgbe_intf.h
@@ -0,0 +1,615 @@
+/*
+ * Copyright 2010 Tilera Corporation. All Rights Reserved.
+ *
+ *   This program is free software; you can redistribute it and/or
+ *   modify it under the terms of the GNU General Public License
+ *   as published by the Free Software Foundation, version 2.
+ *
+ *   This program is distributed in the hope that it will be useful, but
+ *   WITHOUT ANY WARRANTY; without even the implied warranty of
+ *   MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
+ *   NON INFRINGEMENT.  See the GNU General Public License for
+ *   more details.
+ */
+
+/**
+ * @file drv_xgbe_intf.h
+ * Interface to the hypervisor XGBE driver.
+ */
+
+#ifndef __DRV_XGBE_INTF_H__
+#define __DRV_XGBE_INTF_H__
+
+/**
+ * An object for forwarding VAs and PAs to the hypervisor.
+ * @ingroup types
+ *
+ * This allows the supervisor to specify a number of areas of memory to
+ * store packet buffers.
+ */
+typedef struct
+{
+  /** The physical address of the memory. */
+  HV_PhysAddr pa;
+  /** Page table entry for the memory.  This is only used to derive the
+   *  memory's caching mode; the PA bits are ignored. */
+  HV_PTE pte;
+  /** The virtual address of the memory. */
+  HV_VirtAddr va;
+  /** Size (in bytes) of the memory area. */
+  int size;
+
+}
+netio_ipp_address_t;
+
+/** The various pread/pwrite offsets into the hypervisor-level driver.
+ * @ingroup types
+ */
+typedef enum
+{
+  /** Inform the Linux driver of the address of the NetIO arena memory.
+   *  This offset is actually only used to convey information from netio
+   *  to the Linux driver; it never makes it from there to the hypervisor.
+   *  Write-only; takes a uint32_t specifying the VA address. */
+  NETIO_FIXED_ADDR               = 0x5000000000000000ULL,
+
+  /** Inform the Linux driver of the size of the NetIO arena memory.
+   *  This offset is actually only used to convey information from netio
+   *  to the Linux driver; it never makes it from there to the hypervisor.
+   *  Write-only; takes a uint32_t specifying the VA size. */
+  NETIO_FIXED_SIZE               = 0x5100000000000000ULL,
+
+  /** Register current tile with IPP.  Write then read: write, takes a
+   *  netio_input_config_t, read returns a pointer to a netio_queue_impl_t. */
+  NETIO_IPP_INPUT_REGISTER_OFF   = 0x6000000000000000ULL,
+
+  /** Unregister current tile from IPP.  Write-only, takes a dummy argument. */
+  NETIO_IPP_INPUT_UNREGISTER_OFF = 0x6100000000000000ULL,
+
+  /** Start packets flowing.  Write-only, takes a dummy argument. */
+  NETIO_IPP_INPUT_INIT_OFF       = 0x6200000000000000ULL,
+
+  /** Stop packets flowing.  Write-only, takes a dummy argument. */
+  NETIO_IPP_INPUT_UNINIT_OFF     = 0x6300000000000000ULL,
+
+  /** Configure group (typically we group on VLAN).  Write-only: takes an
+   *  array of netio_group_t's, low 24 bits of the offset is the base group
+   *  number times the size of a netio_group_t. */
+  NETIO_IPP_INPUT_GROUP_CFG_OFF  = 0x6400000000000000ULL,
+
+  /** Configure bucket.  Write-only: takes an array of netio_bucket_t's, low
+   *  24 bits of the offset is the base bucket number times the size of a
+   *  netio_bucket_t. */
+  NETIO_IPP_INPUT_BUCKET_CFG_OFF = 0x6500000000000000ULL,
+
+  /** Get/set a parameter.  Read or write: read or write data is the parameter
+   *  value, low 32 bits of the offset is a __netio_getset_offset_t. */
+  NETIO_IPP_PARAM_OFF            = 0x6600000000000000ULL,
+
+  /** Get fast I/O index.  Read-only; returns a 4-byte base index value. */
+  NETIO_IPP_GET_FASTIO_OFF       = 0x6700000000000000ULL,
+
+  /** Configure hijack IP address.  Packets with this IPv4 dest address
+   *  go to bucket NETIO_NUM_BUCKETS - 1.  Write-only: takes an IP address
+   *  in some standard form.  FIXME: Define the form! */
+  NETIO_IPP_INPUT_HIJACK_CFG_OFF  = 0x6800000000000000ULL,
+
+  /**
+   * Offsets beyond this point are reserved for the supervisor (although that
+   * enforcement must be done by the supervisor driver itself).
+   */
+  NETIO_IPP_USER_MAX_OFF         = 0x6FFFFFFFFFFFFFFFULL,
+
+  /** Register I/O memory.  Write-only, takes a netio_ipp_address_t. */
+  NETIO_IPP_IOMEM_REGISTER_OFF   = 0x7000000000000000ULL,
+
+  /** Unregister I/O memory.  Write-only, takes a netio_ipp_address_t. */
+  NETIO_IPP_IOMEM_UNREGISTER_OFF = 0x7100000000000000ULL,
+
+  /* Offsets greater than 0x7FFFFFFF can't be used directly from Linux
+   * userspace code due to limitations in the pread/pwrite syscalls. */
+
+  /** Drain LIPP buffers. */
+  NETIO_IPP_DRAIN_OFF              = 0xFA00000000000000ULL,
+
+  /** Supply a netio_ipp_address_t to be used as shared memory for the
+   *  LEPP command queue. */
+  NETIO_EPP_SHM_OFF              = 0xFB00000000000000ULL,
+
+  /* 0xFC... is currently unused. */
+
+  /** Stop IPP/EPP tiles.  Write-only, takes a dummy argument.  */
+  NETIO_IPP_STOP_SHIM_OFF        = 0xFD00000000000000ULL,
+
+  /** Start IPP/EPP tiles.  Write-only, takes a dummy argument.  */
+  NETIO_IPP_START_SHIM_OFF       = 0xFE00000000000000ULL,
+
+  /** Supply packet arena.  Write-only, takes an array of
+    * netio_ipp_address_t values. */
+  NETIO_IPP_ADDRESS_OFF          = 0xFF00000000000000ULL,
+} netio_hv_offset_t;
+
+/** Extract the base offset from an offset */
+#define NETIO_BASE_OFFSET(off)    ((off) & 0xFF00000000000000ULL)
+/** Extract the local offset from an offset */
+#define NETIO_LOCAL_OFFSET(off)   ((off) & 0x00FFFFFFFFFFFFFFULL)
+
+
+/**
+ * Get/set offset.
+ */
+typedef union
+{
+  struct
+  {
+    uint64_t addr:48;        /**< Class-specific address */
+    unsigned int class:8;    /**< Class (e.g., NETIO_PARAM) */
+    unsigned int opcode:8;   /**< High 8 bits of NETIO_IPP_PARAM_OFF */
+  }
+  bits;                      /**< Bitfields */
+  uint64_t word;             /**< Aggregated value to use as the offset */
+}
+__netio_getset_offset_t;
+
+/**
+ * Fast I/O index offsets (must be contiguous).
+ */
+typedef enum
+{
+  NETIO_FASTIO_ALLOCATE         = 0, /**< Get empty packet buffer */
+  NETIO_FASTIO_FREE_BUFFER      = 1, /**< Give buffer back to IPP */
+  NETIO_FASTIO_RETURN_CREDITS   = 2, /**< Give credits to IPP */
+  NETIO_FASTIO_SEND_PKT_NOCK    = 3, /**< Send a packet, no checksum */
+  NETIO_FASTIO_SEND_PKT_CK      = 4, /**< Send a packet, with checksum */
+  NETIO_FASTIO_SEND_PKT_VEC     = 5, /**< Send a vector of packets */
+  NETIO_FASTIO_SENDV_PKT        = 6, /**< Sendv one packet */
+  NETIO_FASTIO_NUM_INDEX        = 7, /**< Total number of fast I/O indices */
+} netio_fastio_index_t;
+
+/** 3-word return type for Fast I/O call. */
+typedef struct
+{
+  int err;            /**< Error code. */
+  uint32_t val0;      /**< Value.  Meaning depends upon the specific call. */
+  uint32_t val1;      /**< Value.  Meaning depends upon the specific call. */
+} netio_fastio_rv3_t;
+
+/** 0-argument fast I/O call */
+int __netio_fastio0(uint32_t fastio_index);
+/** 1-argument fast I/O call */
+int __netio_fastio1(uint32_t fastio_index, uint32_t arg0);
+/** 3-argument fast I/O call, 2-word return value */
+netio_fastio_rv3_t __netio_fastio3_rv3(uint32_t fastio_index, uint32_t arg0,
+                                       uint32_t arg1, uint32_t arg2);
+/** 4-argument fast I/O call */
+int __netio_fastio4(uint32_t fastio_index, uint32_t arg0, uint32_t arg1,
+                    uint32_t arg2, uint32_t arg3);
+/** 6-argument fast I/O call */
+int __netio_fastio6(uint32_t fastio_index, uint32_t arg0, uint32_t arg1,
+                    uint32_t arg2, uint32_t arg3, uint32_t arg4, uint32_t arg5);
+/** 9-argument fast I/O call */
+int __netio_fastio9(uint32_t fastio_index, uint32_t arg0, uint32_t arg1,
+                    uint32_t arg2, uint32_t arg3, uint32_t arg4, uint32_t arg5,
+                    uint32_t arg6, uint32_t arg7, uint32_t arg8);
+
+/** Allocate an empty packet.
+ * @param fastio_index Fast I/O index.
+ * @param size Size of the packet to allocate.
+ */
+#define __netio_fastio_allocate(fastio_index, size) \
+  __netio_fastio1((fastio_index) + NETIO_FASTIO_ALLOCATE, size)
+
+/** Free a buffer.
+ * @param fastio_index Fast I/O index.
+ * @param handle Handle for the packet to free.
+ */
+#define __netio_fastio_free_buffer(fastio_index, handle) \
+  __netio_fastio1((fastio_index) + NETIO_FASTIO_FREE_BUFFER, handle)
+
+/** Increment our receive credits.
+ * @param fastio_index Fast I/O index.
+ * @param credits Number of credits to add.
+ */
+#define __netio_fastio_return_credits(fastio_index, credits) \
+  __netio_fastio1((fastio_index) + NETIO_FASTIO_RETURN_CREDITS, credits)
+
+/** Send packet, no checksum.
+ * @param fastio_index Fast I/O index.
+ * @param ackflag Nonzero if we want an ack.
+ * @param size Size of the packet.
+ * @param va Virtual address of start of packet.
+ * @param handle Packet handle.
+ */
+#define __netio_fastio_send_pkt_nock(fastio_index, ackflag, size, va, handle) \
+  __netio_fastio4((fastio_index) + NETIO_FASTIO_SEND_PKT_NOCK, ackflag, \
+                  size, va, handle)
+
+/** Send packet, calculate checksum.
+ * @param fastio_index Fast I/O index.
+ * @param ackflag Nonzero if we want an ack.
+ * @param size Size of the packet.
+ * @param va Virtual address of start of packet.
+ * @param handle Packet handle.
+ * @param csum0 Shim checksum header.
+ * @param csum1 Checksum seed.
+ */
+#define __netio_fastio_send_pkt_ck(fastio_index, ackflag, size, va, handle, \
+                                   csum0, csum1) \
+  __netio_fastio6((fastio_index) + NETIO_FASTIO_SEND_PKT_CK, ackflag, \
+                  size, va, handle, csum0, csum1)
+
+
+/** Format for the "csum0" argument to the __netio_fastio_send routines
+ * and LEPP.  Note that this is currently exactly identical to the
+ * ShimProtocolOffloadHeader.
+ */
+typedef union
+{
+  struct
+  {
+    unsigned int start_byte:7;       /**< The first byte to be checksummed */
+    unsigned int count:14;           /**< Number of bytes to be checksummed. */
+    unsigned int destination_byte:7; /**< The byte to write the checksum to. */
+    unsigned int reserved:4;         /**< Reserved. */
+  } bits;                            /**< Decomposed method of access. */
+  unsigned int word;                 /**< To send out the IDN. */
+} __netio_checksum_header_t;
+
+
+/** Sendv packet with 1 or 2 segments.
+ * @param fastio_index Fast I/O index.
+ * @param flags Ack/csum/notify flags in low 3 bits; number of segments minus
+ *        1 in next 2 bits; expected checksum in high 16 bits.
+ * @param confno Confirmation number to request, if notify flag set.
+ * @param csum0 Checksum descriptor; if zero, no checksum.
+ * @param va_F Virtual address of first segment.
+ * @param va_L Virtual address of last segment, if 2 segments.
+ * @param len_F_L Length of first segment in low 16 bits; length of last
+ *        segment, if 2 segments, in high 16 bits.
+ */
+#define __netio_fastio_sendv_pkt_1_2(fastio_index, flags, confno, csum0, \
+                                     va_F, va_L, len_F_L) \
+  __netio_fastio6((fastio_index) + NETIO_FASTIO_SENDV_PKT, flags, confno, \
+                  csum0, va_F, va_L, len_F_L)
+
+/** Send packet on PCIe interface.
+ * @param fastio_index Fast I/O index.
+ * @param flags Ack/csum/notify flags in low 3 bits.
+ * @param confno Confirmation number to request, if notify flag set.
+ * @param csum0 Checksum descriptor; Hard wired 0, not needed for PCIe.
+ * @param va_F Virtual address of the packet buffer.
+ * @param va_L Virtual address of last segment, if 2 segments. Hard wired 0.
+ * @param len_F_L Length of the packet buffer in low 16 bits.
+ */
+#define __netio_fastio_send_pcie_pkt(fastio_index, flags, confno, csum0, \
+                                     va_F, va_L, len_F_L) \
+  __netio_fastio6((fastio_index) + PCIE_FASTIO_SENDV_PKT, flags, confno, \
+                  csum0, va_F, va_L, len_F_L)
+
+/** Sendv packet with 3 or 4 segments.
+ * @param fastio_index Fast I/O index.
+ * @param flags Ack/csum/notify flags in low 3 bits; number of segments minus
+ *        1 in next 2 bits; expected checksum in high 16 bits.
+ * @param confno Confirmation number to request, if notify flag set.
+ * @param csum0 Checksum descriptor; if zero, no checksum.
+ * @param va_F Virtual address of first segment.
+ * @param va_L Virtual address of last segment (third segment if 3 segments,
+ *        fourth segment if 4 segments).
+ * @param len_F_L Length of first segment in low 16 bits; length of last
+ *        segment in high 16 bits.
+ * @param va_M0 Virtual address of "middle 0" segment; this segment is sent
+ *        second when there are three segments, and third if there are four.
+ * @param va_M1 Virtual address of "middle 1" segment; this segment is sent
+ *        second when there are four segments.
+ * @param len_M0_M1 Length of middle 0 segment in low 16 bits; length of middle
+ *        1 segment, if 4 segments, in high 16 bits.
+ */
+#define __netio_fastio_sendv_pkt_3_4(fastio_index, flags, confno, csum0, va_F, \
+                                     va_L, len_F_L, va_M0, va_M1, len_M0_M1) \
+  __netio_fastio9((fastio_index) + NETIO_FASTIO_SENDV_PKT, flags, confno, \
+                  csum0, va_F, va_L, len_F_L, va_M0, va_M1, len_M0_M1)
+
+/** Send vector of packets.
+ * @param fastio_index Fast I/O index.
+ * @param seqno Number of packets transmitted so far on this interface;
+ *        used to decide which packets should be acknowledged.
+ * @param nentries Number of entries in vector.
+ * @param va Virtual address of start of vector entry array.
+ * @return 3-word netio_fastio_rv3_t structure.  The structure's err member
+ *         is an error code, or zero if no error.  The val0 member is the
+ *         updated value of seqno; it has been incremented by 1 for each
+ *         packet sent.  That increment may be less than nentries if an
+ *         error occured, or if some of the entries in the vector contain
+ *         handles equal to NETIO_PKT_HANDLE_NONE.  The val1 member is the
+ *         updated value of nentries; it has been decremented by 1 for each
+ *         vector entry processed.  Again, that decrement may be less than
+ *         nentries (leaving the returned value positive) if an error
+ *         occurred.
+ */
+#define __netio_fastio_send_pkt_vec(fastio_index, seqno, nentries, va) \
+  __netio_fastio3_rv3((fastio_index) + NETIO_FASTIO_SEND_PKT_VEC, seqno, \
+                      nentries, va)
+
+
+/** An egress DMA command for LEPP. */
+typedef struct
+{
+  /** Is this a TSO transfer?
+   *
+   * NOTE: This field is always 0, to distinguish it from
+   * lepp_tso_cmd_t.  It must come first!
+   */
+  uint8_t tso               : 1;
+
+  /** Unused padding bits. */
+  uint8_t _unused           : 3;
+
+  /** Should this packet be sent directly from caches instead of DRAM,
+   * using hash-for-home to locate the packet data?
+   */
+  uint8_t hash_for_home     : 1;
+
+  /** Should we compute a checksum? */
+  uint8_t compute_checksum  : 1;
+
+  /** Is this the final buffer for this packet?
+   *
+   * A single packet can be split over several input buffers (a "gather"
+   * operation).  This flag indicates that this is the last buffer
+   * in a packet.
+   */
+  uint8_t end_of_packet     : 1;
+
+  /** Should LEPP advance 'comp_busy' when this DMA is fully finished? */
+  uint8_t send_completion   : 1;
+
+  /** High bits of Client Physical Address of the start of the buffer
+   *  to be egressed.
+   *
+   *  NOTE: Only 6 bits are actually needed here, as CPAs are
+   *  currently 38 bits.  So two bits could be scavenged from this.
+   */
+  uint8_t cpa_hi;
+
+  /** The number of bytes to be egressed. */
+  uint16_t length;
+
+  /** Low 32 bits of Client Physical Address of the start of the buffer
+   *  to be egressed.
+   */
+  uint32_t cpa_lo;
+
+  /** Checksum information (only used if 'compute_checksum'). */
+  __netio_checksum_header_t checksum_data;
+
+} lepp_cmd_t;
+
+
+/** A chunk of physical memory for a TSO egress. */
+typedef struct
+{
+  /** The low bits of the CPA. */
+  uint32_t cpa_lo;
+  /** The high bits of the CPA. */
+  uint16_t cpa_hi               : 15;
+  /** Should this packet be sent directly from caches instead of DRAM,
+   *  using hash-for-home to locate the packet data?
+   */
+  uint16_t hash_for_home        : 1;
+  /** The length in bytes. */
+  uint16_t length;
+} lepp_frag_t;
+
+
+/** An LEPP command that handles TSO. */
+typedef struct
+{
+  /** Is this a TSO transfer?
+   *
+   *  NOTE: This field is always 1, to distinguish it from
+   *  lepp_cmd_t.  It must come first!
+   */
+  uint8_t tso             : 1;
+
+  /** Unused padding bits. */
+  uint8_t _unused         : 7;
+
+  /** Size of the header[] array in bytes.  It must be in the range
+   *  [40, 127], which are the smallest header for a TCP packet over
+   *  Ethernet and the maximum possible prepend size supported by
+   *  hardware, respectively.  Note that the array storage must be
+   *  padded out to a multiple of four bytes so that the following
+   *  LEPP command is aligned properly.
+   */
+  uint8_t header_size;
+
+  /** Byte offset of the IP header in header[]. */
+  uint8_t ip_offset;
+
+  /** Byte offset of the TCP header in header[]. */
+  uint8_t tcp_offset;
+
+  /** The number of bytes to use for the payload of each packet,
+   *  except of course the last one, which may not have enough bytes.
+   *  This means that each Ethernet packet except the last will have a
+   *  size of header_size + payload_size.
+   */
+  uint16_t payload_size;
+
+  /** The length of the 'frags' array that follows this struct. */
+  uint16_t num_frags;
+
+  /** The actual frags. */
+  lepp_frag_t frags[0 /* Variable-sized; num_frags entries. */];
+
+  /*
+   * The packet header template logically follows frags[],
+   * but you can't declare that in C.
+   *
+   * uint32_t header[header_size_in_words_rounded_up];
+   */
+
+} lepp_tso_cmd_t;
+
+
+/** An LEPP completion ring entry. */
+typedef void* lepp_comp_t;
+
+
+/** Maximum number of frags for one TSO command.  This is adapted from
+ *  linux's "MAX_SKB_FRAGS", and presumably over-estimates by one, for
+ *  our page size of exactly 65536.  We add one for a "body" fragment.
+ */
+#define LEPP_MAX_FRAGS (65536 / HV_PAGE_SIZE_SMALL + 2 + 1)
+
+/** Total number of bytes needed for an lepp_tso_cmd_t. */
+#define LEPP_TSO_CMD_SIZE(num_frags, header_size) \
+  (sizeof(lepp_tso_cmd_t) + \
+   (num_frags) * sizeof(lepp_frag_t) + \
+   (((header_size) + 3) & -4))
+
+/** The size of the lepp "cmd" queue. */
+#define LEPP_CMD_QUEUE_BYTES \
+ (((CHIP_L2_CACHE_SIZE() - 2 * CHIP_L2_LINE_SIZE()) / \
+  (sizeof(lepp_cmd_t) + sizeof(lepp_comp_t))) * sizeof(lepp_cmd_t))
+
+/** The largest possible command that can go in lepp_queue_t::cmds[]. */
+#define LEPP_MAX_CMD_SIZE LEPP_TSO_CMD_SIZE(LEPP_MAX_FRAGS, 128)
+
+/** The largest possible value of lepp_queue_t::cmd_{head, tail} (inclusive).
+ */
+#define LEPP_CMD_LIMIT \
+  (LEPP_CMD_QUEUE_BYTES - LEPP_MAX_CMD_SIZE)
+
+/** The maximum number of completions in an LEPP queue. */
+#define LEPP_COMP_QUEUE_SIZE \
+  ((LEPP_CMD_LIMIT + sizeof(lepp_cmd_t) - 1) / sizeof(lepp_cmd_t))
+
+/** Increment an index modulo the queue size. */
+#define LEPP_QINC(var) \
+  (var = __insn_mnz(var - (LEPP_COMP_QUEUE_SIZE - 1), var + 1))
+
+/** A queue used to convey egress commands from the client to LEPP. */
+typedef struct
+{
+  /** Index of first completion not yet processed by user code.
+   *  If this is equal to comp_busy, there are no such completions.
+   *
+   *  NOTE: This is only read/written by the user.
+   */
+  unsigned int comp_head;
+
+  /** Index of first completion record not yet completed.
+   *  If this is equal to comp_tail, there are no such completions.
+   *  This index gets advanced (modulo LEPP_QUEUE_SIZE) whenever
+   *  a command with the 'completion' bit set is finished.
+   *
+   *  NOTE: This is only written by LEPP, only read by the user.
+   */
+  volatile unsigned int comp_busy;
+
+  /** Index of the first empty slot in the completion ring.
+   *  Entries from this up to but not including comp_head (in ring order)
+   *  can be filled in with completion data.
+   *
+   *  NOTE: This is only read/written by the user.
+   */
+  unsigned int comp_tail;
+
+  /** Byte index of first command enqueued for LEPP but not yet processed.
+   *
+   *  This is always divisible by sizeof(void*) and always <= LEPP_CMD_LIMIT.
+   *
+   *  NOTE: LEPP advances this counter as soon as it no longer needs
+   *  the cmds[] storage for this entry, but the transfer is not actually
+   *  complete (i.e. the buffer pointed to by the command is no longer
+   *  needed) until comp_busy advances.
+   *
+   *  If this is equal to cmd_tail, the ring is empty.
+   *
+   *  NOTE: This is only written by LEPP, only read by the user.
+   */
+  volatile unsigned int cmd_head;
+
+  /** Byte index of first empty slot in the command ring.  This field can
+   *  be incremented up to but not equal to cmd_head (because that would
+   *  mean the ring is empty).
+   *
+   *  This is always divisible by sizeof(void*) and always <= LEPP_CMD_LIMIT.
+   *
+   *  NOTE: This is read/written by the user, only read by LEPP.
+   */
+  volatile unsigned int cmd_tail;
+
+  /** A ring of variable-sized egress DMA commands.
+   *
+   *  NOTE: Only written by the user, only read by LEPP.
+   */
+  char cmds[LEPP_CMD_QUEUE_BYTES]
+    __attribute__((aligned(CHIP_L2_LINE_SIZE())));
+
+  /** A ring of user completion data.
+   *  NOTE: Only read/written by the user.
+   */
+  lepp_comp_t comps[LEPP_COMP_QUEUE_SIZE]
+    __attribute__((aligned(CHIP_L2_LINE_SIZE())));
+} lepp_queue_t;
+
+
+/** An internal helper function for determining the number of entries
+ *  available in a ring buffer, given that there is one sentinel.
+ */
+static inline unsigned int
+_lepp_num_free_slots(unsigned int head, unsigned int tail)
+{
+  /*
+   * One entry is reserved for use as a sentinel, to distinguish
+   * "empty" from "full".  So we compute
+   * (head - tail - 1) % LEPP_QUEUE_SIZE, but without using a slow % operation.
+   */
+  return (head - tail - 1) + ((head <= tail) ? LEPP_COMP_QUEUE_SIZE : 0);
+}
+
+
+/** Returns how many new comp entries can be enqueued. */
+static inline unsigned int
+lepp_num_free_comp_slots(const lepp_queue_t* q)
+{
+  return _lepp_num_free_slots(q->comp_head, q->comp_tail);
+}
+
+static inline int
+lepp_qsub(int v1, int v2)
+{
+  int delta = v1 - v2;
+  return delta + ((delta >> 31) & LEPP_COMP_QUEUE_SIZE);
+}
+
+
+/** FIXME: Check this from linux, via a new "pwrite()" call. */
+#define LIPP_VERSION 1
+
+
+/** We use exactly two bytes of alignment padding. */
+#define LIPP_PACKET_PADDING 2
+
+/** The minimum size of a "small" buffer (including the padding). */
+#define LIPP_SMALL_PACKET_SIZE 128
+
+/*
+ * NOTE: The following two values should total to less than around
+ * 13582, to keep the total size used for "lipp_state_t" below 64K.
+ */
+
+/** The maximum number of "small" buffers.
+ *  This is enough for 53 network cpus with 128 credits.  Note that
+ *  if these are exhausted, we will fall back to using large buffers.
+ */
+#define LIPP_SMALL_BUFFERS 6785
+
+/** The maximum number of "large" buffers.
+ *  This is enough for 53 network cpus with 128 credits.
+ */
+#define LIPP_LARGE_BUFFERS 6785
+
+#endif /* __DRV_XGBE_INTF_H__ */
diff --git a/arch/tile/include/hv/netio_errors.h b/arch/tile/include/hv/netio_errors.h
new file mode 100644
index 000000000000..e1591bff61b5
--- /dev/null
+++ b/arch/tile/include/hv/netio_errors.h
@@ -0,0 +1,122 @@
+/*
+ * Copyright 2010 Tilera Corporation. All Rights Reserved.
+ *
+ *   This program is free software; you can redistribute it and/or
+ *   modify it under the terms of the GNU General Public License
+ *   as published by the Free Software Foundation, version 2.
+ *
+ *   This program is distributed in the hope that it will be useful, but
+ *   WITHOUT ANY WARRANTY; without even the implied warranty of
+ *   MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
+ *   NON INFRINGEMENT.  See the GNU General Public License for
+ *   more details.
+ */
+
+/**
+ * Error codes returned from NetIO routines.
+ */
+
+#ifndef __NETIO_ERRORS_H__
+#define __NETIO_ERRORS_H__
+
+/**
+ * @addtogroup error
+ *
+ * @brief The error codes returned by NetIO functions.
+ *
+ * NetIO functions return 0 (defined as ::NETIO_NO_ERROR) on success, and
+ * a negative value if an error occurs.
+ *
+ * In cases where a NetIO function failed due to a error reported by
+ * system libraries, the error code will be the negation of the
+ * system errno at the time of failure.  The @ref netio_strerror()
+ * function will deliver error strings for both NetIO and system error
+ * codes.
+ *
+ * @{
+ */
+
+/** The set of all NetIO errors. */
+typedef enum
+{
+  /** Operation successfully completed. */
+  NETIO_NO_ERROR        = 0,
+
+  /** A packet was successfully retrieved from an input queue. */
+  NETIO_PKT             = 0,
+
+  /** Largest NetIO error number. */
+  NETIO_ERR_MAX         = -701,
+
+  /** The tile is not registered with the IPP. */
+  NETIO_NOT_REGISTERED  = -701,
+
+  /** No packet was available to retrieve from the input queue. */
+  NETIO_NOPKT           = -702,
+
+  /** The requested function is not implemented. */
+  NETIO_NOT_IMPLEMENTED = -703,
+
+  /** On a registration operation, the target queue already has the maximum
+   *  number of tiles registered for it, and no more may be added.  On a
+   *  packet send operation, the output queue is full and nothing more can
+   *  be queued until some of the queued packets are actually transmitted. */
+  NETIO_QUEUE_FULL      = -704,
+
+  /** The calling process or thread is not bound to exactly one CPU. */
+  NETIO_BAD_AFFINITY    = -705,
+
+  /** Cannot allocate memory on requested controllers. */
+  NETIO_CANNOT_HOME     = -706,
+
+  /** On a registration operation, the IPP specified is not configured
+   *  to support the options requested; for instance, the application
+   *  wants a specific type of tagged headers which the configured IPP
+   *  doesn't support.  Or, the supplied configuration information is
+   *  not self-consistent, or is out of range; for instance, specifying
+   *  both NETIO_RECV and NETIO_NO_RECV, or asking for more than
+   *  NETIO_MAX_SEND_BUFFERS to be preallocated.  On a VLAN or bucket
+   *  configure operation, the number of items, or the base item, was
+   *  out of range.
+   */
+  NETIO_BAD_CONFIG      = -707,
+
+  /** Too many tiles have registered to transmit packets. */
+  NETIO_TOOMANY_XMIT    = -708,
+
+  /** Packet transmission was attempted on a queue which was registered
+      with transmit disabled. */
+  NETIO_UNREG_XMIT      = -709,
+
+  /** This tile is already registered with the IPP. */
+  NETIO_ALREADY_REGISTERED = -710,
+
+  /** The Ethernet link is down. The application should try again later. */
+  NETIO_LINK_DOWN       = -711,
+
+  /** An invalid memory buffer has been specified.  This may be an unmapped
+   * virtual address, or one which does not meet alignment requirements.
+   * For netio_input_register(), this error may be returned when multiple
+   * processes specify different memory regions to be used for NetIO
+   * buffers.  That can happen if these processes specify explicit memory
+   * regions with the ::NETIO_FIXED_BUFFER_VA flag, or if tmc_cmem_init()
+   * has not been called by a common ancestor of the processes.
+   */
+  NETIO_FAULT           = -712,
+
+  /** Cannot combine user-managed shared memory and cache coherence. */
+  NETIO_BAD_CACHE_CONFIG = -713,
+
+  /** Smallest NetIO error number. */
+  NETIO_ERR_MIN         = -713,
+
+#ifndef __DOXYGEN__
+  /** Used internally to mean that no response is needed; never returned to
+   *  an application. */
+  NETIO_NO_RESPONSE     = 1
+#endif
+} netio_error_t;
+
+/** @} */
+
+#endif /* __NETIO_ERRORS_H__ */
diff --git a/arch/tile/include/hv/netio_intf.h b/arch/tile/include/hv/netio_intf.h
new file mode 100644
index 000000000000..8d20972aba2c
--- /dev/null
+++ b/arch/tile/include/hv/netio_intf.h
@@ -0,0 +1,2975 @@
+/*
+ * Copyright 2010 Tilera Corporation. All Rights Reserved.
+ *
+ *   This program is free software; you can redistribute it and/or
+ *   modify it under the terms of the GNU General Public License
+ *   as published by the Free Software Foundation, version 2.
+ *
+ *   This program is distributed in the hope that it will be useful, but
+ *   WITHOUT ANY WARRANTY; without even the implied warranty of
+ *   MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
+ *   NON INFRINGEMENT.  See the GNU General Public License for
+ *   more details.
+ */
+
+/**
+ * NetIO interface structures and macros.
+ */
+
+#ifndef __NETIO_INTF_H__
+#define __NETIO_INTF_H__
+
+#include <hv/netio_errors.h>
+
+#ifdef __KERNEL__
+#include <linux/types.h>
+#else
+#include <stdint.h>
+#endif
+
+#if !defined(__HV__) && !defined(__BOGUX__) && !defined(__KERNEL__)
+#include <assert.h>
+#define netio_assert assert  /**< Enable assertions from macros */
+#else
+#define netio_assert(...) ((void)(0))  /**< Disable assertions from macros */
+#endif
+
+/*
+ * If none of these symbols are defined, we're building libnetio in an
+ * environment where we have pthreads, so we'll enable locking.
+ */
+#if !defined(__HV__) && !defined(__BOGUX__) && !defined(__KERNEL__) && \
+    !defined(__NEWLIB__)
+#define _NETIO_PTHREAD       /**< Include a mutex in netio_queue_t below */
+
+/*
+ * If NETIO_UNLOCKED is defined, we don't do use per-cpu locks on
+ * per-packet NetIO operations.  We still do pthread locking on things
+ * like netio_input_register, though.  This is used for building
+ * libnetio_unlocked.
+ */
+#ifndef NETIO_UNLOCKED
+
+/* Avoid PLT overhead by using our own inlined per-cpu lock. */
+#include <sched.h>
+typedef int _netio_percpu_mutex_t;
+
+static __inline int
+_netio_percpu_mutex_init(_netio_percpu_mutex_t* lock)
+{
+  *lock = 0;
+  return 0;
+}
+
+static __inline int
+_netio_percpu_mutex_lock(_netio_percpu_mutex_t* lock)
+{
+  while (__builtin_expect(__insn_tns(lock), 0))
+    sched_yield();
+  return 0;
+}
+
+static __inline int
+_netio_percpu_mutex_unlock(_netio_percpu_mutex_t* lock)
+{
+  *lock = 0;
+  return 0;
+}
+
+#else /* NETIO_UNLOCKED */
+
+/* Don't do any locking for per-packet NetIO operations. */
+typedef int _netio_percpu_mutex_t;
+#define _netio_percpu_mutex_init(L)
+#define _netio_percpu_mutex_lock(L)
+#define _netio_percpu_mutex_unlock(L)
+
+#endif /* NETIO_UNLOCKED */
+#endif /* !__HV__, !__BOGUX, !__KERNEL__, !__NEWLIB__ */
+
+/** How many tiles can register for a given queue.
+ *  @ingroup setup */
+#define NETIO_MAX_TILES_PER_QUEUE  64
+
+
+/** Largest permissible queue identifier.
+ *  @ingroup setup  */
+#define NETIO_MAX_QUEUE_ID        255
+
+
+#ifndef __DOXYGEN__
+
+/* Metadata packet checksum/ethertype flags. */
+
+/** The L4 checksum has not been calculated. */
+#define _NETIO_PKT_NO_L4_CSUM_SHIFT           0
+#define _NETIO_PKT_NO_L4_CSUM_RMASK           1
+#define _NETIO_PKT_NO_L4_CSUM_MASK \
+         (_NETIO_PKT_NO_L4_CSUM_RMASK << _NETIO_PKT_NO_L4_CSUM_SHIFT)
+
+/** The L3 checksum has not been calculated. */
+#define _NETIO_PKT_NO_L3_CSUM_SHIFT           1
+#define _NETIO_PKT_NO_L3_CSUM_RMASK           1
+#define _NETIO_PKT_NO_L3_CSUM_MASK \
+         (_NETIO_PKT_NO_L3_CSUM_RMASK << _NETIO_PKT_NO_L3_CSUM_SHIFT)
+
+/** The L3 checksum is incorrect (or perhaps has not been calculated). */
+#define _NETIO_PKT_BAD_L3_CSUM_SHIFT          2
+#define _NETIO_PKT_BAD_L3_CSUM_RMASK          1
+#define _NETIO_PKT_BAD_L3_CSUM_MASK \
+         (_NETIO_PKT_BAD_L3_CSUM_RMASK << _NETIO_PKT_BAD_L3_CSUM_SHIFT)
+
+/** The Ethernet packet type is unrecognized. */
+#define _NETIO_PKT_TYPE_UNRECOGNIZED_SHIFT    3
+#define _NETIO_PKT_TYPE_UNRECOGNIZED_RMASK    1
+#define _NETIO_PKT_TYPE_UNRECOGNIZED_MASK \
+         (_NETIO_PKT_TYPE_UNRECOGNIZED_RMASK << \
+          _NETIO_PKT_TYPE_UNRECOGNIZED_SHIFT)
+
+/* Metadata packet type flags. */
+
+/** Where the packet type bits are; this field is the index into
+ *  _netio_pkt_info. */
+#define _NETIO_PKT_TYPE_SHIFT        4
+#define _NETIO_PKT_TYPE_RMASK        0x3F
+
+/** How many VLAN tags the packet has, and, if we have two, which one we
+ *  actually grouped on.  A VLAN within a proprietary (Marvell or Broadcom)
+ *  tag is counted here. */
+#define _NETIO_PKT_VLAN_SHIFT        4
+#define _NETIO_PKT_VLAN_RMASK        0x3
+#define _NETIO_PKT_VLAN_MASK \
+         (_NETIO_PKT_VLAN_RMASK << _NETIO_PKT_VLAN_SHIFT)
+#define _NETIO_PKT_VLAN_NONE         0   /* No VLAN tag. */
+#define _NETIO_PKT_VLAN_ONE          1   /* One VLAN tag. */
+#define _NETIO_PKT_VLAN_TWO_OUTER    2   /* Two VLAN tags, outer one used. */
+#define _NETIO_PKT_VLAN_TWO_INNER    3   /* Two VLAN tags, inner one used. */
+
+/** Which proprietary tags the packet has. */
+#define _NETIO_PKT_TAG_SHIFT         6
+#define _NETIO_PKT_TAG_RMASK         0x3
+#define _NETIO_PKT_TAG_MASK \
+          (_NETIO_PKT_TAG_RMASK << _NETIO_PKT_TAG_SHIFT)
+#define _NETIO_PKT_TAG_NONE          0   /* No proprietary tags. */
+#define _NETIO_PKT_TAG_MRVL          1   /* Marvell HyperG.Stack tags. */
+#define _NETIO_PKT_TAG_MRVL_EXT      2   /* HyperG.Stack extended tags. */
+#define _NETIO_PKT_TAG_BRCM          3   /* Broadcom HiGig tags. */
+
+/** Whether a packet has an LLC + SNAP header. */
+#define _NETIO_PKT_SNAP_SHIFT        8
+#define _NETIO_PKT_SNAP_RMASK        0x1
+#define _NETIO_PKT_SNAP_MASK \
+          (_NETIO_PKT_SNAP_RMASK << _NETIO_PKT_SNAP_SHIFT)
+
+/* NOTE: Bits 9 and 10 are unused. */
+
+/** Length of any custom data before the L2 header, in words. */
+#define _NETIO_PKT_CUSTOM_LEN_SHIFT  11
+#define _NETIO_PKT_CUSTOM_LEN_RMASK  0x1F
+#define _NETIO_PKT_CUSTOM_LEN_MASK \
+          (_NETIO_PKT_CUSTOM_LEN_RMASK << _NETIO_PKT_CUSTOM_LEN_SHIFT)
+
+/** The L4 checksum is incorrect (or perhaps has not been calculated). */
+#define _NETIO_PKT_BAD_L4_CSUM_SHIFT 16
+#define _NETIO_PKT_BAD_L4_CSUM_RMASK 0x1
+#define _NETIO_PKT_BAD_L4_CSUM_MASK \
+          (_NETIO_PKT_BAD_L4_CSUM_RMASK << _NETIO_PKT_BAD_L4_CSUM_SHIFT)
+
+/** Length of the L2 header, in words. */
+#define _NETIO_PKT_L2_LEN_SHIFT  17
+#define _NETIO_PKT_L2_LEN_RMASK  0x1F
+#define _NETIO_PKT_L2_LEN_MASK \
+          (_NETIO_PKT_L2_LEN_RMASK << _NETIO_PKT_L2_LEN_SHIFT)
+
+
+/* Flags in minimal packet metadata. */
+
+/** We need an eDMA checksum on this packet. */
+#define _NETIO_PKT_NEED_EDMA_CSUM_SHIFT            0
+#define _NETIO_PKT_NEED_EDMA_CSUM_RMASK            1
+#define _NETIO_PKT_NEED_EDMA_CSUM_MASK \
+         (_NETIO_PKT_NEED_EDMA_CSUM_RMASK << _NETIO_PKT_NEED_EDMA_CSUM_SHIFT)
+
+/* Data within the packet information table. */
+
+/* Note that, for efficiency, code which uses these fields assumes that none
+ * of the shift values below are zero.  See uses below for an explanation. */
+
+/** Offset within the L2 header of the innermost ethertype (in halfwords). */
+#define _NETIO_PKT_INFO_ETYPE_SHIFT       6
+#define _NETIO_PKT_INFO_ETYPE_RMASK    0x1F
+
+/** Offset within the L2 header of the VLAN tag (in halfwords). */
+#define _NETIO_PKT_INFO_VLAN_SHIFT       11
+#define _NETIO_PKT_INFO_VLAN_RMASK     0x1F
+
+#endif
+
+
+/** The size of a memory buffer representing a small packet.
+ *  @ingroup egress */
+#define SMALL_PACKET_SIZE 256
+
+/** The size of a memory buffer representing a large packet.
+ *  @ingroup egress */
+#define LARGE_PACKET_SIZE 2048
+
+/** The size of a memory buffer representing a jumbo packet.
+ *  @ingroup egress */
+#define JUMBO_PACKET_SIZE (12 * 1024)
+
+
+/* Common ethertypes.
+ * @ingroup ingress */
+/** @{ */
+/** The ethertype of IPv4. */
+#define ETHERTYPE_IPv4 (0x0800)
+/** The ethertype of ARP. */
+#define ETHERTYPE_ARP (0x0806)
+/** The ethertype of VLANs. */
+#define ETHERTYPE_VLAN (0x8100)
+/** The ethertype of a Q-in-Q header. */
+#define ETHERTYPE_Q_IN_Q (0x9100)
+/** The ethertype of IPv6. */
+#define ETHERTYPE_IPv6 (0x86DD)
+/** The ethertype of MPLS. */
+#define ETHERTYPE_MPLS (0x8847)
+/** @} */
+
+
+/** The possible return values of NETIO_PKT_STATUS.
+ * @ingroup ingress
+ */
+typedef enum
+{
+  /** No problems were detected with this packet. */
+  NETIO_PKT_STATUS_OK,
+  /** The packet is undersized; this is expected behavior if the packet's
+    * ethertype is unrecognized, but otherwise the packet is likely corrupt. */
+  NETIO_PKT_STATUS_UNDERSIZE,
+  /** The packet is oversized and some trailing bytes have been discarded.
+      This is expected behavior for short packets, since it's impossible to
+      precisely determine the amount of padding which may have been added to
+      them to make them meet the minimum Ethernet packet size. */
+  NETIO_PKT_STATUS_OVERSIZE,
+  /** The packet was judged to be corrupt by hardware (for instance, it had
+      a bad CRC, or part of it was discarded due to lack of buffer space in
+      the I/O shim) and should be discarded. */
+  NETIO_PKT_STATUS_BAD
+} netio_pkt_status_t;
+
+
+/** Log2 of how many buckets we have. */
+#define NETIO_LOG2_NUM_BUCKETS (10)
+
+/** How many buckets we have.
+ * @ingroup ingress */
+#define NETIO_NUM_BUCKETS (1 << NETIO_LOG2_NUM_BUCKETS)
+
+
+/**
+ * @brief A group-to-bucket identifier.
+ *
+ * @ingroup setup
+ *
+ * This tells us what to do with a given group.
+ */
+typedef union {
+  /** The header broken down into bits. */
+  struct {
+    /** Whether we should balance on L4, if available */
+    unsigned int __balance_on_l4:1;
+    /** Whether we should balance on L3, if available */
+    unsigned int __balance_on_l3:1;
+    /** Whether we should balance on L2, if available */
+    unsigned int __balance_on_l2:1;
+    /** Reserved for future use */
+    unsigned int __reserved:1;
+    /** The base bucket to use to send traffic */
+    unsigned int __bucket_base:NETIO_LOG2_NUM_BUCKETS;
+    /** The mask to apply to the balancing value. This must be one less
+     * than a power of two, e.g. 0x3 or 0xFF.
+     */
+    unsigned int __bucket_mask:NETIO_LOG2_NUM_BUCKETS;
+    /** Pad to 32 bits */
+    unsigned int __padding:(32 - 4 - 2 * NETIO_LOG2_NUM_BUCKETS);
+  } bits;
+  /** To send out the IDN. */
+  unsigned int word;
+}
+netio_group_t;
+
+
+/**
+ * @brief A VLAN-to-bucket identifier.
+ *
+ * @ingroup setup
+ *
+ * This tells us what to do with a given VLAN.
+ */
+typedef netio_group_t netio_vlan_t;
+
+
+/**
+ * A bucket-to-queue mapping.
+ * @ingroup setup
+ */
+typedef unsigned char netio_bucket_t;
+
+
+/**
+ * A packet size can always fit in a netio_size_t.
+ * @ingroup setup
+ */
+typedef unsigned int netio_size_t;
+
+
+/**
+ * @brief Ethernet standard (ingress) packet metadata.
+ *
+ * @ingroup ingress
+ *
+ * This is additional data associated with each packet.
+ * This structure is opaque and accessed through the @ref ingress.
+ *
+ * Also, the buffer population operation currently assumes that standard
+ * metadata is at least as large as minimal metadata, and will need to be
+ * modified if that is no longer the case.
+ */
+typedef struct
+{
+#ifdef __DOXYGEN__
+  /** This structure is opaque. */
+  unsigned char opaque[24];
+#else
+  /** The overall ordinal of the packet */
+  unsigned int __packet_ordinal;
+  /** The ordinal of the packet within the group */
+  unsigned int __group_ordinal;
+  /** The best flow hash IPP could compute. */
+  unsigned int __flow_hash;
+  /** Flags pertaining to checksum calculation, packet type, etc. */
+  unsigned int __flags;
+  /** The first word of "user data". */
+  unsigned int __user_data_0;
+  /** The second word of "user data". */
+  unsigned int __user_data_1;
+#endif
+}
+netio_pkt_metadata_t;
+
+
+/** To ensure that the L3 header is aligned mod 4, the L2 header should be
+ * aligned mod 4 plus 2, since every supported L2 header is 4n + 2 bytes
+ * long.  The standard way to do this is to simply add 2 bytes of padding
+ * before the L2 header.
+ */
+#define NETIO_PACKET_PADDING 2
+
+
+
+/**
+ * @brief Ethernet minimal (egress) packet metadata.
+ *
+ * @ingroup egress
+ *
+ * This structure represents information about packets which have
+ * been processed by @ref netio_populate_buffer() or
+ * @ref netio_populate_prepend_buffer().  This structure is opaque
+ * and accessed through the @ref egress.
+ *
+ * @internal This structure is actually copied into the memory used by
+ * standard metadata, which is assumed to be large enough.
+ */
+typedef struct
+{
+#ifdef __DOXYGEN__
+  /** This structure is opaque. */
+  unsigned char opaque[14];
+#else
+  /** The offset of the L2 header from the start of the packet data. */
+  unsigned short l2_offset;
+  /** The offset of the L3 header from the start of the packet data. */
+  unsigned short l3_offset;
+  /** Where to write the checksum. */
+  unsigned char csum_location;
+  /** Where to start checksumming from. */
+  unsigned char csum_start;
+  /** Flags pertaining to checksum calculation etc. */
+  unsigned short flags;
+  /** The L2 length of the packet. */
+  unsigned short l2_length;
+  /** The checksum with which to seed the checksum generator. */
+  unsigned short csum_seed;
+  /** How much to checksum. */
+  unsigned short csum_length;
+#endif
+}
+netio_pkt_minimal_metadata_t;
+
+
+#ifndef __DOXYGEN__
+
+/**
+ * @brief An I/O notification header.
+ *
+ * This is the first word of data received from an I/O shim in a notification
+ * packet. It contains framing and status information.
+ */
+typedef union
+{
+  unsigned int word; /**< The whole word. */
+  /** The various fields. */
+  struct
+  {
+    unsigned int __channel:7;    /**< Resource channel. */
+    unsigned int __type:4;       /**< Type. */
+    unsigned int __ack:1;        /**< Whether an acknowledgement is needed. */
+    unsigned int __reserved:1;   /**< Reserved. */
+    unsigned int __protocol:1;   /**< A protocol-specific word is added. */
+    unsigned int __status:2;     /**< Status of the transfer. */
+    unsigned int __framing:2;    /**< Framing of the transfer. */
+    unsigned int __transfer_size:14; /**< Transfer size in bytes (total). */
+  } bits;
+}
+__netio_pkt_notif_t;
+
+
+/**
+ * Returns the base address of the packet.
+ */
+#define _NETIO_PKT_HANDLE_BASE(p) \
+  ((unsigned char*)((p).word & 0xFFFFFFC0))
+
+/**
+ * Returns the base address of the packet.
+ */
+#define _NETIO_PKT_BASE(p) \
+  _NETIO_PKT_HANDLE_BASE(p->__packet)
+
+/**
+ * @brief An I/O notification packet (second word)
+ *
+ * This is the second word of data received from an I/O shim in a notification
+ * packet.  This is the virtual address of the packet buffer, plus some flag
+ * bits.  (The virtual address of the packet is always 256-byte aligned so we
+ * have room for 8 bits' worth of flags in the low 8 bits.)
+ *
+ * @internal
+ * NOTE: The low two bits must contain "__queue", so the "packet size"
+ * (SIZE_SMALL, SIZE_LARGE, or SIZE_JUMBO) can be determined quickly.
+ *
+ * If __addr or __offset are moved, _NETIO_PKT_BASE
+ * (defined right below this) must be changed.
+ */
+typedef union
+{
+  unsigned int word; /**< The whole word. */
+  /** The various fields. */
+  struct
+  {
+    /** Which queue the packet will be returned to once it is sent back to
+        the IPP.  This is one of the SIZE_xxx values. */
+    unsigned int __queue:2;
+
+    /** The IPP handle of the sending IPP. */
+    unsigned int __ipp_handle:2;
+
+    /** Reserved for future use. */
+    unsigned int __reserved:1;
+
+    /** If 1, this packet has minimal (egress) metadata; otherwise, it
+        has standard (ingress) metadata. */
+    unsigned int __minimal:1;
+
+    /** Offset of the metadata within the packet.  This value is multiplied
+     *  by 64 and added to the base packet address to get the metadata
+     *  address.  Note that this field is aligned within the word such that
+     *  you can easily extract the metadata address with a 26-bit mask. */
+    unsigned int __offset:2;
+
+    /** The top 24 bits of the packet's virtual address. */
+    unsigned int __addr:24;
+  } bits;
+}
+__netio_pkt_handle_t;
+
+#endif /* !__DOXYGEN__ */
+
+
+/**
+ * @brief A handle for an I/O packet's storage.
+ * @ingroup ingress
+ *
+ * netio_pkt_handle_t encodes the concept of a ::netio_pkt_t with its
+ * packet metadata removed.  It is a much smaller type that exists to
+ * facilitate applications where the full ::netio_pkt_t type is too
+ * large, such as those that cache enormous numbers of packets or wish
+ * to transmit packet descriptors over the UDN.
+ *
+ * Because there is no metadata, most ::netio_pkt_t operations cannot be
+ * performed on a netio_pkt_handle_t.  It supports only
+ * netio_free_handle() (to free the buffer) and
+ * NETIO_PKT_CUSTOM_DATA_H() (to access a pointer to its contents).
+ * The application must acquire any additional metadata it wants from the
+ * original ::netio_pkt_t and record it separately.
+ *
+ * A netio_pkt_handle_t can be extracted from a ::netio_pkt_t by calling
+ * NETIO_PKT_HANDLE().  An invalid handle (analogous to NULL) can be
+ * created by assigning the value ::NETIO_PKT_HANDLE_NONE. A handle can
+ * be tested for validity with NETIO_PKT_HANDLE_IS_VALID().
+ */
+typedef struct
+{
+  unsigned int word; /**< Opaque bits. */
+} netio_pkt_handle_t;
+
+/**
+ * @brief A packet descriptor.
+ *
+ * @ingroup ingress
+ * @ingroup egress
+ *
+ * This data structure represents a packet.  The structure is manipulated
+ * through the @ref ingress and the @ref egress.
+ *
+ * While the contents of a netio_pkt_t are opaque, the structure itself is
+ * portable.  This means that it may be shared between all tiles which have
+ * done a netio_input_register() call for the interface on which the pkt_t
+ * was initially received (via netio_get_packet()) or retrieved (via
+ * netio_get_buffer()).  The contents of a netio_pkt_t can be transmitted to
+ * another tile via shared memory, or via a UDN message, or by other means.
+ * The destination tile may then use the pkt_t as if it had originally been
+ * received locally; it may read or write the packet's data, read its
+ * metadata, free the packet, send the packet, transfer the netio_pkt_t to
+ * yet another tile, and so forth.
+ *
+ * Once a netio_pkt_t has been transferred to a second tile, the first tile
+ * should not reference the original copy; in particular, if more than one
+ * tile frees or sends the same netio_pkt_t, the IPP's packet free lists will
+ * become corrupted.  Note also that each tile which reads or modifies
+ * packet data must obey the memory coherency rules outlined in @ref input.
+ */
+typedef struct
+{
+#ifdef __DOXYGEN__
+  /** This structure is opaque. */
+  unsigned char opaque[32];
+#else
+  /** For an ingress packet (one with standard metadata), this is the
+   *  notification header we got from the I/O shim.  For an egress packet
+   *  (one with minimal metadata), this word is zero if the packet has not
+   *  been populated, and nonzero if it has. */
+  __netio_pkt_notif_t __notif_header;
+
+  /** Virtual address of the packet buffer, plus state flags. */
+  __netio_pkt_handle_t __packet;
+
+  /** Metadata associated with the packet. */
+  netio_pkt_metadata_t __metadata;
+#endif
+}
+netio_pkt_t;
+
+
+#ifndef __DOXYGEN__
+
+#define __NETIO_PKT_NOTIF_HEADER(pkt) ((pkt)->__notif_header)
+#define __NETIO_PKT_IPP_HANDLE(pkt) ((pkt)->__packet.bits.__ipp_handle)
+#define __NETIO_PKT_QUEUE(pkt) ((pkt)->__packet.bits.__queue)
+#define __NETIO_PKT_NOTIF_HEADER_M(mda, pkt) ((pkt)->__notif_header)
+#define __NETIO_PKT_IPP_HANDLE_M(mda, pkt) ((pkt)->__packet.bits.__ipp_handle)
+#define __NETIO_PKT_MINIMAL(pkt) ((pkt)->__packet.bits.__minimal)
+#define __NETIO_PKT_QUEUE_M(mda, pkt) ((pkt)->__packet.bits.__queue)
+#define __NETIO_PKT_FLAGS_M(mda, pkt) ((mda)->__flags)
+
+/* Packet information table, used by the attribute access functions below. */
+extern const uint16_t _netio_pkt_info[];
+
+#endif /* __DOXYGEN__ */
+
+
+#ifndef __DOXYGEN__
+/* These macros are deprecated and will disappear in a future MDE release. */
+#define NETIO_PKT_GOOD_CHECKSUM(pkt) \
+  NETIO_PKT_L4_CSUM_CORRECT(pkt)
+#define NETIO_PKT_GOOD_CHECKSUM_M(mda, pkt) \
+  NETIO_PKT_L4_CSUM_CORRECT_M(mda, pkt)
+#endif /* __DOXYGEN__ */
+
+
+/* Packet attribute access functions. */
+
+/** Return a pointer to the metadata for a packet.
+ * @ingroup ingress
+ *
+ * Calling this function once and passing the result to other retrieval
+ * functions with a "_M" suffix usually improves performance.  This
+ * function must be called on an 'ingress' packet (i.e. one retrieved
+ * by @ref netio_get_packet(), on which @ref netio_populate_buffer() or
+ * @ref netio_populate_prepend_buffer have not been called). Use of this
+ * function on an 'egress' packet will cause an assertion failure.
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return A pointer to the packet's standard metadata.
+ */
+static __inline netio_pkt_metadata_t*
+NETIO_PKT_METADATA(netio_pkt_t* pkt)
+{
+  netio_assert(!pkt->__packet.bits.__minimal);
+  return &pkt->__metadata;
+}
+
+
+/** Return a pointer to the minimal metadata for a packet.
+ * @ingroup egress
+ *
+ * Calling this function once and passing the result to other retrieval
+ * functions with a "_MM" suffix usually improves performance.  This
+ * function must be called on an 'egress' packet (i.e. one on which
+ * @ref netio_populate_buffer() or @ref netio_populate_prepend_buffer()
+ * have been called, or one retrieved by @ref netio_get_buffer()). Use of
+ * this function on an 'ingress' packet will cause an assertion failure.
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return A pointer to the packet's standard metadata.
+ */
+static __inline netio_pkt_minimal_metadata_t*
+NETIO_PKT_MINIMAL_METADATA(netio_pkt_t* pkt)
+{
+  netio_assert(pkt->__packet.bits.__minimal);
+  return (netio_pkt_minimal_metadata_t*) &pkt->__metadata;
+}
+
+
+/** Determine whether a packet has 'minimal' metadata.
+ * @ingroup pktfuncs
+ *
+ * This function will return nonzero if the packet is an 'egress'
+ * packet (i.e. one on which @ref netio_populate_buffer() or
+ * @ref netio_populate_prepend_buffer() have been called, or one
+ * retrieved by @ref netio_get_buffer()), and zero if the packet
+ * is an 'ingress' packet (i.e. one retrieved by @ref netio_get_packet(),
+ * which has not been converted into an 'egress' packet).
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the packet has minimal metadata.
+ */
+static __inline unsigned int
+NETIO_PKT_IS_MINIMAL(netio_pkt_t* pkt)
+{
+  return pkt->__packet.bits.__minimal;
+}
+
+
+/** Return a handle for a packet's storage.
+ * @ingroup pktfuncs
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return A handle for the packet's storage.
+ */
+static __inline netio_pkt_handle_t
+NETIO_PKT_HANDLE(netio_pkt_t* pkt)
+{
+  netio_pkt_handle_t h;
+  h.word = pkt->__packet.word;
+  return h;
+}
+
+
+/** A special reserved value indicating the absence of a packet handle.
+ *
+ * @ingroup pktfuncs
+ */
+#define NETIO_PKT_HANDLE_NONE ((netio_pkt_handle_t) { 0 })
+
+
+/** Test whether a packet handle is valid.
+ *
+ * Applications may wish to use the reserved value NETIO_PKT_HANDLE_NONE
+ * to indicate no packet at all.  This function tests to see if a packet
+ * handle is a real handle, not this special reserved value.
+ *
+ * @ingroup pktfuncs
+ *
+ * @param[in] handle Handle on which to operate.
+ * @return One if the packet handle is valid, else zero.
+ */
+static __inline unsigned int
+NETIO_PKT_HANDLE_IS_VALID(netio_pkt_handle_t handle)
+{
+  return handle.word != 0;
+}
+
+
+
+/** Return a pointer to the start of the packet's custom header.
+ *  A custom header may or may not be present, depending upon the IPP; its
+ *  contents and alignment are also IPP-dependent.  Currently, none of the
+ *  standard IPPs supplied by Tilera produce a custom header.  If present,
+ *  the custom header precedes the L2 header in the packet buffer.
+ * @ingroup ingress
+ *
+ * @param[in] handle Handle on which to operate.
+ * @return A pointer to start of the packet.
+ */
+static __inline unsigned char*
+NETIO_PKT_CUSTOM_DATA_H(netio_pkt_handle_t handle)
+{
+  return _NETIO_PKT_HANDLE_BASE(handle) + NETIO_PACKET_PADDING;
+}
+
+
+/** Return the length of the packet's custom header.
+ *  A custom header may or may not be present, depending upon the IPP; its
+ *  contents and alignment are also IPP-dependent.  Currently, none of the
+ *  standard IPPs supplied by Tilera produce a custom header.  If present,
+ *  the custom header precedes the L2 header in the packet buffer.
+ *
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The length of the packet's custom header, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_CUSTOM_HEADER_LENGTH_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  /*
+   * Note that we effectively need to extract a quantity from the flags word
+   * which is measured in words, and then turn it into bytes by shifting
+   * it left by 2.  We do this all at once by just shifting right two less
+   * bits, and shifting the mask up two bits.
+   */
+  return ((mda->__flags >> (_NETIO_PKT_CUSTOM_LEN_SHIFT - 2)) &
+          (_NETIO_PKT_CUSTOM_LEN_RMASK << 2));
+}
+
+
+/** Return the length of the packet, starting with the custom header.
+ *  A custom header may or may not be present, depending upon the IPP; its
+ *  contents and alignment are also IPP-dependent.  Currently, none of the
+ *  standard IPPs supplied by Tilera produce a custom header.  If present,
+ *  the custom header precedes the L2 header in the packet buffer.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The length of the packet, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_CUSTOM_LENGTH_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return (__NETIO_PKT_NOTIF_HEADER(pkt).bits.__transfer_size -
+          NETIO_PACKET_PADDING);
+}
+
+
+/** Return a pointer to the start of the packet's custom header.
+ *  A custom header may or may not be present, depending upon the IPP; its
+ *  contents and alignment are also IPP-dependent.  Currently, none of the
+ *  standard IPPs supplied by Tilera produce a custom header.  If present,
+ *  the custom header precedes the L2 header in the packet buffer.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return A pointer to start of the packet.
+ */
+static __inline unsigned char*
+NETIO_PKT_CUSTOM_DATA_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return NETIO_PKT_CUSTOM_DATA_H(NETIO_PKT_HANDLE(pkt));
+}
+
+
+/** Return the length of the packet's L2 (Ethernet plus VLAN or SNAP) header.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The length of the packet's L2 header, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_L2_HEADER_LENGTH_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  /*
+   * Note that we effectively need to extract a quantity from the flags word
+   * which is measured in words, and then turn it into bytes by shifting
+   * it left by 2.  We do this all at once by just shifting right two less
+   * bits, and shifting the mask up two bits.  We then add two bytes.
+   */
+  return ((mda->__flags >> (_NETIO_PKT_L2_LEN_SHIFT - 2)) &
+          (_NETIO_PKT_L2_LEN_RMASK << 2)) + 2;
+}
+
+
+/** Return the length of the packet, starting with the L2 (Ethernet) header.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The length of the packet, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_L2_LENGTH_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return (NETIO_PKT_CUSTOM_LENGTH_M(mda, pkt) -
+          NETIO_PKT_CUSTOM_HEADER_LENGTH_M(mda,pkt));
+}
+
+
+/** Return a pointer to the start of the packet's L2 (Ethernet) header.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return A pointer to start of the packet.
+ */
+static __inline unsigned char*
+NETIO_PKT_L2_DATA_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return (NETIO_PKT_CUSTOM_DATA_M(mda, pkt) +
+          NETIO_PKT_CUSTOM_HEADER_LENGTH_M(mda, pkt));
+}
+
+
+/** Retrieve the length of the packet, starting with the L3 (generally,
+ *  the IP) header.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return Length of the packet's L3 header and data, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_L3_LENGTH_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return (NETIO_PKT_L2_LENGTH_M(mda, pkt) -
+          NETIO_PKT_L2_HEADER_LENGTH_M(mda,pkt));
+}
+
+
+/** Return a pointer to the packet's L3 (generally, the IP) header.
+ * @ingroup ingress
+ *
+ * Note that we guarantee word alignment of the L3 header.
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return A pointer to the packet's L3 header.
+ */
+static __inline unsigned char*
+NETIO_PKT_L3_DATA_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return (NETIO_PKT_L2_DATA_M(mda, pkt) +
+          NETIO_PKT_L2_HEADER_LENGTH_M(mda, pkt));
+}
+
+
+/** Return the ordinal of the packet.
+ * @ingroup ingress
+ *
+ * Each packet is given an ordinal number when it is delivered by the IPP.
+ * In the medium term, the ordinal is unique and monotonically increasing,
+ * being incremented by 1 for each packet; the ordinal of the first packet
+ * delivered after the IPP starts is zero.  (Since the ordinal is of finite
+ * size, given enough input packets, it will eventually wrap around to zero;
+ * in the long term, therefore, ordinals are not unique.)  The ordinals
+ * handed out by different IPPs are not disjoint, so two packets from
+ * different IPPs may have identical ordinals.  Packets dropped by the
+ * IPP or by the I/O shim are not assigned ordinals.
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's per-IPP packet ordinal.
+ */
+static __inline unsigned int
+NETIO_PKT_ORDINAL_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return mda->__packet_ordinal;
+}
+
+
+/** Return the per-group ordinal of the packet.
+ * @ingroup ingress
+ *
+ * Each packet is given a per-group ordinal number when it is
+ * delivered by the IPP. By default, the group is the packet's VLAN,
+ * although IPP can be recompiled to use different values.  In
+ * the medium term, the ordinal is unique and monotonically
+ * increasing, being incremented by 1 for each packet; the ordinal of
+ * the first packet distributed to a particular group is zero.
+ * (Since the ordinal is of finite size, given enough input packets,
+ * it will eventually wrap around to zero; in the long term,
+ * therefore, ordinals are not unique.)  The ordinals handed out by
+ * different IPPs are not disjoint, so two packets from different IPPs
+ * may have identical ordinals; similarly, packets distributed to
+ * different groups may have identical ordinals.  Packets dropped by
+ * the IPP or by the I/O shim are not assigned ordinals.
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's per-IPP, per-group ordinal.
+ */
+static __inline unsigned int
+NETIO_PKT_GROUP_ORDINAL_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return mda->__group_ordinal;
+}
+
+
+/** Return the VLAN ID assigned to the packet.
+ * @ingroup ingress
+ *
+ * This value is usually contained within the packet header.
+ *
+ * This value will be zero if the packet does not have a VLAN tag, or if
+ * this value was not extracted from the packet.
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's VLAN ID.
+ */
+static __inline unsigned short
+NETIO_PKT_VLAN_ID_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  int vl = (mda->__flags >> _NETIO_PKT_VLAN_SHIFT) & _NETIO_PKT_VLAN_RMASK;
+  unsigned short* pkt_p;
+  int index;
+  unsigned short val;
+
+  if (vl == _NETIO_PKT_VLAN_NONE)
+    return 0;
+
+  pkt_p = (unsigned short*) NETIO_PKT_L2_DATA_M(mda, pkt);
+  index = (mda->__flags >> _NETIO_PKT_TYPE_SHIFT) & _NETIO_PKT_TYPE_RMASK;
+
+  val = pkt_p[(_netio_pkt_info[index] >> _NETIO_PKT_INFO_VLAN_SHIFT) &
+              _NETIO_PKT_INFO_VLAN_RMASK];
+
+#ifdef __TILECC__
+  return (__insn_bytex(val) >> 16) & 0xFFF;
+#else
+  return (__builtin_bswap32(val) >> 16) & 0xFFF;
+#endif
+}
+
+
+/** Return the ethertype of the packet.
+ * @ingroup ingress
+ *
+ * This value is usually contained within the packet header.
+ *
+ * This value is reliable if @ref NETIO_PKT_ETHERTYPE_RECOGNIZED_M()
+ * returns true, and otherwise, may not be well defined.
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's ethertype.
+ */
+static __inline unsigned short
+NETIO_PKT_ETHERTYPE_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  unsigned short* pkt_p = (unsigned short*) NETIO_PKT_L2_DATA_M(mda, pkt);
+  int index = (mda->__flags >> _NETIO_PKT_TYPE_SHIFT) & _NETIO_PKT_TYPE_RMASK;
+
+  unsigned short val =
+    pkt_p[(_netio_pkt_info[index] >> _NETIO_PKT_INFO_ETYPE_SHIFT) &
+          _NETIO_PKT_INFO_ETYPE_RMASK];
+
+  return __builtin_bswap32(val) >> 16;
+}
+
+
+/** Return the flow hash computed on the packet.
+ * @ingroup ingress
+ *
+ * For TCP and UDP packets, this hash is calculated by hashing together
+ * the "5-tuple" values, specifically the source IP address, destination
+ * IP address, protocol type, source port and destination port.
+ * The hash value is intended to be helpful for millions of distinct
+ * flows.
+ *
+ * For IPv4 or IPv6 packets which are neither TCP nor UDP, the flow hash is
+ * derived by hashing together the source and destination IP addresses.
+ *
+ * For MPLS-encapsulated packets, the flow hash is derived by hashing
+ * the first MPLS label.
+ *
+ * For all other packets the flow hash is computed from the source
+ * and destination Ethernet addresses.
+ *
+ * The hash is symmetric, meaning it produces the same value if the
+ * source and destination are swapped. The only exceptions are
+ * tunneling protocols 0x04 (IP in IP Encapsulation), 0x29 (Simple
+ * Internet Protocol), 0x2F (General Routing Encapsulation) and 0x32
+ * (Encap Security Payload), which use only the destination address
+ * since the source address is not meaningful.
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's 32-bit flow hash.
+ */
+static __inline unsigned int
+NETIO_PKT_FLOW_HASH_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return mda->__flow_hash;
+}
+
+
+/** Return the first word of "user data" for the packet.
+ *
+ * The contents of the user data words depend on the IPP.
+ *
+ * When using the standard ipp1, ipp2, or ipp4 sub-drivers, the first
+ * word of user data contains the least significant bits of the 64-bit
+ * arrival cycle count (see @c get_cycle_count_low()).
+ *
+ * See the <em>System Programmer's Guide</em> for details.
+ *
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's first word of "user data".
+ */
+static __inline unsigned int
+NETIO_PKT_USER_DATA_0_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return mda->__user_data_0;
+}
+
+
+/** Return the second word of "user data" for the packet.
+ *
+ * The contents of the user data words depend on the IPP.
+ *
+ * When using the standard ipp1, ipp2, or ipp4 sub-drivers, the second
+ * word of user data contains the most significant bits of the 64-bit
+ * arrival cycle count (see @c get_cycle_count_high()).
+ *
+ * See the <em>System Programmer's Guide</em> for details.
+ *
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's second word of "user data".
+ */
+static __inline unsigned int
+NETIO_PKT_USER_DATA_1_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return mda->__user_data_1;
+}
+
+
+/** Determine whether the L4 (TCP/UDP) checksum was calculated.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the L4 checksum was calculated.
+ */
+static __inline unsigned int
+NETIO_PKT_L4_CSUM_CALCULATED_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return !(mda->__flags & _NETIO_PKT_NO_L4_CSUM_MASK);
+}
+
+
+/** Determine whether the L4 (TCP/UDP) checksum was calculated and found to
+ *  be correct.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the checksum was calculated and is correct.
+ */
+static __inline unsigned int
+NETIO_PKT_L4_CSUM_CORRECT_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return !(mda->__flags &
+           (_NETIO_PKT_BAD_L4_CSUM_MASK | _NETIO_PKT_NO_L4_CSUM_MASK));
+}
+
+
+/** Determine whether the L3 (IP) checksum was calculated.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the L3 (IP) checksum was calculated.
+*/
+static __inline unsigned int
+NETIO_PKT_L3_CSUM_CALCULATED_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return !(mda->__flags & _NETIO_PKT_NO_L3_CSUM_MASK);
+}
+
+
+/** Determine whether the L3 (IP) checksum was calculated and found to be
+ *  correct.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the checksum was calculated and is correct.
+ */
+static __inline unsigned int
+NETIO_PKT_L3_CSUM_CORRECT_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return !(mda->__flags &
+           (_NETIO_PKT_BAD_L3_CSUM_MASK | _NETIO_PKT_NO_L3_CSUM_MASK));
+}
+
+
+/** Determine whether the ethertype was recognized and L3 packet data was
+ *  processed.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the ethertype was recognized and L3 packet data was
+ *   processed.
+ */
+static __inline unsigned int
+NETIO_PKT_ETHERTYPE_RECOGNIZED_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return !(mda->__flags & _NETIO_PKT_TYPE_UNRECOGNIZED_MASK);
+}
+
+
+/** Retrieve the status of a packet and any errors that may have occurred
+ * during ingress processing (length mismatches, CRC errors, etc.).
+ * @ingroup ingress
+ *
+ * Note that packets for which @ref NETIO_PKT_ETHERTYPE_RECOGNIZED()
+ * returns zero are always reported as underlength, as there is no a priori
+ * means to determine their length.  Normally, applications should use
+ * @ref NETIO_PKT_BAD_M() instead of explicitly checking status with this
+ * function.
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's status.
+ */
+static __inline netio_pkt_status_t
+NETIO_PKT_STATUS_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return (netio_pkt_status_t) __NETIO_PKT_NOTIF_HEADER(pkt).bits.__status;
+}
+
+
+/** Report whether a packet is bad (i.e., was shorter than expected based on
+ *  its headers, or had a bad CRC).
+ * @ingroup ingress
+ *
+ * Note that this function does not verify L3 or L4 checksums.
+ *
+ * @param[in] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the packet is bad and should be discarded.
+ */
+static __inline unsigned int
+NETIO_PKT_BAD_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return ((NETIO_PKT_STATUS_M(mda, pkt) & 1) &&
+          (NETIO_PKT_ETHERTYPE_RECOGNIZED_M(mda, pkt) ||
+           NETIO_PKT_STATUS_M(mda, pkt) == NETIO_PKT_STATUS_BAD));
+}
+
+
+/** Return the length of the packet, starting with the L2 (Ethernet) header.
+ * @ingroup egress
+ *
+ * @param[in] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The length of the packet, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_L2_LENGTH_MM(netio_pkt_minimal_metadata_t* mmd, netio_pkt_t* pkt)
+{
+  return mmd->l2_length;
+}
+
+
+/** Return the length of the L2 (Ethernet) header.
+ * @ingroup egress
+ *
+ * @param[in] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return The length of the packet's L2 header, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_L2_HEADER_LENGTH_MM(netio_pkt_minimal_metadata_t* mmd,
+                              netio_pkt_t* pkt)
+{
+  return mmd->l3_offset - mmd->l2_offset;
+}
+
+
+/** Return the length of the packet, starting with the L3 (IP) header.
+ * @ingroup egress
+ *
+ * @param[in] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return Length of the packet's L3 header and data, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_L3_LENGTH_MM(netio_pkt_minimal_metadata_t* mmd, netio_pkt_t* pkt)
+{
+  return (NETIO_PKT_L2_LENGTH_MM(mmd, pkt) -
+          NETIO_PKT_L2_HEADER_LENGTH_MM(mmd, pkt));
+}
+
+
+/** Return a pointer to the packet's L3 (generally, the IP) header.
+ * @ingroup egress
+ *
+ * Note that we guarantee word alignment of the L3 header.
+ *
+ * @param[in] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return A pointer to the packet's L3 header.
+ */
+static __inline unsigned char*
+NETIO_PKT_L3_DATA_MM(netio_pkt_minimal_metadata_t* mmd, netio_pkt_t* pkt)
+{
+  return _NETIO_PKT_BASE(pkt) + mmd->l3_offset;
+}
+
+
+/** Return a pointer to the packet's L2 (Ethernet) header.
+ * @ingroup egress
+ *
+ * @param[in] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @return A pointer to start of the packet.
+ */
+static __inline unsigned char*
+NETIO_PKT_L2_DATA_MM(netio_pkt_minimal_metadata_t* mmd, netio_pkt_t* pkt)
+{
+  return _NETIO_PKT_BASE(pkt) + mmd->l2_offset;
+}
+
+
+/** Retrieve the status of a packet and any errors that may have occurred
+ * during ingress processing (length mismatches, CRC errors, etc.).
+ * @ingroup ingress
+ *
+ * Note that packets for which @ref NETIO_PKT_ETHERTYPE_RECOGNIZED()
+ * returns zero are always reported as underlength, as there is no a priori
+ * means to determine their length.  Normally, applications should use
+ * @ref NETIO_PKT_BAD() instead of explicitly checking status with this
+ * function.
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's status.
+ */
+static __inline netio_pkt_status_t
+NETIO_PKT_STATUS(netio_pkt_t* pkt)
+{
+  netio_assert(!pkt->__packet.bits.__minimal);
+
+  return (netio_pkt_status_t) __NETIO_PKT_NOTIF_HEADER(pkt).bits.__status;
+}
+
+
+/** Report whether a packet is bad (i.e., was shorter than expected based on
+ *  its headers, or had a bad CRC).
+ * @ingroup ingress
+ *
+ * Note that this function does not verify L3 or L4 checksums.
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the packet is bad and should be discarded.
+ */
+static __inline unsigned int
+NETIO_PKT_BAD(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_BAD_M(mda, pkt);
+}
+
+
+/** Return the length of the packet's custom header.
+ *  A custom header may or may not be present, depending upon the IPP; its
+ *  contents and alignment are also IPP-dependent.  Currently, none of the
+ *  standard IPPs supplied by Tilera produce a custom header.  If present,
+ *  the custom header precedes the L2 header in the packet buffer.
+ * @ingroup pktfuncs
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return The length of the packet's custom header, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_CUSTOM_HEADER_LENGTH(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_CUSTOM_HEADER_LENGTH_M(mda, pkt);
+}
+
+
+/** Return the length of the packet, starting with the custom header.
+ *  A custom header may or may not be present, depending upon the IPP; its
+ *  contents and alignment are also IPP-dependent.  Currently, none of the
+ *  standard IPPs supplied by Tilera produce a custom header.  If present,
+ *  the custom header precedes the L2 header in the packet buffer.
+ * @ingroup pktfuncs
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return  The length of the packet, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_CUSTOM_LENGTH(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_CUSTOM_LENGTH_M(mda, pkt);
+}
+
+
+/** Return a pointer to the packet's custom header.
+ *  A custom header may or may not be present, depending upon the IPP; its
+ *  contents and alignment are also IPP-dependent.  Currently, none of the
+ *  standard IPPs supplied by Tilera produce a custom header.  If present,
+ *  the custom header precedes the L2 header in the packet buffer.
+ * @ingroup pktfuncs
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return A pointer to start of the packet.
+ */
+static __inline unsigned char*
+NETIO_PKT_CUSTOM_DATA(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_CUSTOM_DATA_M(mda, pkt);
+}
+
+
+/** Return the length of the packet's L2 (Ethernet plus VLAN or SNAP) header.
+ * @ingroup pktfuncs
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return The length of the packet's L2 header, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_L2_HEADER_LENGTH(netio_pkt_t* pkt)
+{
+  if (NETIO_PKT_IS_MINIMAL(pkt))
+  {
+    netio_pkt_minimal_metadata_t* mmd = NETIO_PKT_MINIMAL_METADATA(pkt);
+
+    return NETIO_PKT_L2_HEADER_LENGTH_MM(mmd, pkt);
+  }
+  else
+  {
+    netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+    return NETIO_PKT_L2_HEADER_LENGTH_M(mda, pkt);
+  }
+}
+
+
+/** Return the length of the packet, starting with the L2 (Ethernet) header.
+ * @ingroup pktfuncs
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return  The length of the packet, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_L2_LENGTH(netio_pkt_t* pkt)
+{
+  if (NETIO_PKT_IS_MINIMAL(pkt))
+  {
+    netio_pkt_minimal_metadata_t* mmd = NETIO_PKT_MINIMAL_METADATA(pkt);
+
+    return NETIO_PKT_L2_LENGTH_MM(mmd, pkt);
+  }
+  else
+  {
+    netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+    return NETIO_PKT_L2_LENGTH_M(mda, pkt);
+  }
+}
+
+
+/** Return a pointer to the packet's L2 (Ethernet) header.
+ * @ingroup pktfuncs
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return A pointer to start of the packet.
+ */
+static __inline unsigned char*
+NETIO_PKT_L2_DATA(netio_pkt_t* pkt)
+{
+  if (NETIO_PKT_IS_MINIMAL(pkt))
+  {
+    netio_pkt_minimal_metadata_t* mmd = NETIO_PKT_MINIMAL_METADATA(pkt);
+
+    return NETIO_PKT_L2_DATA_MM(mmd, pkt);
+  }
+  else
+  {
+    netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+    return NETIO_PKT_L2_DATA_M(mda, pkt);
+  }
+}
+
+
+/** Retrieve the length of the packet, starting with the L3 (generally, the IP)
+ * header.
+ * @ingroup pktfuncs
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return Length of the packet's L3 header and data, in bytes.
+ */
+static __inline netio_size_t
+NETIO_PKT_L3_LENGTH(netio_pkt_t* pkt)
+{
+  if (NETIO_PKT_IS_MINIMAL(pkt))
+  {
+    netio_pkt_minimal_metadata_t* mmd = NETIO_PKT_MINIMAL_METADATA(pkt);
+
+    return NETIO_PKT_L3_LENGTH_MM(mmd, pkt);
+  }
+  else
+  {
+    netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+    return NETIO_PKT_L3_LENGTH_M(mda, pkt);
+  }
+}
+
+
+/** Return a pointer to the packet's L3 (generally, the IP) header.
+ * @ingroup pktfuncs
+ *
+ * Note that we guarantee word alignment of the L3 header.
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return A pointer to the packet's L3 header.
+ */
+static __inline unsigned char*
+NETIO_PKT_L3_DATA(netio_pkt_t* pkt)
+{
+  if (NETIO_PKT_IS_MINIMAL(pkt))
+  {
+    netio_pkt_minimal_metadata_t* mmd = NETIO_PKT_MINIMAL_METADATA(pkt);
+
+    return NETIO_PKT_L3_DATA_MM(mmd, pkt);
+  }
+  else
+  {
+    netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+    return NETIO_PKT_L3_DATA_M(mda, pkt);
+  }
+}
+
+
+/** Return the ordinal of the packet.
+ * @ingroup ingress
+ *
+ * Each packet is given an ordinal number when it is delivered by the IPP.
+ * In the medium term, the ordinal is unique and monotonically increasing,
+ * being incremented by 1 for each packet; the ordinal of the first packet
+ * delivered after the IPP starts is zero.  (Since the ordinal is of finite
+ * size, given enough input packets, it will eventually wrap around to zero;
+ * in the long term, therefore, ordinals are not unique.)  The ordinals
+ * handed out by different IPPs are not disjoint, so two packets from
+ * different IPPs may have identical ordinals.  Packets dropped by the
+ * IPP or by the I/O shim are not assigned ordinals.
+ *
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's per-IPP packet ordinal.
+ */
+static __inline unsigned int
+NETIO_PKT_ORDINAL(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_ORDINAL_M(mda, pkt);
+}
+
+
+/** Return the per-group ordinal of the packet.
+ * @ingroup ingress
+ *
+ * Each packet is given a per-group ordinal number when it is
+ * delivered by the IPP. By default, the group is the packet's VLAN,
+ * although IPP can be recompiled to use different values.  In
+ * the medium term, the ordinal is unique and monotonically
+ * increasing, being incremented by 1 for each packet; the ordinal of
+ * the first packet distributed to a particular group is zero.
+ * (Since the ordinal is of finite size, given enough input packets,
+ * it will eventually wrap around to zero; in the long term,
+ * therefore, ordinals are not unique.)  The ordinals handed out by
+ * different IPPs are not disjoint, so two packets from different IPPs
+ * may have identical ordinals; similarly, packets distributed to
+ * different groups may have identical ordinals.  Packets dropped by
+ * the IPP or by the I/O shim are not assigned ordinals.
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's per-IPP, per-group ordinal.
+ */
+static __inline unsigned int
+NETIO_PKT_GROUP_ORDINAL(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_GROUP_ORDINAL_M(mda, pkt);
+}
+
+
+/** Return the VLAN ID assigned to the packet.
+ * @ingroup ingress
+ *
+ * This is usually also contained within the packet header.  If the packet
+ * does not have a VLAN tag, the VLAN ID returned by this function is zero.
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's VLAN ID.
+ */
+static __inline unsigned short
+NETIO_PKT_VLAN_ID(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_VLAN_ID_M(mda, pkt);
+}
+
+
+/** Return the ethertype of the packet.
+ * @ingroup ingress
+ *
+ * This value is reliable if @ref NETIO_PKT_ETHERTYPE_RECOGNIZED()
+ * returns true, and otherwise, may not be well defined.
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's ethertype.
+ */
+static __inline unsigned short
+NETIO_PKT_ETHERTYPE(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_ETHERTYPE_M(mda, pkt);
+}
+
+
+/** Return the flow hash computed on the packet.
+ * @ingroup ingress
+ *
+ * For TCP and UDP packets, this hash is calculated by hashing together
+ * the "5-tuple" values, specifically the source IP address, destination
+ * IP address, protocol type, source port and destination port.
+ * The hash value is intended to be helpful for millions of distinct
+ * flows.
+ *
+ * For IPv4 or IPv6 packets which are neither TCP nor UDP, the flow hash is
+ * derived by hashing together the source and destination IP addresses.
+ *
+ * For MPLS-encapsulated packets, the flow hash is derived by hashing
+ * the first MPLS label.
+ *
+ * For all other packets the flow hash is computed from the source
+ * and destination Ethernet addresses.
+ *
+ * The hash is symmetric, meaning it produces the same value if the
+ * source and destination are swapped. The only exceptions are
+ * tunneling protocols 0x04 (IP in IP Encapsulation), 0x29 (Simple
+ * Internet Protocol), 0x2F (General Routing Encapsulation) and 0x32
+ * (Encap Security Payload), which use only the destination address
+ * since the source address is not meaningful.
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's 32-bit flow hash.
+ */
+static __inline unsigned int
+NETIO_PKT_FLOW_HASH(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_FLOW_HASH_M(mda, pkt);
+}
+
+
+/** Return the first word of "user data" for the packet.
+ *
+ * The contents of the user data words depend on the IPP.
+ *
+ * When using the standard ipp1, ipp2, or ipp4 sub-drivers, the first
+ * word of user data contains the least significant bits of the 64-bit
+ * arrival cycle count (see @c get_cycle_count_low()).
+ *
+ * See the <em>System Programmer's Guide</em> for details.
+ *
+ * @ingroup ingress
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's first word of "user data".
+ */
+static __inline unsigned int
+NETIO_PKT_USER_DATA_0(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_USER_DATA_0_M(mda, pkt);
+}
+
+
+/** Return the second word of "user data" for the packet.
+ *
+ * The contents of the user data words depend on the IPP.
+ *
+ * When using the standard ipp1, ipp2, or ipp4 sub-drivers, the second
+ * word of user data contains the most significant bits of the 64-bit
+ * arrival cycle count (see @c get_cycle_count_high()).
+ *
+ * See the <em>System Programmer's Guide</em> for details.
+ *
+ * @ingroup ingress
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return The packet's second word of "user data".
+ */
+static __inline unsigned int
+NETIO_PKT_USER_DATA_1(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_USER_DATA_1_M(mda, pkt);
+}
+
+
+/** Determine whether the L4 (TCP/UDP) checksum was calculated.
+ * @ingroup ingress
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the L4 checksum was calculated.
+ */
+static __inline unsigned int
+NETIO_PKT_L4_CSUM_CALCULATED(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_L4_CSUM_CALCULATED_M(mda, pkt);
+}
+
+
+/** Determine whether the L4 (TCP/UDP) checksum was calculated and found to
+ *  be correct.
+ * @ingroup ingress
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the checksum was calculated and is correct.
+ */
+static __inline unsigned int
+NETIO_PKT_L4_CSUM_CORRECT(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_L4_CSUM_CORRECT_M(mda, pkt);
+}
+
+
+/** Determine whether the L3 (IP) checksum was calculated.
+ * @ingroup ingress
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the L3 (IP) checksum was calculated.
+*/
+static __inline unsigned int
+NETIO_PKT_L3_CSUM_CALCULATED(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_L3_CSUM_CALCULATED_M(mda, pkt);
+}
+
+
+/** Determine whether the L3 (IP) checksum was calculated and found to be
+ *  correct.
+ * @ingroup ingress
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the checksum was calculated and is correct.
+ */
+static __inline unsigned int
+NETIO_PKT_L3_CSUM_CORRECT(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_L3_CSUM_CORRECT_M(mda, pkt);
+}
+
+
+/** Determine whether the Ethertype was recognized and L3 packet data was
+ *  processed.
+ * @ingroup ingress
+ *
+ * @param[in] pkt Packet on which to operate.
+ * @return Nonzero if the Ethertype was recognized and L3 packet data was
+ *   processed.
+ */
+static __inline unsigned int
+NETIO_PKT_ETHERTYPE_RECOGNIZED(netio_pkt_t* pkt)
+{
+  netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+  return NETIO_PKT_ETHERTYPE_RECOGNIZED_M(mda, pkt);
+}
+
+
+/** Set an egress packet's L2 length, using a metadata pointer to speed the
+ * computation.
+ * @ingroup egress
+ *
+ * @param[in,out] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @param[in] len Packet L2 length, in bytes.
+ */
+static __inline void
+NETIO_PKT_SET_L2_LENGTH_MM(netio_pkt_minimal_metadata_t* mmd, netio_pkt_t* pkt,
+                           int len)
+{
+  mmd->l2_length = len;
+}
+
+
+/** Set an egress packet's L2 length.
+ * @ingroup egress
+ *
+ * @param[in,out] pkt Packet on which to operate.
+ * @param[in] len Packet L2 length, in bytes.
+ */
+static __inline void
+NETIO_PKT_SET_L2_LENGTH(netio_pkt_t* pkt, int len)
+{
+  netio_pkt_minimal_metadata_t* mmd = NETIO_PKT_MINIMAL_METADATA(pkt);
+
+  NETIO_PKT_SET_L2_LENGTH_MM(mmd, pkt, len);
+}
+
+
+/** Set an egress packet's L2 header length, using a metadata pointer to
+ *  speed the computation.
+ * @ingroup egress
+ *
+ * It is not normally necessary to call this routine; only the L2 length,
+ * not the header length, is needed to transmit a packet.  It may be useful if
+ * the egress packet will later be processed by code which expects to use
+ * functions like @ref NETIO_PKT_L3_DATA() to get a pointer to the L3 payload.
+ *
+ * @param[in,out] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @param[in] len Packet L2 header length, in bytes.
+ */
+static __inline void
+NETIO_PKT_SET_L2_HEADER_LENGTH_MM(netio_pkt_minimal_metadata_t* mmd,
+                                  netio_pkt_t* pkt, int len)
+{
+  mmd->l3_offset = mmd->l2_offset + len;
+}
+
+
+/** Set an egress packet's L2 header length.
+ * @ingroup egress
+ *
+ * It is not normally necessary to call this routine; only the L2 length,
+ * not the header length, is needed to transmit a packet.  It may be useful if
+ * the egress packet will later be processed by code which expects to use
+ * functions like @ref NETIO_PKT_L3_DATA() to get a pointer to the L3 payload.
+ *
+ * @param[in,out] pkt Packet on which to operate.
+ * @param[in] len Packet L2 header length, in bytes.
+ */
+static __inline void
+NETIO_PKT_SET_L2_HEADER_LENGTH(netio_pkt_t* pkt, int len)
+{
+  netio_pkt_minimal_metadata_t* mmd = NETIO_PKT_MINIMAL_METADATA(pkt);
+
+  NETIO_PKT_SET_L2_HEADER_LENGTH_MM(mmd, pkt, len);
+}
+
+
+/** Set up an egress packet for hardware checksum computation, using a
+ *  metadata pointer to speed the operation.
+ * @ingroup egress
+ *
+ *  NetIO provides the ability to automatically calculate a standard
+ *  16-bit Internet checksum on transmitted packets.  The application
+ *  may specify the point in the packet where the checksum starts, the
+ *  number of bytes to be checksummed, and the two bytes in the packet
+ *  which will be replaced with the completed checksum.  (If the range
+ *  of bytes to be checksummed includes the bytes to be replaced, the
+ *  initial values of those bytes will be included in the checksum.)
+ *
+ *  For some protocols, the packet checksum covers data which is not present
+ *  in the packet, or is at least not contiguous to the main data payload.
+ *  For instance, the TCP checksum includes a "pseudo-header" which includes
+ *  the source and destination IP addresses of the packet.  To accommodate
+ *  this, the checksum engine may be "seeded" with an initial value, which
+ *  the application would need to compute based on the specific protocol's
+ *  requirements.  Note that the seed is given in host byte order (little-
+ *  endian), not network byte order (big-endian); code written to compute a
+ *  pseudo-header checksum in network byte order will need to byte-swap it
+ *  before use as the seed.
+ *
+ *  Note that the checksum is computed as part of the transmission process,
+ *  so it will not be present in the packet upon completion of this routine.
+ *
+ * @param[in,out] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ * @param[in] start Offset within L2 packet of the first byte to include in
+ *   the checksum.
+ * @param[in] length Number of bytes to include in the checksum.
+ *   the checksum.
+ * @param[in] location Offset within L2 packet of the first of the two bytes
+ *   to be replaced with the calculated checksum.
+ * @param[in] seed Initial value of the running checksum before any of the
+ *   packet data is added.
+ */
+static __inline void
+NETIO_PKT_DO_EGRESS_CSUM_MM(netio_pkt_minimal_metadata_t* mmd,
+                            netio_pkt_t* pkt, int start, int length,
+                            int location, uint16_t seed)
+{
+  mmd->csum_start = start;
+  mmd->csum_length = length;
+  mmd->csum_location = location;
+  mmd->csum_seed = seed;
+  mmd->flags |= _NETIO_PKT_NEED_EDMA_CSUM_MASK;
+}
+
+
+/** Set up an egress packet for hardware checksum computation.
+ * @ingroup egress
+ *
+ *  NetIO provides the ability to automatically calculate a standard
+ *  16-bit Internet checksum on transmitted packets.  The application
+ *  may specify the point in the packet where the checksum starts, the
+ *  number of bytes to be checksummed, and the two bytes in the packet
+ *  which will be replaced with the completed checksum.  (If the range
+ *  of bytes to be checksummed includes the bytes to be replaced, the
+ *  initial values of those bytes will be included in the checksum.)
+ *
+ *  For some protocols, the packet checksum covers data which is not present
+ *  in the packet, or is at least not contiguous to the main data payload.
+ *  For instance, the TCP checksum includes a "pseudo-header" which includes
+ *  the source and destination IP addresses of the packet.  To accommodate
+ *  this, the checksum engine may be "seeded" with an initial value, which
+ *  the application would need to compute based on the specific protocol's
+ *  requirements.  Note that the seed is given in host byte order (little-
+ *  endian), not network byte order (big-endian); code written to compute a
+ *  pseudo-header checksum in network byte order will need to byte-swap it
+ *  before use as the seed.
+ *
+ *  Note that the checksum is computed as part of the transmission process,
+ *  so it will not be present in the packet upon completion of this routine.
+ *
+ * @param[in,out] pkt Packet on which to operate.
+ * @param[in] start Offset within L2 packet of the first byte to include in
+ *   the checksum.
+ * @param[in] length Number of bytes to include in the checksum.
+ *   the checksum.
+ * @param[in] location Offset within L2 packet of the first of the two bytes
+ *   to be replaced with the calculated checksum.
+ * @param[in] seed Initial value of the running checksum before any of the
+ *   packet data is added.
+ */
+static __inline void
+NETIO_PKT_DO_EGRESS_CSUM(netio_pkt_t* pkt, int start, int length,
+                         int location, uint16_t seed)
+{
+  netio_pkt_minimal_metadata_t* mmd = NETIO_PKT_MINIMAL_METADATA(pkt);
+
+  NETIO_PKT_DO_EGRESS_CSUM_MM(mmd, pkt, start, length, location, seed);
+}
+
+
+/** Return the number of bytes which could be prepended to a packet, using a
+ *  metadata pointer to speed the operation.
+ *  See @ref netio_populate_prepend_buffer() to get a full description of
+ *  prepending.
+ *
+ * @param[in,out] mda Pointer to packet's standard metadata.
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline int
+NETIO_PKT_PREPEND_AVAIL_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+  return (pkt->__packet.bits.__offset << 6) +
+         NETIO_PKT_CUSTOM_HEADER_LENGTH_M(mda, pkt);
+}
+
+
+/** Return the number of bytes which could be prepended to a packet, using a
+ *  metadata pointer to speed the operation.
+ *  See @ref netio_populate_prepend_buffer() to get a full description of
+ *  prepending.
+ * @ingroup egress
+ *
+ * @param[in,out] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline int
+NETIO_PKT_PREPEND_AVAIL_MM(netio_pkt_minimal_metadata_t* mmd, netio_pkt_t* pkt)
+{
+  return (pkt->__packet.bits.__offset << 6) + mmd->l2_offset;
+}
+
+
+/** Return the number of bytes which could be prepended to a packet.
+ *  See @ref netio_populate_prepend_buffer() to get a full description of
+ *  prepending.
+ * @ingroup egress
+ *
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline int
+NETIO_PKT_PREPEND_AVAIL(netio_pkt_t* pkt)
+{
+  if (NETIO_PKT_IS_MINIMAL(pkt))
+  {
+    netio_pkt_minimal_metadata_t* mmd = NETIO_PKT_MINIMAL_METADATA(pkt);
+
+    return NETIO_PKT_PREPEND_AVAIL_MM(mmd, pkt);
+  }
+  else
+  {
+    netio_pkt_metadata_t* mda = NETIO_PKT_METADATA(pkt);
+
+    return NETIO_PKT_PREPEND_AVAIL_M(mda, pkt);
+  }
+}
+
+
+/** Flush a packet's minimal metadata from the cache, using a metadata pointer
+ *  to speed the operation.
+ * @ingroup egress
+ *
+ * @param[in] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_FLUSH_MINIMAL_METADATA_MM(netio_pkt_minimal_metadata_t* mmd,
+                                    netio_pkt_t* pkt)
+{
+}
+
+
+/** Invalidate a packet's minimal metadata from the cache, using a metadata
+ *  pointer to speed the operation.
+ * @ingroup egress
+ *
+ * @param[in] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_INV_MINIMAL_METADATA_MM(netio_pkt_minimal_metadata_t* mmd,
+                                  netio_pkt_t* pkt)
+{
+}
+
+
+/** Flush and then invalidate a packet's minimal metadata from the cache,
+ *  using a metadata pointer to speed the operation.
+ * @ingroup egress
+ *
+ * @param[in] mmd Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_FLUSH_INV_MINIMAL_METADATA_MM(netio_pkt_minimal_metadata_t* mmd,
+                                        netio_pkt_t* pkt)
+{
+}
+
+
+/** Flush a packet's metadata from the cache, using a metadata pointer
+ *  to speed the operation.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's minimal metadata.
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_FLUSH_METADATA_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+}
+
+
+/** Invalidate a packet's metadata from the cache, using a metadata
+ *  pointer to speed the operation.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's metadata.
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_INV_METADATA_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+}
+
+
+/** Flush and then invalidate a packet's metadata from the cache,
+ *  using a metadata pointer to speed the operation.
+ * @ingroup ingress
+ *
+ * @param[in] mda Pointer to packet's metadata.
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_FLUSH_INV_METADATA_M(netio_pkt_metadata_t* mda, netio_pkt_t* pkt)
+{
+}
+
+
+/** Flush a packet's minimal metadata from the cache.
+ * @ingroup egress
+ *
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_FLUSH_MINIMAL_METADATA(netio_pkt_t* pkt)
+{
+}
+
+
+/** Invalidate a packet's minimal metadata from the cache.
+ * @ingroup egress
+ *
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_INV_MINIMAL_METADATA(netio_pkt_t* pkt)
+{
+}
+
+
+/** Flush and then invalidate a packet's minimal metadata from the cache.
+ * @ingroup egress
+ *
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_FLUSH_INV_MINIMAL_METADATA(netio_pkt_t* pkt)
+{
+}
+
+
+/** Flush a packet's metadata from the cache.
+ * @ingroup ingress
+ *
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_FLUSH_METADATA(netio_pkt_t* pkt)
+{
+}
+
+
+/** Invalidate a packet's metadata from the cache.
+ * @ingroup ingress
+ *
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_INV_METADATA(netio_pkt_t* pkt)
+{
+}
+
+
+/** Flush and then invalidate a packet's metadata from the cache.
+ * @ingroup ingress
+ *
+ * @param[in] pkt Packet on which to operate.
+ */
+static __inline void
+NETIO_PKT_FLUSH_INV_METADATA(netio_pkt_t* pkt)
+{
+}
+
+/** Number of NUMA nodes we can distribute buffers to.
+ * @ingroup setup */
+#define NETIO_NUM_NODE_WEIGHTS  16
+
+/**
+ * @brief An object for specifying the characteristics of NetIO communication
+ * endpoint.
+ *
+ * @ingroup setup
+ *
+ * The @ref netio_input_register() function uses this structure to define
+ * how an application tile will communicate with an IPP.
+ *
+ *
+ * Future updates to NetIO may add new members to this structure,
+ * which can affect the success of the registration operation.  Thus,
+ * if dynamically initializing the structure, applications are urged to
+ * zero it out first, for example:
+ *
+ * @code
+ * netio_input_config_t config;
+ * memset(&config, 0, sizeof (config));
+ * config.flags = NETIO_RECV | NETIO_XMIT_CSUM | NETIO_TAG_NONE;
+ * config.num_receive_packets = NETIO_MAX_RECEIVE_PKTS;
+ * config.queue_id = 0;
+ *     .
+ *     .
+ *     .
+ * @endcode
+ *
+ * since that guarantees that any unused structure members, including
+ * members which did not exist when the application was first developed,
+ * will not have unexpected values.
+ *
+ * If statically initializing the structure, we strongly recommend use of
+ * C99-style named initializers, for example:
+ *
+ * @code
+ * netio_input_config_t config = {
+ *    .flags = NETIO_RECV | NETIO_XMIT_CSUM | NETIO_TAG_NONE,
+ *    .num_receive_packets = NETIO_MAX_RECEIVE_PKTS,
+ *    .queue_id = 0,
+ * },
+ * @endcode
+ *
+ * instead of the old-style structure initialization:
+ *
+ * @code
+ * // Bad example! Currently equivalent to the above, but don't do this.
+ * netio_input_config_t config = {
+ *    NETIO_RECV | NETIO_XMIT_CSUM | NETIO_TAG_NONE, NETIO_MAX_RECEIVE_PKTS, 0
+ * },
+ * @endcode
+ *
+ * since the C99 style requires no changes to the code if elements of the
+ * config structure are rearranged.  (It also makes the initialization much
+ * easier to understand.)
+ *
+ * Except for items which address a particular tile's transmit or receive
+ * characteristics, such as the ::NETIO_RECV flag, applications are advised
+ * to specify the same set of configuration data on all registrations.
+ * This prevents differing results if multiple tiles happen to do their
+ * registration operations in a different order on different invocations of
+ * the application.  This is particularly important for things like link
+ * management flags, and buffer size and homing specifications.
+ *
+ * Unless the ::NETIO_FIXED_BUFFER_VA flag is specified in flags, the NetIO
+ * buffer pool is automatically created and mapped into the application's
+ * virtual address space at an address chosen by the operating system,
+ * using the common memory (cmem) facility in the Tilera Multicore
+ * Components library.  The cmem facility allows multiple processes to gain
+ * access to shared memory which is mapped into each process at an
+ * identical virtual address.  In order for this to work, the processes
+ * must have a common ancestor, which must create the common memory using
+ * tmc_cmem_init().
+ *
+ * In programs using the iLib process creation API, or in programs which use
+ * only one process (which include programs using the pthreads library),
+ * tmc_cmem_init() is called automatically.  All other applications
+ * must call it explicitly, before any child processes which might call
+ * netio_input_register() are created.
+ */
+typedef struct
+{
+  /** Registration characteristics.
+
+      This value determines several characteristics of the registration;
+      flags for different types of behavior are ORed together to make the
+      final flag value.  Generally applications should specify exactly
+      one flag from each of the following categories:
+
+      - Whether the application will be receiving packets on this queue
+        (::NETIO_RECV or ::NETIO_NO_RECV).
+
+      - Whether the application will be transmitting packets on this queue,
+        and if so, whether it will request egress checksum calculation
+        (::NETIO_XMIT, ::NETIO_XMIT_CSUM, or ::NETIO_NO_XMIT).  It is
+        legal to call netio_get_buffer() without one of the XMIT flags,
+        as long as ::NETIO_RECV is specified; in this case, the retrieved
+        buffers must be passed to another tile for transmission.
+
+      - Whether the application expects any vendor-specific tags in
+        its packets' L2 headers (::NETIO_TAG_NONE, ::NETIO_TAG_BRCM,
+        or ::NETIO_TAG_MRVL).  This must match the configuration of the
+        target IPP.
+
+      To accommodate applications written to previous versions of the NetIO
+      interface, none of the flags above are currently required; if omitted,
+      NetIO behaves more or less as if ::NETIO_RECV | ::NETIO_XMIT_CSUM |
+      ::NETIO_TAG_NONE were used.  However, explicit specification of
+      the relevant flags allows NetIO to do a better job of resource
+      allocation, allows earlier detection of certain configuration errors,
+      and may enable advanced features or higher performance in the future,
+      so their use is strongly recommended.
+
+      Note that specifying ::NETIO_NO_RECV along with ::NETIO_NO_XMIT
+      is a special case, intended primarily for use by programs which
+      retrieve network statistics or do link management operations.
+      When these flags are both specified, the resulting queue may not
+      be used with NetIO routines other than netio_get(), netio_set(),
+      and netio_input_unregister().  See @ref link for more information
+      on link management.
+
+      Other flags are optional; their use is described below.
+  */
+  int flags;
+
+  /** Interface name.  This is a string which identifies the specific
+      Ethernet controller hardware to be used.  The format of the string
+      is a device type and a device index, separated by a slash; so,
+      the first 10 Gigabit Ethernet controller is named "xgbe/0", while
+      the second 10/100/1000 Megabit Ethernet controller is named "gbe/1".
+   */
+  const char* interface;
+
+  /** Receive packet queue size.  This specifies the maximum number
+      of ingress packets that can be received on this queue without
+      being retrieved by @ref netio_get_packet().  If the IPP's distribution
+      algorithm calls for a packet to be sent to this queue, and this
+      number of packets are already pending there, the new packet
+      will either be discarded, or sent to another tile registered
+      for the same queue_id (see @ref drops).  This value must
+      be at least ::NETIO_MIN_RECEIVE_PKTS, can always be at least
+      ::NETIO_MAX_RECEIVE_PKTS, and may be larger than that on certain
+      interfaces.
+   */
+  int num_receive_packets;
+
+  /** The queue ID being requested.  Legal values for this range from 0
+      to ::NETIO_MAX_QUEUE_ID, inclusive.  ::NETIO_MAX_QUEUE_ID is always
+      greater than or equal to the number of tiles; this allows one queue
+      for each tile, plus at least one additional queue.  Some applications
+      may wish to use the additional queue as a destination for unwanted
+      packets, since packets delivered to queues for which no tiles have
+      registered are discarded.
+   */
+  unsigned int queue_id;
+
+  /** Maximum number of small send buffers to be held in the local empty
+      buffer cache.  This specifies the size of the area which holds
+      empty small egress buffers requested from the IPP but not yet
+      retrieved via @ref netio_get_buffer().  This value must be greater
+      than zero if the application will ever use @ref netio_get_buffer()
+      to allocate empty small egress buffers; it may be no larger than
+      ::NETIO_MAX_SEND_BUFFERS.  See @ref epp for more details on empty
+      buffer caching.
+   */
+  int num_send_buffers_small_total;
+
+  /** Number of small send buffers to be preallocated at registration.
+      If this value is nonzero, the specified number of empty small egress
+      buffers will be requested from the IPP during the netio_input_register
+      operation; this may speed the execution of @ref netio_get_buffer().
+      This may be no larger than @ref num_send_buffers_small_total.  See @ref
+      epp for more details on empty buffer caching.
+   */
+  int num_send_buffers_small_prealloc;
+
+  /** Maximum number of large send buffers to be held in the local empty
+      buffer cache.  This specifies the size of the area which holds empty
+      large egress buffers requested from the IPP but not yet retrieved via
+      @ref netio_get_buffer().  This value must be greater than zero if the
+      application will ever use @ref netio_get_buffer() to allocate empty
+      large egress buffers; it may be no larger than ::NETIO_MAX_SEND_BUFFERS.
+      See @ref epp for more details on empty buffer caching.
+   */
+  int num_send_buffers_large_total;
+
+  /** Number of large send buffers to be preallocated at registration.
+      If this value is nonzero, the specified number of empty large egress
+      buffers will be requested from the IPP during the netio_input_register
+      operation; this may speed the execution of @ref netio_get_buffer().
+      This may be no larger than @ref num_send_buffers_large_total.  See @ref
+      epp for more details on empty buffer caching.
+   */
+  int num_send_buffers_large_prealloc;
+
+  /** Maximum number of jumbo send buffers to be held in the local empty
+      buffer cache.  This specifies the size of the area which holds empty
+      jumbo egress buffers requested from the IPP but not yet retrieved via
+      @ref netio_get_buffer().  This value must be greater than zero if the
+      application will ever use @ref netio_get_buffer() to allocate empty
+      jumbo egress buffers; it may be no larger than ::NETIO_MAX_SEND_BUFFERS.
+      See @ref epp for more details on empty buffer caching.
+   */
+  int num_send_buffers_jumbo_total;
+
+  /** Number of jumbo send buffers to be preallocated at registration.
+      If this value is nonzero, the specified number of empty jumbo egress
+      buffers will be requested from the IPP during the netio_input_register
+      operation; this may speed the execution of @ref netio_get_buffer().
+      This may be no larger than @ref num_send_buffers_jumbo_total.  See @ref
+      epp for more details on empty buffer caching.
+   */
+  int num_send_buffers_jumbo_prealloc;
+
+  /** Total packet buffer size.  This determines the total size, in bytes,
+      of the NetIO buffer pool.  Note that the maximum number of available
+      buffers of each size is determined during hypervisor configuration
+      (see the <em>System Programmer's Guide</em> for details); this just
+      influences how much host memory is allocated for those buffers.
+
+      The buffer pool is allocated from common memory, which will be
+      automatically initialized if needed.  If your buffer pool is larger
+      than 240 MB, you might need to explicitly call @c tmc_cmem_init(),
+      as described in the Application Libraries Reference Manual (UG227).
+
+      Packet buffers are currently allocated in chunks of 16 MB; this
+      value will be rounded up to the next larger multiple of 16 MB.
+      If this value is zero, a default of 32 MB will be used; this was
+      the value used by previous versions of NetIO.  Note that taking this
+      default also affects the placement of buffers on Linux NUMA nodes.
+      See @ref buffer_node_weights for an explanation of buffer placement.
+
+      In order to successfully allocate packet buffers, Linux must have
+      available huge pages on the relevant Linux NUMA nodes.  See the
+      <em>System Programmer's Guide</em> for information on configuring
+      huge page support in Linux.
+   */
+  uint64_t total_buffer_size;
+
+  /** Buffer placement weighting factors.
+
+      This array specifies the relative amount of buffering to place
+      on each of the available Linux NUMA nodes.  This array is
+      indexed by the NUMA node, and the values in the array are
+      proportional to the amount of buffer space to allocate on that
+      node.
+
+      If memory striping is enabled in the Hypervisor, then there is
+      only one logical NUMA node (node 0). In that case, NetIO will by
+      default ignore the suggested buffer node weights, and buffers
+      will be striped across the physical memory controllers. See
+      UG209 System Programmer's Guide for a description of the
+      hypervisor option that controls memory striping.
+
+      If memory striping is disabled, then there are up to four NUMA
+      nodes, corresponding to the four DDRAM controllers in the TILE
+      processor architecture.  See UG100 Tile Processor Architecture
+      Overview for a diagram showing the location of each of the DDRAM
+      controllers relative to the tile array.
+
+      For instance, if memory striping is disabled, the following
+      configuration strucure:
+
+      @code
+      netio_input_config_t config = {
+            .
+            .
+            .
+        .total_buffer_size = 4 * 16 * 1024 * 1024;
+        .buffer_node_weights = { 1, 0, 1, 0 },
+      },
+      @endcode
+
+      would result in 32 MB of buffers being placed on controller 0, and
+      32 MB on controller 2.  (Since buffers are allocated in units of
+      16 MB, some sets of weights will not be able to be matched exactly.)
+
+      For the weights to be effective, @ref total_buffer_size must be
+      nonzero.  If @ref total_buffer_size is zero, causing the default
+      32 MB of buffer space to be used, then any specified weights will
+      be ignored, and buffers will positioned as they were in previous
+      versions of NetIO:
+
+      - For xgbe/0 and gbe/0, 16 MB of buffers will be placed on controller 1,
+        and the other 16 MB will be placed on controller 2.
+
+      - For xgbe/1 and gbe/1, 16 MB of buffers will be placed on controller 2,
+        and the other 16 MB will be placed on controller 3.
+
+      If @ref total_buffer_size is nonzero, but all weights are zero,
+      then all buffer space will be allocated on Linux NUMA node zero.
+
+      By default, the specified buffer placement is treated as a hint;
+      if sufficient free memory is not available on the specified
+      controllers, the buffers will be allocated elsewhere.  However,
+      if the ::NETIO_STRICT_HOMING flag is specified in @ref flags, then a
+      failure to allocate buffer space exactly as requested will cause the
+      registration operation to fail with an error of ::NETIO_CANNOT_HOME.
+
+      Note that maximal network performance cannot be achieved with
+      only one memory controller.
+   */
+  uint8_t buffer_node_weights[NETIO_NUM_NODE_WEIGHTS];
+
+  /** Fixed virtual address for packet buffers.  Only valid when
+      ::NETIO_FIXED_BUFFER_VA is specified in @ref flags; see the
+      description of that flag for details.
+   */
+  void* fixed_buffer_va;
+
+  /**
+      Maximum number of outstanding send packet requests.  This value is
+      only relevant when an EPP is in use; it determines the number of
+      slots in the EPP's outgoing packet queue which this tile is allowed
+      to consume, and thus the number of packets which may be sent before
+      the sending tile must wait for an acknowledgment from the EPP.
+      Modifying this value is generally only helpful when using @ref
+      netio_send_packet_vector(), where it can help improve performance by
+      allowing a single vector send operation to process more packets.
+      Typically it is not specified, and the default, which divides the
+      outgoing packet slots evenly between all tiles on the chip, is used.
+
+      If a registration asks for more outgoing packet queue slots than are
+      available, ::NETIO_TOOMANY_XMIT will be returned.  The total number
+      of packet queue slots which are available for all tiles for each EPP
+      is subject to change, but is currently ::NETIO_TOTAL_SENDS_OUTSTANDING.
+
+
+      This value is ignored if ::NETIO_XMIT is not specified in flags.
+      If you want to specify a large value here for a specific tile, you are
+      advised to specify NETIO_NO_XMIT on other, non-transmitting tiles so
+      that they do not consume a default number of packet slots.  Any tile
+      transmitting is required to have at least ::NETIO_MIN_SENDS_OUTSTANDING
+      slots allocated to it; values less than that will be silently
+      increased by the NetIO library.
+   */
+  int num_sends_outstanding;
+}
+netio_input_config_t;
+
+
+/** Registration flags; used in the @ref netio_input_config_t structure.
+ * @addtogroup setup
+ */
+/** @{ */
+
+/** Fail a registration request if we can't put packet buffers
+    on the specified memory controllers. */
+#define NETIO_STRICT_HOMING   0x00000002
+
+/** This application expects no tags on its L2 headers. */
+#define NETIO_TAG_NONE        0x00000004
+
+/** This application expects Marvell extended tags on its L2 headers. */
+#define NETIO_TAG_MRVL        0x00000008
+
+/** This application expects Broadcom tags on its L2 headers. */
+#define NETIO_TAG_BRCM        0x00000010
+
+/** This registration may call routines which receive packets. */
+#define NETIO_RECV            0x00000020
+
+/** This registration may not call routines which receive packets. */
+#define NETIO_NO_RECV         0x00000040
+
+/** This registration may call routines which transmit packets. */
+#define NETIO_XMIT            0x00000080
+
+/** This registration may call routines which transmit packets with
+    checksum acceleration. */
+#define NETIO_XMIT_CSUM       0x00000100
+
+/** This registration may not call routines which transmit packets. */
+#define NETIO_NO_XMIT         0x00000200
+
+/** This registration wants NetIO buffers mapped at an application-specified
+    virtual address.
+
+    NetIO buffers are by default created by the TMC common memory facility,
+    which must be configured by a common ancestor of all processes sharing
+    a network interface.  When this flag is specified, NetIO buffers are
+    instead mapped at an address chosen by the application (and specified
+    in @ref netio_input_config_t::fixed_buffer_va).  This allows multiple
+    unrelated but cooperating processes to share a NetIO interface.
+    All processes sharing the same interface must specify this flag,
+    and all must specify the same fixed virtual address.
+
+    @ref netio_input_config_t::fixed_buffer_va must be a
+    multiple of 16 MB, and the packet buffers will occupy @ref
+    netio_input_config_t::total_buffer_size bytes of virtual address
+    space, beginning at that address.  If any of those virtual addresses
+    are currently occupied by other memory objects, like application or
+    shared library code or data, @ref netio_input_register() will return
+    ::NETIO_FAULT.  While it is impossible to provide a fixed_buffer_va
+    which will work for all applications, a good first guess might be to
+    use 0xb0000000 minus @ref netio_input_config_t::total_buffer_size.
+    If that fails, it might be helpful to consult the running application's
+    virtual address description file (/proc/<em>pid</em>/maps) to see
+    which regions of virtual address space are available.
+ */
+#define NETIO_FIXED_BUFFER_VA 0x00000400
+
+/** This registration call will not complete unless the network link
+    is up.  The process will wait several seconds for this to happen (the
+    precise interval is link-dependent), but if the link does not come up,
+    ::NETIO_LINK_DOWN will be returned.  This flag is the default if
+    ::NETIO_NOREQUIRE_LINK_UP is not specified.  Note that this flag by
+    itself does not request that the link be brought up; that can be done
+    with the ::NETIO_AUTO_LINK_UPDN or ::NETIO_AUTO_LINK_UP flags (the
+    latter is the default if no NETIO_AUTO_LINK_xxx flags are specified),
+    or by explicitly setting the link's desired state via netio_set().
+    If the link is not brought up by one of those methods, and this flag
+    is specified, the registration operation will return ::NETIO_LINK_DOWN.
+    This flag is ignored if it is specified along with ::NETIO_NO_XMIT and
+    ::NETIO_NO_RECV.  See @ref link for more information on link
+    management.
+ */
+#define NETIO_REQUIRE_LINK_UP    0x00000800
+
+/** This registration call will complete even if the network link is not up.
+    Whenever the link is not up, packets will not be sent or received:
+    netio_get_packet() will return ::NETIO_NOPKT once all queued packets
+    have been drained, and netio_send_packet() and similar routines will
+    return NETIO_QUEUE_FULL once the outgoing packet queue in the EPP
+    or the I/O shim is full.  See @ref link for more information on link
+    management.
+ */
+#define NETIO_NOREQUIRE_LINK_UP  0x00001000
+
+#ifndef __DOXYGEN__
+/*
+ * These are part of the implementation of the NETIO_AUTO_LINK_xxx flags,
+ * but should not be used directly by applications, and are thus not
+ * documented.
+ */
+#define _NETIO_AUTO_UP        0x00002000
+#define _NETIO_AUTO_DN        0x00004000
+#define _NETIO_AUTO_PRESENT   0x00008000
+#endif
+
+/** Set the desired state of the link to up, allowing any speeds which are
+    supported by the link hardware, as part of this registration operation.
+    Do not take down the link automatically.  This is the default if
+    no other NETIO_AUTO_LINK_xxx flags are specified.  This flag is ignored
+    if it is specified along with ::NETIO_NO_XMIT and ::NETIO_NO_RECV.
+    See @ref link for more information on link management.
+ */
+#define NETIO_AUTO_LINK_UP     (_NETIO_AUTO_PRESENT | _NETIO_AUTO_UP)
+
+/** Set the desired state of the link to up, allowing any speeds which are
+    supported by the link hardware, as part of this registration operation.
+    Set the desired state of the link to down the next time no tiles are
+    registered for packet reception or transmission.  This flag is ignored
+    if it is specified along with ::NETIO_NO_XMIT and ::NETIO_NO_RECV.
+    See @ref link for more information on link management.
+ */
+#define NETIO_AUTO_LINK_UPDN   (_NETIO_AUTO_PRESENT | _NETIO_AUTO_UP | \
+                                _NETIO_AUTO_DN)
+
+/** Set the desired state of the link to down the next time no tiles are
+    registered for packet reception or transmission.  This flag is ignored
+    if it is specified along with ::NETIO_NO_XMIT and ::NETIO_NO_RECV.
+    See @ref link for more information on link management.
+ */
+#define NETIO_AUTO_LINK_DN     (_NETIO_AUTO_PRESENT | _NETIO_AUTO_DN)
+
+/** Do not bring up the link automatically as part of this registration
+    operation.  Do not take down the link automatically.  This flag
+    is ignored if it is specified along with ::NETIO_NO_XMIT and
+    ::NETIO_NO_RECV.  See @ref link for more information on link management.
+  */
+#define NETIO_AUTO_LINK_NONE   _NETIO_AUTO_PRESENT
+
+
+/** Minimum number of receive packets. */
+#define NETIO_MIN_RECEIVE_PKTS            16
+
+/** Lower bound on the maximum number of receive packets; may be higher
+    than this on some interfaces. */
+#define NETIO_MAX_RECEIVE_PKTS           128
+
+/** Maximum number of send buffers, per packet size. */
+#define NETIO_MAX_SEND_BUFFERS            16
+
+/** Number of EPP queue slots, and thus outstanding sends, per EPP. */
+#define NETIO_TOTAL_SENDS_OUTSTANDING   2015
+
+/** Minimum number of EPP queue slots, and thus outstanding sends, per
+ *  transmitting tile. */
+#define NETIO_MIN_SENDS_OUTSTANDING       16
+
+
+/**@}*/
+
+#ifndef __DOXYGEN__
+
+/**
+ * An object for providing Ethernet packets to a process.
+ */
+struct __netio_queue_impl_t;
+
+/**
+ * An object for managing the user end of a NetIO queue.
+ */
+struct __netio_queue_user_impl_t;
+
+#endif /* !__DOXYGEN__ */
+
+
+/** A netio_queue_t describes a NetIO communications endpoint.
+ * @ingroup setup
+ */
+typedef struct
+{
+#ifdef __DOXYGEN__
+  uint8_t opaque[8];                 /**< This is an opaque structure. */
+#else
+  struct __netio_queue_impl_t* __system_part;    /**< The system part. */
+  struct __netio_queue_user_impl_t* __user_part; /**< The user part. */
+#ifdef _NETIO_PTHREAD
+  _netio_percpu_mutex_t lock;                    /**< Queue lock. */
+#endif
+#endif
+}
+netio_queue_t;
+
+
+/**
+ * @brief Packet send context.
+ *
+ * @ingroup egress
+ *
+ * Packet send context for use with netio_send_packet_prepare and _commit.
+ */
+typedef struct
+{
+#ifdef __DOXYGEN__
+  uint8_t opaque[44];   /**< This is an opaque structure. */
+#else
+  uint8_t flags;        /**< Defined below */
+  uint8_t datalen;      /**< Number of valid words pointed to by data. */
+  uint32_t request[9];  /**< Request to be sent to the EPP or shim.  Note
+                             that this is smaller than the 11-word maximum
+                             request size, since some constant values are
+                             not saved in the context. */
+  uint32_t *data;       /**< Data to be sent to the EPP or shim via IDN. */
+#endif
+}
+netio_send_pkt_context_t;
+
+
+#ifndef __DOXYGEN__
+#define SEND_PKT_CTX_USE_EPP   1  /**< We're sending to an EPP. */
+#define SEND_PKT_CTX_SEND_CSUM 2  /**< Request includes a checksum. */
+#endif
+
+/**
+ * @brief Packet vector entry.
+ *
+ * @ingroup egress
+ *
+ * This data structure is used with netio_send_packet_vector() to send multiple
+ * packets with one NetIO call.  The structure should be initialized by
+ * calling netio_pkt_vector_set(), rather than by setting the fields
+ * directly.
+ *
+ * This structure is guaranteed to be a power of two in size, no
+ * bigger than one L2 cache line, and to be aligned modulo its size.
+ */
+typedef struct
+#ifndef __DOXYGEN__
+__attribute__((aligned(8)))
+#endif
+{
+  /** Reserved for use by the user application.  When initialized with
+   *  the netio_set_pkt_vector_entry() function, this field is guaranteed
+   *  to be visible to readers only after all other fields are already
+   *  visible.  This way it can be used as a valid flag or generation
+   *  counter. */
+  uint8_t user_data;
+
+  /* Structure members below this point should not be accessed directly by
+   * applications, as they may change in the future. */
+
+  /** Low 8 bits of the packet address to send.  The high bits are
+   *  acquired from the 'handle' field. */
+  uint8_t buffer_address_low;
+
+  /** Number of bytes to transmit. */
+  uint16_t size;
+
+  /** The raw handle from a netio_pkt_t.  If this is NETIO_PKT_HANDLE_NONE,
+   *  this vector entry will be skipped and no packet will be transmitted. */
+  netio_pkt_handle_t handle;
+}
+netio_pkt_vector_entry_t;
+
+
+/**
+ * @brief Initialize fields in a packet vector entry.
+ *
+ * @ingroup egress
+ *
+ * @param[out] v Pointer to the vector entry to be initialized.
+ * @param[in] pkt Packet to be transmitted when the vector entry is passed to
+ *        netio_send_packet_vector().  Note that the packet's attributes
+ *        (e.g., its L2 offset and length) are captured at the time this
+ *        routine is called; subsequent changes in those attributes will not
+ *        be reflected in the packet which is actually transmitted.
+ *        Changes in the packet's contents, however, will be so reflected.
+ *        If this is NULL, no packet will be transmitted.
+ * @param[in] user_data User data to be set in the vector entry.
+ *        This function guarantees that the "user_data" field will become
+ *        visible to a reader only after all other fields have become visible.
+ *        This allows a structure in a ring buffer to be written and read
+ *        by a polling reader without any locks or other synchronization.
+ */
+static __inline void
+netio_pkt_vector_set(volatile netio_pkt_vector_entry_t* v, netio_pkt_t* pkt,
+                     uint8_t user_data)
+{
+  if (pkt)
+  {
+    if (NETIO_PKT_IS_MINIMAL(pkt))
+    {
+      netio_pkt_minimal_metadata_t* mmd =
+        (netio_pkt_minimal_metadata_t*) &pkt->__metadata;
+      v->buffer_address_low = (uintptr_t) NETIO_PKT_L2_DATA_MM(mmd, pkt) & 0xFF;
+      v->size = NETIO_PKT_L2_LENGTH_MM(mmd, pkt);
+    }
+    else
+    {
+      netio_pkt_metadata_t* mda = &pkt->__metadata;
+      v->buffer_address_low = (uintptr_t) NETIO_PKT_L2_DATA_M(mda, pkt) & 0xFF;
+      v->size = NETIO_PKT_L2_LENGTH_M(mda, pkt);
+    }
+    v->handle.word = pkt->__packet.word;
+  }
+  else
+  {
+    v->handle.word = 0;   /* Set handle to NETIO_PKT_HANDLE_NONE. */
+  }
+
+  __asm__("" : : : "memory");
+
+  v->user_data = user_data;
+}
+
+
+/**
+ * Flags and structures for @ref netio_get() and @ref netio_set().
+ * @ingroup config
+ */
+
+/** @{ */
+/** Parameter class; addr is a NETIO_PARAM_xxx value. */
+#define NETIO_PARAM       0
+/** Interface MAC address. This address is only valid with @ref netio_get().
+ *  The value is a 6-byte MAC address.  Depending upon the overall system
+ *  design, a MAC address may or may not be available for each interface. */
+#define NETIO_PARAM_MAC        0
+
+/** Determine whether to suspend output on the receipt of pause frames.
+ *  If the value is nonzero, the I/O shim will suspend output when a pause
+ *  frame is received.  If the value is zero, pause frames will be ignored. */
+#define NETIO_PARAM_PAUSE_IN   1
+
+/** Determine whether to send pause frames if the I/O shim packet FIFOs are
+ *  nearly full.  If the value is zero, pause frames are not sent.  If
+ *  the value is nonzero, it is the delay value which will be sent in any
+ *  pause frames which are output, in units of 512 bit times. */
+#define NETIO_PARAM_PAUSE_OUT  2
+
+/** Jumbo frame support.  The value is a 4-byte integer.  If the value is
+ *  nonzero, the MAC will accept frames of up to 10240 bytes.  If the value
+ *  is zero, the MAC will only accept frames of up to 1544 bytes. */
+#define NETIO_PARAM_JUMBO      3
+
+/** I/O shim's overflow statistics register.  The value is two 16-bit integers.
+ *  The first 16-bit value (or the low 16 bits, if the value is treated as a
+ *  32-bit number) is the count of packets which were completely dropped and
+ *  not delivered by the shim.  The second 16-bit value (or the high 16 bits,
+ *  if the value is treated as a 32-bit number) is the count of packets
+ *  which were truncated and thus only partially delivered by the shim.  This
+ *  register is automatically reset to zero after it has been read.
+ */
+#define NETIO_PARAM_OVERFLOW   4
+
+/** IPP statistics.  This address is only valid with @ref netio_get().  The
+ *  value is a netio_stat_t structure.  Unlike the I/O shim statistics, the
+ *  IPP statistics are not all reset to zero on read; see the description
+ *  of the netio_stat_t for details. */
+#define NETIO_PARAM_STAT 5
+
+/** Possible link state.  The value is a combination of "NETIO_LINK_xxx"
+ *  flags.  With @ref netio_get(), this will indicate which flags are
+ *  actually supported by the hardware.
+ *
+ *  For historical reasons, specifying this value to netio_set() will have
+ *  the same behavior as using ::NETIO_PARAM_LINK_CONFIG, but this usage is
+ *  discouraged.
+ */
+#define NETIO_PARAM_LINK_POSSIBLE_STATE 6
+
+/** Link configuration. The value is a combination of "NETIO_LINK_xxx" flags.
+ *  With @ref netio_set(), this will attempt to immediately bring up the
+ *  link using whichever of the requested flags are supported by the
+ *  hardware, or take down the link if the flags are zero; if this is
+ *  not possible, an error will be returned.  Many programs will want
+ *  to use ::NETIO_PARAM_LINK_DESIRED_STATE instead.
+ *
+ *  For historical reasons, specifying this value to netio_get() will
+ *  have the same behavior as using ::NETIO_PARAM_LINK_POSSIBLE_STATE,
+ *  but this usage is discouraged.
+ */
+#define NETIO_PARAM_LINK_CONFIG NETIO_PARAM_LINK_POSSIBLE_STATE
+
+/** Current link state. This address is only valid with @ref netio_get().
+ *  The value is zero or more of the "NETIO_LINK_xxx" flags, ORed together.
+ *  If the link is down, the value ANDed with NETIO_LINK_SPEED will be
+ *  zero; if the link is up, the value ANDed with NETIO_LINK_SPEED will
+ *  result in exactly one of the NETIO_LINK_xxx values, indicating the
+ *  current speed. */
+#define NETIO_PARAM_LINK_CURRENT_STATE 7
+
+/** Variant symbol for current state, retained for compatibility with
+ *  pre-MDE-2.1 programs. */
+#define NETIO_PARAM_LINK_STATUS NETIO_PARAM_LINK_CURRENT_STATE
+
+/** Packet Coherence protocol. This address is only valid with @ref netio_get().
+ *  The value is nonzero if the interface is configured for cache-coherent DMA.
+ */
+#define NETIO_PARAM_COHERENT 8
+
+/** Desired link state. The value is a conbination of "NETIO_LINK_xxx"
+ *  flags, which specify the desired state for the link.  With @ref
+ *  netio_set(), this will, in the background, attempt to bring up the link
+ *  using whichever of the requested flags are reasonable, or take down the
+ *  link if the flags are zero.  The actual link up or down operation may
+ *  happen after this call completes.  If the link state changes in the
+ *  future, the system will continue to try to get back to the desired link
+ *  state; for instance, if the link is brought up successfully, and then
+ *  the network cable is disconnected, the link will go down.  However, the
+ *  desired state of the link is still up, so if the cable is reconnected,
+ *  the link will be brought up again.
+ *
+ *  With @ref netio_get(), this will indicate the desired state for the
+ *  link, as set with a previous netio_set() call, or implicitly by a
+ *  netio_input_register() or netio_input_unregister() operation.  This may
+ *  not reflect the current state of the link; to get that, use
+ *  ::NETIO_PARAM_LINK_CURRENT_STATE. */
+#define NETIO_PARAM_LINK_DESIRED_STATE 9
+
+/** NetIO statistics structure.  Retrieved using the ::NETIO_PARAM_STAT
+ *  address passed to @ref netio_get(). */
+typedef struct
+{
+  /** Number of packets which have been received by the IPP and forwarded
+   *  to a tile's receive queue for processing.  This value wraps at its
+   *  maximum, and is not cleared upon read. */
+  uint32_t packets_received;
+
+  /** Number of packets which have been dropped by the IPP, because they could
+   *  not be received, or could not be forwarded to a tile.  The former happens
+   *  when the IPP does not have a free packet buffer of suitable size for an
+   *  incoming frame.  The latter happens when all potential destination tiles
+   *  for a packet, as defined by the group, bucket, and queue configuration,
+   *  have full receive queues.   This value wraps at its maximum, and is not
+   *  cleared upon read. */
+  uint32_t packets_dropped;
+
+  /*
+   * Note: the #defines after each of the following four one-byte values
+   * denote their location within the third word of the netio_stat_t.  They
+   * are intended for use only by the IPP implementation and are thus omitted
+   * from the Doxygen output.
+   */
+
+  /** Number of packets dropped because no worker was able to accept a new
+   *  packet.  This value saturates at its maximum, and is cleared upon
+   *  read. */
+  uint8_t drops_no_worker;
+#ifndef __DOXYGEN__
+#define NETIO_STAT_DROPS_NO_WORKER   0
+#endif
+
+  /** Number of packets dropped because no small buffers were available.
+   *  This value saturates at its maximum, and is cleared upon read. */
+  uint8_t drops_no_smallbuf;
+#ifndef __DOXYGEN__
+#define NETIO_STAT_DROPS_NO_SMALLBUF 1
+#endif
+
+  /** Number of packets dropped because no large buffers were available.
+   *  This value saturates at its maximum, and is cleared upon read. */
+  uint8_t drops_no_largebuf;
+#ifndef __DOXYGEN__
+#define NETIO_STAT_DROPS_NO_LARGEBUF 2
+#endif
+
+  /** Number of packets dropped because no jumbo buffers were available.
+   *  This value saturates at its maximum, and is cleared upon read. */
+  uint8_t drops_no_jumbobuf;
+#ifndef __DOXYGEN__
+#define NETIO_STAT_DROPS_NO_JUMBOBUF 3
+#endif
+}
+netio_stat_t;
+
+
+/** Link can run, should run, or is running at 10 Mbps. */
+#define NETIO_LINK_10M         0x01
+
+/** Link can run, should run, or is running at 100 Mbps. */
+#define NETIO_LINK_100M        0x02
+
+/** Link can run, should run, or is running at 1 Gbps. */
+#define NETIO_LINK_1G          0x04
+
+/** Link can run, should run, or is running at 10 Gbps. */
+#define NETIO_LINK_10G         0x08
+
+/** Link should run at the highest speed supported by the link and by
+ *  the device connected to the link.  Only usable as a value for
+ *  the link's desired state; never returned as a value for the current
+ *  or possible states. */
+#define NETIO_LINK_ANYSPEED    0x10
+
+/** All legal link speeds. */
+#define NETIO_LINK_SPEED  (NETIO_LINK_10M  | \
+                           NETIO_LINK_100M | \
+                           NETIO_LINK_1G   | \
+                           NETIO_LINK_10G  | \
+                           NETIO_LINK_ANYSPEED)
+
+
+/** MAC register class.  Addr is a register offset within the MAC.
+ *  Registers within the XGbE and GbE MACs are documented in the Tile
+ *  Processor I/O Device Guide (UG104). MAC registers start at address
+ *  0x4000, and do not include the MAC_INTERFACE registers. */
+#define NETIO_MAC             1
+
+/** MDIO register class (IEEE 802.3 clause 22 format).  Addr is the "addr"
+ *  member of a netio_mdio_addr_t structure. */
+#define NETIO_MDIO            2
+
+/** MDIO register class (IEEE 802.3 clause 45 format).  Addr is the "addr"
+ *  member of a netio_mdio_addr_t structure. */
+#define NETIO_MDIO_CLAUSE45   3
+
+/** NetIO MDIO address type.  Retrieved or provided using the ::NETIO_MDIO
+ *  address passed to @ref netio_get() or @ref netio_set(). */
+typedef union
+{
+  struct
+  {
+    unsigned int reg:16;  /**< MDIO register offset.  For clause 22 access,
+                               must be less than 32. */
+    unsigned int phy:5;   /**< Which MDIO PHY to access. */
+    unsigned int dev:5;   /**< Which MDIO device to access within that PHY.
+                               Applicable for clause 45 access only; ignored
+                               for clause 22 access. */
+  }
+  bits;                   /**< Container for bitfields. */
+  uint64_t addr;          /**< Value to pass to @ref netio_get() or
+                           *   @ref netio_set(). */
+}
+netio_mdio_addr_t;
+
+/** @} */
+
+#endif /* __NETIO_INTF_H__ */
diff --git a/arch/tile/kernel/Makefile b/arch/tile/kernel/Makefile
index 112b1e248f05..b4c8e8ec45dc 100644
--- a/arch/tile/kernel/Makefile
+++ b/arch/tile/kernel/Makefile
@@ -15,3 +15,4 @@ obj-$(CONFIG_SMP)		+= smpboot.o smp.o tlb.o
 obj-$(CONFIG_MODULES)           += module.o
 obj-$(CONFIG_EARLY_PRINTK)      += early_printk.o
 obj-$(CONFIG_KEXEC)             += machine_kexec.o relocate_kernel.o
+obj-$(CONFIG_PCI)               += pci.o
diff --git a/arch/tile/kernel/compat.c b/arch/tile/kernel/compat.c
index 67617a05e602..dbc213adf5e1 100644
--- a/arch/tile/kernel/compat.c
+++ b/arch/tile/kernel/compat.c
@@ -21,7 +21,6 @@
 #include <linux/kdev_t.h>
 #include <linux/fs.h>
 #include <linux/fcntl.h>
-#include <linux/smp_lock.h>
 #include <linux/uaccess.h>
 #include <linux/signal.h>
 #include <asm/syscalls.h>
diff --git a/arch/tile/kernel/compat_signal.c b/arch/tile/kernel/compat_signal.c
index fb64b99959d4..543d6a33aa26 100644
--- a/arch/tile/kernel/compat_signal.c
+++ b/arch/tile/kernel/compat_signal.c
@@ -15,7 +15,6 @@
 #include <linux/sched.h>
 #include <linux/mm.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/kernel.h>
 #include <linux/signal.h>
 #include <linux/errno.h>
diff --git a/arch/tile/kernel/pci.c b/arch/tile/kernel/pci.c
new file mode 100644
index 000000000000..a1ee25be9ad9
--- /dev/null
+++ b/arch/tile/kernel/pci.c
@@ -0,0 +1,621 @@
+/*
+ * Copyright 2010 Tilera Corporation. All Rights Reserved.
+ *
+ *   This program is free software; you can redistribute it and/or
+ *   modify it under the terms of the GNU General Public License
+ *   as published by the Free Software Foundation, version 2.
+ *
+ *   This program is distributed in the hope that it will be useful, but
+ *   WITHOUT ANY WARRANTY; without even the implied warranty of
+ *   MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
+ *   NON INFRINGEMENT.  See the GNU General Public License for
+ *   more details.
+ */
+
+#include <linux/kernel.h>
+#include <linux/pci.h>
+#include <linux/delay.h>
+#include <linux/string.h>
+#include <linux/init.h>
+#include <linux/capability.h>
+#include <linux/sched.h>
+#include <linux/errno.h>
+#include <linux/bootmem.h>
+#include <linux/irq.h>
+#include <linux/io.h>
+#include <linux/uaccess.h>
+
+#include <asm/processor.h>
+#include <asm/sections.h>
+#include <asm/byteorder.h>
+#include <asm/hv_driver.h>
+#include <hv/drv_pcie_rc_intf.h>
+
+
+/*
+ * Initialization flow and process
+ * -------------------------------
+ *
+ * This files containes the routines to search for PCI buses,
+ * enumerate the buses, and configure any attached devices.
+ *
+ * There are two entry points here:
+ * 1) tile_pci_init
+ *    This sets up the pci_controller structs, and opens the
+ *    FDs to the hypervisor.  This is called from setup_arch() early
+ *    in the boot process.
+ * 2) pcibios_init
+ *    This probes the PCI bus(es) for any attached hardware.  It's
+ *    called by subsys_initcall.  All of the real work is done by the
+ *    generic Linux PCI layer.
+ *
+ */
+
+/*
+ * This flag tells if the platform is TILEmpower that needs
+ * special configuration for the PLX switch chip.
+ */
+int __write_once tile_plx_gen1;
+
+static struct pci_controller controllers[TILE_NUM_PCIE];
+static int num_controllers;
+
+static struct pci_ops tile_cfg_ops;
+
+
+/*
+ * We don't need to worry about the alignment of resources.
+ */
+resource_size_t pcibios_align_resource(void *data, const struct resource *res,
+                            resource_size_t size, resource_size_t align)
+{
+        return res->start;
+}
+EXPORT_SYMBOL(pcibios_align_resource);
+
+/*
+ * Open a FD to the hypervisor PCI device.
+ *
+ * controller_id is the controller number, config type is 0 or 1 for
+ * config0 or config1 operations.
+ */
+static int __init tile_pcie_open(int controller_id, int config_type)
+{
+        char filename[32];
+        int fd;
+
+        sprintf(filename, "pcie/%d/config%d", controller_id, config_type);
+
+        fd = hv_dev_open((HV_VirtAddr)filename, 0);
+
+        return fd;
+}
+
+
+/*
+ * Get the IRQ numbers from the HV and set up the handlers for them.
+ */
+static int __init tile_init_irqs(int controller_id,
+                                 struct pci_controller *controller)
+{
+        char filename[32];
+        int fd;
+        int ret;
+        int x;
+        struct pcie_rc_config rc_config;
+
+        sprintf(filename, "pcie/%d/ctl", controller_id);
+        fd = hv_dev_open((HV_VirtAddr)filename, 0);
+        if (fd < 0) {
+                pr_err("PCI: hv_dev_open(%s) failed\n", filename);
+                return -1;
+        }
+        ret = hv_dev_pread(fd, 0, (HV_VirtAddr)(&rc_config),
+                           sizeof(rc_config), PCIE_RC_CONFIG_MASK_OFF);
+        hv_dev_close(fd);
+        if (ret != sizeof(rc_config)) {
+                pr_err("PCI: wanted %zd bytes, got %d\n",
+                       sizeof(rc_config), ret);
+                return -1;
+        }
+        /* Record irq_base so that we can map INTx to IRQ # later. */
+        controller->irq_base = rc_config.intr;
+
+        for (x = 0; x < 4; x++)
+                tile_irq_activate(rc_config.intr + x,
+                                  TILE_IRQ_HW_CLEAR);
+
+        if (rc_config.plx_gen1)
+                controller->plx_gen1 = 1;
+
+        return 0;
+}
+
+/*
+ * First initialization entry point, called from setup_arch().
+ *
+ * Find valid controllers and fill in pci_controller structs for each
+ * of them.
+ *
+ * Returns the number of controllers discovered.
+ */
+int __init tile_pci_init(void)
+{
+        int i;
+
+        pr_info("PCI: Searching for controllers...\n");
+
+        /* Do any configuration we need before using the PCIe */
+
+        for (i = 0; i < TILE_NUM_PCIE; i++) {
+                int hv_cfg_fd0 = -1;
+                int hv_cfg_fd1 = -1;
+                int hv_mem_fd = -1;
+                char name[32];
+                struct pci_controller *controller;
+
+                /*
+                 * Open the fd to the HV.  If it fails then this
+                 * device doesn't exist.
+                 */
+                hv_cfg_fd0 = tile_pcie_open(i, 0);
+                if (hv_cfg_fd0 < 0)
+                        continue;
+                hv_cfg_fd1 = tile_pcie_open(i, 1);
+                if (hv_cfg_fd1 < 0) {
+                        pr_err("PCI: Couldn't open config fd to HV "
+                            "for controller %d\n", i);
+                        goto err_cont;
+                }
+
+                sprintf(name, "pcie/%d/mem", i);
+                hv_mem_fd = hv_dev_open((HV_VirtAddr)name, 0);
+                if (hv_mem_fd < 0) {
+                        pr_err("PCI: Could not open mem fd to HV!\n");
+                        goto err_cont;
+                }
+
+                pr_info("PCI: Found PCI controller #%d\n", i);
+
+                controller = &controllers[num_controllers];
+
+                if (tile_init_irqs(i, controller)) {
+                        pr_err("PCI: Could not initialize "
+                               "IRQs, aborting.\n");
+                        goto err_cont;
+                }
+
+                controller->index = num_controllers;
+                controller->hv_cfg_fd[0] = hv_cfg_fd0;
+                controller->hv_cfg_fd[1] = hv_cfg_fd1;
+                controller->hv_mem_fd = hv_mem_fd;
+                controller->first_busno = 0;
+                controller->last_busno = 0xff;
+                controller->ops = &tile_cfg_ops;
+
+                num_controllers++;
+                continue;
+
+err_cont:
+                if (hv_cfg_fd0 >= 0)
+                        hv_dev_close(hv_cfg_fd0);
+                if (hv_cfg_fd1 >= 0)
+                        hv_dev_close(hv_cfg_fd1);
+                if (hv_mem_fd >= 0)
+                        hv_dev_close(hv_mem_fd);
+                continue;
+        }
+
+        /*
+         * Before using the PCIe, see if we need to do any platform-specific
+         * configuration, such as the PLX switch Gen 1 issue on TILEmpower.
+         */
+        for (i = 0; i < num_controllers; i++) {
+                struct pci_controller *controller = &controllers[i];
+
+                if (controller->plx_gen1)
+                        tile_plx_gen1 = 1;
+        }
+
+        return num_controllers;
+}
+
+/*
+ * (pin - 1) converts from the PCI standard's [1:4] convention to
+ * a normal [0:3] range.
+ */
+static int tile_map_irq(struct pci_dev *dev, u8 slot, u8 pin)
+{
+        struct pci_controller *controller =
+                (struct pci_controller *)dev->sysdata;
+        return (pin - 1) + controller->irq_base;
+}
+
+
+static void __init fixup_read_and_payload_sizes(void)
+{
+        struct pci_dev *dev = NULL;
+        int smallest_max_payload = 0x1; /* Tile maxes out at 256 bytes. */
+        int max_read_size = 0x2; /* Limit to 512 byte reads. */
+        u16 new_values;
+
+        /* Scan for the smallest maximum payload size. */
+        while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
+                int pcie_caps_offset;
+                u32 devcap;
+                int max_payload;
+
+                pcie_caps_offset = pci_find_capability(dev, PCI_CAP_ID_EXP);
+                if (pcie_caps_offset == 0)
+                        continue;
+
+                pci_read_config_dword(dev, pcie_caps_offset + PCI_EXP_DEVCAP,
+                                      &devcap);
+                max_payload = devcap & PCI_EXP_DEVCAP_PAYLOAD;
+                if (max_payload < smallest_max_payload)
+                        smallest_max_payload = max_payload;
+        }
+
+        /* Now, set the max_payload_size for all devices to that value. */
+        new_values = (max_read_size << 12) | (smallest_max_payload << 5);
+        while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
+                int pcie_caps_offset;
+                u16 devctl;
+
+                pcie_caps_offset = pci_find_capability(dev, PCI_CAP_ID_EXP);
+                if (pcie_caps_offset == 0)
+                        continue;
+
+                pci_read_config_word(dev, pcie_caps_offset + PCI_EXP_DEVCTL,
+                                     &devctl);
+                devctl &= ~(PCI_EXP_DEVCTL_PAYLOAD | PCI_EXP_DEVCTL_READRQ);
+                devctl |= new_values;
+                pci_write_config_word(dev, pcie_caps_offset + PCI_EXP_DEVCTL,
+                                      devctl);
+        }
+}
+
+
+/*
+ * Second PCI initialization entry point, called by subsys_initcall.
+ *
+ * The controllers have been set up by the time we get here, by a call to
+ * tile_pci_init.
+ */
+static int __init pcibios_init(void)
+{
+        int i;
+
+        pr_info("PCI: Probing PCI hardware\n");
+
+        /*
+         * Delay a bit in case devices aren't ready.  Some devices are
+         * known to require at least 20ms here, but we use a more
+         * conservative value.
+         */
+        mdelay(250);
+
+        /* Scan all of the recorded PCI controllers.  */
+        for (i = 0; i < num_controllers; i++) {
+                struct pci_controller *controller = &controllers[i];
+                struct pci_bus *bus;
+
+                pr_info("PCI: initializing controller #%d\n", i);
+
+                /*
+                 * This comes from the generic Linux PCI driver.
+                 *
+                 * It reads the PCI tree for this bus into the Linux
+                 * data structures.
+                 *
+                 * This is inlined in linux/pci.h and calls into
+                 * pci_scan_bus_parented() in probe.c.
+                 */
+                bus = pci_scan_bus(0, controller->ops, controller);
+                controller->root_bus = bus;
+                controller->last_busno = bus->subordinate;
+
+        }
+
+        /* Do machine dependent PCI interrupt routing */
+        pci_fixup_irqs(pci_common_swizzle, tile_map_irq);
+
+        /*
+         * This comes from the generic Linux PCI driver.
+         *
+         * It allocates all of the resources (I/O memory, etc)
+         * associated with the devices read in above.
+         */
+
+        pci_assign_unassigned_resources();
+
+        /* Configure the max_read_size and max_payload_size values. */
+        fixup_read_and_payload_sizes();
+
+        /* Record the I/O resources in the PCI controller structure. */
+        for (i = 0; i < num_controllers; i++) {
+                struct pci_bus *root_bus = controllers[i].root_bus;
+                struct pci_bus *next_bus;
+                struct pci_dev *dev;
+
+                list_for_each_entry(dev, &root_bus->devices, bus_list) {
+                        /* Find the PCI host controller, ie. the 1st bridge. */
+                        if ((dev->class >> 8) == PCI_CLASS_BRIDGE_PCI &&
+                                (PCI_SLOT(dev->devfn) == 0)) {
+                                next_bus = dev->subordinate;
+                                controllers[i].mem_resources[0] =
+                                        *next_bus->resource[0];
+                                controllers[i].mem_resources[1] =
+                                         *next_bus->resource[1];
+                                controllers[i].mem_resources[2] =
+                                         *next_bus->resource[2];
+
+                                break;
+                        }
+                }
+
+        }
+
+        return 0;
+}
+subsys_initcall(pcibios_init);
+
+/*
+ * No bus fixups needed.
+ */
+void __devinit pcibios_fixup_bus(struct pci_bus *bus)
+{
+        /* Nothing needs to be done. */
+}
+
+/*
+ * This can be called from the generic PCI layer, but doesn't need to
+ * do anything.
+ */
+char __devinit *pcibios_setup(char *str)
+{
+        /* Nothing needs to be done. */
+        return str;
+}
+
+/*
+ * This is called from the generic Linux layer.
+ */
+void __init pcibios_update_irq(struct pci_dev *dev, int irq)
+{
+        pci_write_config_byte(dev, PCI_INTERRUPT_LINE, irq);
+}
+
+/*
+ * Enable memory and/or address decoding, as appropriate, for the
+ * device described by the 'dev' struct.
+ *
+ * This is called from the generic PCI layer, and can be called
+ * for bridges or endpoints.
+ */
+int pcibios_enable_device(struct pci_dev *dev, int mask)
+{
+        u16 cmd, old_cmd;
+        u8 header_type;
+        int i;
+        struct resource *r;
+
+        pci_read_config_byte(dev, PCI_HEADER_TYPE, &header_type);
+
+        pci_read_config_word(dev, PCI_COMMAND, &cmd);
+        old_cmd = cmd;
+        if ((header_type & 0x7F) == PCI_HEADER_TYPE_BRIDGE) {
+                /*
+                 * For bridges, we enable both memory and I/O decoding
+                 * in call cases.
+                 */
+                cmd |= PCI_COMMAND_IO;
+                cmd |= PCI_COMMAND_MEMORY;
+        } else {
+                /*
+                 * For endpoints, we enable memory and/or I/O decoding
+                 * only if they have a memory resource of that type.
+                 */
+                for (i = 0; i < 6; i++) {
+                        r = &dev->resource[i];
+                        if (r->flags & IORESOURCE_UNSET) {
+                                pr_err("PCI: Device %s not available "
+                                       "because of resource collisions\n",
+                                       pci_name(dev));
+                                return -EINVAL;
+                        }
+                        if (r->flags & IORESOURCE_IO)
+                                cmd |= PCI_COMMAND_IO;
+                        if (r->flags & IORESOURCE_MEM)
+                                cmd |= PCI_COMMAND_MEMORY;
+                }
+        }
+
+        /*
+         * We only write the command if it changed.
+         */
+        if (cmd != old_cmd)
+                pci_write_config_word(dev, PCI_COMMAND, cmd);
+        return 0;
+}
+
+void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max)
+{
+        unsigned long start = pci_resource_start(dev, bar);
+        unsigned long len = pci_resource_len(dev, bar);
+        unsigned long flags = pci_resource_flags(dev, bar);
+
+        if (!len)
+                return NULL;
+        if (max && len > max)
+                len = max;
+
+        if (!(flags & IORESOURCE_MEM)) {
+                pr_info("PCI: Trying to map invalid resource %#lx\n", flags);
+                start = 0;
+        }
+
+        return (void __iomem *)start;
+}
+EXPORT_SYMBOL(pci_iomap);
+
+
+/****************************************************************
+ *
+ * Tile PCI config space read/write routines
+ *
+ ****************************************************************/
+
+/*
+ * These are the normal read and write ops
+ * These are expanded with macros from  pci_bus_read_config_byte() etc.
+ *
+ * devfn is the combined PCI slot & function.
+ *
+ * offset is in bytes, from the start of config space for the
+ * specified bus & slot.
+ */
+
+static int __devinit tile_cfg_read(struct pci_bus *bus,
+                                   unsigned int devfn,
+                                   int offset,
+                                   int size,
+                                   u32 *val)
+{
+        struct pci_controller *controller = bus->sysdata;
+        int busnum = bus->number & 0xff;
+        int slot = (devfn >> 3) & 0x1f;
+        int function = devfn & 0x7;
+        u32 addr;
+        int config_mode = 1;
+
+        /*
+         * There is no bridge between the Tile and bus 0, so we
+         * use config0 to talk to bus 0.
+         *
+         * If we're talking to a bus other than zero then we
+         * must have found a bridge.
+         */
+        if (busnum == 0) {
+                /*
+                 * We fake an empty slot for (busnum == 0) && (slot > 0),
+                 * since there is only one slot on bus 0.
+                 */
+                if (slot) {
+                        *val = 0xFFFFFFFF;
+                        return 0;
+                }
+                config_mode = 0;
+        }
+
+        addr = busnum << 20;            /* Bus in 27:20 */
+        addr |= slot << 15;             /* Slot (device) in 19:15 */
+        addr |= function << 12;         /* Function is in 14:12 */
+        addr |= (offset & 0xFFF);       /* byte address in 0:11 */
+
+        return hv_dev_pread(controller->hv_cfg_fd[config_mode], 0,
+                            (HV_VirtAddr)(val), size, addr);
+}
+
+
+/*
+ * See tile_cfg_read() for relevent comments.
+ * Note that "val" is the value to write, not a pointer to that value.
+ */
+static int __devinit tile_cfg_write(struct pci_bus *bus,
+                                    unsigned int devfn,
+                                    int offset,
+                                    int size,
+                                    u32 val)
+{
+        struct pci_controller *controller = bus->sysdata;
+        int busnum = bus->number & 0xff;
+        int slot = (devfn >> 3) & 0x1f;
+        int function = devfn & 0x7;
+        u32 addr;
+        int config_mode = 1;
+        HV_VirtAddr valp = (HV_VirtAddr)&val;
+
+        /*
+         * For bus 0 slot 0 we use config 0 accesses.
+         */
+        if (busnum == 0) {
+                /*
+                 * We fake an empty slot for (busnum == 0) && (slot > 0),
+                 * since there is only one slot on bus 0.
+                 */
+                if (slot)
+                        return 0;
+                config_mode = 0;
+        }
+
+        addr = busnum << 20;            /* Bus in 27:20 */
+        addr |= slot << 15;             /* Slot (device) in 19:15 */
+        addr |= function << 12;         /* Function is in 14:12 */
+        addr |= (offset & 0xFFF);       /* byte address in 0:11 */
+
+#ifdef __BIG_ENDIAN
+        /* Point to the correct part of the 32-bit "val". */
+        valp += 4 - size;
+#endif
+
+        return hv_dev_pwrite(controller->hv_cfg_fd[config_mode], 0,
+                             valp, size, addr);
+}
+
+
+static struct pci_ops tile_cfg_ops = {
+        .read =         tile_cfg_read,
+        .write =        tile_cfg_write,
+};
+
+
+/*
+ * In the following, each PCI controller's mem_resources[1]
+ * represents its (non-prefetchable) PCI memory resource.
+ * mem_resources[0] and mem_resources[2] refer to its PCI I/O and
+ * prefetchable PCI memory resources, respectively.
+ * For more details, see pci_setup_bridge() in setup-bus.c.
+ * By comparing the target PCI memory address against the
+ * end address of controller 0, we can determine the controller
+ * that should accept the PCI memory access.
+ */
+#define TILE_READ(size, type)                                           \
+type _tile_read##size(unsigned long addr)                               \
+{                                                                       \
+        type val;                                                       \
+        int idx = 0;                                                    \
+        if (addr > controllers[0].mem_resources[1].end &&               \
+            addr > controllers[0].mem_resources[2].end)                 \
+                idx = 1;                                                \
+        if (hv_dev_pread(controllers[idx].hv_mem_fd, 0,                 \
+                         (HV_VirtAddr)(&val), sizeof(type), addr))      \
+                pr_err("PCI: read %zd bytes at 0x%lX failed\n",         \
+                       sizeof(type), addr);                             \
+        return val;                                                     \
+}                                                                       \
+EXPORT_SYMBOL(_tile_read##size)
+
+TILE_READ(b, u8);
+TILE_READ(w, u16);
+TILE_READ(l, u32);
+TILE_READ(q, u64);
+
+#define TILE_WRITE(size, type)                                          \
+void _tile_write##size(type val, unsigned long addr)                    \
+{                                                                       \
+        int idx = 0;                                                    \
+        if (addr > controllers[0].mem_resources[1].end &&               \
+            addr > controllers[0].mem_resources[2].end)                 \
+                idx = 1;                                                \
+        if (hv_dev_pwrite(controllers[idx].hv_mem_fd, 0,                \
+                          (HV_VirtAddr)(&val), sizeof(type), addr))     \
+                pr_err("PCI: write %zd bytes at 0x%lX failed\n",        \
+                       sizeof(type), addr);                             \
+}                                                                       \
+EXPORT_SYMBOL(_tile_write##size)
+
+TILE_WRITE(b, u8);
+TILE_WRITE(w, u16);
+TILE_WRITE(l, u32);
+TILE_WRITE(q, u64);
diff --git a/arch/tile/kernel/setup.c b/arch/tile/kernel/setup.c
index fb0b3cbeae14..f18573643ed1 100644
--- a/arch/tile/kernel/setup.c
+++ b/arch/tile/kernel/setup.c
@@ -840,7 +840,7 @@ static int __init topology_init(void)
         for_each_online_node(i)
                 register_one_node(i);
 
-        for_each_present_cpu(i)
+        for (i = 0; i < smp_height * smp_width; ++i)
                 register_cpu(&cpu_devices[i], i);
 
         return 0;
diff --git a/arch/tile/kernel/signal.c b/arch/tile/kernel/signal.c
index 687719d4abd1..757407e36696 100644
--- a/arch/tile/kernel/signal.c
+++ b/arch/tile/kernel/signal.c
@@ -16,7 +16,6 @@
 #include <linux/sched.h>
 #include <linux/mm.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/kernel.h>
 #include <linux/signal.h>
 #include <linux/errno.h>
diff --git a/arch/tile/kernel/smpboot.c b/arch/tile/kernel/smpboot.c
index 74d62d098edf..b949edcec200 100644
--- a/arch/tile/kernel/smpboot.c
+++ b/arch/tile/kernel/smpboot.c
@@ -18,7 +18,6 @@
 #include <linux/mm.h>
 #include <linux/sched.h>
 #include <linux/kernel_stat.h>
-#include <linux/smp_lock.h>
 #include <linux/bootmem.h>
 #include <linux/notifier.h>
 #include <linux/cpu.h>
diff --git a/arch/tile/kernel/sys.c b/arch/tile/kernel/sys.c
index 7e764669a022..e2187d24a9b4 100644
--- a/arch/tile/kernel/sys.c
+++ b/arch/tile/kernel/sys.c
@@ -20,7 +20,6 @@
 #include <linux/sched.h>
 #include <linux/mm.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/syscalls.h>
 #include <linux/mman.h>
 #include <linux/file.h>
diff --git a/arch/tile/lib/memchr_32.c b/arch/tile/lib/memchr_32.c
index 6235283b4859..cc3d9badf030 100644
--- a/arch/tile/lib/memchr_32.c
+++ b/arch/tile/lib/memchr_32.c
@@ -18,12 +18,24 @@
 
 void *memchr(const void *s, int c, size_t n)
 {
+        const uint32_t *last_word_ptr;
+        const uint32_t *p;
+        const char *last_byte_ptr;
+        uintptr_t s_int;
+        uint32_t goal, before_mask, v, bits;
+        char *ret;
+
+        if (__builtin_expect(n == 0, 0)) {
+                /* Don't dereference any memory if the array is empty. */
+                return NULL;
+        }
+
         /* Get an aligned pointer. */
-        const uintptr_t s_int = (uintptr_t) s;
-        const uint32_t *p = (const uint32_t *)(s_int & -4);
+        s_int = (uintptr_t) s;
+        p = (const uint32_t *)(s_int & -4);
 
         /* Create four copies of the byte for which we are looking. */
-        const uint32_t goal = 0x01010101 * (uint8_t) c;
+        goal = 0x01010101 * (uint8_t) c;
 
         /* Read the first word, but munge it so that bytes before the array
          * will not match goal.
@@ -31,23 +43,14 @@ void *memchr(const void *s, int c, size_t n)
          * Note that this shift count expression works because we know
          * shift counts are taken mod 32.
          */
-        const uint32_t before_mask = (1 << (s_int << 3)) - 1;
-        uint32_t v = (*p | before_mask) ^ (goal & before_mask);
+        before_mask = (1 << (s_int << 3)) - 1;
+        v = (*p | before_mask) ^ (goal & before_mask);
 
         /* Compute the address of the last byte. */
-        const char *const last_byte_ptr = (const char *)s + n - 1;
+        last_byte_ptr = (const char *)s + n - 1;
 
         /* Compute the address of the word containing the last byte. */
-        const uint32_t *const last_word_ptr =
-            (const uint32_t *)((uintptr_t) last_byte_ptr & -4);
-
-        uint32_t bits;
-        char *ret;
-
-        if (__builtin_expect(n == 0, 0)) {
-                /* Don't dereference any memory if the array is empty. */
-                return NULL;
-        }
+        last_word_ptr = (const uint32_t *)((uintptr_t) last_byte_ptr & -4);
 
         while ((bits = __insn_seqb(v, goal)) == 0) {
                 if (__builtin_expect(p == last_word_ptr, 0)) {
diff --git a/arch/tile/lib/spinlock_32.c b/arch/tile/lib/spinlock_32.c
index 485e24d62c6b..5cd1c4004eca 100644
--- a/arch/tile/lib/spinlock_32.c
+++ b/arch/tile/lib/spinlock_32.c
@@ -167,23 +167,30 @@ void arch_write_lock_slow(arch_rwlock_t *rwlock, u32 val)
          * when we compare them.
          */
         u32 my_ticket_;
+        u32 iterations = 0;
 
-        /* Take out the next ticket; this will also stop would-be readers. */
-        if (val & 1)
-                val = get_rwlock(rwlock);
-        rwlock->lock = __insn_addb(val, 1 << WR_NEXT_SHIFT);
+        /*
+         * Wait until there are no readers, then bump up the next
+         * field and capture the ticket value.
+         */
+        for (;;) {
+                if (!(val & 1)) {
+                        if ((val >> RD_COUNT_SHIFT) == 0)
+                                break;
+                        rwlock->lock = val;
+                }
+                delay_backoff(iterations++);
+                val = __insn_tns((int *)&rwlock->lock);
+        }
 
-        /* Extract my ticket value from the original word. */
+        /* Take out the next ticket and extract my ticket value. */
+        rwlock->lock = __insn_addb(val, 1 << WR_NEXT_SHIFT);
         my_ticket_ = val >> WR_NEXT_SHIFT;
 
-        /*
-         * Wait until the "current" field matches our ticket, and
-         * there are no remaining readers.
-         */
+        /* Wait until the "current" field matches our ticket. */
         for (;;) {
                 u32 curr_ = val >> WR_CURR_SHIFT;
-                u32 readers = val >> RD_COUNT_SHIFT;
-                u32 delta = ((my_ticket_ - curr_) & WR_MASK) + !!readers;
+                u32 delta = ((my_ticket_ - curr_) & WR_MASK);
                 if (likely(delta == 0))
                         break;
 
diff --git a/arch/tile/mm/fault.c b/arch/tile/mm/fault.c
index f295b4ac941d..dcebfc831cd6 100644
--- a/arch/tile/mm/fault.c
+++ b/arch/tile/mm/fault.c
@@ -24,7 +24,6 @@
 #include <linux/mman.h>
 #include <linux/mm.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/interrupt.h>
 #include <linux/init.h>
 #include <linux/tty.h>
diff --git a/arch/tile/mm/hugetlbpage.c b/arch/tile/mm/hugetlbpage.c
index 24688b697a8d..201a582c4137 100644
--- a/arch/tile/mm/hugetlbpage.c
+++ b/arch/tile/mm/hugetlbpage.c
@@ -21,7 +21,6 @@
 #include <linux/mm.h>
 #include <linux/hugetlb.h>
 #include <linux/pagemap.h>
-#include <linux/smp_lock.h>
 #include <linux/slab.h>
 #include <linux/err.h>
 #include <linux/sysctl.h>
diff --git a/arch/um/drivers/line.c b/arch/um/drivers/line.c
index 7f7338c90784..1664cce7b0ac 100644
--- a/arch/um/drivers/line.c
+++ b/arch/um/drivers/line.c
@@ -727,6 +727,9 @@ struct winch {
 
 static void free_winch(struct winch *winch, int free_irq_ok)
 {
+        if (free_irq_ok)
+                free_irq(WINCH_IRQ, winch);
+
         list_del(&winch->list);
 
         if (winch->pid != -1)
@@ -735,8 +738,6 @@ static void free_winch(struct winch *winch, int free_irq_ok)
                 os_close_file(winch->fd);
         if (winch->stack != 0)
                 free_stack(winch->stack, 0);
-        if (free_irq_ok)
-                free_irq(WINCH_IRQ, winch);
         kfree(winch);
 }
 
diff --git a/arch/um/kernel/exec.c b/arch/um/kernel/exec.c
index 340268be00b5..09bd7b585726 100644
--- a/arch/um/kernel/exec.c
+++ b/arch/um/kernel/exec.c
@@ -5,7 +5,6 @@
 
 #include "linux/stddef.h"
 #include "linux/fs.h"
-#include "linux/smp_lock.h"
 #include "linux/ptrace.h"
 #include "linux/sched.h"
 #include "linux/slab.h"
diff --git a/arch/x86/ia32/sys_ia32.c b/arch/x86/ia32/sys_ia32.c
index 849813f398e7..5852519b2d0f 100644
--- a/arch/x86/ia32/sys_ia32.c
+++ b/arch/x86/ia32/sys_ia32.c
@@ -28,7 +28,6 @@
 #include <linux/syscalls.h>
 #include <linux/times.h>
 #include <linux/utsname.h>
-#include <linux/smp_lock.h>
 #include <linux/mm.h>
 #include <linux/uio.h>
 #include <linux/poll.h>
diff --git a/arch/x86/include/asm/fixmap.h b/arch/x86/include/asm/fixmap.h
index 4d293dced62f..9479a037419f 100644
--- a/arch/x86/include/asm/fixmap.h
+++ b/arch/x86/include/asm/fixmap.h
@@ -216,8 +216,8 @@ static inline unsigned long virt_to_fix(const unsigned long vaddr)
 }
 
 /* Return an pointer with offset calculated */
-static inline unsigned long __set_fixmap_offset(enum fixed_addresses idx,
-                                phys_addr_t phys, pgprot_t flags)
+static __always_inline unsigned long
+__set_fixmap_offset(enum fixed_addresses idx, phys_addr_t phys, pgprot_t flags)
 {
         __set_fixmap(idx, phys, flags);
         return fix_to_virt(idx) + (phys & (PAGE_SIZE - 1));
diff --git a/arch/x86/include/asm/xen/interface.h b/arch/x86/include/asm/xen/interface.h
index e8506c1f0c55..1c10c88ee4e1 100644
--- a/arch/x86/include/asm/xen/interface.h
+++ b/arch/x86/include/asm/xen/interface.h
@@ -61,9 +61,9 @@ DEFINE_GUEST_HANDLE(void);
 #define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START)
 #endif
 
-#ifndef machine_to_phys_mapping
-#define machine_to_phys_mapping ((unsigned long *)HYPERVISOR_VIRT_START)
-#endif
+#define MACH2PHYS_VIRT_START  mk_unsigned_long(__MACH2PHYS_VIRT_START)
+#define MACH2PHYS_VIRT_END    mk_unsigned_long(__MACH2PHYS_VIRT_END)
+#define MACH2PHYS_NR_ENTRIES  ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>__MACH2PHYS_SHIFT)
 
 /* Maximum number of virtual CPUs in multi-processor guests. */
 #define MAX_VIRT_CPUS 32
diff --git a/arch/x86/include/asm/xen/interface_32.h b/arch/x86/include/asm/xen/interface_32.h
index 42a7e004ae5c..8413688b2571 100644
--- a/arch/x86/include/asm/xen/interface_32.h
+++ b/arch/x86/include/asm/xen/interface_32.h
@@ -32,6 +32,11 @@
 /* And the trap vector is... */
 #define TRAP_INSTR "int $0x82"
 
+#define __MACH2PHYS_VIRT_START 0xF5800000
+#define __MACH2PHYS_VIRT_END   0xF6800000
+
+#define __MACH2PHYS_SHIFT      2
+
 /*
  * Virtual addresses beyond this are not modifiable by guest OSes. The
  * machine->physical mapping table starts at this address, read-only.
diff --git a/arch/x86/include/asm/xen/interface_64.h b/arch/x86/include/asm/xen/interface_64.h
index 100d2662b97c..839a4811cf98 100644
--- a/arch/x86/include/asm/xen/interface_64.h
+++ b/arch/x86/include/asm/xen/interface_64.h
@@ -39,18 +39,7 @@
 #define __HYPERVISOR_VIRT_END   0xFFFF880000000000
 #define __MACH2PHYS_VIRT_START  0xFFFF800000000000
 #define __MACH2PHYS_VIRT_END    0xFFFF804000000000
-
-#ifndef HYPERVISOR_VIRT_START
-#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START)
-#define HYPERVISOR_VIRT_END   mk_unsigned_long(__HYPERVISOR_VIRT_END)
-#endif
-
-#define MACH2PHYS_VIRT_START  mk_unsigned_long(__MACH2PHYS_VIRT_START)
-#define MACH2PHYS_VIRT_END    mk_unsigned_long(__MACH2PHYS_VIRT_END)
-#define MACH2PHYS_NR_ENTRIES  ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>3)
-#ifndef machine_to_phys_mapping
-#define machine_to_phys_mapping ((unsigned long *)HYPERVISOR_VIRT_START)
-#endif
+#define __MACH2PHYS_SHIFT       3
 
 /*
  * int HYPERVISOR_set_segment_base(unsigned int which, unsigned long base)
diff --git a/arch/x86/include/asm/xen/page.h b/arch/x86/include/asm/xen/page.h
index dd8c1414b3d5..8760cc60a21c 100644
--- a/arch/x86/include/asm/xen/page.h
+++ b/arch/x86/include/asm/xen/page.h
@@ -5,6 +5,7 @@
 #include <linux/types.h>
 #include <linux/spinlock.h>
 #include <linux/pfn.h>
+#include <linux/mm.h>
 
 #include <asm/uaccess.h>
 #include <asm/page.h>
@@ -35,6 +36,8 @@ typedef struct xpaddr {
 #define MAX_DOMAIN_PAGES                                                \
     ((unsigned long)((u64)CONFIG_XEN_MAX_DOMAIN_MEMORY * 1024 * 1024 * 1024 / PAGE_SIZE))
 
+extern unsigned long *machine_to_phys_mapping;
+extern unsigned int   machine_to_phys_order;
 
 extern unsigned long get_phys_to_machine(unsigned long pfn);
 extern bool set_phys_to_machine(unsigned long pfn, unsigned long mfn);
@@ -69,10 +72,8 @@ static inline unsigned long mfn_to_pfn(unsigned long mfn)
         if (xen_feature(XENFEAT_auto_translated_physmap))
                 return mfn;
 
-#if 0
         if (unlikely((mfn >> machine_to_phys_order) != 0))
-                return max_mapnr;
-#endif
+                return ~0;
 
         pfn = 0;
         /*
diff --git a/arch/x86/kernel/cpuid.c b/arch/x86/kernel/cpuid.c
index 1b7b31ab7d86..212a6a42527c 100644
--- a/arch/x86/kernel/cpuid.c
+++ b/arch/x86/kernel/cpuid.c
@@ -33,7 +33,6 @@
 #include <linux/init.h>
 #include <linux/poll.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/major.h>
 #include <linux/fs.h>
 #include <linux/device.h>
diff --git a/arch/x86/kernel/kgdb.c b/arch/x86/kernel/kgdb.c
index ec592caac4b4..cd21b654dec6 100644
--- a/arch/x86/kernel/kgdb.c
+++ b/arch/x86/kernel/kgdb.c
@@ -315,14 +315,18 @@ static void kgdb_remove_all_hw_break(void)
                 if (!breakinfo[i].enabled)
                         continue;
                 bp = *per_cpu_ptr(breakinfo[i].pev, cpu);
-                if (bp->attr.disabled == 1)
+                if (!bp->attr.disabled) {
+                        arch_uninstall_hw_breakpoint(bp);
+                        bp->attr.disabled = 1;
                         continue;
+                }
                 if (dbg_is_early)
                         early_dr7 &= ~encode_dr7(i, breakinfo[i].len,
                                                  breakinfo[i].type);
-                else
-                        arch_uninstall_hw_breakpoint(bp);
-                bp->attr.disabled = 1;
+                else if (hw_break_release_slot(i))
+                        printk(KERN_ERR "KGDB: hw bpt remove failed %lx\n",
+                               breakinfo[i].addr);
+                breakinfo[i].enabled = 0;
         }
 }
 
diff --git a/arch/x86/kernel/msr.c b/arch/x86/kernel/msr.c
index 7bf2dc4c8f70..12fcbe2c143e 100644
--- a/arch/x86/kernel/msr.c
+++ b/arch/x86/kernel/msr.c
@@ -30,7 +30,6 @@
 #include <linux/init.h>
 #include <linux/poll.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/major.h>
 #include <linux/fs.h>
 #include <linux/device.h>
diff --git a/arch/x86/kvm/svm.c b/arch/x86/kvm/svm.c
index 82e144a4e514..1ca12298ffc7 100644
--- a/arch/x86/kvm/svm.c
+++ b/arch/x86/kvm/svm.c
@@ -3395,6 +3395,7 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu)
         vcpu->arch.regs[VCPU_REGS_RIP] = svm->vmcb->save.rip;
 
         load_host_msrs(vcpu);
+        kvm_load_ldt(ldt_selector);
         loadsegment(fs, fs_selector);
 #ifdef CONFIG_X86_64
         load_gs_index(gs_selector);
@@ -3402,7 +3403,6 @@ static void svm_vcpu_run(struct kvm_vcpu *vcpu)
 #else
         loadsegment(gs, gs_selector);
 #endif
-        kvm_load_ldt(ldt_selector);
 
         reload_tss(vcpu);
 
diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c
index 8da0e45ff7c9..ff21fdda0c53 100644
--- a/arch/x86/kvm/vmx.c
+++ b/arch/x86/kvm/vmx.c
@@ -821,10 +821,9 @@ static void vmx_save_host_state(struct kvm_vcpu *vcpu)
 #endif
 
 #ifdef CONFIG_X86_64
-        if (is_long_mode(&vmx->vcpu)) {
-                rdmsrl(MSR_KERNEL_GS_BASE, vmx->msr_host_kernel_gs_base);
+        rdmsrl(MSR_KERNEL_GS_BASE, vmx->msr_host_kernel_gs_base);
+        if (is_long_mode(&vmx->vcpu))
                 wrmsrl(MSR_KERNEL_GS_BASE, vmx->msr_guest_kernel_gs_base);
-        }
 #endif
         for (i = 0; i < vmx->save_nmsrs; ++i)
                 kvm_set_shared_msr(vmx->guest_msrs[i].index,
@@ -839,23 +838,23 @@ static void __vmx_load_host_state(struct vcpu_vmx *vmx)
 
         ++vmx->vcpu.stat.host_state_reload;
         vmx->host_state.loaded = 0;
-        if (vmx->host_state.fs_reload_needed)
-                loadsegment(fs, vmx->host_state.fs_sel);
+#ifdef CONFIG_X86_64
+        if (is_long_mode(&vmx->vcpu))
+                rdmsrl(MSR_KERNEL_GS_BASE, vmx->msr_guest_kernel_gs_base);
+#endif
         if (vmx->host_state.gs_ldt_reload_needed) {
                 kvm_load_ldt(vmx->host_state.ldt_sel);
 #ifdef CONFIG_X86_64
                 load_gs_index(vmx->host_state.gs_sel);
-                wrmsrl(MSR_KERNEL_GS_BASE, current->thread.gs);
 #else
                 loadsegment(gs, vmx->host_state.gs_sel);
 #endif
         }
+        if (vmx->host_state.fs_reload_needed)
+                loadsegment(fs, vmx->host_state.fs_sel);
         reload_tss();
 #ifdef CONFIG_X86_64
-        if (is_long_mode(&vmx->vcpu)) {
-                rdmsrl(MSR_KERNEL_GS_BASE, vmx->msr_guest_kernel_gs_base);
-                wrmsrl(MSR_KERNEL_GS_BASE, vmx->msr_host_kernel_gs_base);
-        }
+        wrmsrl(MSR_KERNEL_GS_BASE, vmx->msr_host_kernel_gs_base);
 #endif
         if (current_thread_info()->status & TS_USEDFPU)
                 clts();
diff --git a/arch/x86/pci/acpi.c b/arch/x86/pci/acpi.c
index 15466c096ba5..0972315c3860 100644
--- a/arch/x86/pci/acpi.c
+++ b/arch/x86/pci/acpi.c
@@ -138,7 +138,6 @@ setup_resource(struct acpi_resource *acpi_res, void *data)
         struct acpi_resource_address64 addr;
         acpi_status status;
         unsigned long flags;
-        struct resource *root, *conflict;
         u64 start, end;
 
         status = resource_to_addr(acpi_res, &addr);
@@ -146,12 +145,10 @@ setup_resource(struct acpi_resource *acpi_res, void *data)
                 return AE_OK;
 
         if (addr.resource_type == ACPI_MEMORY_RANGE) {
-                root = &iomem_resource;
                 flags = IORESOURCE_MEM;
                 if (addr.info.mem.caching == ACPI_PREFETCHABLE_MEMORY)
                         flags |= IORESOURCE_PREFETCH;
         } else if (addr.resource_type == ACPI_IO_RANGE) {
-                root = &ioport_resource;
                 flags = IORESOURCE_IO;
         } else
                 return AE_OK;
@@ -172,25 +169,90 @@ setup_resource(struct acpi_resource *acpi_res, void *data)
                 return AE_OK;
         }
 
-        conflict = insert_resource_conflict(root, res);
-        if (conflict) {
-                dev_err(&info->bridge->dev,
-                        "address space collision: host bridge window %pR "
-                        "conflicts with %s %pR\n",
-                        res, conflict->name, conflict);
-        } else {
-                pci_bus_add_resource(info->bus, res, 0);
-                info->res_num++;
-                if (addr.translation_offset)
-                        dev_info(&info->bridge->dev, "host bridge window %pR "
-                                 "(PCI address [%#llx-%#llx])\n",
-                                 res, res->start - addr.translation_offset,
-                                 res->end - addr.translation_offset);
+        info->res_num++;
+        if (addr.translation_offset)
+                dev_info(&info->bridge->dev, "host bridge window %pR "
+                         "(PCI address [%#llx-%#llx])\n",
+                         res, res->start - addr.translation_offset,
+                         res->end - addr.translation_offset);
+        else
+                dev_info(&info->bridge->dev, "host bridge window %pR\n", res);
+
+        return AE_OK;
+}
+
+static bool resource_contains(struct resource *res, resource_size_t point)
+{
+        if (res->start <= point && point <= res->end)
+                return true;
+        return false;
+}
+
+static void coalesce_windows(struct pci_root_info *info, int type)
+{
+        int i, j;
+        struct resource *res1, *res2;
+
+        for (i = 0; i < info->res_num; i++) {
+                res1 = &info->res[i];
+                if (!(res1->flags & type))
+                        continue;
+
+                for (j = i + 1; j < info->res_num; j++) {
+                        res2 = &info->res[j];
+                        if (!(res2->flags & type))
+                                continue;
+
+                        /*
+                         * I don't like throwing away windows because then
+                         * our resources no longer match the ACPI _CRS, but
+                         * the kernel resource tree doesn't allow overlaps.
+                         */
+                        if (resource_contains(res1, res2->start) ||
+                            resource_contains(res1, res2->end) ||
+                            resource_contains(res2, res1->start) ||
+                            resource_contains(res2, res1->end)) {
+                                res1->start = min(res1->start, res2->start);
+                                res1->end = max(res1->end, res2->end);
+                                dev_info(&info->bridge->dev,
+                                         "host bridge window expanded to %pR; %pR ignored\n",
+                                         res1, res2);
+                                res2->flags = 0;
+                        }
+                }
+        }
+}
+
+static void add_resources(struct pci_root_info *info)
+{
+        int i;
+        struct resource *res, *root, *conflict;
+
+        if (!pci_use_crs)
+                return;
+
+        coalesce_windows(info, IORESOURCE_MEM);
+        coalesce_windows(info, IORESOURCE_IO);
+
+        for (i = 0; i < info->res_num; i++) {
+                res = &info->res[i];
+
+                if (res->flags & IORESOURCE_MEM)
+                        root = &iomem_resource;
+                else if (res->flags & IORESOURCE_IO)
+                        root = &ioport_resource;
                 else
-                        dev_info(&info->bridge->dev,
-                                 "host bridge window %pR\n", res);
+                        continue;
+
+                conflict = insert_resource_conflict(root, res);
+                if (conflict)
+                        dev_err(&info->bridge->dev,
+                                "address space collision: host bridge window %pR "
+                                "conflicts with %s %pR\n",
+                                res, conflict->name, conflict);
+                else
+                        pci_bus_add_resource(info->bus, res, 0);
         }
-        return AE_OK;
 }
 
 static void
@@ -224,6 +286,7 @@ get_current_resources(struct acpi_device *device, int busnum,
         acpi_walk_resources(device->handle, METHOD_NAME__CRS, setup_resource,
                                 &info);
 
+        add_resources(&info);
         return;
 
 name_alloc_fail:
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index 235c0f4d3861..02c710bebf7a 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -75,6 +75,11 @@ DEFINE_PER_CPU(struct vcpu_info, xen_vcpu_info);
 enum xen_domain_type xen_domain_type = XEN_NATIVE;
 EXPORT_SYMBOL_GPL(xen_domain_type);
 
+unsigned long *machine_to_phys_mapping = (void *)MACH2PHYS_VIRT_START;
+EXPORT_SYMBOL(machine_to_phys_mapping);
+unsigned int   machine_to_phys_order;
+EXPORT_SYMBOL(machine_to_phys_order);
+
 struct start_info *xen_start_info;
 EXPORT_SYMBOL_GPL(xen_start_info);
 
@@ -1090,6 +1095,8 @@ static void __init xen_setup_stackprotector(void)
 /* First C function to be called on Xen boot */
 asmlinkage void __init xen_start_kernel(void)
 {
+        struct physdev_set_iopl set_iopl;
+        int rc;
         pgd_t *pgd;
 
         if (!xen_start_info)
@@ -1097,6 +1104,8 @@ asmlinkage void __init xen_start_kernel(void)
 
         xen_domain_type = XEN_PV_DOMAIN;
 
+        xen_setup_machphys_mapping();
+
         /* Install Xen paravirt ops */
         pv_info = xen_info;
         pv_init_ops = xen_init_ops;
@@ -1191,8 +1200,6 @@ asmlinkage void __init xen_start_kernel(void)
         /* Allocate and initialize top and mid mfn levels for p2m structure */
         xen_build_mfn_list_list();
 
-        init_mm.pgd = pgd;
-
         /* keep using Xen gdt for now; no urgent need to change it */
 
 #ifdef CONFIG_X86_32
@@ -1202,10 +1209,18 @@ asmlinkage void __init xen_start_kernel(void)
 #else
         pv_info.kernel_rpl = 0;
 #endif
-
         /* set the limit of our address space */
         xen_reserve_top();
 
+        /* We used to do this in xen_arch_setup, but that is too late on AMD
+         * were early_cpu_init (run before ->arch_setup()) calls early_amd_init
+         * which pokes 0xcf8 port.
+         */
+        set_iopl.iopl = 1;
+        rc = HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
+        if (rc != 0)
+                xen_raw_printk("physdev_op failed %d\n", rc);
+
 #ifdef CONFIG_X86_32
         /* set up basic CPUID stuff */
         cpu_detect(&new_cpu_data);
diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c
index 21ed8d7f75a5..a1feff9e59b6 100644
--- a/arch/x86/xen/mmu.c
+++ b/arch/x86/xen/mmu.c
@@ -2034,6 +2034,20 @@ static __init void xen_map_identity_early(pmd_t *pmd, unsigned long max_pfn)
         set_page_prot(pmd, PAGE_KERNEL_RO);
 }
 
+void __init xen_setup_machphys_mapping(void)
+{
+        struct xen_machphys_mapping mapping;
+        unsigned long machine_to_phys_nr_ents;
+
+        if (HYPERVISOR_memory_op(XENMEM_machphys_mapping, &mapping) == 0) {
+                machine_to_phys_mapping = (unsigned long *)mapping.v_start;
+                machine_to_phys_nr_ents = mapping.max_mfn + 1;
+        } else {
+                machine_to_phys_nr_ents = MACH2PHYS_NR_ENTRIES;
+        }
+        machine_to_phys_order = fls(machine_to_phys_nr_ents - 1);
+}
+
 #ifdef CONFIG_X86_64
 static void convert_pfn_mfn(void *v)
 {
@@ -2119,44 +2133,83 @@ __init pgd_t *xen_setup_kernel_pagetable(pgd_t *pgd,
         return pgd;
 }
 #else   /* !CONFIG_X86_64 */
-static RESERVE_BRK_ARRAY(pmd_t, level2_kernel_pgt, PTRS_PER_PMD);
+static RESERVE_BRK_ARRAY(pmd_t, initial_kernel_pmd, PTRS_PER_PMD);
+static RESERVE_BRK_ARRAY(pmd_t, swapper_kernel_pmd, PTRS_PER_PMD);
+
+static __init void xen_write_cr3_init(unsigned long cr3)
+{
+        unsigned long pfn = PFN_DOWN(__pa(swapper_pg_dir));
+
+        BUG_ON(read_cr3() != __pa(initial_page_table));
+        BUG_ON(cr3 != __pa(swapper_pg_dir));
+
+        /*
+         * We are switching to swapper_pg_dir for the first time (from
+         * initial_page_table) and therefore need to mark that page
+         * read-only and then pin it.
+         *
+         * Xen disallows sharing of kernel PMDs for PAE
+         * guests. Therefore we must copy the kernel PMD from
+         * initial_page_table into a new kernel PMD to be used in
+         * swapper_pg_dir.
+         */
+        swapper_kernel_pmd =
+                extend_brk(sizeof(pmd_t) * PTRS_PER_PMD, PAGE_SIZE);
+        memcpy(swapper_kernel_pmd, initial_kernel_pmd,
+               sizeof(pmd_t) * PTRS_PER_PMD);
+        swapper_pg_dir[KERNEL_PGD_BOUNDARY] =
+                __pgd(__pa(swapper_kernel_pmd) | _PAGE_PRESENT);
+        set_page_prot(swapper_kernel_pmd, PAGE_KERNEL_RO);
+
+        set_page_prot(swapper_pg_dir, PAGE_KERNEL_RO);
+        xen_write_cr3(cr3);
+        pin_pagetable_pfn(MMUEXT_PIN_L3_TABLE, pfn);
+
+        pin_pagetable_pfn(MMUEXT_UNPIN_TABLE,
+                          PFN_DOWN(__pa(initial_page_table)));
+        set_page_prot(initial_page_table, PAGE_KERNEL);
+        set_page_prot(initial_kernel_pmd, PAGE_KERNEL);
+
+        pv_mmu_ops.write_cr3 = &xen_write_cr3;
+}
 
 __init pgd_t *xen_setup_kernel_pagetable(pgd_t *pgd,
                                          unsigned long max_pfn)
 {
         pmd_t *kernel_pmd;
 
-        level2_kernel_pgt = extend_brk(sizeof(pmd_t) * PTRS_PER_PMD, PAGE_SIZE);
+        initial_kernel_pmd =
+                extend_brk(sizeof(pmd_t) * PTRS_PER_PMD, PAGE_SIZE);
 
         max_pfn_mapped = PFN_DOWN(__pa(xen_start_info->pt_base) +
                                   xen_start_info->nr_pt_frames * PAGE_SIZE +
                                   512*1024);
 
         kernel_pmd = m2v(pgd[KERNEL_PGD_BOUNDARY].pgd);
-        memcpy(level2_kernel_pgt, kernel_pmd, sizeof(pmd_t) * PTRS_PER_PMD);
+        memcpy(initial_kernel_pmd, kernel_pmd, sizeof(pmd_t) * PTRS_PER_PMD);
 
-        xen_map_identity_early(level2_kernel_pgt, max_pfn);
+        xen_map_identity_early(initial_kernel_pmd, max_pfn);
 
-        memcpy(swapper_pg_dir, pgd, sizeof(pgd_t) * PTRS_PER_PGD);
-        set_pgd(&swapper_pg_dir[KERNEL_PGD_BOUNDARY],
-                        __pgd(__pa(level2_kernel_pgt) | _PAGE_PRESENT));
+        memcpy(initial_page_table, pgd, sizeof(pgd_t) * PTRS_PER_PGD);
+        initial_page_table[KERNEL_PGD_BOUNDARY] =
+                __pgd(__pa(initial_kernel_pmd) | _PAGE_PRESENT);
 
-        set_page_prot(level2_kernel_pgt, PAGE_KERNEL_RO);
-        set_page_prot(swapper_pg_dir, PAGE_KERNEL_RO);
+        set_page_prot(initial_kernel_pmd, PAGE_KERNEL_RO);
+        set_page_prot(initial_page_table, PAGE_KERNEL_RO);
         set_page_prot(empty_zero_page, PAGE_KERNEL_RO);
 
         pin_pagetable_pfn(MMUEXT_UNPIN_TABLE, PFN_DOWN(__pa(pgd)));
 
-        xen_write_cr3(__pa(swapper_pg_dir));
-
-        pin_pagetable_pfn(MMUEXT_PIN_L3_TABLE, PFN_DOWN(__pa(swapper_pg_dir)));
+        pin_pagetable_pfn(MMUEXT_PIN_L3_TABLE,
+                          PFN_DOWN(__pa(initial_page_table)));
+        xen_write_cr3(__pa(initial_page_table));
 
         memblock_x86_reserve_range(__pa(xen_start_info->pt_base),
                       __pa(xen_start_info->pt_base +
                            xen_start_info->nr_pt_frames * PAGE_SIZE),
                       "XEN PAGETABLES");
 
-        return swapper_pg_dir;
+        return initial_page_table;
 }
 #endif  /* CONFIG_X86_64 */
 
@@ -2290,7 +2343,11 @@ static const struct pv_mmu_ops xen_mmu_ops __initdata = {
         .write_cr2 = xen_write_cr2,
 
         .read_cr3 = xen_read_cr3,
+#ifdef CONFIG_X86_32
+        .write_cr3 = xen_write_cr3_init,
+#else
         .write_cr3 = xen_write_cr3,
+#endif
 
         .flush_tlb_user = xen_flush_tlb,
         .flush_tlb_kernel = xen_flush_tlb,
@@ -2627,7 +2684,8 @@ int xen_remap_domain_mfn_range(struct vm_area_struct *vma,
 
         prot = __pgprot(pgprot_val(prot) | _PAGE_IOMAP);
 
-        vma->vm_flags |= VM_IO | VM_RESERVED | VM_PFNMAP;
+        BUG_ON(!((vma->vm_flags & (VM_PFNMAP | VM_RESERVED | VM_IO)) ==
+                                (VM_PFNMAP | VM_RESERVED | VM_IO)));
 
         rmd.mfn = mfn;
         rmd.prot = prot;
diff --git a/arch/x86/xen/setup.c b/arch/x86/xen/setup.c
index 769c4b01fa32..01afd8a94607 100644
--- a/arch/x86/xen/setup.c
+++ b/arch/x86/xen/setup.c
@@ -23,7 +23,6 @@
 #include <xen/interface/callback.h>
 #include <xen/interface/memory.h>
 #include <xen/interface/physdev.h>
-#include <xen/interface/memory.h>
 #include <xen/features.h>
 
 #include "xen-ops.h"
@@ -248,8 +247,7 @@ char * __init xen_memory_setup(void)
         else
                 extra_pages = 0;
 
-        if (!xen_initial_domain())
-                xen_add_extra_mem(extra_pages);
+        xen_add_extra_mem(extra_pages);
 
         return "Xen";
 }
@@ -337,9 +335,6 @@ void __cpuinit xen_enable_syscall(void)
 
 void __init xen_arch_setup(void)
 {
-        struct physdev_set_iopl set_iopl;
-        int rc;
-
         xen_panic_handler_init();
 
         HYPERVISOR_vm_assist(VMASST_CMD_enable, VMASST_TYPE_4gb_segments);
@@ -356,11 +351,6 @@ void __init xen_arch_setup(void)
         xen_enable_sysenter();
         xen_enable_syscall();
 
-        set_iopl.iopl = 1;
-        rc = HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
-        if (rc != 0)
-                printk(KERN_INFO "physdev_op failed %d\n", rc);
-
 #ifdef CONFIG_ACPI
         if (!(xen_start_info->flags & SIF_INITDOMAIN)) {
                 printk(KERN_INFO "ACPI in unprivileged domain disabled\n");
diff --git a/block/compat_ioctl.c b/block/compat_ioctl.c
index 58c6ee5b010c..cc3eb78e333a 100644
--- a/block/compat_ioctl.c
+++ b/block/compat_ioctl.c
@@ -8,7 +8,6 @@
 #include <linux/hdreg.h>
 #include <linux/slab.h>
 #include <linux/syscalls.h>
-#include <linux/smp_lock.h>
 #include <linux/types.h>
 #include <linux/uaccess.h>
 
diff --git a/block/ioctl.c b/block/ioctl.c
index 3d866d0037f2..a9a302eba01e 100644
--- a/block/ioctl.c
+++ b/block/ioctl.c
@@ -5,7 +5,6 @@
 #include <linux/hdreg.h>
 #include <linux/backing-dev.h>
 #include <linux/buffer_head.h>
-#include <linux/smp_lock.h>
 #include <linux/blktrace_api.h>
 #include <asm/uaccess.h>
 
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index 3f91c01c217f..66aa4bee80a6 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -3166,8 +3166,8 @@ static inline int __ata_scsi_queuecmd(struct scsi_cmnd *scmd,
 
 /**
  *      ata_scsi_queuecmd - Issue SCSI cdb to libata-managed device
+ *      @shost: SCSI host of command to be sent
  *      @cmd: SCSI command to be sent
- *      @done: Completion function, called when command is complete
  *
  *      In some cases, this function translates SCSI commands into
  *      ATA taskfiles, and queues the taskfiles to be sent to
@@ -3177,37 +3177,36 @@ static inline int __ata_scsi_queuecmd(struct scsi_cmnd *scmd,
  *      ATA and ATAPI devices appearing as SCSI devices.
  *
  *      LOCKING:
- *      Releases scsi-layer-held lock, and obtains host lock.
+ *      ATA host lock
  *
  *      RETURNS:
  *      Return value from __ata_scsi_queuecmd() if @cmd can be queued,
  *      0 otherwise.
  */
-int ata_scsi_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
+int ata_scsi_queuecmd(struct Scsi_Host *shost, struct scsi_cmnd *cmd)
 {
         struct ata_port *ap;
         struct ata_device *dev;
         struct scsi_device *scsidev = cmd->device;
-        struct Scsi_Host *shost = scsidev->host;
         int rc = 0;
+        unsigned long irq_flags;
 
         ap = ata_shost_to_port(shost);
 
-        spin_unlock(shost->host_lock);
-        spin_lock(ap->lock);
+        spin_lock_irqsave(ap->lock, irq_flags);
 
         ata_scsi_dump_cdb(ap, cmd);
 
         dev = ata_scsi_find_dev(ap, scsidev);
         if (likely(dev))
-                rc = __ata_scsi_queuecmd(cmd, done, dev);
+                rc = __ata_scsi_queuecmd(cmd, cmd->scsi_done, dev);
         else {
                 cmd->result = (DID_BAD_TARGET << 16);
-                done(cmd);
+                cmd->scsi_done(cmd);
         }
 
-        spin_unlock(ap->lock);
-        spin_lock(shost->host_lock);
+        spin_unlock_irqrestore(ap->lock, irq_flags);
+
         return rc;
 }
 
diff --git a/drivers/ata/sata_via.c b/drivers/ata/sata_via.c
index c21589986c69..8b677bbf2d37 100644
--- a/drivers/ata/sata_via.c
+++ b/drivers/ata/sata_via.c
@@ -538,7 +538,7 @@ static int vt8251_prepare_host(struct pci_dev *pdev, struct ata_host **r_host)
         return 0;
 }
 
-static void svia_configure(struct pci_dev *pdev)
+static void svia_configure(struct pci_dev *pdev, int board_id)
 {
         u8 tmp8;
 
@@ -577,7 +577,7 @@ static void svia_configure(struct pci_dev *pdev)
         }
 
         /*
-         * vt6421 has problems talking to some drives.  The following
+         * vt6420/1 has problems talking to some drives.  The following
          * is the fix from Joseph Chan <JosephChan@via.com.tw>.
          *
          * When host issues HOLD, device may send up to 20DW of data
@@ -596,8 +596,9 @@ static void svia_configure(struct pci_dev *pdev)
          *
          * https://bugzilla.kernel.org/show_bug.cgi?id=15173
          * http://article.gmane.org/gmane.linux.ide/46352
+         * http://thread.gmane.org/gmane.linux.kernel/1062139
          */
-        if (pdev->device == 0x3249) {
+        if (board_id == vt6420 || board_id == vt6421) {
                 pci_read_config_byte(pdev, 0x52, &tmp8);
                 tmp8 |= 1 << 2;
                 pci_write_config_byte(pdev, 0x52, tmp8);
@@ -652,7 +653,7 @@ static int svia_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
         if (rc)
                 return rc;
 
-        svia_configure(pdev);
+        svia_configure(pdev, board_id);
 
         pci_set_master(pdev);
         return ata_host_activate(host, pdev->irq, ata_bmdma_interrupt,
diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c
index 31b526661ec4..ead3e79d6fcf 100644
--- a/drivers/base/power/main.c
+++ b/drivers/base/power/main.c
@@ -475,20 +475,33 @@ End:
  */
 void dpm_resume_noirq(pm_message_t state)
 {
-        struct device *dev;
+        struct list_head list;
         ktime_t starttime = ktime_get();
 
+        INIT_LIST_HEAD(&list);
         mutex_lock(&dpm_list_mtx);
         transition_started = false;
-        list_for_each_entry(dev, &dpm_list, power.entry)
+        while (!list_empty(&dpm_list)) {
+                struct device *dev = to_device(dpm_list.next);
+
+                get_device(dev);
                 if (dev->power.status > DPM_OFF) {
                         int error;
 
                         dev->power.status = DPM_OFF;
+                        mutex_unlock(&dpm_list_mtx);
+
                         error = device_resume_noirq(dev, state);
+
+                        mutex_lock(&dpm_list_mtx);
                         if (error)
                                 pm_dev_err(dev, state, " early", error);
                 }
+                if (!list_empty(&dev->power.entry))
+                        list_move_tail(&dev->power.entry, &list);
+                put_device(dev);
+        }
+        list_splice(&list, &dpm_list);
         mutex_unlock(&dpm_list_mtx);
         dpm_show_time(starttime, state, "early");
         resume_device_irqs();
@@ -789,20 +802,33 @@ End:
  */
 int dpm_suspend_noirq(pm_message_t state)
 {
-        struct device *dev;
+        struct list_head list;
         ktime_t starttime = ktime_get();
         int error = 0;
 
+        INIT_LIST_HEAD(&list);
         suspend_device_irqs();
         mutex_lock(&dpm_list_mtx);
-        list_for_each_entry_reverse(dev, &dpm_list, power.entry) {
+        while (!list_empty(&dpm_list)) {
+                struct device *dev = to_device(dpm_list.prev);
+
+                get_device(dev);
+                mutex_unlock(&dpm_list_mtx);
+
                 error = device_suspend_noirq(dev, state);
+
+                mutex_lock(&dpm_list_mtx);
                 if (error) {
                         pm_dev_err(dev, state, " late", error);
+                        put_device(dev);
                         break;
                 }
                 dev->power.status = DPM_OFF_IRQ;
+                if (!list_empty(&dev->power.entry))
+                        list_move(&dev->power.entry, &list);
+                put_device(dev);
         }
+        list_splice_tail(&list, &dpm_list);
         mutex_unlock(&dpm_list_mtx);
         if (error)
                 dpm_resume_noirq(resume_event(state));
diff --git a/drivers/block/cciss_scsi.c b/drivers/block/cciss_scsi.c
index 575495f3c4b8..727d0225b7d0 100644
--- a/drivers/block/cciss_scsi.c
+++ b/drivers/block/cciss_scsi.c
@@ -62,8 +62,8 @@ static int cciss_scsi_proc_info(
                 int length,        /* length of data in buffer */
                 int func);         /* 0 == read, 1 == write */
 
-static int cciss_scsi_queue_command (struct scsi_cmnd *cmd,
-                void (* done)(struct scsi_cmnd *));
+static int cciss_scsi_queue_command (struct Scsi_Host *h,
+                                     struct scsi_cmnd *cmd);
 static int cciss_eh_device_reset_handler(struct scsi_cmnd *);
 static int cciss_eh_abort_handler(struct scsi_cmnd *);
 
@@ -1406,7 +1406,7 @@ static void cciss_scatter_gather(ctlr_info_t *h, CommandList_struct *c,
 
 
 static int
-cciss_scsi_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd *))
+cciss_scsi_queue_command_lck(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 {
         ctlr_info_t *h;
         int rc;
@@ -1504,6 +1504,8 @@ cciss_scsi_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd
         return 0;
 }
 
+static DEF_SCSI_QCMD(cciss_scsi_queue_command)
+
 static void cciss_unregister_scsi(ctlr_info_t *h)
 {
         struct cciss_scsi_adapter_data_t *sa;
diff --git a/drivers/block/drbd/drbd_receiver.c b/drivers/block/drbd/drbd_receiver.c
index d299fe9e78c8..89d8a7cc4054 100644
--- a/drivers/block/drbd/drbd_receiver.c
+++ b/drivers/block/drbd/drbd_receiver.c
@@ -36,7 +36,6 @@
 #include <linux/memcontrol.h>
 #include <linux/mm_inline.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/pkt_sched.h>
 #define __KERNEL_SYSCALLS__
 #include <linux/unistd.h>
diff --git a/drivers/block/drbd/drbd_worker.c b/drivers/block/drbd/drbd_worker.c
index b0551ba7ad0c..47d223c2409c 100644
--- a/drivers/block/drbd/drbd_worker.c
+++ b/drivers/block/drbd/drbd_worker.c
@@ -26,7 +26,6 @@
 #include <linux/module.h>
 #include <linux/drbd.h>
 #include <linux/sched.h>
-#include <linux/smp_lock.h>
 #include <linux/wait.h>
 #include <linux/mm.h>
 #include <linux/memcontrol.h>
diff --git a/drivers/char/agp/frontend.c b/drivers/char/agp/frontend.c
index 43412c03969e..3cb4539a98b2 100644
--- a/drivers/char/agp/frontend.c
+++ b/drivers/char/agp/frontend.c
@@ -39,7 +39,6 @@
 #include <linux/mm.h>
 #include <linux/fs.h>
 #include <linux/sched.h>
-#include <linux/smp_lock.h>
 #include <asm/uaccess.h>
 #include <asm/pgtable.h>
 #include "agp.h"
diff --git a/drivers/char/amiserial.c b/drivers/char/amiserial.c
index c0bd6f472c52..6ee3348bc3e4 100644
--- a/drivers/char/amiserial.c
+++ b/drivers/char/amiserial.c
@@ -81,7 +81,6 @@ static char *serial_version = "4.30";
 #include <linux/mm.h>
 #include <linux/seq_file.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/init.h>
 #include <linux/bitops.h>
 #include <linux/platform_device.h>
diff --git a/drivers/char/briq_panel.c b/drivers/char/briq_panel.c
index f6718f05dad4..095ab90535ce 100644
--- a/drivers/char/briq_panel.c
+++ b/drivers/char/briq_panel.c
@@ -6,7 +6,6 @@
 
 #include <linux/module.h>
 
-#include <linux/smp_lock.h>
 #include <linux/types.h>
 #include <linux/errno.h>
 #include <linux/tty.h>
diff --git a/drivers/char/hpet.c b/drivers/char/hpet.c
index 55b8667f739f..7066e801b9d3 100644
--- a/drivers/char/hpet.c
+++ b/drivers/char/hpet.c
@@ -14,7 +14,6 @@
 #include <linux/interrupt.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
-#include <linux/smp_lock.h>
 #include <linux/types.h>
 #include <linux/miscdevice.h>
 #include <linux/major.h>
diff --git a/drivers/char/hw_random/core.c b/drivers/char/hw_random/core.c
index 788da05190cc..2016aad85203 100644
--- a/drivers/char/hw_random/core.c
+++ b/drivers/char/hw_random/core.c
@@ -37,7 +37,6 @@
 #include <linux/kernel.h>
 #include <linux/fs.h>
 #include <linux/sched.h>
-#include <linux/smp_lock.h>
 #include <linux/init.h>
 #include <linux/miscdevice.h>
 #include <linux/delay.h>
diff --git a/drivers/char/i8k.c b/drivers/char/i8k.c
index f0863bec186f..d72433f2d310 100644
--- a/drivers/char/i8k.c
+++ b/drivers/char/i8k.c
@@ -120,7 +120,7 @@ static int i8k_smm(struct smm_regs *regs)
         int eax = regs->eax;
 
 #if defined(CONFIG_X86_64)
-        asm("pushq %%rax\n\t"
+        asm volatile("pushq %%rax\n\t"
                 "movl 0(%%rax),%%edx\n\t"
                 "pushq %%rdx\n\t"
                 "movl 4(%%rax),%%ebx\n\t"
@@ -142,11 +142,11 @@ static int i8k_smm(struct smm_regs *regs)
                 "lahf\n\t"
                 "shrl $8,%%eax\n\t"
                 "andl $1,%%eax\n"
-                :"=a"(rc), "+m" (*regs)
+                :"=a"(rc)
                 :    "a"(regs)
                 :    "%ebx", "%ecx", "%edx", "%esi", "%edi", "memory");
 #else
-        asm("pushl %%eax\n\t"
+        asm volatile("pushl %%eax\n\t"
             "movl 0(%%eax),%%edx\n\t"
             "push %%edx\n\t"
             "movl 4(%%eax),%%ebx\n\t"
@@ -168,7 +168,7 @@ static int i8k_smm(struct smm_regs *regs)
             "lahf\n\t"
             "shrl $8,%%eax\n\t"
             "andl $1,%%eax\n"
-            :"=a"(rc), "+m" (*regs)
+            :"=a"(rc)
             :    "a"(regs)
             :    "%ebx", "%ecx", "%edx", "%esi", "%edi", "memory");
 #endif
diff --git a/drivers/char/istallion.c b/drivers/char/istallion.c
index 667abd23ad6a..7c6de4c92458 100644
--- a/drivers/char/istallion.c
+++ b/drivers/char/istallion.c
@@ -21,7 +21,6 @@
 #include <linux/module.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/interrupt.h>
 #include <linux/tty.h>
 #include <linux/tty_flip.h>
diff --git a/drivers/char/serial167.c b/drivers/char/serial167.c
index f646725bd567..748c3b0ecd89 100644
--- a/drivers/char/serial167.c
+++ b/drivers/char/serial167.c
@@ -52,7 +52,6 @@
 #include <linux/interrupt.h>
 #include <linux/serial.h>
 #include <linux/serialP.h>
-#include <linux/smp_lock.h>
 #include <linux/string.h>
 #include <linux/fcntl.h>
 #include <linux/ptrace.h>
diff --git a/drivers/char/specialix.c b/drivers/char/specialix.c
index 9f8495b4fc8f..a7616d226a49 100644
--- a/drivers/char/specialix.c
+++ b/drivers/char/specialix.c
@@ -87,7 +87,6 @@
 #include <linux/tty_flip.h>
 #include <linux/mm.h>
 #include <linux/serial.h>
-#include <linux/smp_lock.h>
 #include <linux/fcntl.h>
 #include <linux/major.h>
 #include <linux/delay.h>
diff --git a/drivers/char/stallion.c b/drivers/char/stallion.c
index 4bef6ab83622..461a5a045517 100644
--- a/drivers/char/stallion.c
+++ b/drivers/char/stallion.c
@@ -40,7 +40,6 @@
 #include <linux/stallion.h>
 #include <linux/ioport.h>
 #include <linux/init.h>
-#include <linux/smp_lock.h>
 #include <linux/device.h>
 #include <linux/delay.h>
 #include <linux/ctype.h>
diff --git a/drivers/char/sx.c b/drivers/char/sx.c
index e53f16865397..a786326cea2f 100644
--- a/drivers/char/sx.c
+++ b/drivers/char/sx.c
@@ -216,7 +216,6 @@
 #include <linux/eisa.h>
 #include <linux/pci.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/init.h>
 #include <linux/miscdevice.h>
 #include <linux/bitops.h>
diff --git a/drivers/char/uv_mmtimer.c b/drivers/char/uv_mmtimer.c
index 493b47a0d511..956ebe2080a5 100644
--- a/drivers/char/uv_mmtimer.c
+++ b/drivers/char/uv_mmtimer.c
@@ -23,7 +23,6 @@
 #include <linux/interrupt.h>
 #include <linux/time.h>
 #include <linux/math64.h>
-#include <linux/smp_lock.h>
 
 #include <asm/genapic.h>
 #include <asm/uv/uv_hub.h>
diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c
index 6c1b676643a9..896a2ced1d27 100644
--- a/drivers/char/virtio_console.c
+++ b/drivers/char/virtio_console.c
@@ -1547,31 +1547,16 @@ static int init_vqs(struct ports_device *portdev)
         nr_queues = use_multiport(portdev) ? (nr_ports + 1) * 2 : 2;
 
         vqs = kmalloc(nr_queues * sizeof(struct virtqueue *), GFP_KERNEL);
-        if (!vqs) {
-                err = -ENOMEM;
-                goto fail;
-        }
         io_callbacks = kmalloc(nr_queues * sizeof(vq_callback_t *), GFP_KERNEL);
-        if (!io_callbacks) {
-                err = -ENOMEM;
-                goto free_vqs;
-        }
         io_names = kmalloc(nr_queues * sizeof(char *), GFP_KERNEL);
-        if (!io_names) {
-                err = -ENOMEM;
-                goto free_callbacks;
-        }
         portdev->in_vqs = kmalloc(nr_ports * sizeof(struct virtqueue *),
                                   GFP_KERNEL);
-        if (!portdev->in_vqs) {
-                err = -ENOMEM;
-                goto free_names;
-        }
         portdev->out_vqs = kmalloc(nr_ports * sizeof(struct virtqueue *),
                                    GFP_KERNEL);
-        if (!portdev->out_vqs) {
+        if (!vqs || !io_callbacks || !io_names || !portdev->in_vqs ||
+                        !portdev->out_vqs) {
                 err = -ENOMEM;
-                goto free_invqs;
+                goto free;
         }
 
         /*
@@ -1605,7 +1590,7 @@ static int init_vqs(struct ports_device *portdev)
                                               io_callbacks,
                                               (const char **)io_names);
         if (err)
-                goto free_outvqs;
+                goto free;
 
         j = 0;
         portdev->in_vqs[0] = vqs[0];
@@ -1621,23 +1606,19 @@ static int init_vqs(struct ports_device *portdev)
                         portdev->out_vqs[i] = vqs[j + 1];
                 }
         }
-        kfree(io_callbacks);
         kfree(io_names);
+        kfree(io_callbacks);
         kfree(vqs);
 
         return 0;
 
-free_names:
-        kfree(io_names);
-free_callbacks:
-        kfree(io_callbacks);
-free_outvqs:
+free:
         kfree(portdev->out_vqs);
-free_invqs:
         kfree(portdev->in_vqs);
-free_vqs:
+        kfree(io_names);
+        kfree(io_callbacks);
         kfree(vqs);
-fail:
+
         return err;
 }
 
diff --git a/drivers/firewire/sbp2.c b/drivers/firewire/sbp2.c
index bfae4b309791..afa576a75a8e 100644
--- a/drivers/firewire/sbp2.c
+++ b/drivers/firewire/sbp2.c
@@ -1468,7 +1468,7 @@ static int sbp2_map_scatterlist(struct sbp2_command_orb *orb,
 
 /* SCSI stack integration */
 
-static int sbp2_scsi_queuecommand(struct scsi_cmnd *cmd, scsi_done_fn_t done)
+static int sbp2_scsi_queuecommand_lck(struct scsi_cmnd *cmd, scsi_done_fn_t done)
 {
         struct sbp2_logical_unit *lu = cmd->device->hostdata;
         struct fw_device *device = target_device(lu->tgt);
@@ -1534,6 +1534,8 @@ static int sbp2_scsi_queuecommand(struct scsi_cmnd *cmd, scsi_done_fn_t done)
         return retval;
 }
 
+static DEF_SCSI_QCMD(sbp2_scsi_queuecommand)
+
 static int sbp2_scsi_slave_alloc(struct scsi_device *sdev)
 {
         struct sbp2_logical_unit *lu = sdev->hostdata;
diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c
index b744dad5c237..a39794bac04b 100644
--- a/drivers/gpu/drm/drm_fops.c
+++ b/drivers/gpu/drm/drm_fops.c
@@ -37,7 +37,6 @@
 #include "drmP.h"
 #include <linux/poll.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 
 /* from BKL pushdown: note that nothing else serializes idr_find() */
 DEFINE_MUTEX(drm_global_mutex);
diff --git a/drivers/gpu/drm/i915/i915_drv.c b/drivers/gpu/drm/i915/i915_drv.c
index 80745f85902c..f737960712e6 100644
--- a/drivers/gpu/drm/i915/i915_drv.c
+++ b/drivers/gpu/drm/i915/i915_drv.c
@@ -150,7 +150,8 @@ static const struct intel_device_info intel_ironlake_d_info = {
 
 static const struct intel_device_info intel_ironlake_m_info = {
         .gen = 5, .is_mobile = 1,
-        .need_gfx_hws = 1, .has_fbc = 1, .has_rc6 = 1, .has_hotplug = 1,
+        .need_gfx_hws = 1, .has_rc6 = 1, .has_hotplug = 1,
+        .has_fbc = 0, /* disabled due to buggy hardware */
         .has_bsd_ring = 1,
 };
 
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 90414ae86afc..409826da3099 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -1045,6 +1045,8 @@ void i915_gem_clflush_object(struct drm_gem_object *obj);
 int i915_gem_object_set_domain(struct drm_gem_object *obj,
                                uint32_t read_domains,
                                uint32_t write_domain);
+int i915_gem_object_flush_gpu(struct drm_i915_gem_object *obj,
+                              bool interruptible);
 int i915_gem_init_ringbuffer(struct drm_device *dev);
 void i915_gem_cleanup_ringbuffer(struct drm_device *dev);
 int i915_gem_do_init(struct drm_device *dev, unsigned long start,
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index ef188e391406..17b1cba3b5f1 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -547,6 +547,19 @@ i915_gem_pread_ioctl(struct drm_device *dev, void *data,
         struct drm_i915_gem_object *obj_priv;
         int ret = 0;
 
+        if (args->size == 0)
+                return 0;
+
+        if (!access_ok(VERIFY_WRITE,
+                       (char __user *)(uintptr_t)args->data_ptr,
+                       args->size))
+                return -EFAULT;
+
+        ret = fault_in_pages_writeable((char __user *)(uintptr_t)args->data_ptr,
+                                       args->size);
+        if (ret)
+                return -EFAULT;
+
         ret = i915_mutex_lock_interruptible(dev);
         if (ret)
                 return ret;
@@ -564,23 +577,6 @@ i915_gem_pread_ioctl(struct drm_device *dev, void *data,
                 goto out;
         }
 
-        if (args->size == 0)
-                goto out;
-
-        if (!access_ok(VERIFY_WRITE,
-                       (char __user *)(uintptr_t)args->data_ptr,
-                       args->size)) {
-                ret = -EFAULT;
-                goto out;
-        }
-
-        ret = fault_in_pages_writeable((char __user *)(uintptr_t)args->data_ptr,
-                                       args->size);
-        if (ret) {
-                ret = -EFAULT;
-                goto out;
-        }
-
         ret = i915_gem_object_get_pages_or_evict(obj);
         if (ret)
                 goto out;
@@ -981,7 +977,20 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
         struct drm_i915_gem_pwrite *args = data;
         struct drm_gem_object *obj;
         struct drm_i915_gem_object *obj_priv;
-        int ret = 0;
+        int ret;
+
+        if (args->size == 0)
+                return 0;
+
+        if (!access_ok(VERIFY_READ,
+                       (char __user *)(uintptr_t)args->data_ptr,
+                       args->size))
+                return -EFAULT;
+
+        ret = fault_in_pages_readable((char __user *)(uintptr_t)args->data_ptr,
+                                      args->size);
+        if (ret)
+                return -EFAULT;
 
         ret = i915_mutex_lock_interruptible(dev);
         if (ret)
@@ -994,30 +1003,12 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
         }
         obj_priv = to_intel_bo(obj);
 
-
         /* Bounds check destination. */
         if (args->offset > obj->size || args->size > obj->size - args->offset) {
                 ret = -EINVAL;
                 goto out;
         }
 
-        if (args->size == 0)
-                goto out;
-
-        if (!access_ok(VERIFY_READ,
-                       (char __user *)(uintptr_t)args->data_ptr,
-                       args->size)) {
-                ret = -EFAULT;
-                goto out;
-        }
-
-        ret = fault_in_pages_readable((char __user *)(uintptr_t)args->data_ptr,
-                                      args->size);
-        if (ret) {
-                ret = -EFAULT;
-                goto out;
-        }
-
         /* We can only do the GTT pwrite on untiled buffers, as otherwise
          * it would end up going through the fenced access, and we'll get
          * different detiling behavior between reading and writing.
@@ -2907,6 +2898,20 @@ i915_gem_object_set_to_display_plane(struct drm_gem_object *obj,
         return 0;
 }
 
+int
+i915_gem_object_flush_gpu(struct drm_i915_gem_object *obj,
+                          bool interruptible)
+{
+        if (!obj->active)
+                return 0;
+
+        if (obj->base.write_domain & I915_GEM_GPU_DOMAINS)
+                i915_gem_flush_ring(obj->base.dev, NULL, obj->ring,
+                                    0, obj->base.write_domain);
+
+        return i915_gem_object_wait_rendering(&obj->base, interruptible);
+}
+
 /**
  * Moves a single object to the CPU read, and possibly write domain.
  *
diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c
index c55c77043357..8df574316063 100644
--- a/drivers/gpu/drm/i915/intel_crt.c
+++ b/drivers/gpu/drm/i915/intel_crt.c
@@ -34,6 +34,25 @@
 #include "i915_drm.h"
 #include "i915_drv.h"
 
+/* Here's the desired hotplug mode */
+#define ADPA_HOTPLUG_BITS (ADPA_CRT_HOTPLUG_PERIOD_128 |                \
+                           ADPA_CRT_HOTPLUG_WARMUP_10MS |               \
+                           ADPA_CRT_HOTPLUG_SAMPLE_4S |                 \
+                           ADPA_CRT_HOTPLUG_VOLTAGE_50 |                \
+                           ADPA_CRT_HOTPLUG_VOLREF_325MV |              \
+                           ADPA_CRT_HOTPLUG_ENABLE)
+
+struct intel_crt {
+        struct intel_encoder base;
+        bool force_hotplug_required;
+};
+
+static struct intel_crt *intel_attached_crt(struct drm_connector *connector)
+{
+        return container_of(intel_attached_encoder(connector),
+                            struct intel_crt, base);
+}
+
 static void intel_crt_dpms(struct drm_encoder *encoder, int mode)
 {
         struct drm_device *dev = encoder->dev;
@@ -129,7 +148,7 @@ static void intel_crt_mode_set(struct drm_encoder *encoder,
                            dpll_md & ~DPLL_MD_UDI_MULTIPLIER_MASK);
         }
 
-        adpa = 0;
+        adpa = ADPA_HOTPLUG_BITS;
         if (adjusted_mode->flags & DRM_MODE_FLAG_PHSYNC)
                 adpa |= ADPA_HSYNC_ACTIVE_HIGH;
         if (adjusted_mode->flags & DRM_MODE_FLAG_PVSYNC)
@@ -157,53 +176,44 @@ static void intel_crt_mode_set(struct drm_encoder *encoder,
 static bool intel_ironlake_crt_detect_hotplug(struct drm_connector *connector)
 {
         struct drm_device *dev = connector->dev;
+        struct intel_crt *crt = intel_attached_crt(connector);
         struct drm_i915_private *dev_priv = dev->dev_private;
-        u32 adpa, temp;
+        u32 adpa;
         bool ret;
-        bool turn_off_dac = false;
 
-        temp = adpa = I915_READ(PCH_ADPA);
+        /* The first time through, trigger an explicit detection cycle */
+        if (crt->force_hotplug_required) {
+                bool turn_off_dac = HAS_PCH_SPLIT(dev);
+                u32 save_adpa;
 
-        if (HAS_PCH_SPLIT(dev))
-                turn_off_dac = true;
-
-        adpa &= ~ADPA_CRT_HOTPLUG_MASK;
-        if (turn_off_dac)
-                adpa &= ~ADPA_DAC_ENABLE;
-
-        /* disable HPD first */
-        I915_WRITE(PCH_ADPA, adpa);
-        (void)I915_READ(PCH_ADPA);
-
-        adpa |= (ADPA_CRT_HOTPLUG_PERIOD_128 |
-                        ADPA_CRT_HOTPLUG_WARMUP_10MS |
-                        ADPA_CRT_HOTPLUG_SAMPLE_4S |
-                        ADPA_CRT_HOTPLUG_VOLTAGE_50 | /* default */
-                        ADPA_CRT_HOTPLUG_VOLREF_325MV |
-                        ADPA_CRT_HOTPLUG_ENABLE |
-                        ADPA_CRT_HOTPLUG_FORCE_TRIGGER);
-
-        DRM_DEBUG_KMS("pch crt adpa 0x%x", adpa);
-        I915_WRITE(PCH_ADPA, adpa);
-
-        if (wait_for((I915_READ(PCH_ADPA) & ADPA_CRT_HOTPLUG_FORCE_TRIGGER) == 0,
-                     1000))
-                DRM_DEBUG_KMS("timed out waiting for FORCE_TRIGGER");
-
-        if (turn_off_dac) {
-                /* Make sure hotplug is enabled */
-                I915_WRITE(PCH_ADPA, temp | ADPA_CRT_HOTPLUG_ENABLE);
-                (void)I915_READ(PCH_ADPA);
+                crt->force_hotplug_required = 0;
+
+                save_adpa = adpa = I915_READ(PCH_ADPA);
+                DRM_DEBUG_KMS("trigger hotplug detect cycle: adpa=0x%x\n", adpa);
+
+                adpa |= ADPA_CRT_HOTPLUG_FORCE_TRIGGER;
+                if (turn_off_dac)
+                        adpa &= ~ADPA_DAC_ENABLE;
+
+                I915_WRITE(PCH_ADPA, adpa);
+
+                if (wait_for((I915_READ(PCH_ADPA) & ADPA_CRT_HOTPLUG_FORCE_TRIGGER) == 0,
+                             1000))
+                        DRM_DEBUG_KMS("timed out waiting for FORCE_TRIGGER");
+
+                if (turn_off_dac) {
+                        I915_WRITE(PCH_ADPA, save_adpa);
+                        POSTING_READ(PCH_ADPA);
+                }
         }
 
         /* Check the status to see if both blue and green are on now */
         adpa = I915_READ(PCH_ADPA);
-        adpa &= ADPA_CRT_HOTPLUG_MONITOR_MASK;
-        if ((adpa == ADPA_CRT_HOTPLUG_MONITOR_COLOR) ||
-                (adpa == ADPA_CRT_HOTPLUG_MONITOR_MONO))
+        if ((adpa & ADPA_CRT_HOTPLUG_MONITOR_MASK) != 0)
                 ret = true;
         else
                 ret = false;
+        DRM_DEBUG_KMS("ironlake hotplug adpa=0x%x, result %d\n", adpa, ret);
 
         return ret;
 }
@@ -277,13 +287,12 @@ static bool intel_crt_ddc_probe(struct drm_i915_private *dev_priv, int ddc_bus)
         return i2c_transfer(&dev_priv->gmbus[ddc_bus].adapter, msgs, 1) == 1;
 }
 
-static bool intel_crt_detect_ddc(struct drm_encoder *encoder)
+static bool intel_crt_detect_ddc(struct intel_crt *crt)
 {
-        struct intel_encoder *intel_encoder = to_intel_encoder(encoder);
-        struct drm_i915_private *dev_priv = encoder->dev->dev_private;
+        struct drm_i915_private *dev_priv = crt->base.base.dev->dev_private;
 
         /* CRT should always be at 0, but check anyway */
-        if (intel_encoder->type != INTEL_OUTPUT_ANALOG)
+        if (crt->base.type != INTEL_OUTPUT_ANALOG)
                 return false;
 
         if (intel_crt_ddc_probe(dev_priv, dev_priv->crt_ddc_pin)) {
@@ -291,7 +300,7 @@ static bool intel_crt_detect_ddc(struct drm_encoder *encoder)
                 return true;
         }
 
-        if (intel_ddc_probe(intel_encoder, dev_priv->crt_ddc_pin)) {
+        if (intel_ddc_probe(&crt->base, dev_priv->crt_ddc_pin)) {
                 DRM_DEBUG_KMS("CRT detected via DDC:0x50 [EDID]\n");
                 return true;
         }
@@ -300,9 +309,9 @@ static bool intel_crt_detect_ddc(struct drm_encoder *encoder)
 }
 
 static enum drm_connector_status
-intel_crt_load_detect(struct drm_crtc *crtc, struct intel_encoder *intel_encoder)
+intel_crt_load_detect(struct drm_crtc *crtc, struct intel_crt *crt)
 {
-        struct drm_encoder *encoder = &intel_encoder->base;
+        struct drm_encoder *encoder = &crt->base.base;
         struct drm_device *dev = encoder->dev;
         struct drm_i915_private *dev_priv = dev->dev_private;
         struct intel_crtc *intel_crtc = to_intel_crtc(crtc);
@@ -434,7 +443,7 @@ static enum drm_connector_status
 intel_crt_detect(struct drm_connector *connector, bool force)
 {
         struct drm_device *dev = connector->dev;
-        struct intel_encoder *encoder = intel_attached_encoder(connector);
+        struct intel_crt *crt = intel_attached_crt(connector);
         struct drm_crtc *crtc;
         int dpms_mode;
         enum drm_connector_status status;
@@ -443,28 +452,31 @@ intel_crt_detect(struct drm_connector *connector, bool force)
                 if (intel_crt_detect_hotplug(connector)) {
                         DRM_DEBUG_KMS("CRT detected via hotplug\n");
                         return connector_status_connected;
-                } else
+                } else {
+                        DRM_DEBUG_KMS("CRT not detected via hotplug\n");
                         return connector_status_disconnected;
+                }
         }
 
-        if (intel_crt_detect_ddc(&encoder->base))
+        if (intel_crt_detect_ddc(crt))
                 return connector_status_connected;
 
         if (!force)
                 return connector->status;
 
         /* for pre-945g platforms use load detect */
-        if (encoder->base.crtc && encoder->base.crtc->enabled) {
-                status = intel_crt_load_detect(encoder->base.crtc, encoder);
+        crtc = crt->base.base.crtc;
+        if (crtc && crtc->enabled) {
+                status = intel_crt_load_detect(crtc, crt);
         } else {
-                crtc = intel_get_load_detect_pipe(encoder, connector,
+                crtc = intel_get_load_detect_pipe(&crt->base, connector,
                                                   NULL, &dpms_mode);
                 if (crtc) {
-                        if (intel_crt_detect_ddc(&encoder->base))
+                        if (intel_crt_detect_ddc(crt))
                                 status = connector_status_connected;
                         else
-                                status = intel_crt_load_detect(crtc, encoder);
-                        intel_release_load_detect_pipe(encoder,
+                                status = intel_crt_load_detect(crtc, crt);
+                        intel_release_load_detect_pipe(&crt->base,
                                                        connector, dpms_mode);
                 } else
                         status = connector_status_unknown;
@@ -536,17 +548,17 @@ static const struct drm_encoder_funcs intel_crt_enc_funcs = {
 void intel_crt_init(struct drm_device *dev)
 {
         struct drm_connector *connector;
-        struct intel_encoder *intel_encoder;
+        struct intel_crt *crt;
         struct intel_connector *intel_connector;
         struct drm_i915_private *dev_priv = dev->dev_private;
 
-        intel_encoder = kzalloc(sizeof(struct intel_encoder), GFP_KERNEL);
-        if (!intel_encoder)
+        crt = kzalloc(sizeof(struct intel_crt), GFP_KERNEL);
+        if (!crt)
                 return;
 
         intel_connector = kzalloc(sizeof(struct intel_connector), GFP_KERNEL);
         if (!intel_connector) {
-                kfree(intel_encoder);
+                kfree(crt);
                 return;
         }
 
@@ -554,20 +566,20 @@ void intel_crt_init(struct drm_device *dev)
         drm_connector_init(dev, &intel_connector->base,
                            &intel_crt_connector_funcs, DRM_MODE_CONNECTOR_VGA);
 
-        drm_encoder_init(dev, &intel_encoder->base, &intel_crt_enc_funcs,
+        drm_encoder_init(dev, &crt->base.base, &intel_crt_enc_funcs,
                          DRM_MODE_ENCODER_DAC);
 
-        intel_connector_attach_encoder(intel_connector, intel_encoder);
+        intel_connector_attach_encoder(intel_connector, &crt->base);
 
-        intel_encoder->type = INTEL_OUTPUT_ANALOG;
-        intel_encoder->clone_mask = (1 << INTEL_SDVO_NON_TV_CLONE_BIT) |
-                                   (1 << INTEL_ANALOG_CLONE_BIT) |
-                                   (1 << INTEL_SDVO_LVDS_CLONE_BIT);
-        intel_encoder->crtc_mask = (1 << 0) | (1 << 1);
+        crt->base.type = INTEL_OUTPUT_ANALOG;
+        crt->base.clone_mask = (1 << INTEL_SDVO_NON_TV_CLONE_BIT |
+                                1 << INTEL_ANALOG_CLONE_BIT |
+                                1 << INTEL_SDVO_LVDS_CLONE_BIT);
+        crt->base.crtc_mask = (1 << 0) | (1 << 1);
         connector->interlace_allowed = 1;
         connector->doublescan_allowed = 0;
 
-        drm_encoder_helper_add(&intel_encoder->base, &intel_crt_helper_funcs);
+        drm_encoder_helper_add(&crt->base.base, &intel_crt_helper_funcs);
         drm_connector_helper_add(connector, &intel_crt_connector_helper_funcs);
 
         drm_sysfs_connector_add(connector);
@@ -577,5 +589,22 @@ void intel_crt_init(struct drm_device *dev)
         else
                 connector->polled = DRM_CONNECTOR_POLL_CONNECT;
 
+        /*
+         * Configure the automatic hotplug detection stuff
+         */
+        crt->force_hotplug_required = 0;
+        if (HAS_PCH_SPLIT(dev)) {
+                u32 adpa;
+
+                adpa = I915_READ(PCH_ADPA);
+                adpa &= ~ADPA_CRT_HOTPLUG_MASK;
+                adpa |= ADPA_HOTPLUG_BITS;
+                I915_WRITE(PCH_ADPA, adpa);
+                POSTING_READ(PCH_ADPA);
+
+                DRM_DEBUG_KMS("pch crt adpa set to 0x%x\n", adpa);
+                crt->force_hotplug_required = 1;
+        }
+
         dev_priv->hotplug_supported_mask |= CRT_HOTPLUG_INT_STATUS;
 }
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 48d8fd686ea9..bee24b1a58e8 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -1611,6 +1611,18 @@ intel_pipe_set_base(struct drm_crtc *crtc, int x, int y,
 
                 wait_event(dev_priv->pending_flip_queue,
                            atomic_read(&obj_priv->pending_flip) == 0);
+
+                /* Big Hammer, we also need to ensure that any pending
+                 * MI_WAIT_FOR_EVENT inside a user batch buffer on the
+                 * current scanout is retired before unpinning the old
+                 * framebuffer.
+                 */
+                ret = i915_gem_object_flush_gpu(obj_priv, false);
+                if (ret) {
+                        i915_gem_object_unpin(to_intel_framebuffer(crtc->fb)->obj);
+                        mutex_unlock(&dev->struct_mutex);
+                        return ret;
+                }
         }
 
         ret = intel_pipe_set_base_atomic(crtc, crtc->fb, x, y,
diff --git a/drivers/gpu/drm/i915/intel_i2c.c b/drivers/gpu/drm/i915/intel_i2c.c
index 2be4f728ed0c..3dba086e7eea 100644
--- a/drivers/gpu/drm/i915/intel_i2c.c
+++ b/drivers/gpu/drm/i915/intel_i2c.c
@@ -160,7 +160,7 @@ intel_gpio_create(struct drm_i915_private *dev_priv, u32 pin)
         };
         struct intel_gpio *gpio;
 
-        if (pin < 1 || pin > 7)
+        if (pin >= ARRAY_SIZE(map_pin_to_reg) || !map_pin_to_reg[pin])
                 return NULL;
 
         gpio = kzalloc(sizeof(struct intel_gpio), GFP_KERNEL);
@@ -172,7 +172,8 @@ intel_gpio_create(struct drm_i915_private *dev_priv, u32 pin)
                 gpio->reg += PCH_GPIOA - GPIOA;
         gpio->dev_priv = dev_priv;
 
-        snprintf(gpio->adapter.name, I2C_NAME_SIZE, "GPIO%c", "?BACDEF?"[pin]);
+        snprintf(gpio->adapter.name, sizeof(gpio->adapter.name),
+                 "i915 GPIO%c", "?BACDE?F"[pin]);
         gpio->adapter.owner = THIS_MODULE;
         gpio->adapter.algo_data = &gpio->algo;
         gpio->adapter.dev.parent = &dev_priv->dev->pdev->dev;
@@ -349,7 +350,7 @@ int intel_setup_gmbus(struct drm_device *dev)
                 "panel",
                 "dpc",
                 "dpb",
-                "reserved"
+                "reserved",
                 "dpd",
         };
         struct drm_i915_private *dev_priv = dev->dev_private;
@@ -366,8 +367,8 @@ int intel_setup_gmbus(struct drm_device *dev)
                 bus->adapter.owner = THIS_MODULE;
                 bus->adapter.class = I2C_CLASS_DDC;
                 snprintf(bus->adapter.name,
-                         I2C_NAME_SIZE,
-                         "gmbus %s",
+                         sizeof(bus->adapter.name),
+                         "i915 gmbus %s",
                          names[i]);
 
                 bus->adapter.dev.parent = &dev->pdev->dev;
diff --git a/drivers/gpu/drm/nouveau/nouveau_backlight.c b/drivers/gpu/drm/nouveau/nouveau_backlight.c
index 406228f4a2a0..b14c81110575 100644
--- a/drivers/gpu/drm/nouveau/nouveau_backlight.c
+++ b/drivers/gpu/drm/nouveau/nouveau_backlight.c
@@ -31,6 +31,7 @@
  */
 
 #include <linux/backlight.h>
+#include <linux/acpi.h>
 
 #include "drmP.h"
 #include "nouveau_drv.h"
@@ -136,6 +137,14 @@ int nouveau_backlight_init(struct drm_device *dev)
 {
         struct drm_nouveau_private *dev_priv = dev->dev_private;
 
+#ifdef CONFIG_ACPI
+        if (acpi_video_backlight_support()) {
+                NV_INFO(dev, "ACPI backlight interface available, "
+                             "not registering our own\n");
+                return 0;
+        }
+#endif
+
         switch (dev_priv->card_type) {
         case NV_40:
                 return nouveau_nv40_backlight_init(dev);
diff --git a/drivers/gpu/drm/nouveau/nouveau_bios.c b/drivers/gpu/drm/nouveau/nouveau_bios.c
index 5f21030a293b..b2293576f278 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bios.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bios.c
@@ -6829,7 +6829,7 @@ nouveau_bios_posted(struct drm_device *dev)
         struct drm_nouveau_private *dev_priv = dev->dev_private;
         unsigned htotal;
 
-        if (dev_priv->chipset >= NV_50) {
+        if (dev_priv->card_type >= NV_50) {
                 if (NVReadVgaCrtc(dev, 0, 0x00) == 0 &&
                     NVReadVgaCrtc(dev, 0, 0x1a) == 0)
                         return false;
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.c b/drivers/gpu/drm/nouveau/nouveau_bo.c
index 80353e2b8409..c41e1c200ef5 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.c
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.c
@@ -143,8 +143,10 @@ nouveau_bo_new(struct drm_device *dev, struct nouveau_channel *chan,
         nvbo->no_vm = no_vm;
         nvbo->tile_mode = tile_mode;
         nvbo->tile_flags = tile_flags;
+        nvbo->bo.bdev = &dev_priv->ttm.bdev;
 
-        nouveau_bo_fixup_align(dev, tile_mode, tile_flags, &align, &size);
+        nouveau_bo_fixup_align(dev, tile_mode, nouveau_bo_tile_layout(nvbo),
+                               &align, &size);
         align >>= PAGE_SHIFT;
 
         nouveau_bo_placement_set(nvbo, flags, 0);
@@ -176,6 +178,31 @@ set_placement_list(uint32_t *pl, unsigned *n, uint32_t type, uint32_t flags)
                 pl[(*n)++] = TTM_PL_FLAG_SYSTEM | flags;
 }
 
+static void
+set_placement_range(struct nouveau_bo *nvbo, uint32_t type)
+{
+        struct drm_nouveau_private *dev_priv = nouveau_bdev(nvbo->bo.bdev);
+
+        if (dev_priv->card_type == NV_10 &&
+            nvbo->tile_mode && (type & TTM_PL_FLAG_VRAM)) {
+                /*
+                 * Make sure that the color and depth buffers are handled
+                 * by independent memory controller units. Up to a 9x
+                 * speed up when alpha-blending and depth-test are enabled
+                 * at the same time.
+                 */
+                int vram_pages = dev_priv->vram_size >> PAGE_SHIFT;
+
+                if (nvbo->tile_flags & NOUVEAU_GEM_TILE_ZETA) {
+                        nvbo->placement.fpfn = vram_pages / 2;
+                        nvbo->placement.lpfn = ~0;
+                } else {
+                        nvbo->placement.fpfn = 0;
+                        nvbo->placement.lpfn = vram_pages / 2;
+                }
+        }
+}
+
 void
 nouveau_bo_placement_set(struct nouveau_bo *nvbo, uint32_t type, uint32_t busy)
 {
@@ -190,6 +217,8 @@ nouveau_bo_placement_set(struct nouveau_bo *nvbo, uint32_t type, uint32_t busy)
         pl->busy_placement = nvbo->busy_placements;
         set_placement_list(nvbo->busy_placements, &pl->num_busy_placement,
                            type | busy, flags);
+
+        set_placement_range(nvbo, type);
 }
 
 int
@@ -525,7 +554,8 @@ nv50_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo,
                 stride  = 16 * 4;
                 height  = amount / stride;
 
-                if (new_mem->mem_type == TTM_PL_VRAM && nvbo->tile_flags) {
+                if (new_mem->mem_type == TTM_PL_VRAM &&
+                    nouveau_bo_tile_layout(nvbo)) {
                         ret = RING_SPACE(chan, 8);
                         if (ret)
                                 return ret;
@@ -546,7 +576,8 @@ nv50_bo_move_m2mf(struct nouveau_channel *chan, struct ttm_buffer_object *bo,
                         BEGIN_RING(chan, NvSubM2MF, 0x0200, 1);
                         OUT_RING  (chan, 1);
                 }
-                if (old_mem->mem_type == TTM_PL_VRAM && nvbo->tile_flags) {
+                if (old_mem->mem_type == TTM_PL_VRAM &&
+                    nouveau_bo_tile_layout(nvbo)) {
                         ret = RING_SPACE(chan, 8);
                         if (ret)
                                 return ret;
@@ -753,7 +784,8 @@ nouveau_bo_vm_bind(struct ttm_buffer_object *bo, struct ttm_mem_reg *new_mem,
         if (dev_priv->card_type == NV_50) {
                 ret = nv50_mem_vm_bind_linear(dev,
                                               offset + dev_priv->vm_vram_base,
-                                              new_mem->size, nvbo->tile_flags,
+                                              new_mem->size,
+                                              nouveau_bo_tile_layout(nvbo),
                                               offset);
                 if (ret)
                         return ret;
@@ -894,7 +926,8 @@ nouveau_ttm_fault_reserve_notify(struct ttm_buffer_object *bo)
          * nothing to do here.
          */
         if (bo->mem.mem_type != TTM_PL_VRAM) {
-                if (dev_priv->card_type < NV_50 || !nvbo->tile_flags)
+                if (dev_priv->card_type < NV_50 ||
+                    !nouveau_bo_tile_layout(nvbo))
                         return 0;
         }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.c b/drivers/gpu/drm/nouveau/nouveau_connector.c
index 0871495096fa..52c356e9a3d1 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.c
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.c
@@ -281,7 +281,7 @@ detect_analog:
         nv_encoder = find_encoder_by_type(connector, OUTPUT_ANALOG);
         if (!nv_encoder && !nouveau_tv_disable)
                 nv_encoder = find_encoder_by_type(connector, OUTPUT_TV);
-        if (nv_encoder) {
+        if (nv_encoder && force) {
                 struct drm_encoder *encoder = to_drm_encoder(nv_encoder);
                 struct drm_encoder_helper_funcs *helper =
                                                 encoder->helper_private;
@@ -641,11 +641,28 @@ nouveau_connector_get_modes(struct drm_connector *connector)
         return ret;
 }
 
+static unsigned
+get_tmds_link_bandwidth(struct drm_connector *connector)
+{
+        struct nouveau_connector *nv_connector = nouveau_connector(connector);
+        struct drm_nouveau_private *dev_priv = connector->dev->dev_private;
+        struct dcb_entry *dcb = nv_connector->detected_encoder->dcb;
+
+        if (dcb->location != DCB_LOC_ON_CHIP ||
+            dev_priv->chipset >= 0x46)
+                return 165000;
+        else if (dev_priv->chipset >= 0x40)
+                return 155000;
+        else if (dev_priv->chipset >= 0x18)
+                return 135000;
+        else
+                return 112000;
+}
+
 static int
 nouveau_connector_mode_valid(struct drm_connector *connector,
                              struct drm_display_mode *mode)
 {
-        struct drm_nouveau_private *dev_priv = connector->dev->dev_private;
         struct nouveau_connector *nv_connector = nouveau_connector(connector);
         struct nouveau_encoder *nv_encoder = nv_connector->detected_encoder;
         struct drm_encoder *encoder = to_drm_encoder(nv_encoder);
@@ -663,11 +680,9 @@ nouveau_connector_mode_valid(struct drm_connector *connector,
                 max_clock = 400000;
                 break;
         case OUTPUT_TMDS:
-                if ((dev_priv->card_type >= NV_50 && !nouveau_duallink) ||
-                    !nv_encoder->dcb->duallink_possible)
-                        max_clock = 165000;
-                else
-                        max_clock = 330000;
+                max_clock = get_tmds_link_bandwidth(connector);
+                if (nouveau_duallink && nv_encoder->dcb->duallink_possible)
+                        max_clock *= 2;
                 break;
         case OUTPUT_ANALOG:
                 max_clock = nv_encoder->dcb->crtconf.maxfreq;
@@ -709,44 +724,6 @@ nouveau_connector_best_encoder(struct drm_connector *connector)
         return NULL;
 }
 
-void
-nouveau_connector_set_polling(struct drm_connector *connector)
-{
-        struct drm_device *dev = connector->dev;
-        struct drm_nouveau_private *dev_priv = dev->dev_private;
-        struct drm_crtc *crtc;
-        bool spare_crtc = false;
-
-        list_for_each_entry(crtc, &dev->mode_config.crtc_list, head)
-                spare_crtc |= !crtc->enabled;
-
-        connector->polled = 0;
-
-        switch (connector->connector_type) {
-        case DRM_MODE_CONNECTOR_VGA:
-        case DRM_MODE_CONNECTOR_TV:
-                if (dev_priv->card_type >= NV_50 ||
-                    (nv_gf4_disp_arch(dev) && spare_crtc))
-                        connector->polled = DRM_CONNECTOR_POLL_CONNECT;
-                break;
-
-        case DRM_MODE_CONNECTOR_DVII:
-        case DRM_MODE_CONNECTOR_DVID:
-        case DRM_MODE_CONNECTOR_HDMIA:
-        case DRM_MODE_CONNECTOR_DisplayPort:
-        case DRM_MODE_CONNECTOR_eDP:
-                if (dev_priv->card_type >= NV_50)
-                        connector->polled = DRM_CONNECTOR_POLL_HPD;
-                else if (connector->connector_type == DRM_MODE_CONNECTOR_DVID ||
-                         spare_crtc)
-                        connector->polled = DRM_CONNECTOR_POLL_CONNECT;
-                break;
-
-        default:
-                break;
-        }
-}
-
 static const struct drm_connector_helper_funcs
 nouveau_connector_helper_funcs = {
         .get_modes = nouveau_connector_get_modes,
@@ -872,6 +849,7 @@ nouveau_connector_create(struct drm_device *dev, int index)
                                         dev->mode_config.scaling_mode_property,
                                         nv_connector->scaling_mode);
                 }
+                connector->polled = DRM_CONNECTOR_POLL_CONNECT;
                 /* fall-through */
         case DCB_CONNECTOR_TV_0:
         case DCB_CONNECTOR_TV_1:
@@ -888,11 +866,16 @@ nouveau_connector_create(struct drm_device *dev, int index)
                                 dev->mode_config.dithering_mode_property,
                                 nv_connector->use_dithering ?
                                 DRM_MODE_DITHERING_ON : DRM_MODE_DITHERING_OFF);
+
+                if (dcb->type != DCB_CONNECTOR_LVDS) {
+                        if (dev_priv->card_type >= NV_50)
+                                connector->polled = DRM_CONNECTOR_POLL_HPD;
+                        else
+                                connector->polled = DRM_CONNECTOR_POLL_CONNECT;
+                }
                 break;
         }
 
-        nouveau_connector_set_polling(connector);
-
         drm_sysfs_connector_add(connector);
         dcb->drm = connector;
         return dcb->drm;
diff --git a/drivers/gpu/drm/nouveau/nouveau_connector.h b/drivers/gpu/drm/nouveau/nouveau_connector.h
index c21ed6b16f88..711b1e9203af 100644
--- a/drivers/gpu/drm/nouveau/nouveau_connector.h
+++ b/drivers/gpu/drm/nouveau/nouveau_connector.h
@@ -52,9 +52,6 @@ static inline struct nouveau_connector *nouveau_connector(
 struct drm_connector *
 nouveau_connector_create(struct drm_device *, int index);
 
-void
-nouveau_connector_set_polling(struct drm_connector *);
-
 int
 nouveau_connector_bpp(struct drm_connector *);
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.h b/drivers/gpu/drm/nouveau/nouveau_drv.h
index 3a07e580d27a..1c7db64c03bf 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drv.h
+++ b/drivers/gpu/drm/nouveau/nouveau_drv.h
@@ -100,6 +100,9 @@ struct nouveau_bo {
         int pin_refcnt;
 };
 
+#define nouveau_bo_tile_layout(nvbo)                            \
+        ((nvbo)->tile_flags & NOUVEAU_GEM_TILE_LAYOUT_MASK)
+
 static inline struct nouveau_bo *
 nouveau_bo(struct ttm_buffer_object *bo)
 {
@@ -304,6 +307,7 @@ struct nouveau_fifo_engine {
         void (*destroy_context)(struct nouveau_channel *);
         int  (*load_context)(struct nouveau_channel *);
         int  (*unload_context)(struct drm_device *);
+        void (*tlb_flush)(struct drm_device *dev);
 };
 
 struct nouveau_pgraph_object_method {
@@ -336,6 +340,7 @@ struct nouveau_pgraph_engine {
         void (*destroy_context)(struct nouveau_channel *);
         int  (*load_context)(struct nouveau_channel *);
         int  (*unload_context)(struct drm_device *);
+        void (*tlb_flush)(struct drm_device *dev);
 
         void (*set_region_tiling)(struct drm_device *dev, int i, uint32_t addr,
                                   uint32_t size, uint32_t pitch);
@@ -485,13 +490,13 @@ enum nv04_fp_display_regs {
 };
 
 struct nv04_crtc_reg {
-        unsigned char MiscOutReg;     /* */
+        unsigned char MiscOutReg;
         uint8_t CRTC[0xa0];
         uint8_t CR58[0x10];
         uint8_t Sequencer[5];
         uint8_t Graphics[9];
         uint8_t Attribute[21];
-        unsigned char DAC[768];       /* Internal Colorlookuptable */
+        unsigned char DAC[768];
 
         /* PCRTC regs */
         uint32_t fb_start;
@@ -539,43 +544,9 @@ struct nv04_output_reg {
 };
 
 struct nv04_mode_state {
-        uint32_t bpp;
-        uint32_t width;
-        uint32_t height;
-        uint32_t interlace;
-        uint32_t repaint0;
-        uint32_t repaint1;
-        uint32_t screen;
-        uint32_t scale;
-        uint32_t dither;
-        uint32_t extra;
-        uint32_t fifo;
-        uint32_t pixel;
-        uint32_t horiz;
-        int arbitration0;
-        int arbitration1;
-        uint32_t pll;
-        uint32_t pllB;
-        uint32_t vpll;
-        uint32_t vpll2;
-        uint32_t vpllB;
-        uint32_t vpll2B;
+        struct nv04_crtc_reg crtc_reg[2];
         uint32_t pllsel;
         uint32_t sel_clk;
-        uint32_t general;
-        uint32_t crtcOwner;
-        uint32_t head;
-        uint32_t head2;
-        uint32_t cursorConfig;
-        uint32_t cursor0;
-        uint32_t cursor1;
-        uint32_t cursor2;
-        uint32_t timingH;
-        uint32_t timingV;
-        uint32_t displayV;
-        uint32_t crtcSync;
-
-        struct nv04_crtc_reg crtc_reg[2];
 };
 
 enum nouveau_card_type {
@@ -613,6 +584,12 @@ struct drm_nouveau_private {
         struct work_struct irq_work;
         struct work_struct hpd_work;
 
+        struct {
+                spinlock_t lock;
+                uint32_t hpd0_bits;
+                uint32_t hpd1_bits;
+        } hpd_state;
+
         struct list_head vbl_waiting;
 
         struct {
@@ -1045,6 +1022,7 @@ extern int  nv50_fifo_create_context(struct nouveau_channel *);
 extern void nv50_fifo_destroy_context(struct nouveau_channel *);
 extern int  nv50_fifo_load_context(struct nouveau_channel *);
 extern int  nv50_fifo_unload_context(struct drm_device *);
+extern void nv50_fifo_tlb_flush(struct drm_device *dev);
 
 /* nvc0_fifo.c */
 extern int  nvc0_fifo_init(struct drm_device *);
@@ -1122,6 +1100,8 @@ extern int  nv50_graph_load_context(struct nouveau_channel *);
 extern int  nv50_graph_unload_context(struct drm_device *);
 extern void nv50_graph_context_switch(struct drm_device *);
 extern int  nv50_grctx_init(struct nouveau_grctx *);
+extern void nv50_graph_tlb_flush(struct drm_device *dev);
+extern void nv86_graph_tlb_flush(struct drm_device *dev);
 
 /* nvc0_graph.c */
 extern int  nvc0_graph_init(struct drm_device *);
@@ -1239,7 +1219,6 @@ extern u16 nouveau_bo_rd16(struct nouveau_bo *nvbo, unsigned index);
 extern void nouveau_bo_wr16(struct nouveau_bo *nvbo, unsigned index, u16 val);
 extern u32 nouveau_bo_rd32(struct nouveau_bo *nvbo, unsigned index);
 extern void nouveau_bo_wr32(struct nouveau_bo *nvbo, unsigned index, u32 val);
-extern int nouveau_bo_sync_gpu(struct nouveau_bo *, struct nouveau_channel *);
 
 /* nouveau_fence.c */
 struct nouveau_fence;
diff --git a/drivers/gpu/drm/nouveau/nouveau_fence.c b/drivers/gpu/drm/nouveau/nouveau_fence.c
index 441b12420bb1..ab1bbfbf266e 100644
--- a/drivers/gpu/drm/nouveau/nouveau_fence.c
+++ b/drivers/gpu/drm/nouveau/nouveau_fence.c
@@ -249,6 +249,7 @@ alloc_semaphore(struct drm_device *dev)
 {
         struct drm_nouveau_private *dev_priv = dev->dev_private;
         struct nouveau_semaphore *sema;
+        int ret;
 
         if (!USE_SEMA(dev))
                 return NULL;
@@ -257,10 +258,14 @@ alloc_semaphore(struct drm_device *dev)
         if (!sema)
                 goto fail;
 
+        ret = drm_mm_pre_get(&dev_priv->fence.heap);
+        if (ret)
+                goto fail;
+
         spin_lock(&dev_priv->fence.lock);
         sema->mem = drm_mm_search_free(&dev_priv->fence.heap, 4, 0, 0);
         if (sema->mem)
-                sema->mem = drm_mm_get_block(sema->mem, 4, 0);
+                sema->mem = drm_mm_get_block_atomic(sema->mem, 4, 0);
         spin_unlock(&dev_priv->fence.lock);
 
         if (!sema->mem)
diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c
index 5c4c929d7f74..9a1fdcf400c2 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.c
@@ -107,23 +107,29 @@ nouveau_gem_info(struct drm_gem_object *gem, struct drm_nouveau_gem_info *rep)
 }
 
 static bool
-nouveau_gem_tile_flags_valid(struct drm_device *dev, uint32_t tile_flags) {
-        switch (tile_flags) {
-        case 0x0000:
-        case 0x1800:
-        case 0x2800:
-        case 0x4800:
-        case 0x7000:
-        case 0x7400:
-        case 0x7a00:
-        case 0xe000:
-                break;
-        default:
-                NV_ERROR(dev, "bad page flags: 0x%08x\n", tile_flags);
-                return false;
+nouveau_gem_tile_flags_valid(struct drm_device *dev, uint32_t tile_flags)
+{
+        struct drm_nouveau_private *dev_priv = dev->dev_private;
+
+        if (dev_priv->card_type >= NV_50) {
+                switch (tile_flags & NOUVEAU_GEM_TILE_LAYOUT_MASK) {
+                case 0x0000:
+                case 0x1800:
+                case 0x2800:
+                case 0x4800:
+                case 0x7000:
+                case 0x7400:
+                case 0x7a00:
+                case 0xe000:
+                        return true;
+                }
+        } else {
+                if (!(tile_flags & NOUVEAU_GEM_TILE_LAYOUT_MASK))
+                        return true;
         }
 
-        return true;
+        NV_ERROR(dev, "bad page flags: 0x%08x\n", tile_flags);
+        return false;
 }
 
 int
diff --git a/drivers/gpu/drm/nouveau/nouveau_hw.c b/drivers/gpu/drm/nouveau/nouveau_hw.c
index bed669a54a2d..b9672a05c411 100644
--- a/drivers/gpu/drm/nouveau/nouveau_hw.c
+++ b/drivers/gpu/drm/nouveau/nouveau_hw.c
@@ -519,11 +519,11 @@ nouveau_hw_fix_bad_vpll(struct drm_device *dev, int head)
 
         struct pll_lims pll_lim;
         struct nouveau_pll_vals pv;
-        uint32_t pllreg = head ? NV_RAMDAC_VPLL2 : NV_PRAMDAC_VPLL_COEFF;
+        enum pll_types pll = head ? PLL_VPLL1 : PLL_VPLL0;
 
-        if (get_pll_limits(dev, pllreg, &pll_lim))
+        if (get_pll_limits(dev, pll, &pll_lim))
                 return;
-        nouveau_hw_get_pllvals(dev, pllreg, &pv);
+        nouveau_hw_get_pllvals(dev, pll, &pv);
 
         if (pv.M1 >= pll_lim.vco1.min_m && pv.M1 <= pll_lim.vco1.max_m &&
             pv.N1 >= pll_lim.vco1.min_n && pv.N1 <= pll_lim.vco1.max_n &&
@@ -536,7 +536,7 @@ nouveau_hw_fix_bad_vpll(struct drm_device *dev, int head)
         pv.M1 = pll_lim.vco1.max_m;
         pv.N1 = pll_lim.vco1.min_n;
         pv.log2P = pll_lim.max_usable_log2p;
-        nouveau_hw_setpll(dev, pllreg, &pv);
+        nouveau_hw_setpll(dev, pll_lim.reg, &pv);
 }
 
 /*
diff --git a/drivers/gpu/drm/nouveau/nouveau_hw.h b/drivers/gpu/drm/nouveau/nouveau_hw.h
index 869130f83602..2989090b9434 100644
--- a/drivers/gpu/drm/nouveau/nouveau_hw.h
+++ b/drivers/gpu/drm/nouveau/nouveau_hw.h
@@ -416,6 +416,25 @@ nv_fix_nv40_hw_cursor(struct drm_device *dev, int head)
 }
 
 static inline void
+nv_set_crtc_base(struct drm_device *dev, int head, uint32_t offset)
+{
+        struct drm_nouveau_private *dev_priv = dev->dev_private;
+
+        NVWriteCRTC(dev, head, NV_PCRTC_START, offset);
+
+        if (dev_priv->card_type == NV_04) {
+                /*
+                 * Hilarious, the 24th bit doesn't want to stick to
+                 * PCRTC_START...
+                 */
+                int cre_heb = NVReadVgaCrtc(dev, head, NV_CIO_CRE_HEB__INDEX);
+
+                NVWriteVgaCrtc(dev, head, NV_CIO_CRE_HEB__INDEX,
+                               (cre_heb & ~0x40) | ((offset >> 18) & 0x40));
+        }
+}
+
+static inline void
 nv_show_cursor(struct drm_device *dev, int head, bool show)
 {
         struct drm_nouveau_private *dev_priv = dev->dev_private;
diff --git a/drivers/gpu/drm/nouveau/nouveau_i2c.c b/drivers/gpu/drm/nouveau/nouveau_i2c.c
index fdd7e3de79c8..cb389d014326 100644
--- a/drivers/gpu/drm/nouveau/nouveau_i2c.c
+++ b/drivers/gpu/drm/nouveau/nouveau_i2c.c
@@ -256,7 +256,7 @@ nouveau_i2c_find(struct drm_device *dev, int index)
         if (index >= DCB_MAX_NUM_I2C_ENTRIES)
                 return NULL;
 
-        if (dev_priv->chipset >= NV_50 && (i2c->entry & 0x00000100)) {
+        if (dev_priv->card_type >= NV_50 && (i2c->entry & 0x00000100)) {
                 uint32_t reg = 0xe500, val;
 
                 if (i2c->port_type == 6) {
diff --git a/drivers/gpu/drm/nouveau/nouveau_irq.c b/drivers/gpu/drm/nouveau/nouveau_irq.c
index 6fd51a51c608..7bfd9e6c9d67 100644
--- a/drivers/gpu/drm/nouveau/nouveau_irq.c
+++ b/drivers/gpu/drm/nouveau/nouveau_irq.c
@@ -42,6 +42,13 @@
 #include "nouveau_connector.h"
 #include "nv50_display.h"
 
+static DEFINE_RATELIMIT_STATE(nouveau_ratelimit_state, 3 * HZ, 20);
+
+static int nouveau_ratelimit(void)
+{
+        return __ratelimit(&nouveau_ratelimit_state);
+}
+
 void
 nouveau_irq_preinstall(struct drm_device *dev)
 {
@@ -53,6 +60,7 @@ nouveau_irq_preinstall(struct drm_device *dev)
         if (dev_priv->card_type >= NV_50) {
                 INIT_WORK(&dev_priv->irq_work, nv50_display_irq_handler_bh);
                 INIT_WORK(&dev_priv->hpd_work, nv50_display_irq_hotplug_bh);
+                spin_lock_init(&dev_priv->hpd_state.lock);
                 INIT_LIST_HEAD(&dev_priv->vbl_waiting);
         }
 }
@@ -202,8 +210,8 @@ nouveau_fifo_irq_handler(struct drm_device *dev)
                 }
 
                 if (status & NV_PFIFO_INTR_DMA_PUSHER) {
-                        u32 get = nv_rd32(dev, 0x003244);
-                        u32 put = nv_rd32(dev, 0x003240);
+                        u32 dma_get = nv_rd32(dev, 0x003244);
+                        u32 dma_put = nv_rd32(dev, 0x003240);
                         u32 push = nv_rd32(dev, 0x003220);
                         u32 state = nv_rd32(dev, 0x003228);
 
@@ -213,16 +221,18 @@ nouveau_fifo_irq_handler(struct drm_device *dev)
                                 u32 ib_get = nv_rd32(dev, 0x003334);
                                 u32 ib_put = nv_rd32(dev, 0x003330);
 
-                                NV_INFO(dev, "PFIFO_DMA_PUSHER - Ch %d Get 0x%02x%08x "
+                                if (nouveau_ratelimit())
+                                        NV_INFO(dev, "PFIFO_DMA_PUSHER - Ch %d Get 0x%02x%08x "
                                              "Put 0x%02x%08x IbGet 0x%08x IbPut 0x%08x "
                                              "State 0x%08x Push 0x%08x\n",
-                                        chid, ho_get, get, ho_put, put, ib_get, ib_put,
-                                        state, push);
+                                                chid, ho_get, dma_get, ho_put,
+                                                dma_put, ib_get, ib_put, state,
+                                                push);
 
                                 /* METHOD_COUNT, in DMA_STATE on earlier chipsets */
                                 nv_wr32(dev, 0x003364, 0x00000000);
-                                if (get != put || ho_get != ho_put) {
-                                        nv_wr32(dev, 0x003244, put);
+                                if (dma_get != dma_put || ho_get != ho_put) {
+                                        nv_wr32(dev, 0x003244, dma_put);
                                         nv_wr32(dev, 0x003328, ho_put);
                                 } else
                                 if (ib_get != ib_put) {
@@ -231,10 +241,10 @@ nouveau_fifo_irq_handler(struct drm_device *dev)
                         } else {
                                 NV_INFO(dev, "PFIFO_DMA_PUSHER - Ch %d Get 0x%08x "
                                              "Put 0x%08x State 0x%08x Push 0x%08x\n",
-                                        chid, get, put, state, push);
+                                        chid, dma_get, dma_put, state, push);
 
-                                if (get != put)
-                                        nv_wr32(dev, 0x003244, put);
+                                if (dma_get != dma_put)
+                                        nv_wr32(dev, 0x003244, dma_put);
                         }
 
                         nv_wr32(dev, 0x003228, 0x00000000);
@@ -266,8 +276,9 @@ nouveau_fifo_irq_handler(struct drm_device *dev)
                 }
 
                 if (status) {
-                        NV_INFO(dev, "PFIFO_INTR 0x%08x - Ch %d\n",
-                                status, chid);
+                        if (nouveau_ratelimit())
+                                NV_INFO(dev, "PFIFO_INTR 0x%08x - Ch %d\n",
+                                        status, chid);
                         nv_wr32(dev, NV03_PFIFO_INTR_0, status);
                         status = 0;
                 }
@@ -544,13 +555,6 @@ nouveau_pgraph_intr_notify(struct drm_device *dev, uint32_t nsource)
                 nouveau_graph_dump_trap_info(dev, "PGRAPH_NOTIFY", &trap);
 }
 
-static DEFINE_RATELIMIT_STATE(nouveau_ratelimit_state, 3 * HZ, 20);
-
-static int nouveau_ratelimit(void)
-{
-        return __ratelimit(&nouveau_ratelimit_state);
-}
-
 
 static inline void
 nouveau_pgraph_intr_error(struct drm_device *dev, uint32_t nsource)
diff --git a/drivers/gpu/drm/nouveau/nouveau_mem.c b/drivers/gpu/drm/nouveau/nouveau_mem.c
index a163c7c612e7..fe4a30dc4b42 100644
--- a/drivers/gpu/drm/nouveau/nouveau_mem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_mem.c
@@ -33,9 +33,9 @@
 #include "drmP.h"
 #include "drm.h"
 #include "drm_sarea.h"
-#include "nouveau_drv.h"
 
-#define MIN(a,b) a < b ? a : b
+#include "nouveau_drv.h"
+#include "nouveau_pm.h"
 
 /*
  * NV10-NV40 tiling helpers
@@ -175,11 +175,10 @@ nv50_mem_vm_bind_linear(struct drm_device *dev, uint64_t virt, uint32_t size,
                         }
                 }
         }
-        dev_priv->engine.instmem.flush(dev);
 
-        nv50_vm_flush(dev, 5);
-        nv50_vm_flush(dev, 0);
-        nv50_vm_flush(dev, 4);
+        dev_priv->engine.instmem.flush(dev);
+        dev_priv->engine.fifo.tlb_flush(dev);
+        dev_priv->engine.graph.tlb_flush(dev);
         nv50_vm_flush(dev, 6);
         return 0;
 }
@@ -209,11 +208,10 @@ nv50_mem_vm_unbind(struct drm_device *dev, uint64_t virt, uint32_t size)
                         pte++;
                 }
         }
-        dev_priv->engine.instmem.flush(dev);
 
-        nv50_vm_flush(dev, 5);
-        nv50_vm_flush(dev, 0);
-        nv50_vm_flush(dev, 4);
+        dev_priv->engine.instmem.flush(dev);
+        dev_priv->engine.fifo.tlb_flush(dev);
+        dev_priv->engine.graph.tlb_flush(dev);
         nv50_vm_flush(dev, 6);
 }
 
@@ -653,6 +651,7 @@ nouveau_mem_gart_init(struct drm_device *dev)
 void
 nouveau_mem_timing_init(struct drm_device *dev)
 {
+        /* cards < NVC0 only */
         struct drm_nouveau_private *dev_priv = dev->dev_private;
         struct nouveau_pm_engine *pm = &dev_priv->engine.pm;
         struct nouveau_pm_memtimings *memtimings = &pm->memtimings;
@@ -719,14 +718,14 @@ nouveau_mem_timing_init(struct drm_device *dev)
                 tUNK_19 = 1;
                 tUNK_20 = 0;
                 tUNK_21 = 0;
-                switch (MIN(recordlen,21)) {
-                case 21:
+                switch (min(recordlen, 22)) {
+                case 22:
                         tUNK_21 = entry[21];
-                case 20:
+                case 21:
                         tUNK_20 = entry[20];
-                case 19:
+                case 20:
                         tUNK_19 = entry[19];
-                case 18:
+                case 19:
                         tUNK_18 = entry[18];
                 default:
                         tUNK_0  = entry[0];
@@ -756,24 +755,30 @@ nouveau_mem_timing_init(struct drm_device *dev)
                 timing->reg_100228 = (tUNK_12 << 16 | tUNK_11 << 8 | tUNK_10);
                 if(recordlen > 19) {
                         timing->reg_100228 += (tUNK_19 - 1) << 24;
-                } else {
+                }/* I cannot back-up this else-statement right now
+                         else {
                         timing->reg_100228 += tUNK_12 << 24;
-                }
+                }*/
 
                 /* XXX: reg_10022c */
+                timing->reg_10022c = tUNK_2 - 1;
 
                 timing->reg_100230 = (tUNK_20 << 24 | tUNK_21 << 16 |
                                       tUNK_13 << 8  | tUNK_13);
 
                 /* XXX: +6? */
                 timing->reg_100234 = (tRAS << 24 | (tUNK_19 + 6) << 8 | tRC);
-                if(tUNK_10 > tUNK_11) {
-                        timing->reg_100234 += tUNK_10 << 16;
-                } else {
-                        timing->reg_100234 += tUNK_11 << 16;
+                timing->reg_100234 += max(tUNK_10,tUNK_11) << 16;
+
+                /* XXX; reg_100238, reg_10023c
+                 * reg: 0x00??????
+                 * reg_10023c:
+                 *      0 for pre-NV50 cards
+                 *      0x????0202 for NV50+ cards (empirical evidence) */
+                if(dev_priv->card_type >= NV_50) {
+                        timing->reg_10023c = 0x202;
                 }
 
-                /* XXX; reg_100238, reg_10023c */
                 NV_DEBUG(dev, "Entry %d: 220: %08x %08x %08x %08x\n", i,
                          timing->reg_100220, timing->reg_100224,
                          timing->reg_100228, timing->reg_10022c);
diff --git a/drivers/gpu/drm/nouveau/nouveau_object.c b/drivers/gpu/drm/nouveau/nouveau_object.c
index 896cf8634144..dd572adca02a 100644
--- a/drivers/gpu/drm/nouveau/nouveau_object.c
+++ b/drivers/gpu/drm/nouveau/nouveau_object.c
@@ -129,7 +129,7 @@ nouveau_gpuobj_new(struct drm_device *dev, struct nouveau_channel *chan,
                         if (ramin == NULL) {
                                 spin_unlock(&dev_priv->ramin_lock);
                                 nouveau_gpuobj_ref(NULL, &gpuobj);
-                                return ret;
+                                return -ENOMEM;
                         }
 
                         ramin = drm_mm_get_block_atomic(ramin, size, align);
diff --git a/drivers/gpu/drm/nouveau/nouveau_pm.c b/drivers/gpu/drm/nouveau/nouveau_pm.c
index 1c99c55d6d46..9f7b158f5825 100644
--- a/drivers/gpu/drm/nouveau/nouveau_pm.c
+++ b/drivers/gpu/drm/nouveau/nouveau_pm.c
@@ -284,6 +284,7 @@ nouveau_sysfs_fini(struct drm_device *dev)
         }
 }
 
+#ifdef CONFIG_HWMON
 static ssize_t
 nouveau_hwmon_show_temp(struct device *d, struct device_attribute *a, char *buf)
 {
@@ -395,10 +396,12 @@ static struct attribute *hwmon_attributes[] = {
 static const struct attribute_group hwmon_attrgroup = {
         .attrs = hwmon_attributes,
 };
+#endif
 
 static int
 nouveau_hwmon_init(struct drm_device *dev)
 {
+#ifdef CONFIG_HWMON
         struct drm_nouveau_private *dev_priv = dev->dev_private;
         struct nouveau_pm_engine *pm = &dev_priv->engine.pm;
         struct device *hwmon_dev;
@@ -425,13 +428,14 @@ nouveau_hwmon_init(struct drm_device *dev)
         }
 
         pm->hwmon = hwmon_dev;
-
+#endif
         return 0;
 }
 
 static void
 nouveau_hwmon_fini(struct drm_device *dev)
 {
+#ifdef CONFIG_HWMON
         struct drm_nouveau_private *dev_priv = dev->dev_private;
         struct nouveau_pm_engine *pm = &dev_priv->engine.pm;
 
@@ -439,6 +443,7 @@ nouveau_hwmon_fini(struct drm_device *dev)
                 sysfs_remove_group(&pm->hwmon->kobj, &hwmon_attrgroup);
                 hwmon_device_unregister(pm->hwmon);
         }
+#endif
 }
 
 int
diff --git a/drivers/gpu/drm/nouveau/nouveau_ramht.c b/drivers/gpu/drm/nouveau/nouveau_ramht.c
index 7f16697cc96c..2d8580927ca4 100644
--- a/drivers/gpu/drm/nouveau/nouveau_ramht.c
+++ b/drivers/gpu/drm/nouveau/nouveau_ramht.c
@@ -153,26 +153,42 @@ nouveau_ramht_insert(struct nouveau_channel *chan, u32 handle,
         return -ENOMEM;
 }
 
+static struct nouveau_ramht_entry *
+nouveau_ramht_remove_entry(struct nouveau_channel *chan, u32 handle)
+{
+        struct nouveau_ramht *ramht = chan ? chan->ramht : NULL;
+        struct nouveau_ramht_entry *entry;
+        unsigned long flags;
+
+        if (!ramht)
+                return NULL;
+
+        spin_lock_irqsave(&ramht->lock, flags);
+        list_for_each_entry(entry, &ramht->entries, head) {
+                if (entry->channel == chan &&
+                    (!handle || entry->handle == handle)) {
+                        list_del(&entry->head);
+                        spin_unlock_irqrestore(&ramht->lock, flags);
+
+                        return entry;
+                }
+        }
+        spin_unlock_irqrestore(&ramht->lock, flags);
+
+        return NULL;
+}
+
 static void
-nouveau_ramht_remove_locked(struct nouveau_channel *chan, u32 handle)
+nouveau_ramht_remove_hash(struct nouveau_channel *chan, u32 handle)
 {
         struct drm_device *dev = chan->dev;
         struct drm_nouveau_private *dev_priv = dev->dev_private;
         struct nouveau_instmem_engine *instmem = &dev_priv->engine.instmem;
         struct nouveau_gpuobj *ramht = chan->ramht->gpuobj;
-        struct nouveau_ramht_entry *entry, *tmp;
+        unsigned long flags;
         u32 co, ho;
 
-        list_for_each_entry_safe(entry, tmp, &chan->ramht->entries, head) {
-                if (entry->channel != chan || entry->handle != handle)
-                        continue;
-
-                nouveau_gpuobj_ref(NULL, &entry->gpuobj);
-                list_del(&entry->head);
-                kfree(entry);
-                break;
-        }
-
+        spin_lock_irqsave(&chan->ramht->lock, flags);
         co = ho = nouveau_ramht_hash_handle(chan, handle);
         do {
                 if (nouveau_ramht_entry_valid(dev, ramht, co) &&
@@ -184,7 +200,7 @@ nouveau_ramht_remove_locked(struct nouveau_channel *chan, u32 handle)
                         nv_wo32(ramht, co + 0, 0x00000000);
                         nv_wo32(ramht, co + 4, 0x00000000);
                         instmem->flush(dev);
-                        return;
+                        goto out;
                 }
 
                 co += 8;
@@ -194,17 +210,22 @@ nouveau_ramht_remove_locked(struct nouveau_channel *chan, u32 handle)
 
         NV_ERROR(dev, "RAMHT entry not found. ch=%d, handle=0x%08x\n",
                  chan->id, handle);
+out:
+        spin_unlock_irqrestore(&chan->ramht->lock, flags);
 }
 
 void
 nouveau_ramht_remove(struct nouveau_channel *chan, u32 handle)
 {
-        struct nouveau_ramht *ramht = chan->ramht;
-        unsigned long flags;
+        struct nouveau_ramht_entry *entry;
 
-        spin_lock_irqsave(&ramht->lock, flags);
-        nouveau_ramht_remove_locked(chan, handle);
-        spin_unlock_irqrestore(&ramht->lock, flags);
+        entry = nouveau_ramht_remove_entry(chan, handle);
+        if (!entry)
+                return;
+
+        nouveau_ramht_remove_hash(chan, entry->handle);
+        nouveau_gpuobj_ref(NULL, &entry->gpuobj);
+        kfree(entry);
 }
 
 struct nouveau_gpuobj *
@@ -265,23 +286,19 @@ void
 nouveau_ramht_ref(struct nouveau_ramht *ref, struct nouveau_ramht **ptr,
                   struct nouveau_channel *chan)
 {
-        struct nouveau_ramht_entry *entry, *tmp;
+        struct nouveau_ramht_entry *entry;
         struct nouveau_ramht *ramht;
-        unsigned long flags;
 
         if (ref)
                 kref_get(&ref->refcount);
 
         ramht = *ptr;
         if (ramht) {
-                spin_lock_irqsave(&ramht->lock, flags);
-                list_for_each_entry_safe(entry, tmp, &ramht->entries, head) {
-                        if (entry->channel != chan)
-                                continue;
-
-                        nouveau_ramht_remove_locked(chan, entry->handle);
+                while ((entry = nouveau_ramht_remove_entry(chan, 0))) {
+                        nouveau_ramht_remove_hash(chan, entry->handle);
+                        nouveau_gpuobj_ref(NULL, &entry->gpuobj);
+                        kfree(entry);
                 }
-                spin_unlock_irqrestore(&ramht->lock, flags);
 
                 kref_put(&ramht->refcount, nouveau_ramht_del);
         }
diff --git a/drivers/gpu/drm/nouveau/nouveau_sgdma.c b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
index 288bacac7e5a..d4ac97007038 100644
--- a/drivers/gpu/drm/nouveau/nouveau_sgdma.c
+++ b/drivers/gpu/drm/nouveau/nouveau_sgdma.c
@@ -120,8 +120,8 @@ nouveau_sgdma_bind(struct ttm_backend *be, struct ttm_mem_reg *mem)
         dev_priv->engine.instmem.flush(nvbe->dev);
 
         if (dev_priv->card_type == NV_50) {
-                nv50_vm_flush(dev, 5); /* PGRAPH */
-                nv50_vm_flush(dev, 0); /* PFIFO */
+                dev_priv->engine.fifo.tlb_flush(dev);
+                dev_priv->engine.graph.tlb_flush(dev);
         }
 
         nvbe->bound = true;
@@ -162,8 +162,8 @@ nouveau_sgdma_unbind(struct ttm_backend *be)
         dev_priv->engine.instmem.flush(nvbe->dev);
 
         if (dev_priv->card_type == NV_50) {
-                nv50_vm_flush(dev, 5);
-                nv50_vm_flush(dev, 0);
+                dev_priv->engine.fifo.tlb_flush(dev);
+                dev_priv->engine.graph.tlb_flush(dev);
         }
 
         nvbe->bound = false;
@@ -224,7 +224,11 @@ nouveau_sgdma_init(struct drm_device *dev)
         int i, ret;
 
         if (dev_priv->card_type < NV_50) {
-                aper_size = (64 * 1024 * 1024);
+                if(dev_priv->ramin_rsvd_vram < 2 * 1024 * 1024)
+                        aper_size = 64 * 1024 * 1024;
+                else
+                        aper_size = 512 * 1024 * 1024;
+
                 obj_size  = (aper_size >> NV_CTXDMA_PAGE_SHIFT) * 4;
                 obj_size += 8; /* ctxdma header */
         } else {
diff --git a/drivers/gpu/drm/nouveau/nouveau_state.c b/drivers/gpu/drm/nouveau/nouveau_state.c
index ed7757f14083..049f755567e5 100644
--- a/drivers/gpu/drm/nouveau/nouveau_state.c
+++ b/drivers/gpu/drm/nouveau/nouveau_state.c
@@ -354,6 +354,15 @@ static int nouveau_init_engine_ptrs(struct drm_device *dev)
                 engine->graph.destroy_context   = nv50_graph_destroy_context;
                 engine->graph.load_context      = nv50_graph_load_context;
                 engine->graph.unload_context    = nv50_graph_unload_context;
+                if (dev_priv->chipset != 0x86)
+                        engine->graph.tlb_flush = nv50_graph_tlb_flush;
+                else {
+                        /* from what i can see nvidia do this on every
+                         * pre-NVA3 board except NVAC, but, we've only
+                         * ever seen problems on NV86
+                         */
+                        engine->graph.tlb_flush = nv86_graph_tlb_flush;
+                }
                 engine->fifo.channels           = 128;
                 engine->fifo.init               = nv50_fifo_init;
                 engine->fifo.takedown           = nv50_fifo_takedown;
@@ -365,6 +374,7 @@ static int nouveau_init_engine_ptrs(struct drm_device *dev)
                 engine->fifo.destroy_context    = nv50_fifo_destroy_context;
                 engine->fifo.load_context       = nv50_fifo_load_context;
                 engine->fifo.unload_context     = nv50_fifo_unload_context;
+                engine->fifo.tlb_flush          = nv50_fifo_tlb_flush;
                 engine->display.early_init      = nv50_display_early_init;
                 engine->display.late_takedown   = nv50_display_late_takedown;
                 engine->display.create          = nv50_display_create;
@@ -1041,6 +1051,9 @@ int nouveau_ioctl_getparam(struct drm_device *dev, void *data,
         case NOUVEAU_GETPARAM_PTIMER_TIME:
                 getparam->value = dev_priv->engine.timer.read(dev);
                 break;
+        case NOUVEAU_GETPARAM_HAS_BO_USAGE:
+                getparam->value = 1;
+                break;
         case NOUVEAU_GETPARAM_GRAPH_UNITS:
                 /* NV40 and NV50 versions are quite different, but register
                  * address is the same. User is supposed to know the card
@@ -1051,7 +1064,7 @@ int nouveau_ioctl_getparam(struct drm_device *dev, void *data,
                 }
                 /* FALLTHRU */
         default:
-                NV_ERROR(dev, "unknown parameter %lld\n", getparam->param);
+                NV_DEBUG(dev, "unknown parameter %lld\n", getparam->param);
                 return -EINVAL;
         }
 
@@ -1066,7 +1079,7 @@ nouveau_ioctl_setparam(struct drm_device *dev, void *data,
 
         switch (setparam->param) {
         default:
-                NV_ERROR(dev, "unknown parameter %lld\n", setparam->param);
+                NV_DEBUG(dev, "unknown parameter %lld\n", setparam->param);
                 return -EINVAL;
         }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_temp.c b/drivers/gpu/drm/nouveau/nouveau_temp.c
index 16bbbf1eff63..7ecc4adc1e45 100644
--- a/drivers/gpu/drm/nouveau/nouveau_temp.c
+++ b/drivers/gpu/drm/nouveau/nouveau_temp.c
@@ -191,7 +191,7 @@ nv40_temp_get(struct drm_device *dev)
         int offset = sensor->offset_mult / sensor->offset_div;
         int core_temp;
 
-        if (dev_priv->chipset >= 0x50) {
+        if (dev_priv->card_type >= NV_50) {
                 core_temp = nv_rd32(dev, 0x20008);
         } else {
                 core_temp = nv_rd32(dev, 0x0015b4) & 0x1fff;
diff --git a/drivers/gpu/drm/nouveau/nv04_crtc.c b/drivers/gpu/drm/nouveau/nv04_crtc.c
index c71abc2a34d5..40e180741629 100644
--- a/drivers/gpu/drm/nouveau/nv04_crtc.c
+++ b/drivers/gpu/drm/nouveau/nv04_crtc.c
@@ -158,7 +158,6 @@ nv_crtc_dpms(struct drm_crtc *crtc, int mode)
 {
         struct nouveau_crtc *nv_crtc = nouveau_crtc(crtc);
         struct drm_device *dev = crtc->dev;
-        struct drm_connector *connector;
         unsigned char seq1 = 0, crtc17 = 0;
         unsigned char crtc1A;
 
@@ -213,10 +212,6 @@ nv_crtc_dpms(struct drm_crtc *crtc, int mode)
         NVVgaSeqReset(dev, nv_crtc->index, false);
 
         NVWriteVgaCrtc(dev, nv_crtc->index, NV_CIO_CRE_RPC1_INDEX, crtc1A);
-
-        /* Update connector polling modes */
-        list_for_each_entry(connector, &dev->mode_config.connector_list, head)
-                nouveau_connector_set_polling(connector);
 }
 
 static bool
@@ -831,7 +826,7 @@ nv04_crtc_do_mode_set_base(struct drm_crtc *crtc,
         /* Update the framebuffer location. */
         regp->fb_start = nv_crtc->fb.offset & ~3;
         regp->fb_start += (y * drm_fb->pitch) + (x * drm_fb->bits_per_pixel / 8);
-        NVWriteCRTC(dev, nv_crtc->index, NV_PCRTC_START, regp->fb_start);
+        nv_set_crtc_base(dev, nv_crtc->index, regp->fb_start);
 
         /* Update the arbitration parameters. */
         nouveau_calc_arb(dev, crtc->mode.clock, drm_fb->bits_per_pixel,
diff --git a/drivers/gpu/drm/nouveau/nv04_dfp.c b/drivers/gpu/drm/nouveau/nv04_dfp.c
index c936403b26e2..ef23550407b5 100644
--- a/drivers/gpu/drm/nouveau/nv04_dfp.c
+++ b/drivers/gpu/drm/nouveau/nv04_dfp.c
@@ -185,14 +185,15 @@ static bool nv04_dfp_mode_fixup(struct drm_encoder *encoder,
         struct nouveau_encoder *nv_encoder = nouveau_encoder(encoder);
         struct nouveau_connector *nv_connector = nouveau_encoder_connector_get(nv_encoder);
 
-        /* For internal panels and gpu scaling on DVI we need the native mode */
-        if (nv_connector->scaling_mode != DRM_MODE_SCALE_NONE) {
-                if (!nv_connector->native_mode)
-                        return false;
+        if (!nv_connector->native_mode ||
+            nv_connector->scaling_mode == DRM_MODE_SCALE_NONE ||
+            mode->hdisplay > nv_connector->native_mode->hdisplay ||
+            mode->vdisplay > nv_connector->native_mode->vdisplay) {
+                nv_encoder->mode = *adjusted_mode;
+
+        } else {
                 nv_encoder->mode = *nv_connector->native_mode;
                 adjusted_mode->clock = nv_connector->native_mode->clock;
-        } else {
-                nv_encoder->mode = *adjusted_mode;
         }
 
         return true;
diff --git a/drivers/gpu/drm/nouveau/nv04_pm.c b/drivers/gpu/drm/nouveau/nv04_pm.c
index 6a6eb697d38e..eb1c70dd82ed 100644
--- a/drivers/gpu/drm/nouveau/nv04_pm.c
+++ b/drivers/gpu/drm/nouveau/nv04_pm.c
@@ -76,6 +76,15 @@ nv04_pm_clock_set(struct drm_device *dev, void *pre_state)
                 reg += 4;
 
         nouveau_hw_setpll(dev, reg, &state->calc);
+
+        if (dev_priv->card_type < NV_30 && reg == NV_PRAMDAC_MPLL_COEFF) {
+                if (dev_priv->card_type == NV_20)
+                        nv_mask(dev, 0x1002c4, 0, 1 << 20);
+
+                /* Reset the DLLs */
+                nv_mask(dev, 0x1002c0, 0, 1 << 8);
+        }
+
         kfree(state);
 }
 
diff --git a/drivers/gpu/drm/nouveau/nv50_calc.c b/drivers/gpu/drm/nouveau/nv50_calc.c
index 2cdc2bfe7179..de81151648f8 100644
--- a/drivers/gpu/drm/nouveau/nv50_calc.c
+++ b/drivers/gpu/drm/nouveau/nv50_calc.c
@@ -51,24 +51,28 @@ nv50_calc_pll2(struct drm_device *dev, struct pll_lims *pll, int clk,
                int *N, int *fN, int *M, int *P)
 {
         fixed20_12 fb_div, a, b;
+        u32 refclk = pll->refclk / 10;
+        u32 max_vco_freq = pll->vco1.maxfreq / 10;
+        u32 max_vco_inputfreq = pll->vco1.max_inputfreq / 10;
+        clk /= 10;
 
-        *P = pll->vco1.maxfreq / clk;
+        *P = max_vco_freq / clk;
         if (*P > pll->max_p)
                 *P = pll->max_p;
         if (*P < pll->min_p)
                 *P = pll->min_p;
 
-        /* *M = ceil(refclk / pll->vco.max_inputfreq); */
-        a.full = dfixed_const(pll->refclk);
-        b.full = dfixed_const(pll->vco1.max_inputfreq);
+        /* *M = floor((refclk + max_vco_inputfreq) / max_vco_inputfreq); */
+        a.full = dfixed_const(refclk + max_vco_inputfreq);
+        b.full = dfixed_const(max_vco_inputfreq);
         a.full = dfixed_div(a, b);
-        a.full = dfixed_ceil(a);
+        a.full = dfixed_floor(a);
         *M = dfixed_trunc(a);
 
         /* fb_div = (vco * *M) / refclk; */
         fb_div.full = dfixed_const(clk * *P);
         fb_div.full = dfixed_mul(fb_div, a);
-        a.full = dfixed_const(pll->refclk);
+        a.full = dfixed_const(refclk);
         fb_div.full = dfixed_div(fb_div, a);
 
         /* *N = floor(fb_div); */
diff --git a/drivers/gpu/drm/nouveau/nv50_crtc.c b/drivers/gpu/drm/nouveau/nv50_crtc.c
index 16380d52cd88..56476d0c6de8 100644
--- a/drivers/gpu/drm/nouveau/nv50_crtc.c
+++ b/drivers/gpu/drm/nouveau/nv50_crtc.c
@@ -546,7 +546,7 @@ nv50_crtc_do_mode_set_base(struct drm_crtc *crtc,
         }
 
         nv_crtc->fb.offset = fb->nvbo->bo.offset - dev_priv->vm_vram_base;
-        nv_crtc->fb.tile_flags = fb->nvbo->tile_flags;
+        nv_crtc->fb.tile_flags = nouveau_bo_tile_layout(fb->nvbo);
         nv_crtc->fb.cpp = drm_fb->bits_per_pixel / 8;
         if (!nv_crtc->fb.blanked && dev_priv->chipset != 0x50) {
                 ret = RING_SPACE(evo, 2);
@@ -578,7 +578,7 @@ nv50_crtc_do_mode_set_base(struct drm_crtc *crtc,
                                   fb->nvbo->tile_mode);
         }
         if (dev_priv->chipset == 0x50)
-                OUT_RING(evo, (fb->nvbo->tile_flags << 8) | format);
+                OUT_RING(evo, (nv_crtc->fb.tile_flags << 8) | format);
         else
                 OUT_RING(evo, format);
 
diff --git a/drivers/gpu/drm/nouveau/nv50_display.c b/drivers/gpu/drm/nouveau/nv50_display.c
index 55c9663ef2bf..f624c611ddea 100644
--- a/drivers/gpu/drm/nouveau/nv50_display.c
+++ b/drivers/gpu/drm/nouveau/nv50_display.c
@@ -1032,11 +1032,18 @@ nv50_display_irq_hotplug_bh(struct work_struct *work)
         struct drm_connector *connector;
         const uint32_t gpio_reg[4] = { 0xe104, 0xe108, 0xe280, 0xe284 };
         uint32_t unplug_mask, plug_mask, change_mask;
-        uint32_t hpd0, hpd1 = 0;
+        uint32_t hpd0, hpd1;
 
-        hpd0 = nv_rd32(dev, 0xe054) & nv_rd32(dev, 0xe050);
+        spin_lock_irq(&dev_priv->hpd_state.lock);
+        hpd0 = dev_priv->hpd_state.hpd0_bits;
+        dev_priv->hpd_state.hpd0_bits = 0;
+        hpd1 = dev_priv->hpd_state.hpd1_bits;
+        dev_priv->hpd_state.hpd1_bits = 0;
+        spin_unlock_irq(&dev_priv->hpd_state.lock);
+
+        hpd0 &= nv_rd32(dev, 0xe050);
         if (dev_priv->chipset >= 0x90)
-                hpd1 = nv_rd32(dev, 0xe074) & nv_rd32(dev, 0xe070);
+                hpd1 &= nv_rd32(dev, 0xe070);
 
         plug_mask   = (hpd0 & 0x0000ffff) | (hpd1 << 16);
         unplug_mask = (hpd0 >> 16) | (hpd1 & 0xffff0000);
@@ -1078,10 +1085,6 @@ nv50_display_irq_hotplug_bh(struct work_struct *work)
                         helper->dpms(connector->encoder, DRM_MODE_DPMS_OFF);
         }
 
-        nv_wr32(dev, 0xe054, nv_rd32(dev, 0xe054));
-        if (dev_priv->chipset >= 0x90)
-                nv_wr32(dev, 0xe074, nv_rd32(dev, 0xe074));
-
         drm_helper_hpd_irq_event(dev);
 }
 
@@ -1092,8 +1095,22 @@ nv50_display_irq_handler(struct drm_device *dev)
         uint32_t delayed = 0;
 
         if (nv_rd32(dev, NV50_PMC_INTR_0) & NV50_PMC_INTR_0_HOTPLUG) {
-                if (!work_pending(&dev_priv->hpd_work))
-                        queue_work(dev_priv->wq, &dev_priv->hpd_work);
+                uint32_t hpd0_bits, hpd1_bits = 0;
+
+                hpd0_bits = nv_rd32(dev, 0xe054);
+                nv_wr32(dev, 0xe054, hpd0_bits);
+
+                if (dev_priv->chipset >= 0x90) {
+                        hpd1_bits = nv_rd32(dev, 0xe074);
+                        nv_wr32(dev, 0xe074, hpd1_bits);
+                }
+
+                spin_lock(&dev_priv->hpd_state.lock);
+                dev_priv->hpd_state.hpd0_bits |= hpd0_bits;
+                dev_priv->hpd_state.hpd1_bits |= hpd1_bits;
+                spin_unlock(&dev_priv->hpd_state.lock);
+
+                queue_work(dev_priv->wq, &dev_priv->hpd_work);
         }
 
         while (nv_rd32(dev, NV50_PMC_INTR_0) & NV50_PMC_INTR_0_DISPLAY) {
diff --git a/drivers/gpu/drm/nouveau/nv50_fifo.c b/drivers/gpu/drm/nouveau/nv50_fifo.c
index a46a961102f3..1da65bd60c10 100644
--- a/drivers/gpu/drm/nouveau/nv50_fifo.c
+++ b/drivers/gpu/drm/nouveau/nv50_fifo.c
@@ -464,3 +464,8 @@ nv50_fifo_unload_context(struct drm_device *dev)
         return 0;
 }
 
+void
+nv50_fifo_tlb_flush(struct drm_device *dev)
+{
+        nv50_vm_flush(dev, 5);
+}
diff --git a/drivers/gpu/drm/nouveau/nv50_graph.c b/drivers/gpu/drm/nouveau/nv50_graph.c
index cbf5ae2f67d4..8b669d0af610 100644
--- a/drivers/gpu/drm/nouveau/nv50_graph.c
+++ b/drivers/gpu/drm/nouveau/nv50_graph.c
@@ -402,3 +402,55 @@ struct nouveau_pgraph_object_class nv50_graph_grclass[] = {
         { 0x8597, false, NULL }, /* tesla (nva3, nva5, nva8) */
         {}
 };
+
+void
+nv50_graph_tlb_flush(struct drm_device *dev)
+{
+        nv50_vm_flush(dev, 0);
+}
+
+void
+nv86_graph_tlb_flush(struct drm_device *dev)
+{
+        struct drm_nouveau_private *dev_priv = dev->dev_private;
+        struct nouveau_timer_engine *ptimer = &dev_priv->engine.timer;
+        bool idle, timeout = false;
+        unsigned long flags;
+        u64 start;
+        u32 tmp;
+
+        spin_lock_irqsave(&dev_priv->context_switch_lock, flags);
+        nv_mask(dev, 0x400500, 0x00000001, 0x00000000);
+
+        start = ptimer->read(dev);
+        do {
+                idle = true;
+
+                for (tmp = nv_rd32(dev, 0x400380); tmp && idle; tmp >>= 3) {
+                        if ((tmp & 7) == 1)
+                                idle = false;
+                }
+
+                for (tmp = nv_rd32(dev, 0x400384); tmp && idle; tmp >>= 3) {
+                        if ((tmp & 7) == 1)
+                                idle = false;
+                }
+
+                for (tmp = nv_rd32(dev, 0x400388); tmp && idle; tmp >>= 3) {
+                        if ((tmp & 7) == 1)
+                                idle = false;
+                }
+        } while (!idle && !(timeout = ptimer->read(dev) - start > 2000000000));
+
+        if (timeout) {
+                NV_ERROR(dev, "PGRAPH TLB flush idle timeout fail: "
+                              "0x%08x 0x%08x 0x%08x 0x%08x\n",
+                         nv_rd32(dev, 0x400700), nv_rd32(dev, 0x400380),
+                         nv_rd32(dev, 0x400384), nv_rd32(dev, 0x400388));
+        }
+
+        nv50_vm_flush(dev, 0);
+
+        nv_mask(dev, 0x400500, 0x00000001, 0x00000001);
+        spin_unlock_irqrestore(&dev_priv->context_switch_lock, flags);
+}
diff --git a/drivers/gpu/drm/nouveau/nv50_instmem.c b/drivers/gpu/drm/nouveau/nv50_instmem.c
index a53fc974332b..b773229b7647 100644
--- a/drivers/gpu/drm/nouveau/nv50_instmem.c
+++ b/drivers/gpu/drm/nouveau/nv50_instmem.c
@@ -402,7 +402,6 @@ nv50_instmem_bind(struct drm_device *dev, struct nouveau_gpuobj *gpuobj)
         }
         dev_priv->engine.instmem.flush(dev);
 
-        nv50_vm_flush(dev, 4);
         nv50_vm_flush(dev, 6);
 
         gpuobj->im_bound = 1;
diff --git a/drivers/gpu/drm/radeon/evergreen.c b/drivers/gpu/drm/radeon/evergreen.c
index 488c36c8f5e6..4dc5b4714c5a 100644
--- a/drivers/gpu/drm/radeon/evergreen.c
+++ b/drivers/gpu/drm/radeon/evergreen.c
@@ -1650,7 +1650,36 @@ static void evergreen_gpu_init(struct radeon_device *rdev)
                 }
         }
 
-        rdev->config.evergreen.tile_config = gb_addr_config;
+        /* setup tiling info dword.  gb_addr_config is not adequate since it does
+         * not have bank info, so create a custom tiling dword.
+         * bits 3:0   num_pipes
+         * bits 7:4   num_banks
+         * bits 11:8  group_size
+         * bits 15:12 row_size
+         */
+        rdev->config.evergreen.tile_config = 0;
+        switch (rdev->config.evergreen.max_tile_pipes) {
+        case 1:
+        default:
+                rdev->config.evergreen.tile_config |= (0 << 0);
+                break;
+        case 2:
+                rdev->config.evergreen.tile_config |= (1 << 0);
+                break;
+        case 4:
+                rdev->config.evergreen.tile_config |= (2 << 0);
+                break;
+        case 8:
+                rdev->config.evergreen.tile_config |= (3 << 0);
+                break;
+        }
+        rdev->config.evergreen.tile_config |=
+                ((mc_arb_ramcfg & NOOFBANK_MASK) >> NOOFBANK_SHIFT) << 4;
+        rdev->config.evergreen.tile_config |=
+                ((mc_arb_ramcfg & BURSTLENGTH_MASK) >> BURSTLENGTH_SHIFT) << 8;
+        rdev->config.evergreen.tile_config |=
+                ((gb_addr_config & 0x30000000) >> 28) << 12;
+
         WREG32(GB_BACKEND_MAP, gb_backend_map);
         WREG32(GB_ADDR_CONFIG, gb_addr_config);
         WREG32(DMIF_ADDR_CONFIG, gb_addr_config);
diff --git a/drivers/gpu/drm/radeon/evergreen_blit_kms.c b/drivers/gpu/drm/radeon/evergreen_blit_kms.c
index ac3b6dde23db..e0e590110dd4 100644
--- a/drivers/gpu/drm/radeon/evergreen_blit_kms.c
+++ b/drivers/gpu/drm/radeon/evergreen_blit_kms.c
@@ -459,7 +459,7 @@ int evergreen_blit_init(struct radeon_device *rdev)
         obj_size += evergreen_ps_size * 4;
         obj_size = ALIGN(obj_size, 256);
 
-        r = radeon_bo_create(rdev, NULL, obj_size, true, RADEON_GEM_DOMAIN_VRAM,
+        r = radeon_bo_create(rdev, NULL, obj_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
                                 &rdev->r600_blit.shader_obj);
         if (r) {
                 DRM_ERROR("evergreen failed to allocate shader\n");
diff --git a/drivers/gpu/drm/radeon/r600.c b/drivers/gpu/drm/radeon/r600.c
index 0f806cc7dc75..a3552594ccc4 100644
--- a/drivers/gpu/drm/radeon/r600.c
+++ b/drivers/gpu/drm/radeon/r600.c
@@ -2718,7 +2718,7 @@ static int r600_ih_ring_alloc(struct radeon_device *rdev)
         /* Allocate ring buffer */
         if (rdev->ih.ring_obj == NULL) {
                 r = radeon_bo_create(rdev, NULL, rdev->ih.ring_size,
-                                     true,
+                                     PAGE_SIZE, true,
                                      RADEON_GEM_DOMAIN_GTT,
                                      &rdev->ih.ring_obj);
                 if (r) {
diff --git a/drivers/gpu/drm/radeon/r600_blit_kms.c b/drivers/gpu/drm/radeon/r600_blit_kms.c
index 8362974ef41a..86e5aa07f0db 100644
--- a/drivers/gpu/drm/radeon/r600_blit_kms.c
+++ b/drivers/gpu/drm/radeon/r600_blit_kms.c
@@ -501,7 +501,7 @@ int r600_blit_init(struct radeon_device *rdev)
         obj_size += r6xx_ps_size * 4;
         obj_size = ALIGN(obj_size, 256);
 
-        r = radeon_bo_create(rdev, NULL, obj_size, true, RADEON_GEM_DOMAIN_VRAM,
+        r = radeon_bo_create(rdev, NULL, obj_size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
                                 &rdev->r600_blit.shader_obj);
         if (r) {
                 DRM_ERROR("r600 failed to allocate shader\n");
diff --git a/drivers/gpu/drm/radeon/r600_cs.c b/drivers/gpu/drm/radeon/r600_cs.c
index 37cc2aa9f923..9bebac1ec006 100644
--- a/drivers/gpu/drm/radeon/r600_cs.c
+++ b/drivers/gpu/drm/radeon/r600_cs.c
@@ -50,6 +50,7 @@ struct r600_cs_track {
         u32                     nsamples;
         u32                     cb_color_base_last[8];
         struct radeon_bo        *cb_color_bo[8];
+        u64                     cb_color_bo_mc[8];
         u32                     cb_color_bo_offset[8];
         struct radeon_bo        *cb_color_frag_bo[8];
         struct radeon_bo        *cb_color_tile_bo[8];
@@ -67,6 +68,7 @@ struct r600_cs_track {
         u32                     db_depth_size;
         u32                     db_offset;
         struct radeon_bo        *db_bo;
+        u64                     db_bo_mc;
 };
 
 static inline int r600_bpe_from_format(u32 *bpe, u32 format)
@@ -140,6 +142,68 @@ static inline int r600_bpe_from_format(u32 *bpe, u32 format)
         return 0;
 }
 
+struct array_mode_checker {
+        int array_mode;
+        u32 group_size;
+        u32 nbanks;
+        u32 npipes;
+        u32 nsamples;
+        u32 bpe;
+};
+
+/* returns alignment in pixels for pitch/height/depth and bytes for base */
+static inline int r600_get_array_mode_alignment(struct array_mode_checker *values,
+                                                u32 *pitch_align,
+                                                u32 *height_align,
+                                                u32 *depth_align,
+                                                u64 *base_align)
+{
+        u32 tile_width = 8;
+        u32 tile_height = 8;
+        u32 macro_tile_width = values->nbanks;
+        u32 macro_tile_height = values->npipes;
+        u32 tile_bytes = tile_width * tile_height * values->bpe * values->nsamples;
+        u32 macro_tile_bytes = macro_tile_width * macro_tile_height * tile_bytes;
+
+        switch (values->array_mode) {
+        case ARRAY_LINEAR_GENERAL:
+                /* technically tile_width/_height for pitch/height */
+                *pitch_align = 1; /* tile_width */
+                *height_align = 1; /* tile_height */
+                *depth_align = 1;
+                *base_align = 1;
+                break;
+        case ARRAY_LINEAR_ALIGNED:
+                *pitch_align = max((u32)64, (u32)(values->group_size / values->bpe));
+                *height_align = tile_height;
+                *depth_align = 1;
+                *base_align = values->group_size;
+                break;
+        case ARRAY_1D_TILED_THIN1:
+                *pitch_align = max((u32)tile_width,
+                                   (u32)(values->group_size /
+                                         (tile_height * values->bpe * values->nsamples)));
+                *height_align = tile_height;
+                *depth_align = 1;
+                *base_align = values->group_size;
+                break;
+        case ARRAY_2D_TILED_THIN1:
+                *pitch_align = max((u32)macro_tile_width,
+                                  (u32)(((values->group_size / tile_height) /
+                                         (values->bpe * values->nsamples)) *
+                                        values->nbanks)) * tile_width;
+                *height_align = macro_tile_height * tile_height;
+                *depth_align = 1;
+                *base_align = max(macro_tile_bytes,
+                                  (*pitch_align) * values->bpe * (*height_align) * values->nsamples);
+                break;
+        default:
+                return -EINVAL;
+        }
+
+        return 0;
+}
+
 static void r600_cs_track_init(struct r600_cs_track *track)
 {
         int i;
@@ -153,10 +217,12 @@ static void r600_cs_track_init(struct r600_cs_track *track)
                 track->cb_color_info[i] = 0;
                 track->cb_color_bo[i] = NULL;
                 track->cb_color_bo_offset[i] = 0xFFFFFFFF;
+                track->cb_color_bo_mc[i] = 0xFFFFFFFF;
         }
         track->cb_target_mask = 0xFFFFFFFF;
         track->cb_shader_mask = 0xFFFFFFFF;
         track->db_bo = NULL;
+        track->db_bo_mc = 0xFFFFFFFF;
         /* assume the biggest format and that htile is enabled */
         track->db_depth_info = 7 | (1 << 25);
         track->db_depth_view = 0xFFFFC000;
@@ -168,7 +234,10 @@ static void r600_cs_track_init(struct r600_cs_track *track)
 static inline int r600_cs_track_validate_cb(struct radeon_cs_parser *p, int i)
 {
         struct r600_cs_track *track = p->track;
-        u32 bpe = 0, pitch, slice_tile_max, size, tmp, height, pitch_align;
+        u32 bpe = 0, slice_tile_max, size, tmp;
+        u32 height, height_align, pitch, pitch_align, depth_align;
+        u64 base_offset, base_align;
+        struct array_mode_checker array_check;
         volatile u32 *ib = p->ib->ptr;
         unsigned array_mode;
 
@@ -183,60 +252,40 @@ static inline int r600_cs_track_validate_cb(struct radeon_cs_parser *p, int i)
                         i, track->cb_color_info[i]);
                 return -EINVAL;
         }
-        /* pitch is the number of 8x8 tiles per row */
-        pitch = G_028060_PITCH_TILE_MAX(track->cb_color_size[i]) + 1;
+        /* pitch in pixels */
+        pitch = (G_028060_PITCH_TILE_MAX(track->cb_color_size[i]) + 1) * 8;
         slice_tile_max = G_028060_SLICE_TILE_MAX(track->cb_color_size[i]) + 1;
         slice_tile_max *= 64;
-        height = slice_tile_max / (pitch * 8);
+        height = slice_tile_max / pitch;
         if (height > 8192)
                 height = 8192;
         array_mode = G_0280A0_ARRAY_MODE(track->cb_color_info[i]);
+
+        base_offset = track->cb_color_bo_mc[i] + track->cb_color_bo_offset[i];
+        array_check.array_mode = array_mode;
+        array_check.group_size = track->group_size;
+        array_check.nbanks = track->nbanks;
+        array_check.npipes = track->npipes;
+        array_check.nsamples = track->nsamples;
+        array_check.bpe = bpe;
+        if (r600_get_array_mode_alignment(&array_check,
+                                          &pitch_align, &height_align, &depth_align, &base_align)) {
+                dev_warn(p->dev, "%s invalid tiling %d for %d (0x%08X)\n", __func__,
+                         G_0280A0_ARRAY_MODE(track->cb_color_info[i]), i,
+                         track->cb_color_info[i]);
+                return -EINVAL;
+        }
         switch (array_mode) {
         case V_0280A0_ARRAY_LINEAR_GENERAL:
-                /* technically height & 0x7 */
                 break;
         case V_0280A0_ARRAY_LINEAR_ALIGNED:
-                pitch_align = max((u32)64, (u32)(track->group_size / bpe)) / 8;
-                if (!IS_ALIGNED(pitch, pitch_align)) {
-                        dev_warn(p->dev, "%s:%d cb pitch (%d) invalid\n",
-                                 __func__, __LINE__, pitch);
-                        return -EINVAL;
-                }
-                if (!IS_ALIGNED(height, 8)) {
-                        dev_warn(p->dev, "%s:%d cb height (%d) invalid\n",
-                                 __func__, __LINE__, height);
-                        return -EINVAL;
-                }
                 break;
         case V_0280A0_ARRAY_1D_TILED_THIN1:
-                pitch_align = max((u32)8, (u32)(track->group_size / (8 * bpe * track->nsamples))) / 8;
-                if (!IS_ALIGNED(pitch, pitch_align)) {
-                        dev_warn(p->dev, "%s:%d cb pitch (%d) invalid\n",
-                                 __func__, __LINE__, pitch);
-                        return -EINVAL;
-                }
                 /* avoid breaking userspace */
                 if (height > 7)
                         height &= ~0x7;
-                if (!IS_ALIGNED(height, 8)) {
-                        dev_warn(p->dev, "%s:%d cb height (%d) invalid\n",
-                                 __func__, __LINE__, height);
-                        return -EINVAL;
-                }
                 break;
         case V_0280A0_ARRAY_2D_TILED_THIN1:
-                pitch_align = max((u32)track->nbanks,
-                                  (u32)(((track->group_size / 8) / (bpe * track->nsamples)) * track->nbanks)) / 8;
-                if (!IS_ALIGNED(pitch, pitch_align)) {
-                        dev_warn(p->dev, "%s:%d cb pitch (%d) invalid\n",
-                                __func__, __LINE__, pitch);
-                        return -EINVAL;
-                }
-                if (!IS_ALIGNED((height / 8), track->npipes)) {
-                        dev_warn(p->dev, "%s:%d cb height (%d) invalid\n",
-                                 __func__, __LINE__, height);
-                        return -EINVAL;
-                }
                 break;
         default:
                 dev_warn(p->dev, "%s invalid tiling %d for %d (0x%08X)\n", __func__,
@@ -244,13 +293,29 @@ static inline int r600_cs_track_validate_cb(struct radeon_cs_parser *p, int i)
                         track->cb_color_info[i]);
                 return -EINVAL;
         }
+
+        if (!IS_ALIGNED(pitch, pitch_align)) {
+                dev_warn(p->dev, "%s:%d cb pitch (%d) invalid\n",
+                         __func__, __LINE__, pitch);
+                return -EINVAL;
+        }
+        if (!IS_ALIGNED(height, height_align)) {
+                dev_warn(p->dev, "%s:%d cb height (%d) invalid\n",
+                         __func__, __LINE__, height);
+                return -EINVAL;
+        }
+        if (!IS_ALIGNED(base_offset, base_align)) {
+                dev_warn(p->dev, "%s offset[%d] 0x%llx not aligned\n", __func__, i, base_offset);
+                return -EINVAL;
+        }
+
         /* check offset */
-        tmp = height * pitch * 8 * bpe;
+        tmp = height * pitch * bpe;
         if ((tmp + track->cb_color_bo_offset[i]) > radeon_bo_size(track->cb_color_bo[i])) {
                 if (array_mode == V_0280A0_ARRAY_LINEAR_GENERAL) {
                         /* the initial DDX does bad things with the CB size occasionally */
                         /* it rounds up height too far for slice tile max but the BO is smaller */
-                        tmp = (height - 7) * 8 * bpe;
+                        tmp = (height - 7) * pitch * bpe;
                         if ((tmp + track->cb_color_bo_offset[i]) > radeon_bo_size(track->cb_color_bo[i])) {
                                 dev_warn(p->dev, "%s offset[%d] %d %d %lu too big\n", __func__, i, track->cb_color_bo_offset[i], tmp, radeon_bo_size(track->cb_color_bo[i]));
                                 return -EINVAL;
@@ -260,15 +325,11 @@ static inline int r600_cs_track_validate_cb(struct radeon_cs_parser *p, int i)
                         return -EINVAL;
                 }
         }
-        if (!IS_ALIGNED(track->cb_color_bo_offset[i], track->group_size)) {
-                dev_warn(p->dev, "%s offset[%d] %d not aligned\n", __func__, i, track->cb_color_bo_offset[i]);
-                return -EINVAL;
-        }
         /* limit max tile */
-        tmp = (height * pitch * 8) >> 6;
+        tmp = (height * pitch) >> 6;
         if (tmp < slice_tile_max)
                 slice_tile_max = tmp;
-        tmp = S_028060_PITCH_TILE_MAX(pitch - 1) |
+        tmp = S_028060_PITCH_TILE_MAX((pitch / 8) - 1) |
                 S_028060_SLICE_TILE_MAX(slice_tile_max - 1);
         ib[track->cb_color_size_idx[i]] = tmp;
         return 0;
@@ -310,7 +371,12 @@ static int r600_cs_track_check(struct radeon_cs_parser *p)
         /* Check depth buffer */
         if (G_028800_STENCIL_ENABLE(track->db_depth_control) ||
                 G_028800_Z_ENABLE(track->db_depth_control)) {
-                u32 nviews, bpe, ntiles, pitch, pitch_align, height, size, slice_tile_max;
+                u32 nviews, bpe, ntiles, size, slice_tile_max;
+                u32 height, height_align, pitch, pitch_align, depth_align;
+                u64 base_offset, base_align;
+                struct array_mode_checker array_check;
+                int array_mode;
+
                 if (track->db_bo == NULL) {
                         dev_warn(p->dev, "z/stencil with no depth buffer\n");
                         return -EINVAL;
@@ -353,41 +419,34 @@ static int r600_cs_track_check(struct radeon_cs_parser *p)
                         ib[track->db_depth_size_idx] = S_028000_SLICE_TILE_MAX(tmp - 1) | (track->db_depth_size & 0x3FF);
                 } else {
                         size = radeon_bo_size(track->db_bo);
-                        pitch = G_028000_PITCH_TILE_MAX(track->db_depth_size) + 1;
+                        /* pitch in pixels */
+                        pitch = (G_028000_PITCH_TILE_MAX(track->db_depth_size) + 1) * 8;
                         slice_tile_max = G_028000_SLICE_TILE_MAX(track->db_depth_size) + 1;
                         slice_tile_max *= 64;
-                        height = slice_tile_max / (pitch * 8);
+                        height = slice_tile_max / pitch;
                         if (height > 8192)
                                 height = 8192;
-                        switch (G_028010_ARRAY_MODE(track->db_depth_info)) {
+                        base_offset = track->db_bo_mc + track->db_offset;
+                        array_mode = G_028010_ARRAY_MODE(track->db_depth_info);
+                        array_check.array_mode = array_mode;
+                        array_check.group_size = track->group_size;
+                        array_check.nbanks = track->nbanks;
+                        array_check.npipes = track->npipes;
+                        array_check.nsamples = track->nsamples;
+                        array_check.bpe = bpe;
+                        if (r600_get_array_mode_alignment(&array_check,
+                                                          &pitch_align, &height_align, &depth_align, &base_align)) {
+                                dev_warn(p->dev, "%s invalid tiling %d (0x%08X)\n", __func__,
+                                         G_028010_ARRAY_MODE(track->db_depth_info),
+                                         track->db_depth_info);
+                                return -EINVAL;
+                        }
+                        switch (array_mode) {
                         case V_028010_ARRAY_1D_TILED_THIN1:
-                                pitch_align = (max((u32)8, (u32)(track->group_size / (8 * bpe))) / 8);
-                                if (!IS_ALIGNED(pitch, pitch_align)) {
-                                        dev_warn(p->dev, "%s:%d db pitch (%d) invalid\n",
-                                                 __func__, __LINE__, pitch);
-                                        return -EINVAL;
-                                }
                                 /* don't break userspace */
                                 height &= ~0x7;
-                                if (!IS_ALIGNED(height, 8)) {
-                                        dev_warn(p->dev, "%s:%d db height (%d) invalid\n",
-                                                 __func__, __LINE__, height);
-                                        return -EINVAL;
-                                }
                                 break;
                         case V_028010_ARRAY_2D_TILED_THIN1:
-                                pitch_align = max((u32)track->nbanks,
-                                                  (u32)(((track->group_size / 8) / bpe) * track->nbanks)) / 8;
-                                if (!IS_ALIGNED(pitch, pitch_align)) {
-                                        dev_warn(p->dev, "%s:%d db pitch (%d) invalid\n",
-                                                 __func__, __LINE__, pitch);
-                                        return -EINVAL;
-                                }
-                                if (!IS_ALIGNED((height / 8), track->npipes)) {
-                                        dev_warn(p->dev, "%s:%d db height (%d) invalid\n",
-                                                 __func__, __LINE__, height);
-                                        return -EINVAL;
-                                }
                                 break;
                         default:
                                 dev_warn(p->dev, "%s invalid tiling %d (0x%08X)\n", __func__,
@@ -395,15 +454,27 @@ static int r600_cs_track_check(struct radeon_cs_parser *p)
                                          track->db_depth_info);
                                 return -EINVAL;
                         }
-                        if (!IS_ALIGNED(track->db_offset, track->group_size)) {
-                                dev_warn(p->dev, "%s offset[%d] %d not aligned\n", __func__, i, track->db_offset);
+
+                        if (!IS_ALIGNED(pitch, pitch_align)) {
+                                dev_warn(p->dev, "%s:%d db pitch (%d) invalid\n",
+                                         __func__, __LINE__, pitch);
+                                return -EINVAL;
+                        }
+                        if (!IS_ALIGNED(height, height_align)) {
+                                dev_warn(p->dev, "%s:%d db height (%d) invalid\n",
+                                         __func__, __LINE__, height);
                                 return -EINVAL;
                         }
+                        if (!IS_ALIGNED(base_offset, base_align)) {
+                                dev_warn(p->dev, "%s offset[%d] 0x%llx not aligned\n", __func__, i, base_offset);
+                                return -EINVAL;
+                        }
+
                         ntiles = G_028000_SLICE_TILE_MAX(track->db_depth_size) + 1;
                         nviews = G_028004_SLICE_MAX(track->db_depth_view) + 1;
                         tmp = ntiles * bpe * 64 * nviews;
                         if ((tmp + track->db_offset) > radeon_bo_size(track->db_bo)) {
-                                dev_warn(p->dev, "z/stencil buffer too small (0x%08X %d %d %d -> %d have %ld)\n",
+                                dev_warn(p->dev, "z/stencil buffer too small (0x%08X %d %d %d -> %u have %lu)\n",
                                                 track->db_depth_size, ntiles, nviews, bpe, tmp + track->db_offset,
                                                 radeon_bo_size(track->db_bo));
                                 return -EINVAL;
@@ -954,6 +1025,7 @@ static inline int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx
                 ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
                 track->cb_color_base_last[tmp] = ib[idx];
                 track->cb_color_bo[tmp] = reloc->robj;
+                track->cb_color_bo_mc[tmp] = reloc->lobj.gpu_offset;
                 break;
         case DB_DEPTH_BASE:
                 r = r600_cs_packet_next_reloc(p, &reloc);
@@ -965,6 +1037,7 @@ static inline int r600_cs_check_reg(struct radeon_cs_parser *p, u32 reg, u32 idx
                 track->db_offset = radeon_get_ib_value(p, idx) << 8;
                 ib[idx] += (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
                 track->db_bo = reloc->robj;
+                track->db_bo_mc = reloc->lobj.gpu_offset;
                 break;
         case DB_HTILE_DATA_BASE:
         case SQ_PGM_START_FS:
@@ -1086,16 +1159,25 @@ static void r600_texture_size(unsigned nfaces, unsigned blevel, unsigned nlevels
 static inline int r600_check_texture_resource(struct radeon_cs_parser *p,  u32 idx,
                                               struct radeon_bo *texture,
                                               struct radeon_bo *mipmap,
+                                              u64 base_offset,
+                                              u64 mip_offset,
                                               u32 tiling_flags)
 {
         struct r600_cs_track *track = p->track;
         u32 nfaces, nlevels, blevel, w0, h0, d0, bpe = 0;
-        u32 word0, word1, l0_size, mipmap_size, pitch, pitch_align;
+        u32 word0, word1, l0_size, mipmap_size;
+        u32 height_align, pitch, pitch_align, depth_align;
+        u64 base_align;
+        struct array_mode_checker array_check;
 
         /* on legacy kernel we don't perform advanced check */
         if (p->rdev == NULL)
                 return 0;
 
+        /* convert to bytes */
+        base_offset <<= 8;
+        mip_offset <<= 8;
+
         word0 = radeon_get_ib_value(p, idx + 0);
         if (tiling_flags & RADEON_TILING_MACRO)
                 word0 |= S_038000_TILE_MODE(V_038000_ARRAY_2D_TILED_THIN1);
@@ -1128,46 +1210,38 @@ static inline int r600_check_texture_resource(struct radeon_cs_parser *p,  u32 i
                 return -EINVAL;
         }
 
-        pitch = G_038000_PITCH(word0) + 1;
-        switch (G_038000_TILE_MODE(word0)) {
-        case V_038000_ARRAY_LINEAR_GENERAL:
-                pitch_align = 1;
-                /* XXX check height align */
-                break;
-        case V_038000_ARRAY_LINEAR_ALIGNED:
-                pitch_align = max((u32)64, (u32)(track->group_size / bpe)) / 8;
-                if (!IS_ALIGNED(pitch, pitch_align)) {
-                        dev_warn(p->dev, "%s:%d tex pitch (%d) invalid\n",
-                                 __func__, __LINE__, pitch);
-                        return -EINVAL;
-                }
-                /* XXX check height align */
-                break;
-        case V_038000_ARRAY_1D_TILED_THIN1:
-                pitch_align = max((u32)8, (u32)(track->group_size / (8 * bpe))) / 8;
-                if (!IS_ALIGNED(pitch, pitch_align)) {
-                        dev_warn(p->dev, "%s:%d tex pitch (%d) invalid\n",
-                                 __func__, __LINE__, pitch);
-                        return -EINVAL;
-                }
-                /* XXX check height align */
-                break;
-        case V_038000_ARRAY_2D_TILED_THIN1:
-                pitch_align = max((u32)track->nbanks,
-                                  (u32)(((track->group_size / 8) / bpe) * track->nbanks)) / 8;
-                if (!IS_ALIGNED(pitch, pitch_align)) {
-                        dev_warn(p->dev, "%s:%d tex pitch (%d) invalid\n",
-                                __func__, __LINE__, pitch);
-                        return -EINVAL;
-                }
-                /* XXX check height align */
-                break;
-        default:
-                dev_warn(p->dev, "%s invalid tiling %d (0x%08X)\n", __func__,
-                         G_038000_TILE_MODE(word0), word0);
+        /* pitch in texels */
+        pitch = (G_038000_PITCH(word0) + 1) * 8;
+        array_check.array_mode = G_038000_TILE_MODE(word0);
+        array_check.group_size = track->group_size;
+        array_check.nbanks = track->nbanks;
+        array_check.npipes = track->npipes;
+        array_check.nsamples = 1;
+        array_check.bpe = bpe;
+        if (r600_get_array_mode_alignment(&array_check,
+                                          &pitch_align, &height_align, &depth_align, &base_align)) {
+                dev_warn(p->dev, "%s:%d tex array mode (%d) invalid\n",
+                         __func__, __LINE__, G_038000_TILE_MODE(word0));
+                return -EINVAL;
+        }
+
+        /* XXX check height as well... */
+
+        if (!IS_ALIGNED(pitch, pitch_align)) {
+                dev_warn(p->dev, "%s:%d tex pitch (%d) invalid\n",
+                         __func__, __LINE__, pitch);
+                return -EINVAL;
+        }
+        if (!IS_ALIGNED(base_offset, base_align)) {
+                dev_warn(p->dev, "%s:%d tex base offset (0x%llx) invalid\n",
+                         __func__, __LINE__, base_offset);
+                return -EINVAL;
+        }
+        if (!IS_ALIGNED(mip_offset, base_align)) {
+                dev_warn(p->dev, "%s:%d tex mip offset (0x%llx) invalid\n",
+                         __func__, __LINE__, mip_offset);
                 return -EINVAL;
         }
-        /* XXX check offset align */
 
         word0 = radeon_get_ib_value(p, idx + 4);
         word1 = radeon_get_ib_value(p, idx + 5);
@@ -1402,7 +1476,10 @@ static int r600_packet3_check(struct radeon_cs_parser *p,
                                 mip_offset = (u32)((reloc->lobj.gpu_offset >> 8) & 0xffffffff);
                                 mipmap = reloc->robj;
                                 r = r600_check_texture_resource(p,  idx+(i*7)+1,
-                                                                texture, mipmap, reloc->lobj.tiling_flags);
+                                                                texture, mipmap,
+                                                                base_offset + radeon_get_ib_value(p, idx+1+(i*7)+2),
+                                                                mip_offset + radeon_get_ib_value(p, idx+1+(i*7)+3),
+                                                                reloc->lobj.tiling_flags);
                                 if (r)
                                         return r;
                                 ib[idx+1+(i*7)+2] += base_offset;
diff --git a/drivers/gpu/drm/radeon/r600d.h b/drivers/gpu/drm/radeon/r600d.h
index 966a793e225b..bff4dc4f410f 100644
--- a/drivers/gpu/drm/radeon/r600d.h
+++ b/drivers/gpu/drm/radeon/r600d.h
@@ -51,6 +51,12 @@
 #define PTE_READABLE                            (1 << 5)
 #define PTE_WRITEABLE                           (1 << 6)
 
+/* tiling bits */
+#define     ARRAY_LINEAR_GENERAL              0x00000000
+#define     ARRAY_LINEAR_ALIGNED              0x00000001
+#define     ARRAY_1D_TILED_THIN1              0x00000002
+#define     ARRAY_2D_TILED_THIN1              0x00000004
+
 /* Registers */
 #define ARB_POP                                         0x2418
 #define         ENABLE_TC128                                    (1 << 30)
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index 73f600d39ad4..3a7095743d44 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -1262,6 +1262,10 @@ void r100_pll_errata_after_index(struct radeon_device *rdev);
                 (rdev->family == CHIP_RS400) ||                 \
                 (rdev->family == CHIP_RS480))
 #define ASIC_IS_AVIVO(rdev) ((rdev->family >= CHIP_RS600))
+#define ASIC_IS_DCE2(rdev) ((rdev->family == CHIP_RS600)  ||    \
+                            (rdev->family == CHIP_RS690)  ||    \
+                            (rdev->family == CHIP_RS740)  ||    \
+                            (rdev->family >= CHIP_R600))
 #define ASIC_IS_DCE3(rdev) ((rdev->family >= CHIP_RV620))
 #define ASIC_IS_DCE32(rdev) ((rdev->family >= CHIP_RV730))
 #define ASIC_IS_DCE4(rdev) ((rdev->family >= CHIP_CEDAR))
diff --git a/drivers/gpu/drm/radeon/radeon_benchmark.c b/drivers/gpu/drm/radeon/radeon_benchmark.c
index 7932dc4d6b90..c558685cc637 100644
--- a/drivers/gpu/drm/radeon/radeon_benchmark.c
+++ b/drivers/gpu/drm/radeon/radeon_benchmark.c
@@ -41,7 +41,7 @@ void radeon_benchmark_move(struct radeon_device *rdev, unsigned bsize,
 
         size = bsize;
         n = 1024;
-        r = radeon_bo_create(rdev, NULL, size, true, sdomain, &sobj);
+        r = radeon_bo_create(rdev, NULL, size, PAGE_SIZE, true, sdomain, &sobj);
         if (r) {
                 goto out_cleanup;
         }
@@ -53,7 +53,7 @@ void radeon_benchmark_move(struct radeon_device *rdev, unsigned bsize,
         if (r) {
                 goto out_cleanup;
         }
-        r = radeon_bo_create(rdev, NULL, size, true, ddomain, &dobj);
+        r = radeon_bo_create(rdev, NULL, size, PAGE_SIZE, true, ddomain, &dobj);
         if (r) {
                 goto out_cleanup;
         }
diff --git a/drivers/gpu/drm/radeon/radeon_combios.c b/drivers/gpu/drm/radeon/radeon_combios.c
index 7b7ea269549c..3bddea5b5295 100644
--- a/drivers/gpu/drm/radeon/radeon_combios.c
+++ b/drivers/gpu/drm/radeon/radeon_combios.c
@@ -571,6 +571,7 @@ static struct radeon_i2c_bus_rec combios_setup_i2c_bus(struct radeon_device *rde
         }
 
         if (clk_mask && data_mask) {
+                /* system specific masks */
                 i2c.mask_clk_mask = clk_mask;
                 i2c.mask_data_mask = data_mask;
                 i2c.a_clk_mask = clk_mask;
@@ -579,7 +580,19 @@ static struct radeon_i2c_bus_rec combios_setup_i2c_bus(struct radeon_device *rde
                 i2c.en_data_mask = data_mask;
                 i2c.y_clk_mask = clk_mask;
                 i2c.y_data_mask = data_mask;
+        } else if ((ddc_line == RADEON_GPIOPAD_MASK) ||
+                   (ddc_line == RADEON_MDGPIO_MASK)) {
+                /* default gpiopad masks */
+                i2c.mask_clk_mask = (0x20 << 8);
+                i2c.mask_data_mask = 0x80;
+                i2c.a_clk_mask = (0x20 << 8);
+                i2c.a_data_mask = 0x80;
+                i2c.en_clk_mask = (0x20 << 8);
+                i2c.en_data_mask = 0x80;
+                i2c.y_clk_mask = (0x20 << 8);
+                i2c.y_data_mask = 0x80;
         } else {
+                /* default masks for ddc pads */
                 i2c.mask_clk_mask = RADEON_GPIO_EN_1;
                 i2c.mask_data_mask = RADEON_GPIO_EN_0;
                 i2c.a_clk_mask = RADEON_GPIO_A_1;
diff --git a/drivers/gpu/drm/radeon/radeon_connectors.c b/drivers/gpu/drm/radeon/radeon_connectors.c
index fe6c74780f18..3bef9f6d66fd 100644
--- a/drivers/gpu/drm/radeon/radeon_connectors.c
+++ b/drivers/gpu/drm/radeon/radeon_connectors.c
@@ -1008,9 +1008,21 @@ static void radeon_dp_connector_destroy(struct drm_connector *connector)
 static int radeon_dp_get_modes(struct drm_connector *connector)
 {
         struct radeon_connector *radeon_connector = to_radeon_connector(connector);
+        struct radeon_connector_atom_dig *radeon_dig_connector = radeon_connector->con_priv;
         int ret;
 
+        if (connector->connector_type == DRM_MODE_CONNECTOR_eDP) {
+                if (!radeon_dig_connector->edp_on)
+                        atombios_set_edp_panel_power(connector,
+                                                     ATOM_TRANSMITTER_ACTION_POWER_ON);
+        }
         ret = radeon_ddc_get_modes(radeon_connector);
+        if (connector->connector_type == DRM_MODE_CONNECTOR_eDP) {
+                if (!radeon_dig_connector->edp_on)
+                        atombios_set_edp_panel_power(connector,
+                                                     ATOM_TRANSMITTER_ACTION_POWER_OFF);
+        }
+
         return ret;
 }
 
@@ -1029,8 +1041,14 @@ radeon_dp_detect(struct drm_connector *connector, bool force)
         if (connector->connector_type == DRM_MODE_CONNECTOR_eDP) {
                 /* eDP is always DP */
                 radeon_dig_connector->dp_sink_type = CONNECTOR_OBJECT_ID_DISPLAYPORT;
+                if (!radeon_dig_connector->edp_on)
+                        atombios_set_edp_panel_power(connector,
+                                                     ATOM_TRANSMITTER_ACTION_POWER_ON);
                 if (radeon_dp_getdpcd(radeon_connector))
                         ret = connector_status_connected;
+                if (!radeon_dig_connector->edp_on)
+                        atombios_set_edp_panel_power(connector,
+                                                     ATOM_TRANSMITTER_ACTION_POWER_OFF);
         } else {
                 radeon_dig_connector->dp_sink_type = radeon_dp_getsinktype(radeon_connector);
                 if (radeon_dig_connector->dp_sink_type == CONNECTOR_OBJECT_ID_DISPLAYPORT) {
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 8adfedfe547f..d8ac1849180d 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -180,7 +180,7 @@ int radeon_wb_init(struct radeon_device *rdev)
         int r;
 
         if (rdev->wb.wb_obj == NULL) {
-                r = radeon_bo_create(rdev, NULL, RADEON_GPU_PAGE_SIZE, true,
+                r = radeon_bo_create(rdev, NULL, RADEON_GPU_PAGE_SIZE, PAGE_SIZE, true,
                                 RADEON_GEM_DOMAIN_GTT, &rdev->wb.wb_obj);
                 if (r) {
                         dev_warn(rdev->dev, "(%d) create WB bo failed\n", r);
diff --git a/drivers/gpu/drm/radeon/radeon_encoders.c b/drivers/gpu/drm/radeon/radeon_encoders.c
index f678257c42e6..041943df966b 100644
--- a/drivers/gpu/drm/radeon/radeon_encoders.c
+++ b/drivers/gpu/drm/radeon/radeon_encoders.c
@@ -176,6 +176,7 @@ static inline bool radeon_encoder_is_digital(struct drm_encoder *encoder)
                 return false;
         }
 }
+
 void
 radeon_link_encoder_connector(struct drm_device *dev)
 {
@@ -228,6 +229,27 @@ radeon_get_connector_for_encoder(struct drm_encoder *encoder)
         return NULL;
 }
 
+struct drm_encoder *radeon_atom_get_external_encoder(struct drm_encoder *encoder)
+{
+        struct drm_device *dev = encoder->dev;
+        struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
+        struct drm_encoder *other_encoder;
+        struct radeon_encoder *other_radeon_encoder;
+
+        if (radeon_encoder->is_ext_encoder)
+                return NULL;
+
+        list_for_each_entry(other_encoder, &dev->mode_config.encoder_list, head) {
+                if (other_encoder == encoder)
+                        continue;
+                other_radeon_encoder = to_radeon_encoder(other_encoder);
+                if (other_radeon_encoder->is_ext_encoder &&
+                    (radeon_encoder->devices & other_radeon_encoder->devices))
+                        return other_encoder;
+        }
+        return NULL;
+}
+
 void radeon_panel_mode_fixup(struct drm_encoder *encoder,
                              struct drm_display_mode *adjusted_mode)
 {
@@ -426,52 +448,49 @@ atombios_tv_setup(struct drm_encoder *encoder, int action)
 
 }
 
-void
-atombios_external_tmds_setup(struct drm_encoder *encoder, int action)
-{
-        struct drm_device *dev = encoder->dev;
-        struct radeon_device *rdev = dev->dev_private;
-        struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-        ENABLE_EXTERNAL_TMDS_ENCODER_PS_ALLOCATION args;
-        int index = 0;
-
-        memset(&args, 0, sizeof(args));
-
-        index = GetIndexIntoMasterTable(COMMAND, DVOEncoderControl);
-
-        args.sXTmdsEncoder.ucEnable = action;
-
-        if (radeon_encoder->pixel_clock > 165000)
-                args.sXTmdsEncoder.ucMisc = PANEL_ENCODER_MISC_DUAL;
-
-        /*if (pScrn->rgbBits == 8)*/
-        args.sXTmdsEncoder.ucMisc |= (1 << 1);
-
-        atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
-
-}
+union dvo_encoder_control {
+        ENABLE_EXTERNAL_TMDS_ENCODER_PS_ALLOCATION ext_tmds;
+        DVO_ENCODER_CONTROL_PS_ALLOCATION dvo;
+        DVO_ENCODER_CONTROL_PS_ALLOCATION_V3 dvo_v3;
+};
 
-static void
-atombios_ddia_setup(struct drm_encoder *encoder, int action)
+void
+atombios_dvo_setup(struct drm_encoder *encoder, int action)
 {
         struct drm_device *dev = encoder->dev;
         struct radeon_device *rdev = dev->dev_private;
         struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
-        DVO_ENCODER_CONTROL_PS_ALLOCATION args;
-        int index = 0;
+        union dvo_encoder_control args;
+        int index = GetIndexIntoMasterTable(COMMAND, DVOEncoderControl);
 
         memset(&args, 0, sizeof(args));
 
-        index = GetIndexIntoMasterTable(COMMAND, DVOEncoderControl);
+        if (ASIC_IS_DCE3(rdev)) {
+                /* DCE3+ */
+                args.dvo_v3.ucAction = action;
+                args.dvo_v3.usPixelClock = cpu_to_le16(radeon_encoder->pixel_clock / 10);
+                args.dvo_v3.ucDVOConfig = 0; /* XXX */
+        } else if (ASIC_IS_DCE2(rdev)) {
+                /* DCE2 (pre-DCE3 R6xx, RS600/690/740 */
+                args.dvo.sDVOEncoder.ucAction = action;
+                args.dvo.sDVOEncoder.usPixelClock = cpu_to_le16(radeon_encoder->pixel_clock / 10);
+                /* DFP1, CRT1, TV1 depending on the type of port */
+                args.dvo.sDVOEncoder.ucDeviceType = ATOM_DEVICE_DFP1_INDEX;
+
+                if (radeon_encoder->pixel_clock > 165000)
+                        args.dvo.sDVOEncoder.usDevAttr.sDigAttrib.ucAttribute |= PANEL_ENCODER_MISC_DUAL;
+        } else {
+                /* R4xx, R5xx */
+                args.ext_tmds.sXTmdsEncoder.ucEnable = action;
 
-        args.sDVOEncoder.ucAction = action;
-        args.sDVOEncoder.usPixelClock = cpu_to_le16(radeon_encoder->pixel_clock / 10);
+                if (radeon_encoder->pixel_clock > 165000)
+                        args.ext_tmds.sXTmdsEncoder.ucMisc |= PANEL_ENCODER_MISC_DUAL;
 
-        if (radeon_encoder->pixel_clock > 165000)
-                args.sDVOEncoder.usDevAttr.sDigAttrib.ucAttribute = PANEL_ENCODER_MISC_DUAL;
+                /*if (pScrn->rgbBits == 8)*/
+                args.ext_tmds.sXTmdsEncoder.ucMisc |= ATOM_PANEL_MISC_888RGB;
+        }
 
         atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
-
 }
 
 union lvds_encoder_control {
@@ -532,14 +551,14 @@ atombios_digital_setup(struct drm_encoder *encoder, int action)
                                 if (dig->lcd_misc & ATOM_PANEL_MISC_DUAL)
                                         args.v1.ucMisc |= PANEL_ENCODER_MISC_DUAL;
                                 if (dig->lcd_misc & ATOM_PANEL_MISC_888RGB)
-                                        args.v1.ucMisc |= (1 << 1);
+                                        args.v1.ucMisc |= ATOM_PANEL_MISC_888RGB;
                         } else {
                                 if (dig->linkb)
                                         args.v1.ucMisc |= PANEL_ENCODER_MISC_TMDS_LINKB;
                                 if (radeon_encoder->pixel_clock > 165000)
                                         args.v1.ucMisc |= PANEL_ENCODER_MISC_DUAL;
                                 /*if (pScrn->rgbBits == 8) */
-                                args.v1.ucMisc |= (1 << 1);
+                                args.v1.ucMisc |= ATOM_PANEL_MISC_888RGB;
                         }
                         break;
                 case 2:
@@ -595,6 +614,7 @@ atombios_digital_setup(struct drm_encoder *encoder, int action)
 int
 atombios_get_encoder_mode(struct drm_encoder *encoder)
 {
+        struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
         struct drm_device *dev = encoder->dev;
         struct radeon_device *rdev = dev->dev_private;
         struct drm_connector *connector;
@@ -602,9 +622,20 @@ atombios_get_encoder_mode(struct drm_encoder *encoder)
         struct radeon_connector_atom_dig *dig_connector;
 
         connector = radeon_get_connector_for_encoder(encoder);
-        if (!connector)
-                return 0;
-
+        if (!connector) {
+                switch (radeon_encoder->encoder_id) {
+                case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
+                case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
+                case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
+                case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_LVTMA:
+                case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
+                        return ATOM_ENCODER_MODE_DVI;
+                case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC1:
+                case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC2:
+                default:
+                        return ATOM_ENCODER_MODE_CRT;
+                }
+        }
         radeon_connector = to_radeon_connector(connector);
 
         switch (connector->connector_type) {
@@ -834,6 +865,9 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t
         memset(&args, 0, sizeof(args));
 
         switch (radeon_encoder->encoder_id) {
+        case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
+                index = GetIndexIntoMasterTable(COMMAND, DVOOutputControl);
+                break;
         case ENCODER_OBJECT_ID_INTERNAL_UNIPHY:
         case ENCODER_OBJECT_ID_INTERNAL_UNIPHY1:
         case ENCODER_OBJECT_ID_INTERNAL_UNIPHY2:
@@ -978,6 +1012,105 @@ atombios_dig_transmitter_setup(struct drm_encoder *encoder, int action, uint8_t
         atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
 }
 
+void
+atombios_set_edp_panel_power(struct drm_connector *connector, int action)
+{
+        struct radeon_connector *radeon_connector = to_radeon_connector(connector);
+        struct drm_device *dev = radeon_connector->base.dev;
+        struct radeon_device *rdev = dev->dev_private;
+        union dig_transmitter_control args;
+        int index = GetIndexIntoMasterTable(COMMAND, UNIPHYTransmitterControl);
+        uint8_t frev, crev;
+
+        if (connector->connector_type != DRM_MODE_CONNECTOR_eDP)
+                return;
+
+        if (!ASIC_IS_DCE4(rdev))
+                return;
+
+        if ((action != ATOM_TRANSMITTER_ACTION_POWER_ON) ||
+            (action != ATOM_TRANSMITTER_ACTION_POWER_OFF))
+                return;
+
+        if (!atom_parse_cmd_header(rdev->mode_info.atom_context, index, &frev, &crev))
+                return;
+
+        memset(&args, 0, sizeof(args));
+
+        args.v1.ucAction = action;
+
+        atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+}
+
+union external_encoder_control {
+        EXTERNAL_ENCODER_CONTROL_PS_ALLOCATION v1;
+};
+
+static void
+atombios_external_encoder_setup(struct drm_encoder *encoder,
+                                struct drm_encoder *ext_encoder,
+                                int action)
+{
+        struct drm_device *dev = encoder->dev;
+        struct radeon_device *rdev = dev->dev_private;
+        struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
+        union external_encoder_control args;
+        struct drm_connector *connector = radeon_get_connector_for_encoder(encoder);
+        int index = GetIndexIntoMasterTable(COMMAND, ExternalEncoderControl);
+        u8 frev, crev;
+        int dp_clock = 0;
+        int dp_lane_count = 0;
+        int connector_object_id = 0;
+
+        if (connector) {
+                struct radeon_connector *radeon_connector = to_radeon_connector(connector);
+                struct radeon_connector_atom_dig *dig_connector =
+                        radeon_connector->con_priv;
+
+                dp_clock = dig_connector->dp_clock;
+                dp_lane_count = dig_connector->dp_lane_count;
+                connector_object_id =
+                        (radeon_connector->connector_object_id & OBJECT_ID_MASK) >> OBJECT_ID_SHIFT;
+        }
+
+        memset(&args, 0, sizeof(args));
+
+        if (!atom_parse_cmd_header(rdev->mode_info.atom_context, index, &frev, &crev))
+                return;
+
+        switch (frev) {
+        case 1:
+                /* no params on frev 1 */
+                break;
+        case 2:
+                switch (crev) {
+                case 1:
+                case 2:
+                        args.v1.sDigEncoder.ucAction = action;
+                        args.v1.sDigEncoder.usPixelClock = cpu_to_le16(radeon_encoder->pixel_clock / 10);
+                        args.v1.sDigEncoder.ucEncoderMode = atombios_get_encoder_mode(encoder);
+
+                        if (args.v1.sDigEncoder.ucEncoderMode == ATOM_ENCODER_MODE_DP) {
+                                if (dp_clock == 270000)
+                                        args.v1.sDigEncoder.ucConfig |= ATOM_ENCODER_CONFIG_DPLINKRATE_2_70GHZ;
+                                args.v1.sDigEncoder.ucLaneNum = dp_lane_count;
+                        } else if (radeon_encoder->pixel_clock > 165000)
+                                args.v1.sDigEncoder.ucLaneNum = 8;
+                        else
+                                args.v1.sDigEncoder.ucLaneNum = 4;
+                        break;
+                default:
+                        DRM_ERROR("Unknown table version: %d, %d\n", frev, crev);
+                        return;
+                }
+                break;
+        default:
+                DRM_ERROR("Unknown table version: %d, %d\n", frev, crev);
+                return;
+        }
+        atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+}
+
 static void
 atombios_yuv_setup(struct drm_encoder *encoder, bool enable)
 {
@@ -1021,6 +1154,7 @@ radeon_atom_encoder_dpms(struct drm_encoder *encoder, int mode)
         struct drm_device *dev = encoder->dev;
         struct radeon_device *rdev = dev->dev_private;
         struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
+        struct drm_encoder *ext_encoder = radeon_atom_get_external_encoder(encoder);
         DISPLAY_DEVICE_OUTPUT_CONTROL_PS_ALLOCATION args;
         int index = 0;
         bool is_dig = false;
@@ -1043,9 +1177,14 @@ radeon_atom_encoder_dpms(struct drm_encoder *encoder, int mode)
                 break;
         case ENCODER_OBJECT_ID_INTERNAL_DVO1:
         case ENCODER_OBJECT_ID_INTERNAL_DDI:
-        case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
                 index = GetIndexIntoMasterTable(COMMAND, DVOOutputControl);
                 break;
+        case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
+                if (ASIC_IS_DCE3(rdev))
+                        is_dig = true;
+                else
+                        index = GetIndexIntoMasterTable(COMMAND, DVOOutputControl);
+                break;
         case ENCODER_OBJECT_ID_INTERNAL_LVDS:
                 index = GetIndexIntoMasterTable(COMMAND, LCD1OutputControl);
                 break;
@@ -1082,34 +1221,85 @@ radeon_atom_encoder_dpms(struct drm_encoder *encoder, int mode)
                         if (atombios_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_DP) {
                                 struct drm_connector *connector = radeon_get_connector_for_encoder(encoder);
 
+                                if (connector &&
+                                    (connector->connector_type == DRM_MODE_CONNECTOR_eDP)) {
+                                        struct radeon_connector *radeon_connector = to_radeon_connector(connector);
+                                        struct radeon_connector_atom_dig *radeon_dig_connector =
+                                                radeon_connector->con_priv;
+                                        atombios_set_edp_panel_power(connector,
+                                                                     ATOM_TRANSMITTER_ACTION_POWER_ON);
+                                        radeon_dig_connector->edp_on = true;
+                                }
                                 dp_link_train(encoder, connector);
                                 if (ASIC_IS_DCE4(rdev))
                                         atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_DP_VIDEO_ON);
                         }
+                        if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
+                                atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_LCD_BLON, 0, 0);
                         break;
                 case DRM_MODE_DPMS_STANDBY:
                 case DRM_MODE_DPMS_SUSPEND:
                 case DRM_MODE_DPMS_OFF:
                         atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_DISABLE_OUTPUT, 0, 0);
                         if (atombios_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_DP) {
+                                struct drm_connector *connector = radeon_get_connector_for_encoder(encoder);
+
                                 if (ASIC_IS_DCE4(rdev))
                                         atombios_dig_encoder_setup(encoder, ATOM_ENCODER_CMD_DP_VIDEO_OFF);
+                                if (connector &&
+                                    (connector->connector_type == DRM_MODE_CONNECTOR_eDP)) {
+                                        struct radeon_connector *radeon_connector = to_radeon_connector(connector);
+                                        struct radeon_connector_atom_dig *radeon_dig_connector =
+                                                radeon_connector->con_priv;
+                                        atombios_set_edp_panel_power(connector,
+                                                                     ATOM_TRANSMITTER_ACTION_POWER_OFF);
+                                        radeon_dig_connector->edp_on = false;
+                                }
                         }
+                        if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
+                                atombios_dig_transmitter_setup(encoder, ATOM_TRANSMITTER_ACTION_LCD_BLOFF, 0, 0);
                         break;
                 }
         } else {
                 switch (mode) {
                 case DRM_MODE_DPMS_ON:
                         args.ucAction = ATOM_ENABLE;
+                        atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+                        if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) {
+                                args.ucAction = ATOM_LCD_BLON;
+                                atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+                        }
                         break;
                 case DRM_MODE_DPMS_STANDBY:
                 case DRM_MODE_DPMS_SUSPEND:
                 case DRM_MODE_DPMS_OFF:
                         args.ucAction = ATOM_DISABLE;
+                        atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+                        if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT)) {
+                                args.ucAction = ATOM_LCD_BLOFF;
+                                atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
+                        }
                         break;
                 }
-                atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
         }
+
+        if (ext_encoder) {
+                int action;
+
+                switch (mode) {
+                case DRM_MODE_DPMS_ON:
+                default:
+                        action = ATOM_ENABLE;
+                        break;
+                case DRM_MODE_DPMS_STANDBY:
+                case DRM_MODE_DPMS_SUSPEND:
+                case DRM_MODE_DPMS_OFF:
+                        action = ATOM_DISABLE;
+                        break;
+                }
+                atombios_external_encoder_setup(encoder, ext_encoder, action);
+        }
+
         radeon_atombios_encoder_dpms_scratch_regs(encoder, (mode == DRM_MODE_DPMS_ON) ? true : false);
 
 }
@@ -1242,7 +1432,7 @@ atombios_set_encoder_crtc_source(struct drm_encoder *encoder)
                 break;
         default:
                 DRM_ERROR("Unknown table version: %d, %d\n", frev, crev);
-                break;
+                return;
         }
 
         atom_execute_table(rdev->mode_info.atom_context, index, (uint32_t *)&args);
@@ -1357,6 +1547,7 @@ radeon_atom_encoder_mode_set(struct drm_encoder *encoder,
         struct drm_device *dev = encoder->dev;
         struct radeon_device *rdev = dev->dev_private;
         struct radeon_encoder *radeon_encoder = to_radeon_encoder(encoder);
+        struct drm_encoder *ext_encoder = radeon_atom_get_external_encoder(encoder);
 
         radeon_encoder->pixel_clock = adjusted_mode->clock;
 
@@ -1400,11 +1591,9 @@ radeon_atom_encoder_mode_set(struct drm_encoder *encoder,
                 }
                 break;
         case ENCODER_OBJECT_ID_INTERNAL_DDI:
-                atombios_ddia_setup(encoder, ATOM_ENABLE);
-                break;
         case ENCODER_OBJECT_ID_INTERNAL_DVO1:
         case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
-                atombios_external_tmds_setup(encoder, ATOM_ENABLE);
+                atombios_dvo_setup(encoder, ATOM_ENABLE);
                 break;
         case ENCODER_OBJECT_ID_INTERNAL_DAC1:
         case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC1:
@@ -1419,6 +1608,11 @@ radeon_atom_encoder_mode_set(struct drm_encoder *encoder,
                 }
                 break;
         }
+
+        if (ext_encoder) {
+                atombios_external_encoder_setup(encoder, ext_encoder, ATOM_ENABLE);
+        }
+
         atombios_apply_encoder_quirks(encoder, adjusted_mode);
 
         if (atombios_get_encoder_mode(encoder) == ATOM_ENCODER_MODE_HDMI) {
@@ -1595,11 +1789,9 @@ static void radeon_atom_encoder_disable(struct drm_encoder *encoder)
                 }
                 break;
         case ENCODER_OBJECT_ID_INTERNAL_DDI:
-                atombios_ddia_setup(encoder, ATOM_DISABLE);
-                break;
         case ENCODER_OBJECT_ID_INTERNAL_DVO1:
         case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DVO1:
-                atombios_external_tmds_setup(encoder, ATOM_DISABLE);
+                atombios_dvo_setup(encoder, ATOM_DISABLE);
                 break;
         case ENCODER_OBJECT_ID_INTERNAL_DAC1:
         case ENCODER_OBJECT_ID_INTERNAL_KLDSCP_DAC1:
@@ -1621,6 +1813,53 @@ disable_done:
         radeon_encoder->active_device = 0;
 }
 
+/* these are handled by the primary encoders */
+static void radeon_atom_ext_prepare(struct drm_encoder *encoder)
+{
+
+}
+
+static void radeon_atom_ext_commit(struct drm_encoder *encoder)
+{
+
+}
+
+static void
+radeon_atom_ext_mode_set(struct drm_encoder *encoder,
+                         struct drm_display_mode *mode,
+                         struct drm_display_mode *adjusted_mode)
+{
+
+}
+
+static void radeon_atom_ext_disable(struct drm_encoder *encoder)
+{
+
+}
+
+static void
+radeon_atom_ext_dpms(struct drm_encoder *encoder, int mode)
+{
+
+}
+
+static bool radeon_atom_ext_mode_fixup(struct drm_encoder *encoder,
+                                       struct drm_display_mode *mode,
+                                       struct drm_display_mode *adjusted_mode)
+{
+        return true;
+}
+
+static const struct drm_encoder_helper_funcs radeon_atom_ext_helper_funcs = {
+        .dpms = radeon_atom_ext_dpms,
+        .mode_fixup = radeon_atom_ext_mode_fixup,
+        .prepare = radeon_atom_ext_prepare,
+        .mode_set = radeon_atom_ext_mode_set,
+        .commit = radeon_atom_ext_commit,
+        .disable = radeon_atom_ext_disable,
+        /* no detect for TMDS/LVDS yet */
+};
+
 static const struct drm_encoder_helper_funcs radeon_atom_dig_helper_funcs = {
         .dpms = radeon_atom_encoder_dpms,
         .mode_fixup = radeon_atom_mode_fixup,
@@ -1730,6 +1969,7 @@ radeon_add_atom_encoder(struct drm_device *dev, uint32_t encoder_enum, uint32_t
         radeon_encoder->devices = supported_device;
         radeon_encoder->rmx_type = RMX_OFF;
         radeon_encoder->underscan_type = UNDERSCAN_OFF;
+        radeon_encoder->is_ext_encoder = false;
 
         switch (radeon_encoder->encoder_id) {
         case ENCODER_OBJECT_ID_INTERNAL_LVDS:
@@ -1771,6 +2011,9 @@ radeon_add_atom_encoder(struct drm_device *dev, uint32_t encoder_enum, uint32_t
                         radeon_encoder->rmx_type = RMX_FULL;
                         drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_LVDS);
                         radeon_encoder->enc_priv = radeon_atombios_get_lvds_info(radeon_encoder);
+                } else if (radeon_encoder->devices & (ATOM_DEVICE_CRT_SUPPORT)) {
+                        drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_DAC);
+                        radeon_encoder->enc_priv = radeon_atombios_set_dig_info(radeon_encoder);
                 } else {
                         drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_TMDS);
                         radeon_encoder->enc_priv = radeon_atombios_set_dig_info(radeon_encoder);
@@ -1779,5 +2022,22 @@ radeon_add_atom_encoder(struct drm_device *dev, uint32_t encoder_enum, uint32_t
                 }
                 drm_encoder_helper_add(encoder, &radeon_atom_dig_helper_funcs);
                 break;
+        case ENCODER_OBJECT_ID_SI170B:
+        case ENCODER_OBJECT_ID_CH7303:
+        case ENCODER_OBJECT_ID_EXTERNAL_SDVOA:
+        case ENCODER_OBJECT_ID_EXTERNAL_SDVOB:
+        case ENCODER_OBJECT_ID_TITFP513:
+        case ENCODER_OBJECT_ID_VT1623:
+        case ENCODER_OBJECT_ID_HDMI_SI1930:
+                /* these are handled by the primary encoders */
+                radeon_encoder->is_ext_encoder = true;
+                if (radeon_encoder->devices & (ATOM_DEVICE_LCD_SUPPORT))
+                        drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_LVDS);
+                else if (radeon_encoder->devices & (ATOM_DEVICE_CRT_SUPPORT))
+                        drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_DAC);
+                else
+                        drm_encoder_init(dev, encoder, &radeon_atom_enc_funcs, DRM_MODE_ENCODER_TMDS);
+                drm_encoder_helper_add(encoder, &radeon_atom_ext_helper_funcs);
+                break;
         }
 }
diff --git a/drivers/gpu/drm/radeon/radeon_gart.c b/drivers/gpu/drm/radeon/radeon_gart.c
index e65b90317fab..65016117d95f 100644
--- a/drivers/gpu/drm/radeon/radeon_gart.c
+++ b/drivers/gpu/drm/radeon/radeon_gart.c
@@ -79,8 +79,8 @@ int radeon_gart_table_vram_alloc(struct radeon_device *rdev)
 
         if (rdev->gart.table.vram.robj == NULL) {
                 r = radeon_bo_create(rdev, NULL, rdev->gart.table_size,
-                                        true, RADEON_GEM_DOMAIN_VRAM,
-                                        &rdev->gart.table.vram.robj);
+                                     PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
+                                     &rdev->gart.table.vram.robj);
                 if (r) {
                         return r;
                 }
diff --git a/drivers/gpu/drm/radeon/radeon_gem.c b/drivers/gpu/drm/radeon/radeon_gem.c
index d1e595d91723..df95eb83dac6 100644
--- a/drivers/gpu/drm/radeon/radeon_gem.c
+++ b/drivers/gpu/drm/radeon/radeon_gem.c
@@ -67,7 +67,7 @@ int radeon_gem_object_create(struct radeon_device *rdev, int size,
         if (alignment < PAGE_SIZE) {
                 alignment = PAGE_SIZE;
         }
-        r = radeon_bo_create(rdev, gobj, size, kernel, initial_domain, &robj);
+        r = radeon_bo_create(rdev, gobj, size, alignment, kernel, initial_domain, &robj);
         if (r) {
                 if (r != -ERESTARTSYS)
                         DRM_ERROR("Failed to allocate GEM object (%d, %d, %u, %d)\n",
diff --git a/drivers/gpu/drm/radeon/radeon_i2c.c b/drivers/gpu/drm/radeon/radeon_i2c.c
index 0cfbba02c4d0..ded2a45bc95c 100644
--- a/drivers/gpu/drm/radeon/radeon_i2c.c
+++ b/drivers/gpu/drm/radeon/radeon_i2c.c
@@ -896,7 +896,8 @@ struct radeon_i2c_chan *radeon_i2c_create(struct drm_device *dev,
              ((rdev->family <= CHIP_RS480) ||
               ((rdev->family >= CHIP_RV515) && (rdev->family <= CHIP_R580))))) {
                 /* set the radeon hw i2c adapter */
-                sprintf(i2c->adapter.name, "Radeon i2c hw bus %s", name);
+                snprintf(i2c->adapter.name, sizeof(i2c->adapter.name),
+                         "Radeon i2c hw bus %s", name);
                 i2c->adapter.algo = &radeon_i2c_algo;
                 ret = i2c_add_adapter(&i2c->adapter);
                 if (ret) {
@@ -905,7 +906,8 @@ struct radeon_i2c_chan *radeon_i2c_create(struct drm_device *dev,
                 }
         } else {
                 /* set the radeon bit adapter */
-                sprintf(i2c->adapter.name, "Radeon i2c bit bus %s", name);
+                snprintf(i2c->adapter.name, sizeof(i2c->adapter.name),
+                         "Radeon i2c bit bus %s", name);
                 i2c->adapter.algo_data = &i2c->algo.bit;
                 i2c->algo.bit.pre_xfer = pre_xfer;
                 i2c->algo.bit.post_xfer = post_xfer;
@@ -946,6 +948,8 @@ struct radeon_i2c_chan *radeon_i2c_create_dp(struct drm_device *dev,
         i2c->rec = *rec;
         i2c->adapter.owner = THIS_MODULE;
         i2c->dev = dev;
+        snprintf(i2c->adapter.name, sizeof(i2c->adapter.name),
+                 "Radeon aux bus %s", name);
         i2c_set_adapdata(&i2c->adapter, i2c);
         i2c->adapter.algo_data = &i2c->algo.dp;
         i2c->algo.dp.aux_ch = radeon_dp_i2c_aux_ch;
diff --git a/drivers/gpu/drm/radeon/radeon_irq.c b/drivers/gpu/drm/radeon/radeon_irq.c
index 2f349a300195..465746bd51b7 100644
--- a/drivers/gpu/drm/radeon/radeon_irq.c
+++ b/drivers/gpu/drm/radeon/radeon_irq.c
@@ -76,7 +76,7 @@ int radeon_enable_vblank(struct drm_device *dev, int crtc)
                 default:
                         DRM_ERROR("tried to enable vblank on non-existent crtc %d\n",
                                   crtc);
-                        return EINVAL;
+                        return -EINVAL;
                 }
         } else {
                 switch (crtc) {
@@ -89,7 +89,7 @@ int radeon_enable_vblank(struct drm_device *dev, int crtc)
                 default:
                         DRM_ERROR("tried to enable vblank on non-existent crtc %d\n",
                                   crtc);
-                        return EINVAL;
+                        return -EINVAL;
                 }
         }
 
diff --git a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c
index 0b8397000f4c..59f834ba283d 100644
--- a/drivers/gpu/drm/radeon/radeon_legacy_encoders.c
+++ b/drivers/gpu/drm/radeon/radeon_legacy_encoders.c
@@ -670,7 +670,7 @@ static void radeon_legacy_tmds_ext_mode_set(struct drm_encoder *encoder,
 
         if (rdev->is_atom_bios) {
                 radeon_encoder->pixel_clock = adjusted_mode->clock;
-                atombios_external_tmds_setup(encoder, ATOM_ENABLE);
+                atombios_dvo_setup(encoder, ATOM_ENABLE);
                 fp2_gen_cntl = RREG32(RADEON_FP2_GEN_CNTL);
         } else {
                 fp2_gen_cntl = RREG32(RADEON_FP2_GEN_CNTL);
diff --git a/drivers/gpu/drm/radeon/radeon_mode.h b/drivers/gpu/drm/radeon/radeon_mode.h
index 680f57644e86..e301c6f9e059 100644
--- a/drivers/gpu/drm/radeon/radeon_mode.h
+++ b/drivers/gpu/drm/radeon/radeon_mode.h
@@ -375,6 +375,7 @@ struct radeon_encoder {
         int hdmi_config_offset;
         int hdmi_audio_workaround;
         int hdmi_buffer_status;
+        bool is_ext_encoder;
 };
 
 struct radeon_connector_atom_dig {
@@ -385,6 +386,7 @@ struct radeon_connector_atom_dig {
         u8 dp_sink_type;
         int dp_clock;
         int dp_lane_count;
+        bool edp_on;
 };
 
 struct radeon_gpio_rec {
@@ -523,9 +525,10 @@ struct drm_encoder *radeon_encoder_legacy_primary_dac_add(struct drm_device *dev
 struct drm_encoder *radeon_encoder_legacy_tv_dac_add(struct drm_device *dev, int bios_index, int with_tv);
 struct drm_encoder *radeon_encoder_legacy_tmds_int_add(struct drm_device *dev, int bios_index);
 struct drm_encoder *radeon_encoder_legacy_tmds_ext_add(struct drm_device *dev, int bios_index);
-extern void atombios_external_tmds_setup(struct drm_encoder *encoder, int action);
+extern void atombios_dvo_setup(struct drm_encoder *encoder, int action);
 extern void atombios_digital_setup(struct drm_encoder *encoder, int action);
 extern int atombios_get_encoder_mode(struct drm_encoder *encoder);
+extern void atombios_set_edp_panel_power(struct drm_connector *connector, int action);
 extern void radeon_encoder_set_active_device(struct drm_encoder *encoder);
 
 extern void radeon_crtc_load_lut(struct drm_crtc *crtc);
diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c
index 8eb183466015..1d067743fee0 100644
--- a/drivers/gpu/drm/radeon/radeon_object.c
+++ b/drivers/gpu/drm/radeon/radeon_object.c
@@ -86,11 +86,12 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain)
 }
 
 int radeon_bo_create(struct radeon_device *rdev, struct drm_gem_object *gobj,
-                        unsigned long size, bool kernel, u32 domain,
-                        struct radeon_bo **bo_ptr)
+                     unsigned long size, int byte_align, bool kernel, u32 domain,
+                     struct radeon_bo **bo_ptr)
 {
         struct radeon_bo *bo;
         enum ttm_bo_type type;
+        int page_align = roundup(byte_align, PAGE_SIZE) >> PAGE_SHIFT;
         int r;
 
         if (unlikely(rdev->mman.bdev.dev_mapping == NULL)) {
@@ -115,7 +116,7 @@ retry:
         /* Kernel allocation are uninterruptible */
         mutex_lock(&rdev->vram_mutex);
         r = ttm_bo_init(&rdev->mman.bdev, &bo->tbo, size, type,
-                        &bo->placement, 0, 0, !kernel, NULL, size,
+                        &bo->placement, page_align, 0, !kernel, NULL, size,
                         &radeon_ttm_bo_destroy);
         mutex_unlock(&rdev->vram_mutex);
         if (unlikely(r != 0)) {
diff --git a/drivers/gpu/drm/radeon/radeon_object.h b/drivers/gpu/drm/radeon/radeon_object.h
index 3481bc7f6f58..d143702b244a 100644
--- a/drivers/gpu/drm/radeon/radeon_object.h
+++ b/drivers/gpu/drm/radeon/radeon_object.h
@@ -137,9 +137,10 @@ static inline int radeon_bo_wait(struct radeon_bo *bo, u32 *mem_type,
 }
 
 extern int radeon_bo_create(struct radeon_device *rdev,
-                                struct drm_gem_object *gobj, unsigned long size,
-                                bool kernel, u32 domain,
-                                struct radeon_bo **bo_ptr);
+                            struct drm_gem_object *gobj, unsigned long size,
+                            int byte_align,
+                            bool kernel, u32 domain,
+                            struct radeon_bo **bo_ptr);
 extern int radeon_bo_kmap(struct radeon_bo *bo, void **ptr);
 extern void radeon_bo_kunmap(struct radeon_bo *bo);
 extern void radeon_bo_unref(struct radeon_bo **bo);
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index 6ea798ce8218..06e79822a2bf 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -176,8 +176,8 @@ int radeon_ib_pool_init(struct radeon_device *rdev)
         INIT_LIST_HEAD(&rdev->ib_pool.bogus_ib);
         /* Allocate 1M object buffer */
         r = radeon_bo_create(rdev, NULL,  RADEON_IB_POOL_SIZE*64*1024,
-                                true, RADEON_GEM_DOMAIN_GTT,
-                                &rdev->ib_pool.robj);
+                             PAGE_SIZE, true, RADEON_GEM_DOMAIN_GTT,
+                             &rdev->ib_pool.robj);
         if (r) {
                 DRM_ERROR("radeon: failed to ib pool (%d).\n", r);
                 return r;
@@ -332,7 +332,7 @@ int radeon_ring_init(struct radeon_device *rdev, unsigned ring_size)
         rdev->cp.ring_size = ring_size;
         /* Allocate ring buffer */
         if (rdev->cp.ring_obj == NULL) {
-                r = radeon_bo_create(rdev, NULL, rdev->cp.ring_size, true,
+                r = radeon_bo_create(rdev, NULL, rdev->cp.ring_size, PAGE_SIZE, true,
                                         RADEON_GEM_DOMAIN_GTT,
                                         &rdev->cp.ring_obj);
                 if (r) {
diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c
index 313c96bc09da..5b44f652145c 100644
--- a/drivers/gpu/drm/radeon/radeon_test.c
+++ b/drivers/gpu/drm/radeon/radeon_test.c
@@ -52,7 +52,7 @@ void radeon_test_moves(struct radeon_device *rdev)
                 goto out_cleanup;
         }
 
-        r = radeon_bo_create(rdev, NULL, size, true, RADEON_GEM_DOMAIN_VRAM,
+        r = radeon_bo_create(rdev, NULL, size, PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
                                 &vram_obj);
         if (r) {
                 DRM_ERROR("Failed to create VRAM object\n");
@@ -71,7 +71,7 @@ void radeon_test_moves(struct radeon_device *rdev)
                 void **gtt_start, **gtt_end;
                 void **vram_start, **vram_end;
 
-                r = radeon_bo_create(rdev, NULL, size, true,
+                r = radeon_bo_create(rdev, NULL, size, PAGE_SIZE, true,
                                          RADEON_GEM_DOMAIN_GTT, gtt_obj + i);
                 if (r) {
                         DRM_ERROR("Failed to create GTT object %d\n", i);
diff --git a/drivers/gpu/drm/radeon/radeon_ttm.c b/drivers/gpu/drm/radeon/radeon_ttm.c
index 01c2c736a1da..1272e4b6a1d4 100644
--- a/drivers/gpu/drm/radeon/radeon_ttm.c
+++ b/drivers/gpu/drm/radeon/radeon_ttm.c
@@ -529,7 +529,7 @@ int radeon_ttm_init(struct radeon_device *rdev)
                 DRM_ERROR("Failed initializing VRAM heap.\n");
                 return r;
         }
-        r = radeon_bo_create(rdev, NULL, 256 * 1024, true,
+        r = radeon_bo_create(rdev, NULL, 256 * 1024, PAGE_SIZE, true,
                                 RADEON_GEM_DOMAIN_VRAM,
                                 &rdev->stollen_vga_memory);
         if (r) {
diff --git a/drivers/gpu/drm/radeon/rv770.c b/drivers/gpu/drm/radeon/rv770.c
index 245374e2b778..4dfead8cee33 100644
--- a/drivers/gpu/drm/radeon/rv770.c
+++ b/drivers/gpu/drm/radeon/rv770.c
@@ -915,8 +915,8 @@ static int rv770_vram_scratch_init(struct radeon_device *rdev)
 
         if (rdev->vram_scratch.robj == NULL) {
                 r = radeon_bo_create(rdev, NULL, RADEON_GPU_PAGE_SIZE,
-                                        true, RADEON_GEM_DOMAIN_VRAM,
-                                        &rdev->vram_scratch.robj);
+                                     PAGE_SIZE, true, RADEON_GEM_DOMAIN_VRAM,
+                                     &rdev->vram_scratch.robj);
                 if (r) {
                         return r;
                 }
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index 3ca77dc03915..148a322d8f5d 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -224,6 +224,9 @@ int ttm_bo_reserve_locked(struct ttm_buffer_object *bo,
         int ret;
 
         while (unlikely(atomic_cmpxchg(&bo->reserved, 0, 1) != 0)) {
+                /**
+                 * Deadlock avoidance for multi-bo reserving.
+                 */
                 if (use_sequence && bo->seq_valid &&
                         (sequence - bo->val_seq < (1 << 31))) {
                         return -EAGAIN;
@@ -241,6 +244,14 @@ int ttm_bo_reserve_locked(struct ttm_buffer_object *bo,
         }
 
         if (use_sequence) {
+                /**
+                 * Wake up waiters that may need to recheck for deadlock,
+                 * if we decreased the sequence number.
+                 */
+                if (unlikely((bo->val_seq - sequence < (1 << 31))
+                             || !bo->seq_valid))
+                        wake_up_all(&bo->event_queue);
+
                 bo->val_seq = sequence;
                 bo->seq_valid = true;
         } else {
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
index 36e129f0023f..5408b1b7996f 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
@@ -862,7 +862,7 @@ int vmw_dmabuf_alloc_ioctl(struct drm_device *dev, void *data,
                               &vmw_vram_sys_placement, true,
                               &vmw_user_dmabuf_destroy);
         if (unlikely(ret != 0))
-                return ret;
+                goto out_no_dmabuf;
 
         tmp = ttm_bo_reference(&vmw_user_bo->dma.base);
         ret = ttm_base_object_init(vmw_fpriv(file_priv)->tfile,
@@ -870,19 +870,21 @@ int vmw_dmabuf_alloc_ioctl(struct drm_device *dev, void *data,
                                    false,
                                    ttm_buffer_type,
                                    &vmw_user_dmabuf_release, NULL);
-        if (unlikely(ret != 0)) {
-                ttm_bo_unref(&tmp);
-        } else {
+        if (unlikely(ret != 0))
+                goto out_no_base_object;
+        else {
                 rep->handle = vmw_user_bo->base.hash.key;
                 rep->map_handle = vmw_user_bo->dma.base.addr_space_offset;
                 rep->cur_gmr_id = vmw_user_bo->base.hash.key;
                 rep->cur_gmr_offset = 0;
         }
-        ttm_bo_unref(&tmp);
 
+out_no_base_object:
+        ttm_bo_unref(&tmp);
+out_no_dmabuf:
         ttm_read_unlock(&vmaster->lock);
 
-        return 0;
+        return ret;
 }
 
 int vmw_dmabuf_unref_ioctl(struct drm_device *dev, void *data,
diff --git a/drivers/hid/hidraw.c b/drivers/hid/hidraw.c
index 8a4b32dca9f7..e1f07483691f 100644
--- a/drivers/hid/hidraw.c
+++ b/drivers/hid/hidraw.c
@@ -32,7 +32,6 @@
 #include <linux/hid.h>
 #include <linux/mutex.h>
 #include <linux/sched.h>
-#include <linux/smp_lock.h>
 
 #include <linux/hidraw.h>
 
diff --git a/drivers/hid/usbhid/hiddev.c b/drivers/hid/usbhid/hiddev.c
index fedd88df9a18..984feb351a5a 100644
--- a/drivers/hid/usbhid/hiddev.c
+++ b/drivers/hid/usbhid/hiddev.c
@@ -29,7 +29,6 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 #include <linux/init.h>
-#include <linux/smp_lock.h>
 #include <linux/input.h>
 #include <linux/usb.h>
 #include <linux/hid.h>
diff --git a/drivers/hwmon/amc6821.c b/drivers/hwmon/amc6821.c
index fa9708c2d723..4033974d1bb3 100644
--- a/drivers/hwmon/amc6821.c
+++ b/drivers/hwmon/amc6821.c
@@ -4,7 +4,7 @@
         Copyright (C) 2009 T. Mertelj <tomaz.mertelj@guest.arnes.si>
 
         Based on max6650.c:
-        Copyright (C) 2007 Hans J. Koch <hjk@linutronix.de>
+        Copyright (C) 2007 Hans J. Koch <hjk@hansjkoch.de>
 
         This program is free software; you can redistribute it and/or modify
         it under the terms of the GNU General Public License as published by
diff --git a/drivers/hwmon/i5k_amb.c b/drivers/hwmon/i5k_amb.c
index 937983407e2a..c4c40be0edbf 100644
--- a/drivers/hwmon/i5k_amb.c
+++ b/drivers/hwmon/i5k_amb.c
@@ -497,12 +497,14 @@ static unsigned long chipset_ids[] = {
         0
 };
 
+#ifdef MODULE
 static struct pci_device_id i5k_amb_ids[] __devinitdata = {
         { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_5000_ERR) },
         { PCI_DEVICE(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_5400_ERR) },
         { 0, }
 };
 MODULE_DEVICE_TABLE(pci, i5k_amb_ids);
+#endif
 
 static int __devinit i5k_amb_probe(struct platform_device *pdev)
 {
diff --git a/drivers/hwmon/lis3lv02d_i2c.c b/drivers/hwmon/lis3lv02d_i2c.c
index 9f4bae07f719..8853afce85ce 100644
--- a/drivers/hwmon/lis3lv02d_i2c.c
+++ b/drivers/hwmon/lis3lv02d_i2c.c
@@ -186,7 +186,7 @@ static int __devexit lis3lv02d_i2c_remove(struct i2c_client *client)
         return 0;
 }
 
-#ifdef CONFIG_PM
+#ifdef CONFIG_PM_SLEEP
 static int lis3lv02d_i2c_suspend(struct device *dev)
 {
         struct i2c_client *client = container_of(dev, struct i2c_client, dev);
@@ -213,12 +213,9 @@ static int lis3lv02d_i2c_resume(struct device *dev)
 
         return 0;
 }
-#else
-#define lis3lv02d_i2c_suspend   NULL
-#define lis3lv02d_i2c_resume    NULL
-#define lis3lv02d_i2c_shutdown  NULL
-#endif
+#endif /* CONFIG_PM_SLEEP */
 
+#ifdef CONFIG_PM_RUNTIME
 static int lis3_i2c_runtime_suspend(struct device *dev)
 {
         struct i2c_client *client = container_of(dev, struct i2c_client, dev);
@@ -236,6 +233,7 @@ static int lis3_i2c_runtime_resume(struct device *dev)
         lis3lv02d_poweron(lis3);
         return 0;
 }
+#endif /* CONFIG_PM_RUNTIME */
 
 static const struct i2c_device_id lis3lv02d_id[] = {
         {"lis3lv02d", 0 },
diff --git a/drivers/hwmon/lm93.c b/drivers/hwmon/lm93.c
index 6669255aadcf..c9ed14eba5a6 100644
--- a/drivers/hwmon/lm93.c
+++ b/drivers/hwmon/lm93.c
@@ -20,7 +20,7 @@
     Adapted to 2.6.20 by Carsten Emde <cbe@osadl.org>
         Copyright (c) 2006 Carsten Emde, Open Source Automation Development Lab
 
-    Modified for mainline integration by Hans J. Koch <hjk@linutronix.de>
+    Modified for mainline integration by Hans J. Koch <hjk@hansjkoch.de>
         Copyright (c) 2007 Hans J. Koch, Linutronix GmbH
 
     This program is free software; you can redistribute it and/or modify
@@ -2629,7 +2629,7 @@ static void __exit lm93_exit(void)
 }
 
 MODULE_AUTHOR("Mark M. Hoffman <mhoffman@lightlink.com>, "
-                "Hans J. Koch <hjk@linutronix.de");
+                "Hans J. Koch <hjk@hansjkoch.de>");
 MODULE_DESCRIPTION("LM93 driver");
 MODULE_LICENSE("GPL");
 
diff --git a/drivers/hwmon/lm95241.c b/drivers/hwmon/lm95241.c
index 464340f25496..4546d82f024a 100644
--- a/drivers/hwmon/lm95241.c
+++ b/drivers/hwmon/lm95241.c
@@ -128,9 +128,12 @@ static ssize_t set_interval(struct device *dev, struct device_attribute *attr,
 {
         struct i2c_client *client = to_i2c_client(dev);
         struct lm95241_data *data = i2c_get_clientdata(client);
+        unsigned long val;
 
-        strict_strtol(buf, 10, &data->interval);
-        data->interval = data->interval * HZ / 1000;
+        if (strict_strtoul(buf, 10, &val) < 0)
+                return -EINVAL;
+
+        data->interval = val * HZ / 1000;
 
         return count;
 }
@@ -188,7 +191,9 @@ static ssize_t set_type##flag(struct device *dev, \
         struct lm95241_data *data = i2c_get_clientdata(client); \
 \
         long val; \
-        strict_strtol(buf, 10, &val); \
+\
+        if (strict_strtol(buf, 10, &val) < 0) \
+                return -EINVAL; \
 \
         if ((val == 1) || (val == 2)) { \
 \
@@ -227,7 +232,9 @@ static ssize_t set_min##flag(struct device *dev, \
         struct lm95241_data *data = i2c_get_clientdata(client); \
 \
         long val; \
-        strict_strtol(buf, 10, &val); \
+\
+        if (strict_strtol(buf, 10, &val) < 0) \
+                return -EINVAL;\
 \
         mutex_lock(&data->update_lock); \
 \
@@ -256,7 +263,9 @@ static ssize_t set_max##flag(struct device *dev, \
         struct lm95241_data *data = i2c_get_clientdata(client); \
 \
         long val; \
-        strict_strtol(buf, 10, &val); \
+\
+        if (strict_strtol(buf, 10, &val) < 0) \
+                return -EINVAL; \
 \
         mutex_lock(&data->update_lock); \
 \
diff --git a/drivers/hwmon/max6650.c b/drivers/hwmon/max6650.c
index a0160ee5caef..9a11532ecae8 100644
--- a/drivers/hwmon/max6650.c
+++ b/drivers/hwmon/max6650.c
@@ -2,7 +2,7 @@
  * max6650.c - Part of lm_sensors, Linux kernel modules for hardware
  *             monitoring.
  *
- * (C) 2007 by Hans J. Koch <hjk@linutronix.de>
+ * (C) 2007 by Hans J. Koch <hjk@hansjkoch.de>
  *
  * based on code written by John Morris <john.morris@spirentcom.com>
  * Copyright (c) 2003 Spirent Communications
diff --git a/drivers/hwmon/w83795.c b/drivers/hwmon/w83795.c
index 1d840aa83782..cdbc7448491e 100644
--- a/drivers/hwmon/w83795.c
+++ b/drivers/hwmon/w83795.c
@@ -165,10 +165,14 @@ static const u8 IN_LSB_SHIFT_IDX[][2] = {
 
 #define W83795_REG_VID_CTRL             0x6A
 
+#define W83795_REG_ALARM_CTRL           0x40
+#define ALARM_CTRL_RTSACS               (1 << 7)
 #define W83795_REG_ALARM(index)         (0x41 + (index))
+#define W83795_REG_CLR_CHASSIS          0x4D
 #define W83795_REG_BEEP(index)          (0x50 + (index))
 
-#define W83795_REG_CLR_CHASSIS          0x4D
+#define W83795_REG_OVT_CFG              0x58
+#define OVT_CFG_SEL                     (1 << 7)
 
 
 #define W83795_REG_FCMS1                0x201
@@ -178,6 +182,14 @@ static const u8 IN_LSB_SHIFT_IDX[][2] = {
 
 #define W83795_REG_TSS(index)           (0x209 + (index))
 
+#define TSS_MAP_RESERVED                0xff
+static const u8 tss_map[4][6] = {
+        { 0,  1,  2,  3,  4,  5},
+        { 6,  7,  8,  9,  0,  1},
+        {10, 11, 12, 13,  2,  3},
+        { 4,  5,  4,  5, TSS_MAP_RESERVED, TSS_MAP_RESERVED},
+};
+
 #define PWM_OUTPUT                      0
 #define PWM_FREQ                        1
 #define PWM_START                       2
@@ -369,6 +381,7 @@ struct w83795_data {
         u8 setup_pwm[3];        /* Register value */
 
         u8 alarms[6];           /* Register value */
+        u8 enable_beep;
         u8 beeps[6];            /* Register value */
 
         char valid;
@@ -499,8 +512,11 @@ static void w83795_update_limits(struct i2c_client *client)
         }
 
         /* Read beep settings */
-        for (i = 0; i < ARRAY_SIZE(data->beeps); i++)
-                data->beeps[i] = w83795_read(client, W83795_REG_BEEP(i));
+        if (data->enable_beep) {
+                for (i = 0; i < ARRAY_SIZE(data->beeps); i++)
+                        data->beeps[i] =
+                                w83795_read(client, W83795_REG_BEEP(i));
+        }
 
         data->valid_limits = 1;
 }
@@ -577,6 +593,7 @@ static struct w83795_data *w83795_update_device(struct device *dev)
         struct i2c_client *client = to_i2c_client(dev);
         struct w83795_data *data = i2c_get_clientdata(client);
         u16 tmp;
+        u8 intrusion;
         int i;
 
         mutex_lock(&data->update_lock);
@@ -648,9 +665,24 @@ static struct w83795_data *w83795_update_device(struct device *dev)
                     w83795_read(client, W83795_REG_PWM(i, PWM_OUTPUT));
         }
 
-        /* update alarm */
+        /* Update intrusion and alarms
+         * It is important to read intrusion first, because reading from
+         * register SMI STS6 clears the interrupt status temporarily. */
+        tmp = w83795_read(client, W83795_REG_ALARM_CTRL);
+        /* Switch to interrupt status for intrusion if needed */
+        if (tmp & ALARM_CTRL_RTSACS)
+                w83795_write(client, W83795_REG_ALARM_CTRL,
+                             tmp & ~ALARM_CTRL_RTSACS);
+        intrusion = w83795_read(client, W83795_REG_ALARM(5)) & (1 << 6);
+        /* Switch to real-time alarms */
+        w83795_write(client, W83795_REG_ALARM_CTRL, tmp | ALARM_CTRL_RTSACS);
         for (i = 0; i < ARRAY_SIZE(data->alarms); i++)
                 data->alarms[i] = w83795_read(client, W83795_REG_ALARM(i));
+        data->alarms[5] |= intrusion;
+        /* Restore original configuration if needed */
+        if (!(tmp & ALARM_CTRL_RTSACS))
+                w83795_write(client, W83795_REG_ALARM_CTRL,
+                             tmp & ~ALARM_CTRL_RTSACS);
 
         data->last_updated = jiffies;
         data->valid = 1;
@@ -730,6 +762,10 @@ store_chassis_clear(struct device *dev,
         val = w83795_read(client, W83795_REG_CLR_CHASSIS);
         val |= 0x80;
         w83795_write(client, W83795_REG_CLR_CHASSIS, val);
+
+        /* Clear status and force cache refresh */
+        w83795_read(client, W83795_REG_ALARM(5));
+        data->valid = 0;
         mutex_unlock(&data->update_lock);
         return count;
 }
@@ -857,20 +893,20 @@ show_pwm_enable(struct device *dev, struct device_attribute *attr, char *buf)
         int index = sensor_attr->index;
         u8 tmp;
 
-        if (1 == (data->pwm_fcms[0] & (1 << index))) {
+        /* Speed cruise mode */
+        if (data->pwm_fcms[0] & (1 << index)) {
                 tmp = 2;
                 goto out;
         }
+        /* Thermal cruise or SmartFan IV mode */
         for (tmp = 0; tmp < 6; tmp++) {
                 if (data->pwm_tfmr[tmp] & (1 << index)) {
                         tmp = 3;
                         goto out;
                 }
         }
-        if (data->pwm_fomc & (1 << index))
-                tmp = 0;
-        else
-                tmp = 1;
+        /* Manual mode */
+        tmp = 1;
 
 out:
         return sprintf(buf, "%u\n", tmp);
@@ -890,23 +926,21 @@ store_pwm_enable(struct device *dev, struct device_attribute *attr,
 
         if (strict_strtoul(buf, 10, &val) < 0)
                 return -EINVAL;
-        if (val > 2)
+        if (val < 1 || val > 2)
                 return -EINVAL;
 
         mutex_lock(&data->update_lock);
         switch (val) {
-        case 0:
         case 1:
+                /* Clear speed cruise mode bits */
                 data->pwm_fcms[0] &= ~(1 << index);
                 w83795_write(client, W83795_REG_FCMS1, data->pwm_fcms[0]);
+                /* Clear thermal cruise mode bits */
                 for (i = 0; i < 6; i++) {
                         data->pwm_tfmr[i] &= ~(1 << index);
                         w83795_write(client, W83795_REG_TFMR(i),
                                 data->pwm_tfmr[i]);
                 }
-                data->pwm_fomc |= 1 << index;
-                data->pwm_fomc ^= val << index;
-                w83795_write(client, W83795_REG_FOMC, data->pwm_fomc);
                 break;
         case 2:
                 data->pwm_fcms[0] |= (1 << index);
@@ -918,23 +952,60 @@ store_pwm_enable(struct device *dev, struct device_attribute *attr,
 }
 
 static ssize_t
+show_pwm_mode(struct device *dev, struct device_attribute *attr, char *buf)
+{
+        struct w83795_data *data = w83795_update_pwm_config(dev);
+        int index = to_sensor_dev_attr_2(attr)->index;
+        unsigned int mode;
+
+        if (data->pwm_fomc & (1 << index))
+                mode = 0;       /* DC */
+        else
+                mode = 1;       /* PWM */
+
+        return sprintf(buf, "%u\n", mode);
+}
+
+/*
+ * Check whether a given temperature source can ever be useful.
+ * Returns the number of selectable temperature channels which are
+ * enabled.
+ */
+static int w83795_tss_useful(const struct w83795_data *data, int tsrc)
+{
+        int useful = 0, i;
+
+        for (i = 0; i < 4; i++) {
+                if (tss_map[i][tsrc] == TSS_MAP_RESERVED)
+                        continue;
+                if (tss_map[i][tsrc] < 6)       /* Analog */
+                        useful += (data->has_temp >> tss_map[i][tsrc]) & 1;
+                else                            /* Digital */
+                        useful += (data->has_dts >> (tss_map[i][tsrc] - 6)) & 1;
+        }
+
+        return useful;
+}
+
+static ssize_t
 show_temp_src(struct device *dev, struct device_attribute *attr, char *buf)
 {
         struct sensor_device_attribute_2 *sensor_attr =
             to_sensor_dev_attr_2(attr);
         struct w83795_data *data = w83795_update_pwm_config(dev);
         int index = sensor_attr->index;
-        u8 val = index / 2;
-        u8 tmp = data->temp_src[val];
+        u8 tmp = data->temp_src[index / 2];
 
         if (index & 1)
-                val = 4;
+                tmp >>= 4;      /* Pick high nibble */
         else
-                val = 0;
-        tmp >>= val;
-        tmp &= 0x0f;
+                tmp &= 0x0f;    /* Pick low nibble */
 
-        return sprintf(buf, "%u\n", tmp);
+        /* Look-up the actual temperature channel number */
+        if (tmp >= 4 || tss_map[tmp][index] == TSS_MAP_RESERVED)
+                return -EINVAL;         /* Shouldn't happen */
+
+        return sprintf(buf, "%u\n", (unsigned int)tss_map[tmp][index] + 1);
 }
 
 static ssize_t
@@ -946,12 +1017,21 @@ store_temp_src(struct device *dev, struct device_attribute *attr,
         struct sensor_device_attribute_2 *sensor_attr =
             to_sensor_dev_attr_2(attr);
         int index = sensor_attr->index;
-        unsigned long tmp;
+        int tmp;
+        unsigned long channel;
         u8 val = index / 2;
 
-        if (strict_strtoul(buf, 10, &tmp) < 0)
+        if (strict_strtoul(buf, 10, &channel) < 0 ||
+            channel < 1 || channel > 14)
+                return -EINVAL;
+
+        /* Check if request can be fulfilled */
+        for (tmp = 0; tmp < 4; tmp++) {
+                if (tss_map[tmp][index] == channel - 1)
+                        break;
+        }
+        if (tmp == 4)   /* No match */
                 return -EINVAL;
-        tmp = SENSORS_LIMIT(tmp, 0, 15);
 
         mutex_lock(&data->update_lock);
         if (index & 1) {
@@ -1515,7 +1595,7 @@ store_sf_setup(struct device *dev, struct device_attribute *attr,
 
 #define NOT_USED                        -1
 
-/* Don't change the attribute order, _max and _min are accessed by index
+/* Don't change the attribute order, _max, _min and _beep are accessed by index
  * somewhere else in the code */
 #define SENSOR_ATTR_IN(index) {                                         \
         SENSOR_ATTR_2(in##index##_input, S_IRUGO, show_in, NULL,        \
@@ -1530,6 +1610,8 @@ store_sf_setup(struct device *dev, struct device_attribute *attr,
                 show_alarm_beep, store_beep, BEEP_ENABLE,               \
                 index + ((index > 14) ? 1 : 0)) }
 
+/* Don't change the attribute order, _beep is accessed by index
+ * somewhere else in the code */
 #define SENSOR_ATTR_FAN(index) {                                        \
         SENSOR_ATTR_2(fan##index##_input, S_IRUGO, show_fan,            \
                 NULL, FAN_INPUT, index - 1), \
@@ -1553,9 +1635,13 @@ store_sf_setup(struct device *dev, struct device_attribute *attr,
                 show_pwm, store_pwm, PWM_FREQ, index - 1),       \
         SENSOR_ATTR_2(pwm##index##_enable, S_IWUSR | S_IRUGO,           \
                 show_pwm_enable, store_pwm_enable, NOT_USED, index - 1), \
+        SENSOR_ATTR_2(pwm##index##_mode, S_IRUGO,                       \
+                show_pwm_mode, NULL, NOT_USED, index - 1),              \
         SENSOR_ATTR_2(fan##index##_target, S_IWUSR | S_IRUGO, \
                 show_fanin, store_fanin, FANIN_TARGET, index - 1) }
 
+/* Don't change the attribute order, _beep is accessed by index
+ * somewhere else in the code */
 #define SENSOR_ATTR_DTS(index) {                                        \
         SENSOR_ATTR_2(temp##index##_type, S_IRUGO ,             \
                 show_dts_mode, NULL, NOT_USED, index - 7),      \
@@ -1574,6 +1660,8 @@ store_sf_setup(struct device *dev, struct device_attribute *attr,
         SENSOR_ATTR_2(temp##index##_beep, S_IWUSR | S_IRUGO,            \
                 show_alarm_beep, store_beep, BEEP_ENABLE, index + 17) }
 
+/* Don't change the attribute order, _beep is accessed by index
+ * somewhere else in the code */
 #define SENSOR_ATTR_TEMP(index) {                                       \
         SENSOR_ATTR_2(temp##index##_type, S_IRUGO | (index < 4 ? S_IWUSR : 0), \
                 show_temp_mode, store_temp_mode, NOT_USED, index - 1),  \
@@ -1593,8 +1681,6 @@ store_sf_setup(struct device *dev, struct device_attribute *attr,
         SENSOR_ATTR_2(temp##index##_beep, S_IWUSR | S_IRUGO,            \
                 show_alarm_beep, store_beep, BEEP_ENABLE,               \
                 index + (index > 4 ? 11 : 17)),                         \
-        SENSOR_ATTR_2(temp##index##_source_sel, S_IWUSR | S_IRUGO,      \
-                show_temp_src, store_temp_src, NOT_USED, index - 1),    \
         SENSOR_ATTR_2(temp##index##_pwm_enable, S_IWUSR | S_IRUGO,      \
                 show_temp_pwm_enable, store_temp_pwm_enable,            \
                 TEMP_PWM_ENABLE, index - 1),                            \
@@ -1680,7 +1766,7 @@ static const struct sensor_device_attribute_2 w83795_fan[][4] = {
         SENSOR_ATTR_FAN(14),
 };
 
-static const struct sensor_device_attribute_2 w83795_temp[][29] = {
+static const struct sensor_device_attribute_2 w83795_temp[][28] = {
         SENSOR_ATTR_TEMP(1),
         SENSOR_ATTR_TEMP(2),
         SENSOR_ATTR_TEMP(3),
@@ -1700,7 +1786,7 @@ static const struct sensor_device_attribute_2 w83795_dts[][8] = {
         SENSOR_ATTR_DTS(14),
 };
 
-static const struct sensor_device_attribute_2 w83795_pwm[][7] = {
+static const struct sensor_device_attribute_2 w83795_pwm[][8] = {
         SENSOR_ATTR_PWM(1),
         SENSOR_ATTR_PWM(2),
         SENSOR_ATTR_PWM(3),
@@ -1711,13 +1797,24 @@ static const struct sensor_device_attribute_2 w83795_pwm[][7] = {
         SENSOR_ATTR_PWM(8),
 };
 
+static const struct sensor_device_attribute_2 w83795_tss[6] = {
+        SENSOR_ATTR_2(temp1_source_sel, S_IWUSR | S_IRUGO,
+                      show_temp_src, store_temp_src, NOT_USED, 0),
+        SENSOR_ATTR_2(temp2_source_sel, S_IWUSR | S_IRUGO,
+                      show_temp_src, store_temp_src, NOT_USED, 1),
+        SENSOR_ATTR_2(temp3_source_sel, S_IWUSR | S_IRUGO,
+                      show_temp_src, store_temp_src, NOT_USED, 2),
+        SENSOR_ATTR_2(temp4_source_sel, S_IWUSR | S_IRUGO,
+                      show_temp_src, store_temp_src, NOT_USED, 3),
+        SENSOR_ATTR_2(temp5_source_sel, S_IWUSR | S_IRUGO,
+                      show_temp_src, store_temp_src, NOT_USED, 4),
+        SENSOR_ATTR_2(temp6_source_sel, S_IWUSR | S_IRUGO,
+                      show_temp_src, store_temp_src, NOT_USED, 5),
+};
+
 static const struct sensor_device_attribute_2 sda_single_files[] = {
         SENSOR_ATTR_2(intrusion0_alarm, S_IWUSR | S_IRUGO, show_alarm_beep,
                       store_chassis_clear, ALARM_STATUS, 46),
-        SENSOR_ATTR_2(intrusion0_beep, S_IWUSR | S_IRUGO, show_alarm_beep,
-                      store_beep, BEEP_ENABLE, 46),
-        SENSOR_ATTR_2(beep_enable, S_IWUSR | S_IRUGO, show_alarm_beep,
-                      store_beep, BEEP_ENABLE, 47),
 #ifdef CONFIG_SENSORS_W83795_FANCTRL
         SENSOR_ATTR_2(speed_cruise_tolerance, S_IWUSR | S_IRUGO, show_fanin,
                 store_fanin, FANIN_TOL, NOT_USED),
@@ -1730,6 +1827,13 @@ static const struct sensor_device_attribute_2 sda_single_files[] = {
 #endif
 };
 
+static const struct sensor_device_attribute_2 sda_beep_files[] = {
+        SENSOR_ATTR_2(intrusion0_beep, S_IWUSR | S_IRUGO, show_alarm_beep,
+                      store_beep, BEEP_ENABLE, 46),
+        SENSOR_ATTR_2(beep_enable, S_IWUSR | S_IRUGO, show_alarm_beep,
+                      store_beep, BEEP_ENABLE, 47),
+};
+
 /*
  * Driver interface
  */
@@ -1859,6 +1963,8 @@ static int w83795_handle_files(struct device *dev, int (*fn)(struct device *,
                 if (!(data->has_in & (1 << i)))
                         continue;
                 for (j = 0; j < ARRAY_SIZE(w83795_in[0]); j++) {
+                        if (j == 4 && !data->enable_beep)
+                                continue;
                         err = fn(dev, &w83795_in[i][j].dev_attr);
                         if (err)
                                 return err;
@@ -1869,18 +1975,37 @@ static int w83795_handle_files(struct device *dev, int (*fn)(struct device *,
                 if (!(data->has_fan & (1 << i)))
                         continue;
                 for (j = 0; j < ARRAY_SIZE(w83795_fan[0]); j++) {
+                        if (j == 3 && !data->enable_beep)
+                                continue;
                         err = fn(dev, &w83795_fan[i][j].dev_attr);
                         if (err)
                                 return err;
                 }
         }
 
+        for (i = 0; i < ARRAY_SIZE(w83795_tss); i++) {
+                j = w83795_tss_useful(data, i);
+                if (!j)
+                        continue;
+                err = fn(dev, &w83795_tss[i].dev_attr);
+                if (err)
+                        return err;
+        }
+
         for (i = 0; i < ARRAY_SIZE(sda_single_files); i++) {
                 err = fn(dev, &sda_single_files[i].dev_attr);
                 if (err)
                         return err;
         }
 
+        if (data->enable_beep) {
+                for (i = 0; i < ARRAY_SIZE(sda_beep_files); i++) {
+                        err = fn(dev, &sda_beep_files[i].dev_attr);
+                        if (err)
+                                return err;
+                }
+        }
+
 #ifdef CONFIG_SENSORS_W83795_FANCTRL
         for (i = 0; i < data->has_pwm; i++) {
                 for (j = 0; j < ARRAY_SIZE(w83795_pwm[0]); j++) {
@@ -1899,6 +2024,8 @@ static int w83795_handle_files(struct device *dev, int (*fn)(struct device *,
 #else
                 for (j = 0; j < 8; j++) {
 #endif
+                        if (j == 7 && !data->enable_beep)
+                                continue;
                         err = fn(dev, &w83795_temp[i][j].dev_attr);
                         if (err)
                                 return err;
@@ -1910,6 +2037,8 @@ static int w83795_handle_files(struct device *dev, int (*fn)(struct device *,
                         if (!(data->has_dts & (1 << i)))
                                 continue;
                         for (j = 0; j < ARRAY_SIZE(w83795_dts[0]); j++) {
+                                if (j == 7 && !data->enable_beep)
+                                        continue;
                                 err = fn(dev, &w83795_dts[i][j].dev_attr);
                                 if (err)
                                         return err;
@@ -2049,6 +2178,18 @@ static int w83795_probe(struct i2c_client *client,
         else
                 data->has_pwm = 2;
 
+        /* Check if BEEP pin is available */
+        if (data->chip_type == w83795g) {
+                /* The W83795G has a dedicated BEEP pin */
+                data->enable_beep = 1;
+        } else {
+                /* The W83795ADG has a shared pin for OVT# and BEEP, so you
+                 * can't have both */
+                tmp = w83795_read(client, W83795_REG_OVT_CFG);
+                if ((tmp & OVT_CFG_SEL) == 0)
+                        data->enable_beep = 1;
+        }
+
         err = w83795_handle_files(dev, device_create_file);
         if (err)
                 goto exit_remove;
diff --git a/drivers/i2c/i2c-core.c b/drivers/i2c/i2c-core.c
index d231f683f576..6b4cc567645b 100644
--- a/drivers/i2c/i2c-core.c
+++ b/drivers/i2c/i2c-core.c
@@ -848,6 +848,18 @@ static int i2c_register_adapter(struct i2c_adapter *adap)
                 goto out_list;
         }
 
+        /* Sanity checks */
+        if (unlikely(adap->name[0] == '\0')) {
+                pr_err("i2c-core: Attempt to register an adapter with "
+                       "no name!\n");
+                return -EINVAL;
+        }
+        if (unlikely(!adap->algo)) {
+                pr_err("i2c-core: Attempt to register adapter '%s' with "
+                       "no algo!\n", adap->name);
+                return -EINVAL;
+        }
+
         rt_mutex_init(&adap->bus_lock);
         mutex_init(&adap->userspace_clients_lock);
         INIT_LIST_HEAD(&adap->userspace_clients);
diff --git a/drivers/i2c/i2c-mux.c b/drivers/i2c/i2c-mux.c
index d32a4843fc3a..d7a4833be416 100644
--- a/drivers/i2c/i2c-mux.c
+++ b/drivers/i2c/i2c-mux.c
@@ -120,7 +120,6 @@ struct i2c_adapter *i2c_add_mux_adapter(struct i2c_adapter *parent,
         snprintf(priv->adap.name, sizeof(priv->adap.name),
                  "i2c-%d-mux (chan_id %d)", i2c_adapter_id(parent), chan_id);
         priv->adap.owner = THIS_MODULE;
-        priv->adap.id = parent->id;
         priv->adap.algo = &priv->algo;
         priv->adap.algo_data = priv;
         priv->adap.dev.parent = &parent->dev;
diff --git a/drivers/infiniband/hw/ipath/ipath_file_ops.c b/drivers/infiniband/hw/ipath/ipath_file_ops.c
index 6078992da3f0..9292a15ad7c4 100644
--- a/drivers/infiniband/hw/ipath/ipath_file_ops.c
+++ b/drivers/infiniband/hw/ipath/ipath_file_ops.c
@@ -40,7 +40,6 @@
 #include <linux/highmem.h>
 #include <linux/io.h>
 #include <linux/jiffies.h>
-#include <linux/smp_lock.h>
 #include <asm/pgtable.h>
 
 #include "ipath_kernel.h"
diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index cfc1d65c4577..1e1e347a7715 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -1123,7 +1123,7 @@ static void srp_send_completion(struct ib_cq *cq, void *target_ptr)
         }
 }
 
-static int srp_queuecommand(struct scsi_cmnd *scmnd,
+static int srp_queuecommand_lck(struct scsi_cmnd *scmnd,
                             void (*done)(struct scsi_cmnd *))
 {
         struct srp_target_port *target = host_to_target(scmnd->device->host);
@@ -1196,6 +1196,8 @@ err:
         return SCSI_MLQUEUE_HOST_BUSY;
 }
 
+static DEF_SCSI_QCMD(srp_queuecommand)
+
 static int srp_alloc_iu_bufs(struct srp_target_port *target)
 {
         int i;
diff --git a/drivers/input/input.c b/drivers/input/input.c
index 7f26ca6ecf75..db409d6bd5d2 100644
--- a/drivers/input/input.c
+++ b/drivers/input/input.c
@@ -24,7 +24,6 @@
 #include <linux/device.h>
 #include <linux/mutex.h>
 #include <linux/rcupdate.h>
-#include <linux/smp_lock.h>
 #include "input-compat.h"
 
 MODULE_AUTHOR("Vojtech Pavlik <vojtech@suse.cz>");
@@ -753,7 +752,7 @@ static int input_default_setkeycode(struct input_dev *dev,
         if (index >= dev->keycodemax)
                 return -EINVAL;
 
-        if (dev->keycodesize < sizeof(dev->keycode) &&
+        if (dev->keycodesize < sizeof(ke->keycode) &&
                         (ke->keycode >> (dev->keycodesize * 8)))
                 return -EINVAL;
 
diff --git a/drivers/input/serio/serio_raw.c b/drivers/input/serio/serio_raw.c
index cd82bb125915..b7ba4597f7f0 100644
--- a/drivers/input/serio/serio_raw.c
+++ b/drivers/input/serio/serio_raw.c
@@ -11,7 +11,6 @@
 
 #include <linux/sched.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/poll.h>
 #include <linux/module.h>
 #include <linux/serio.h>
diff --git a/drivers/input/tablet/aiptek.c b/drivers/input/tablet/aiptek.c
index 57b25b84d1fc..0a619c558bfb 100644
--- a/drivers/input/tablet/aiptek.c
+++ b/drivers/input/tablet/aiptek.c
@@ -1097,7 +1097,7 @@ store_tabletPointerMode(struct device *dev, struct device_attribute *attr, const
 }
 
 static DEVICE_ATTR(pointer_mode,
-                   S_IRUGO | S_IWUGO,
+                   S_IRUGO | S_IWUSR,
                    show_tabletPointerMode, store_tabletPointerMode);
 
 /***********************************************************************
@@ -1134,7 +1134,7 @@ store_tabletCoordinateMode(struct device *dev, struct device_attribute *attr, co
 }
 
 static DEVICE_ATTR(coordinate_mode,
-                   S_IRUGO | S_IWUGO,
+                   S_IRUGO | S_IWUSR,
                    show_tabletCoordinateMode, store_tabletCoordinateMode);
 
 /***********************************************************************
@@ -1176,7 +1176,7 @@ store_tabletToolMode(struct device *dev, struct device_attribute *attr, const ch
 }
 
 static DEVICE_ATTR(tool_mode,
-                   S_IRUGO | S_IWUGO,
+                   S_IRUGO | S_IWUSR,
                    show_tabletToolMode, store_tabletToolMode);
 
 /***********************************************************************
@@ -1219,7 +1219,7 @@ store_tabletXtilt(struct device *dev, struct device_attribute *attr, const char
 }
 
 static DEVICE_ATTR(xtilt,
-                   S_IRUGO | S_IWUGO, show_tabletXtilt, store_tabletXtilt);
+                   S_IRUGO | S_IWUSR, show_tabletXtilt, store_tabletXtilt);
 
 /***********************************************************************
  * support routines for the 'ytilt' file. Note that this file
@@ -1261,7 +1261,7 @@ store_tabletYtilt(struct device *dev, struct device_attribute *attr, const char
 }
 
 static DEVICE_ATTR(ytilt,
-                   S_IRUGO | S_IWUGO, show_tabletYtilt, store_tabletYtilt);
+                   S_IRUGO | S_IWUSR, show_tabletYtilt, store_tabletYtilt);
 
 /***********************************************************************
  * support routines for the 'jitter' file. Note that this file
@@ -1288,7 +1288,7 @@ store_tabletJitterDelay(struct device *dev, struct device_attribute *attr, const
 }
 
 static DEVICE_ATTR(jitter,
-                   S_IRUGO | S_IWUGO,
+                   S_IRUGO | S_IWUSR,
                    show_tabletJitterDelay, store_tabletJitterDelay);
 
 /***********************************************************************
@@ -1317,7 +1317,7 @@ store_tabletProgrammableDelay(struct device *dev, struct device_attribute *attr,
 }
 
 static DEVICE_ATTR(delay,
-                   S_IRUGO | S_IWUGO,
+                   S_IRUGO | S_IWUSR,
                    show_tabletProgrammableDelay, store_tabletProgrammableDelay);
 
 /***********************************************************************
@@ -1406,7 +1406,7 @@ store_tabletStylusUpper(struct device *dev, struct device_attribute *attr, const
 }
 
 static DEVICE_ATTR(stylus_upper,
-                   S_IRUGO | S_IWUGO,
+                   S_IRUGO | S_IWUSR,
                    show_tabletStylusUpper, store_tabletStylusUpper);
 
 /***********************************************************************
@@ -1437,7 +1437,7 @@ store_tabletStylusLower(struct device *dev, struct device_attribute *attr, const
 }
 
 static DEVICE_ATTR(stylus_lower,
-                   S_IRUGO | S_IWUGO,
+                   S_IRUGO | S_IWUSR,
                    show_tabletStylusLower, store_tabletStylusLower);
 
 /***********************************************************************
@@ -1475,7 +1475,7 @@ store_tabletMouseLeft(struct device *dev, struct device_attribute *attr, const c
 }
 
 static DEVICE_ATTR(mouse_left,
-                   S_IRUGO | S_IWUGO,
+                   S_IRUGO | S_IWUSR,
                    show_tabletMouseLeft, store_tabletMouseLeft);
 
 /***********************************************************************
@@ -1505,7 +1505,7 @@ store_tabletMouseMiddle(struct device *dev, struct device_attribute *attr, const
 }
 
 static DEVICE_ATTR(mouse_middle,
-                   S_IRUGO | S_IWUGO,
+                   S_IRUGO | S_IWUSR,
                    show_tabletMouseMiddle, store_tabletMouseMiddle);
 
 /***********************************************************************
@@ -1535,7 +1535,7 @@ store_tabletMouseRight(struct device *dev, struct device_attribute *attr, const
 }
 
 static DEVICE_ATTR(mouse_right,
-                   S_IRUGO | S_IWUGO,
+                   S_IRUGO | S_IWUSR,
                    show_tabletMouseRight, store_tabletMouseRight);
 
 /***********************************************************************
@@ -1567,7 +1567,7 @@ store_tabletWheel(struct device *dev, struct device_attribute *attr, const char
 }
 
 static DEVICE_ATTR(wheel,
-                   S_IRUGO | S_IWUGO, show_tabletWheel, store_tabletWheel);
+                   S_IRUGO | S_IWUSR, show_tabletWheel, store_tabletWheel);
 
 /***********************************************************************
  * support routines for the 'execute' file. Note that this file
@@ -1600,7 +1600,7 @@ store_tabletExecute(struct device *dev, struct device_attribute *attr, const cha
 }
 
 static DEVICE_ATTR(execute,
-                   S_IRUGO | S_IWUGO, show_tabletExecute, store_tabletExecute);
+                   S_IRUGO | S_IWUSR, show_tabletExecute, store_tabletExecute);
 
 /***********************************************************************
  * support routines for the 'odm_code' file. Note that this file
diff --git a/drivers/leds/leds-lp5521.c b/drivers/leds/leds-lp5521.c
index 3782f31f06d2..33facd0c45d1 100644
--- a/drivers/leds/leds-lp5521.c
+++ b/drivers/leds/leds-lp5521.c
@@ -125,11 +125,22 @@ struct lp5521_chip {
         u8                      num_leds;
 };
 
-#define cdev_to_led(c)          container_of(c, struct lp5521_led, cdev)
-#define engine_to_lp5521(eng)   container_of((eng), struct lp5521_chip, \
-                                                engines[(eng)->id - 1])
-#define led_to_lp5521(led)      container_of((led), struct lp5521_chip, \
-                                                leds[(led)->id])
+static inline struct lp5521_led *cdev_to_led(struct led_classdev *cdev)
+{
+        return container_of(cdev, struct lp5521_led, cdev);
+}
+
+static inline struct lp5521_chip *engine_to_lp5521(struct lp5521_engine *engine)
+{
+        return container_of(engine, struct lp5521_chip,
+                            engines[engine->id - 1]);
+}
+
+static inline struct lp5521_chip *led_to_lp5521(struct lp5521_led *led)
+{
+        return container_of(led, struct lp5521_chip,
+                            leds[led->id]);
+}
 
 static void lp5521_led_brightness_work(struct work_struct *work);
 
@@ -185,14 +196,17 @@ static int lp5521_load_program(struct lp5521_engine *eng, const u8 *pattern)
 
         /* move current engine to direct mode and remember the state */
         ret = lp5521_set_engine_mode(eng, LP5521_CMD_DIRECT);
-        usleep_range(1000, 10000);
+        /* Mode change requires min 500 us delay. 1 - 2 ms  with margin */
+        usleep_range(1000, 2000);
         ret |= lp5521_read(client, LP5521_REG_OP_MODE, &mode);
 
         /* For loading, all the engines to load mode */
         lp5521_write(client, LP5521_REG_OP_MODE, LP5521_CMD_DIRECT);
-        usleep_range(1000, 10000);
+        /* Mode change requires min 500 us delay. 1 - 2 ms  with margin */
+        usleep_range(1000, 2000);
         lp5521_write(client, LP5521_REG_OP_MODE, LP5521_CMD_LOAD);
-        usleep_range(1000, 10000);
+        /* Mode change requires min 500 us delay. 1 - 2 ms  with margin */
+        usleep_range(1000, 2000);
 
         addr = LP5521_PROG_MEM_BASE + eng->prog_page * LP5521_PROG_MEM_SIZE;
         i2c_smbus_write_i2c_block_data(client,
@@ -231,10 +245,6 @@ static int lp5521_configure(struct i2c_client *client,
 
         lp5521_init_engine(chip, attr_group);
 
-        lp5521_write(client, LP5521_REG_RESET, 0xff);
-
-        usleep_range(10000, 20000);
-
         /* Set all PWMs to direct control mode */
         ret = lp5521_write(client, LP5521_REG_OP_MODE, 0x3F);
 
@@ -251,8 +261,8 @@ static int lp5521_configure(struct i2c_client *client,
         ret |= lp5521_write(client, LP5521_REG_ENABLE,
                         LP5521_MASTER_ENABLE | LP5521_LOGARITHMIC_PWM |
                         LP5521_EXEC_RUN);
-        /* enable takes 500us */
-        usleep_range(500, 20000);
+        /* enable takes 500us. 1 - 2 ms leaves some margin */
+        usleep_range(1000, 2000);
 
         return ret;
 }
@@ -305,7 +315,8 @@ static int lp5521_detect(struct i2c_client *client)
                         LP5521_MASTER_ENABLE | LP5521_LOGARITHMIC_PWM);
         if (ret)
                 return ret;
-        usleep_range(1000, 10000);
+        /* enable takes 500us. 1 - 2 ms leaves some margin */
+        usleep_range(1000, 2000);
         ret = lp5521_read(client, LP5521_REG_ENABLE, &buf);
         if (ret)
                 return ret;
@@ -693,11 +704,16 @@ static int lp5521_probe(struct i2c_client *client,
 
         if (pdata->enable) {
                 pdata->enable(0);
-                usleep_range(1000, 10000);
+                usleep_range(1000, 2000); /* Keep enable down at least 1ms */
                 pdata->enable(1);
-                usleep_range(1000, 10000); /* Spec says min 500us */
+                usleep_range(1000, 2000); /* 500us abs min. */
         }
 
+        lp5521_write(client, LP5521_REG_RESET, 0xff);
+        usleep_range(10000, 20000); /*
+                                     * Exact value is not available. 10 - 20ms
+                                     * appears to be enough for reset.
+                                     */
         ret = lp5521_detect(client);
 
         if (ret) {
diff --git a/drivers/leds/leds-lp5523.c b/drivers/leds/leds-lp5523.c
index 1e11fcc08b28..0cc4ead2fd8b 100644
--- a/drivers/leds/leds-lp5523.c
+++ b/drivers/leds/leds-lp5523.c
@@ -134,15 +134,18 @@ struct lp5523_chip {
         u8                      num_leds;
 };
 
-#define cdev_to_led(c)          container_of(c, struct lp5523_led, cdev)
+static inline struct lp5523_led *cdev_to_led(struct led_classdev *cdev)
+{
+        return container_of(cdev, struct lp5523_led, cdev);
+}
 
-static struct lp5523_chip *engine_to_lp5523(struct lp5523_engine *engine)
+static inline struct lp5523_chip *engine_to_lp5523(struct lp5523_engine *engine)
 {
         return container_of(engine, struct lp5523_chip,
                             engines[engine->id - 1]);
 }
 
-static struct lp5523_chip *led_to_lp5523(struct lp5523_led *led)
+static inline struct lp5523_chip *led_to_lp5523(struct lp5523_led *led)
 {
         return container_of(led, struct lp5523_chip,
                             leds[led->id]);
@@ -200,13 +203,9 @@ static int lp5523_configure(struct i2c_client *client)
                 { 0x9c, 0x50, 0x9c, 0xd0, 0x9d, 0x80, 0xd8, 0x00, 0},
         };
 
-        lp5523_write(client, LP5523_REG_RESET, 0xff);
-
-        usleep_range(10000, 100000);
-
         ret |= lp5523_write(client, LP5523_REG_ENABLE, LP5523_ENABLE);
-        /* Chip startup time after reset is 500 us */
-        usleep_range(1000, 10000);
+        /* Chip startup time is 500 us, 1 - 2 ms gives some margin */
+        usleep_range(1000, 2000);
 
         ret |= lp5523_write(client, LP5523_REG_CONFIG,
                             LP5523_AUTO_INC | LP5523_PWR_SAVE |
@@ -243,8 +242,8 @@ static int lp5523_configure(struct i2c_client *client)
                 return -1;
         }
 
-        /* Wait 3ms and check the engine status */
-        usleep_range(3000, 20000);
+        /* Let the programs run for couple of ms and check the engine status */
+        usleep_range(3000, 6000);
         lp5523_read(client, LP5523_REG_STATUS, &status);
         status &= LP5523_ENG_STATUS_MASK;
 
@@ -449,10 +448,10 @@ static ssize_t lp5523_selftest(struct device *dev,
         /* Measure VDD (i.e. VBAT) first (channel 16 corresponds to VDD) */
         lp5523_write(chip->client, LP5523_REG_LED_TEST_CTRL,
                                     LP5523_EN_LEDTEST | 16);
-        usleep_range(3000, 10000);
+        usleep_range(3000, 6000); /* ADC conversion time is typically 2.7 ms */
         ret = lp5523_read(chip->client, LP5523_REG_STATUS, &status);
         if (!(status & LP5523_LEDTEST_DONE))
-                usleep_range(3000, 10000);
+                usleep_range(3000, 6000); /* Was not ready. Wait little bit */
 
         ret |= lp5523_read(chip->client, LP5523_REG_LED_TEST_ADC, &vdd);
         vdd--;  /* There may be some fluctuation in measurement */
@@ -468,16 +467,16 @@ static ssize_t lp5523_selftest(struct device *dev,
                         chip->pdata->led_config[i].led_current);
 
                 lp5523_write(chip->client, LP5523_REG_LED_PWM_BASE + i, 0xff);
-                /* let current stabilize 2ms before measurements start */
-                usleep_range(2000, 10000);
+                /* let current stabilize 2 - 4ms before measurements start */
+                usleep_range(2000, 4000);
                 lp5523_write(chip->client,
                              LP5523_REG_LED_TEST_CTRL,
                              LP5523_EN_LEDTEST | i);
-                /* ledtest takes 2.7ms */
-                usleep_range(3000, 10000);
+                /* ADC conversion time is 2.7 ms typically */
+                usleep_range(3000, 6000);
                 ret = lp5523_read(chip->client, LP5523_REG_STATUS, &status);
                 if (!(status & LP5523_LEDTEST_DONE))
-                        usleep_range(3000, 10000);
+                        usleep_range(3000, 6000);/* Was not ready. Wait. */
                 ret |= lp5523_read(chip->client, LP5523_REG_LED_TEST_ADC, &adc);
 
                 if (adc >= vdd || adc < LP5523_ADC_SHORTCIRC_LIM)
@@ -930,11 +929,16 @@ static int lp5523_probe(struct i2c_client *client,
 
         if (pdata->enable) {
                 pdata->enable(0);
-                usleep_range(1000, 10000);
+                usleep_range(1000, 2000); /* Keep enable down at least 1ms */
                 pdata->enable(1);
-                usleep_range(1000, 10000); /* Spec says min 500us */
+                usleep_range(1000, 2000); /* 500us abs min. */
         }
 
+        lp5523_write(client, LP5523_REG_RESET, 0xff);
+        usleep_range(10000, 20000); /*
+                                     * Exact value is not available. 10 - 20ms
+                                     * appears to be enough for reset.
+                                     */
         ret = lp5523_detect(client);
         if (ret)
                 goto fail2;
diff --git a/drivers/leds/leds-ss4200.c b/drivers/leds/leds-ss4200.c
index a688293abd0b..614ebebaaa28 100644
--- a/drivers/leds/leds-ss4200.c
+++ b/drivers/leds/leds-ss4200.c
@@ -102,6 +102,7 @@ static struct dmi_system_id __initdata nas_led_whitelist[] = {
                         DMI_MATCH(DMI_PRODUCT_VERSION, "1.00.00")
                 }
         },
+        {}
 };
 
 /*
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 324a3663fcda..84c46a161927 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -1337,7 +1337,7 @@ super_90_rdev_size_change(mdk_rdev_t *rdev, sector_t num_sectors)
         md_super_write(rdev->mddev, rdev, rdev->sb_start, rdev->sb_size,
                        rdev->sb_page);
         md_super_wait(rdev->mddev);
-        return num_sectors / 2; /* kB for sysfs */
+        return num_sectors;
 }
 
 
@@ -1704,7 +1704,7 @@ super_1_rdev_size_change(mdk_rdev_t *rdev, sector_t num_sectors)
         md_super_write(rdev->mddev, rdev, rdev->sb_start, rdev->sb_size,
                        rdev->sb_page);
         md_super_wait(rdev->mddev);
-        return num_sectors / 2; /* kB for sysfs */
+        return num_sectors;
 }
 
 static struct super_type super_types[] = {
@@ -4338,6 +4338,8 @@ static int md_alloc(dev_t dev, char *name)
         if (mddev->kobj.sd &&
             sysfs_create_group(&mddev->kobj, &md_bitmap_group))
                 printk(KERN_DEBUG "pointless warning\n");
+
+        blk_queue_flush(mddev->queue, REQ_FLUSH | REQ_FUA);
  abort:
         mutex_unlock(&disks_mutex);
         if (!error && mddev->kobj.sd) {
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 45f8324196ec..845cf95b612c 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -1161,6 +1161,7 @@ static int raid1_remove_disk(mddev_t *mddev, int number)
                  * is not possible.
                  */
                 if (!test_bit(Faulty, &rdev->flags) &&
+                    !mddev->recovery_disabled &&
                     mddev->degraded < conf->raid_disks) {
                         err = -EBUSY;
                         goto abort;
diff --git a/drivers/media/common/saa7146_i2c.c b/drivers/media/common/saa7146_i2c.c
index 3d88542612ea..74ee172b5bc9 100644
--- a/drivers/media/common/saa7146_i2c.c
+++ b/drivers/media/common/saa7146_i2c.c
@@ -391,7 +391,6 @@ static int saa7146_i2c_xfer(struct i2c_adapter* adapter, struct i2c_msg *msg, in
 
 /*****************************************************************************/
 /* i2c-adapter helper functions                                              */
-#include <linux/i2c-id.h>
 
 /* exported algorithm data */
 static struct i2c_algorithm saa7146_algo = {
diff --git a/drivers/media/dvb/dvb-core/dvb_ca_en50221.c b/drivers/media/dvb/dvb-core/dvb_ca_en50221.c
index 4d0646da6087..7ea517b7e186 100644
--- a/drivers/media/dvb/dvb-core/dvb_ca_en50221.c
+++ b/drivers/media/dvb/dvb-core/dvb_ca_en50221.c
@@ -36,7 +36,6 @@
 #include <linux/delay.h>
 #include <linux/spinlock.h>
 #include <linux/sched.h>
-#include <linux/smp_lock.h>
 #include <linux/kthread.h>
 
 #include "dvb_ca_en50221.h"
diff --git a/drivers/media/dvb/dvb-core/dvb_frontend.c b/drivers/media/dvb/dvb-core/dvb_frontend.c
index 1589d5a5cb46..cad6634610ea 100644
--- a/drivers/media/dvb/dvb-core/dvb_frontend.c
+++ b/drivers/media/dvb/dvb-core/dvb_frontend.c
@@ -36,7 +36,6 @@
 #include <linux/list.h>
 #include <linux/freezer.h>
 #include <linux/jiffies.h>
-#include <linux/smp_lock.h>
 #include <linux/kthread.h>
 #include <asm/processor.h>
 
diff --git a/drivers/media/dvb/ngene/ngene-core.c b/drivers/media/dvb/ngene/ngene-core.c
index 4caeb163a666..3a7ef71087be 100644
--- a/drivers/media/dvb/ngene/ngene-core.c
+++ b/drivers/media/dvb/ngene/ngene-core.c
@@ -34,7 +34,6 @@
 #include <linux/io.h>
 #include <asm/div64.h>
 #include <linux/pci.h>
-#include <linux/smp_lock.h>
 #include <linux/timer.h>
 #include <linux/byteorder/generic.h>
 #include <linux/firmware.h>
diff --git a/drivers/media/dvb/ngene/ngene-dvb.c b/drivers/media/dvb/ngene/ngene-dvb.c
index 48f980b21d66..3832e5983c19 100644
--- a/drivers/media/dvb/ngene/ngene-dvb.c
+++ b/drivers/media/dvb/ngene/ngene-dvb.c
@@ -35,7 +35,6 @@
 #include <linux/io.h>
 #include <asm/div64.h>
 #include <linux/pci.h>
-#include <linux/smp_lock.h>
 #include <linux/timer.h>
 #include <linux/byteorder/generic.h>
 #include <linux/firmware.h>
diff --git a/drivers/media/dvb/ngene/ngene-i2c.c b/drivers/media/dvb/ngene/ngene-i2c.c
index c3ae956714e7..d28554f8ce99 100644
--- a/drivers/media/dvb/ngene/ngene-i2c.c
+++ b/drivers/media/dvb/ngene/ngene-i2c.c
@@ -37,7 +37,6 @@
 #include <asm/div64.h>
 #include <linux/pci.h>
 #include <linux/pci_ids.h>
-#include <linux/smp_lock.h>
 #include <linux/timer.h>
 #include <linux/byteorder/generic.h>
 #include <linux/firmware.h>
diff --git a/drivers/media/radio/radio-mr800.c b/drivers/media/radio/radio-mr800.c
index b540e8072e92..e6b2d085a449 100644
--- a/drivers/media/radio/radio-mr800.c
+++ b/drivers/media/radio/radio-mr800.c
@@ -58,7 +58,6 @@
 #include <linux/module.h>
 #include <linux/init.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/input.h>
 #include <linux/videodev2.h>
 #include <media/v4l2-device.h>
diff --git a/drivers/media/radio/si470x/radio-si470x.h b/drivers/media/radio/si470x/radio-si470x.h
index ea12782359a0..b9914d7a0c9f 100644
--- a/drivers/media/radio/si470x/radio-si470x.h
+++ b/drivers/media/radio/si470x/radio-si470x.h
@@ -31,7 +31,6 @@
 #include <linux/init.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/input.h>
 #include <linux/version.h>
 #include <linux/videodev2.h>
diff --git a/drivers/media/video/bt8xx/bttv-driver.c b/drivers/media/video/bt8xx/bttv-driver.c
index 3da6e80e1041..a529619e51f6 100644
--- a/drivers/media/video/bt8xx/bttv-driver.c
+++ b/drivers/media/video/bt8xx/bttv-driver.c
@@ -42,7 +42,6 @@
 #include <linux/fs.h>
 #include <linux/kernel.h>
 #include <linux/sched.h>
-#include <linux/smp_lock.h>
 #include <linux/interrupt.h>
 #include <linux/kdev_t.h>
 #include "bttvp.h"
diff --git a/drivers/media/video/cx88/cx88-blackbird.c b/drivers/media/video/cx88/cx88-blackbird.c
index 417d1d5c73c4..d7c94848249e 100644
--- a/drivers/media/video/cx88/cx88-blackbird.c
+++ b/drivers/media/video/cx88/cx88-blackbird.c
@@ -33,7 +33,6 @@
 #include <linux/delay.h>
 #include <linux/device.h>
 #include <linux/firmware.h>
-#include <linux/smp_lock.h>
 #include <media/v4l2-common.h>
 #include <media/v4l2-ioctl.h>
 #include <media/cx2341x.h>
diff --git a/drivers/media/video/cx88/cx88-video.c b/drivers/media/video/cx88/cx88-video.c
index d2f159daa8b5..88b51194f917 100644
--- a/drivers/media/video/cx88/cx88-video.c
+++ b/drivers/media/video/cx88/cx88-video.c
@@ -31,7 +31,6 @@
 #include <linux/kmod.h>
 #include <linux/kernel.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/interrupt.h>
 #include <linux/dma-mapping.h>
 #include <linux/delay.h>
diff --git a/drivers/media/video/imx074.c b/drivers/media/video/imx074.c
index 380e459f899d..27b5dfdfbb93 100644
--- a/drivers/media/video/imx074.c
+++ b/drivers/media/video/imx074.c
@@ -451,7 +451,6 @@ static int imx074_probe(struct i2c_client *client,
         ret = imx074_video_probe(icd, client);
         if (ret < 0) {
                 icd->ops = NULL;
-                i2c_set_clientdata(client, NULL);
                 kfree(priv);
                 return ret;
         }
@@ -468,7 +467,6 @@ static int imx074_remove(struct i2c_client *client)
         icd->ops = NULL;
         if (icl->free_bus)
                 icl->free_bus(icl);
-        i2c_set_clientdata(client, NULL);
         client->driver = NULL;
         kfree(priv);
 
diff --git a/drivers/media/video/ir-kbd-i2c.c b/drivers/media/video/ir-kbd-i2c.c
index 5a000c65ae98..ce4a75375909 100644
--- a/drivers/media/video/ir-kbd-i2c.c
+++ b/drivers/media/video/ir-kbd-i2c.c
@@ -44,7 +44,6 @@
 #include <linux/errno.h>
 #include <linux/slab.h>
 #include <linux/i2c.h>
-#include <linux/i2c-id.h>
 #include <linux/workqueue.h>
 
 #include <media/ir-core.h>
diff --git a/drivers/media/video/ov6650.c b/drivers/media/video/ov6650.c
index 31f19373bbae..cf93de988068 100644
--- a/drivers/media/video/ov6650.c
+++ b/drivers/media/video/ov6650.c
@@ -1174,7 +1174,6 @@ static int ov6650_probe(struct i2c_client *client,
 
         if (ret) {
                 icd->ops = NULL;
-                i2c_set_clientdata(client, NULL);
                 kfree(priv);
         }
 
@@ -1185,7 +1184,6 @@ static int ov6650_remove(struct i2c_client *client)
 {
         struct ov6650 *priv = to_ov6650(client);
 
-        i2c_set_clientdata(client, NULL);
         kfree(priv);
         return 0;
 }
diff --git a/drivers/media/video/pwc/pwc-if.c b/drivers/media/video/pwc/pwc-if.c
index e62beb4efdb4..f3dc89da4c4e 100644
--- a/drivers/media/video/pwc/pwc-if.c
+++ b/drivers/media/video/pwc/pwc-if.c
@@ -62,7 +62,6 @@
 #include <linux/module.h>
 #include <linux/poll.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #ifdef CONFIG_USB_PWC_INPUT_EVDEV
 #include <linux/usb/input.h>
 #endif
diff --git a/drivers/media/video/s2255drv.c b/drivers/media/video/s2255drv.c
index f5a46c458717..a845753665c1 100644
--- a/drivers/media/video/s2255drv.c
+++ b/drivers/media/video/s2255drv.c
@@ -49,7 +49,6 @@
 #include <linux/videodev2.h>
 #include <linux/version.h>
 #include <linux/mm.h>
-#include <linux/smp_lock.h>
 #include <media/videobuf-vmalloc.h>
 #include <media/v4l2-common.h>
 #include <media/v4l2-device.h>
diff --git a/drivers/media/video/saa7134/saa7134-empress.c b/drivers/media/video/saa7134/saa7134-empress.c
index 1467a30a434f..b890aafe7d64 100644
--- a/drivers/media/video/saa7134/saa7134-empress.c
+++ b/drivers/media/video/saa7134/saa7134-empress.c
@@ -21,7 +21,6 @@
 #include <linux/list.h>
 #include <linux/module.h>
 #include <linux/kernel.h>
-#include <linux/smp_lock.h>
 #include <linux/delay.h>
 
 #include "saa7134-reg.h"
diff --git a/drivers/media/video/saa7164/saa7164.h b/drivers/media/video/saa7164/saa7164.h
index 1d9c5cbbbc52..041ae8e20f68 100644
--- a/drivers/media/video/saa7164/saa7164.h
+++ b/drivers/media/video/saa7164/saa7164.h
@@ -58,7 +58,6 @@
 #include <media/tveeprom.h>
 #include <media/videobuf-dma-sg.h>
 #include <media/videobuf-dvb.h>
-#include <linux/smp_lock.h>
 #include <dvb_demux.h>
 #include <dvb_frontend.h>
 #include <dvb_net.h>
diff --git a/drivers/media/video/usbvision/usbvision-video.c b/drivers/media/video/usbvision/usbvision-video.c
index db6b828594f5..011c0c386995 100644
--- a/drivers/media/video/usbvision/usbvision-video.c
+++ b/drivers/media/video/usbvision/usbvision-video.c
@@ -50,7 +50,6 @@
 #include <linux/list.h>
 #include <linux/timer.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/mm.h>
 #include <linux/highmem.h>
 #include <linux/vmalloc.h>
diff --git a/drivers/media/video/v4l2-compat-ioctl32.c b/drivers/media/video/v4l2-compat-ioctl32.c
index 86294ed35c9b..e30e8dfb6205 100644
--- a/drivers/media/video/v4l2-compat-ioctl32.c
+++ b/drivers/media/video/v4l2-compat-ioctl32.c
@@ -18,7 +18,6 @@
 #include <linux/videodev.h>
 #include <linux/videodev2.h>
 #include <linux/module.h>
-#include <linux/smp_lock.h>
 #include <media/v4l2-ioctl.h>
 
 #ifdef CONFIG_COMPAT
diff --git a/drivers/message/fusion/mptfc.c b/drivers/message/fusion/mptfc.c
index e15220ff52fc..d784c36707c0 100644
--- a/drivers/message/fusion/mptfc.c
+++ b/drivers/message/fusion/mptfc.c
@@ -97,8 +97,7 @@ static u8	mptfcInternalCtx = MPT_MAX_PROTOCOL_DRIVERS;
 
 static int mptfc_target_alloc(struct scsi_target *starget);
 static int mptfc_slave_alloc(struct scsi_device *sdev);
-static int mptfc_qcmd(struct scsi_cmnd *SCpnt,
-                      void (*done)(struct scsi_cmnd *));
+static int mptfc_qcmd(struct Scsi_Host *shost, struct scsi_cmnd *SCpnt);
 static void mptfc_target_destroy(struct scsi_target *starget);
 static void mptfc_set_rport_loss_tmo(struct fc_rport *rport, uint32_t timeout);
 static void __devexit mptfc_remove(struct pci_dev *pdev);
@@ -650,7 +649,7 @@ mptfc_slave_alloc(struct scsi_device *sdev)
 }
 
 static int
-mptfc_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
+mptfc_qcmd_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
 {
         struct mptfc_rport_info *ri;
         struct fc_rport *rport = starget_to_rport(scsi_target(SCpnt->device));
@@ -681,6 +680,8 @@ mptfc_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
         return mptscsih_qcmd(SCpnt,done);
 }
 
+static DEF_SCSI_QCMD(mptfc_qcmd)
+
 /*
  *      mptfc_display_port_link_speed - displaying link speed
  *      @ioc: Pointer to MPT_ADAPTER structure
diff --git a/drivers/message/fusion/mptsas.c b/drivers/message/fusion/mptsas.c
index 83a5115f0251..d48c2c6058e1 100644
--- a/drivers/message/fusion/mptsas.c
+++ b/drivers/message/fusion/mptsas.c
@@ -1889,7 +1889,7 @@ mptsas_slave_alloc(struct scsi_device *sdev)
 }
 
 static int
-mptsas_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
+mptsas_qcmd_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
 {
         MPT_SCSI_HOST   *hd;
         MPT_ADAPTER     *ioc;
@@ -1913,6 +1913,8 @@ mptsas_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
         return mptscsih_qcmd(SCpnt,done);
 }
 
+static DEF_SCSI_QCMD(mptsas_qcmd)
+
 /**
  *      mptsas_mptsas_eh_timed_out - resets the scsi_cmnd timeout
  *              if the device under question is currently in the
diff --git a/drivers/message/fusion/mptspi.c b/drivers/message/fusion/mptspi.c
index 0e2803155ae2..6d9568d2ec59 100644
--- a/drivers/message/fusion/mptspi.c
+++ b/drivers/message/fusion/mptspi.c
@@ -780,7 +780,7 @@ static int mptspi_slave_configure(struct scsi_device *sdev)
 }
 
 static int
-mptspi_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
+mptspi_qcmd_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
 {
         struct _MPT_SCSI_HOST *hd = shost_priv(SCpnt->device->host);
         VirtDevice      *vdevice = SCpnt->device->hostdata;
@@ -805,6 +805,8 @@ mptspi_qcmd(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
         return mptscsih_qcmd(SCpnt,done);
 }
 
+static DEF_SCSI_QCMD(mptspi_qcmd)
+
 static void mptspi_slave_destroy(struct scsi_device *sdev)
 {
         struct scsi_target *starget = scsi_target(sdev);
diff --git a/drivers/message/i2o/i2o_scsi.c b/drivers/message/i2o/i2o_scsi.c
index ea6b2197da8a..97bdf82ec905 100644
--- a/drivers/message/i2o/i2o_scsi.c
+++ b/drivers/message/i2o/i2o_scsi.c
@@ -506,7 +506,7 @@ static struct i2o_driver i2o_scsi_driver = {
  *      Locks: takes the controller lock on error path only
  */
 
-static int i2o_scsi_queuecommand(struct scsi_cmnd *SCpnt,
+static int i2o_scsi_queuecommand_lck(struct scsi_cmnd *SCpnt,
                                  void (*done) (struct scsi_cmnd *))
 {
         struct i2o_controller *c;
@@ -688,7 +688,9 @@ static int i2o_scsi_queuecommand(struct scsi_cmnd *SCpnt,
 
       exit:
         return rc;
-};
+}
+
+static DEF_SCSI_QCMD(i2o_scsi_queuecommand)
 
 /**
  *      i2o_scsi_abort - abort a running command
diff --git a/drivers/misc/apds9802als.c b/drivers/misc/apds9802als.c
index 0ed09358027e..644d4cd071cc 100644
--- a/drivers/misc/apds9802als.c
+++ b/drivers/misc/apds9802als.c
@@ -251,7 +251,6 @@ static int apds9802als_probe(struct i2c_client *client,
 
         return res;
 als_error1:
-        i2c_set_clientdata(client, NULL);
         kfree(data);
         return res;
 }
diff --git a/drivers/misc/isl29020.c b/drivers/misc/isl29020.c
index ca47e6285075..307aada5fffe 100644
--- a/drivers/misc/isl29020.c
+++ b/drivers/misc/isl29020.c
@@ -183,9 +183,7 @@ static int  isl29020_probe(struct i2c_client *client,
 
 static int isl29020_remove(struct i2c_client *client)
 {
-        struct als_data *data = i2c_get_clientdata(client);
         sysfs_remove_group(&client->dev.kobj, &m_als_gr);
-        kfree(data);
         return 0;
 }
 
@@ -245,6 +243,6 @@ static void  __exit sensor_isl29020_exit(void)
 module_init(sensor_isl29020_init);
 module_exit(sensor_isl29020_exit);
 
-MODULE_AUTHOR("Kalhan Trisal <kalhan.trisal@intel.com");
+MODULE_AUTHOR("Kalhan Trisal <kalhan.trisal@intel.com>");
 MODULE_DESCRIPTION("Intersil isl29020 ALS Driver");
 MODULE_LICENSE("GPL v2");
diff --git a/drivers/misc/sgi-xp/xpc_partition.c b/drivers/misc/sgi-xp/xpc_partition.c
index d551f09ccb79..6956f7e7d439 100644
--- a/drivers/misc/sgi-xp/xpc_partition.c
+++ b/drivers/misc/sgi-xp/xpc_partition.c
@@ -439,18 +439,23 @@ xpc_discovery(void)
          * nodes that can comprise an access protection grouping. The access
          * protection is in regards to memory, IOI and IPI.
          */
-        max_regions = 64;
         region_size = xp_region_size;
 
-        switch (region_size) {
-        case 128:
-                max_regions *= 2;
-        case 64:
-                max_regions *= 2;
-        case 32:
-                max_regions *= 2;
-                region_size = 16;
-                DBUG_ON(!is_shub2());
+        if (is_uv())
+                max_regions = 256;
+        else {
+                max_regions = 64;
+
+                switch (region_size) {
+                case 128:
+                        max_regions *= 2;
+                case 64:
+                        max_regions *= 2;
+                case 32:
+                        max_regions *= 2;
+                        region_size = 16;
+                        DBUG_ON(!is_shub2());
+                }
         }
 
         for (region = 0; region < max_regions; region++) {
diff --git a/drivers/net/3c59x.c b/drivers/net/3c59x.c
index e1da258bbfb7..0a92436f0538 100644
--- a/drivers/net/3c59x.c
+++ b/drivers/net/3c59x.c
@@ -699,7 +699,8 @@ DEFINE_WINDOW_IO(32)
 #define DEVICE_PCI(dev) NULL
 #endif
 
-#define VORTEX_PCI(vp) (((vp)->gendev) ? DEVICE_PCI((vp)->gendev) : NULL)
+#define VORTEX_PCI(vp)                                                  \
+        ((struct pci_dev *) (((vp)->gendev) ? DEVICE_PCI((vp)->gendev) : NULL))
 
 #ifdef CONFIG_EISA
 #define DEVICE_EISA(dev) (((dev)->bus == &eisa_bus_type) ? to_eisa_device((dev)) : NULL)
@@ -707,7 +708,8 @@ DEFINE_WINDOW_IO(32)
 #define DEVICE_EISA(dev) NULL
 #endif
 
-#define VORTEX_EISA(vp) (((vp)->gendev) ? DEVICE_EISA((vp)->gendev) : NULL)
+#define VORTEX_EISA(vp)                                                 \
+        ((struct eisa_device *) (((vp)->gendev) ? DEVICE_EISA((vp)->gendev) : NULL))
 
 /* The action to take with a media selection timer tick.
    Note that we deviate from the 3Com order by checking 10base2 before AUI.
diff --git a/drivers/net/8139cp.c b/drivers/net/8139cp.c
index ac422cd332ea..dd16e83933a2 100644
--- a/drivers/net/8139cp.c
+++ b/drivers/net/8139cp.c
@@ -490,13 +490,11 @@ static inline unsigned int cp_rx_csum_ok (u32 status)
 {
         unsigned int protocol = (status >> 16) & 0x3;
 
-        if (likely((protocol == RxProtoTCP) && (!(status & TCPFail))))
+        if (((protocol == RxProtoTCP) && !(status & TCPFail)) ||
+            ((protocol == RxProtoUDP) && !(status & UDPFail)))
                 return 1;
-        else if ((protocol == RxProtoUDP) && (!(status & UDPFail)))
-                return 1;
-        else if ((protocol == RxProtoIP) && (!(status & IPFail)))
-                return 1;
-        return 0;
+        else
+                return 0;
 }
 
 static int cp_rx_poll(struct napi_struct *napi, int budget)
diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
index f6668cdaac85..43db398437b7 100644
--- a/drivers/net/Kconfig
+++ b/drivers/net/Kconfig
@@ -2945,6 +2945,18 @@ source "drivers/s390/net/Kconfig"
 
 source "drivers/net/caif/Kconfig"
 
+config TILE_NET
+        tristate "Tilera GBE/XGBE network driver support"
+        depends on TILE
+        default y
+        select CRC32
+        help
+          This is a standard Linux network device driver for the
+          on-chip Tilera Gigabit Ethernet and XAUI interfaces.
+
+          To compile this driver as a module, choose M here: the module
+          will be called tile_net.
+
 config XEN_NETDEV_FRONTEND
         tristate "Xen network device frontend driver"
         depends on XEN
diff --git a/drivers/net/Makefile b/drivers/net/Makefile
index 652fc6b98039..b90738d13994 100644
--- a/drivers/net/Makefile
+++ b/drivers/net/Makefile
@@ -301,3 +301,4 @@ obj-$(CONFIG_CAIF) += caif/
 
 obj-$(CONFIG_OCTEON_MGMT_ETHERNET) += octeon/
 obj-$(CONFIG_PCH_GBE) += pch_gbe/
+obj-$(CONFIG_TILE_NET) += tile/
diff --git a/drivers/net/atl1c/atl1c_hw.c b/drivers/net/atl1c/atl1c_hw.c
index 919080b2c3a5..1bf672009948 100644
--- a/drivers/net/atl1c/atl1c_hw.c
+++ b/drivers/net/atl1c/atl1c_hw.c
@@ -82,7 +82,7 @@ static int atl1c_get_permanent_address(struct atl1c_hw *hw)
         addr[0] = addr[1] = 0;
         AT_READ_REG(hw, REG_OTP_CTRL, &otp_ctrl_data);
         if (atl1c_check_eeprom_exist(hw)) {
-                if (hw->nic_type == athr_l1c || hw->nic_type == athr_l2c_b) {
+                if (hw->nic_type == athr_l1c || hw->nic_type == athr_l2c) {
                         /* Enable OTP CLK */
                         if (!(otp_ctrl_data & OTP_CTRL_CLK_EN)) {
                                 otp_ctrl_data |= OTP_CTRL_CLK_EN;
diff --git a/drivers/net/benet/be_main.c b/drivers/net/benet/be_main.c
index c36cd2ffbadc..93354eee2cfd 100644
--- a/drivers/net/benet/be_main.c
+++ b/drivers/net/benet/be_main.c
@@ -2458,6 +2458,12 @@ int be_load_fw(struct be_adapter *adapter, u8 *func)
         int status, i = 0, num_imgs = 0;
         const u8 *p;
 
+        if (!netif_running(adapter->netdev)) {
+                dev_err(&adapter->pdev->dev,
+                        "Firmware load not allowed (interface is down)\n");
+                return -EPERM;
+        }
+
         strcpy(fw_file, func);
 
         status = request_firmware(&fw, fw_file, &adapter->pdev->dev);
diff --git a/drivers/net/bnx2x/bnx2x_main.c b/drivers/net/bnx2x/bnx2x_main.c
index e9ad16f00b56..9709b8569666 100644
--- a/drivers/net/bnx2x/bnx2x_main.c
+++ b/drivers/net/bnx2x/bnx2x_main.c
@@ -9064,7 +9064,7 @@ static int __devinit bnx2x_init_one(struct pci_dev *pdev,
         default:
                 pr_err("Unknown board_type (%ld), aborting\n",
                            ent->driver_data);
-                return ENODEV;
+                return -ENODEV;
         }
 
         cid_count += CNIC_CONTEXT_USE;
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index bdb68a600382..71a169740d05 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -878,8 +878,10 @@ static void __bond_resend_igmp_join_requests(struct net_device *dev)
         rcu_read_lock();
         in_dev = __in_dev_get_rcu(dev);
         if (in_dev) {
+                read_lock(&in_dev->mc_list_lock);
                 for (im = in_dev->mc_list; im; im = im->next)
                         ip_mc_rejoin_group(im);
+                read_unlock(&in_dev->mc_list_lock);
         }
 
         rcu_read_unlock();
diff --git a/drivers/net/caif/caif_spi.c b/drivers/net/caif/caif_spi.c
index 8b4cea57a6c5..20da1996d354 100644
--- a/drivers/net/caif/caif_spi.c
+++ b/drivers/net/caif/caif_spi.c
@@ -635,8 +635,8 @@ int cfspi_spi_probe(struct platform_device *pdev)
 
         ndev = alloc_netdev(sizeof(struct cfspi),
                         "cfspi%d", cfspi_setup);
-        if (!dev)
-                return -ENODEV;
+        if (!ndev)
+                return -ENOMEM;
 
         cfspi = netdev_priv(ndev);
         netif_stop_queue(ndev);
diff --git a/drivers/net/e1000/e1000_main.c b/drivers/net/e1000/e1000_main.c
index 4686c3983fc3..4d62f7bfa036 100644
--- a/drivers/net/e1000/e1000_main.c
+++ b/drivers/net/e1000/e1000_main.c
@@ -31,7 +31,7 @@
 
 char e1000_driver_name[] = "e1000";
 static char e1000_driver_string[] = "Intel(R) PRO/1000 Network Driver";
-#define DRV_VERSION "7.3.21-k6-NAPI"
+#define DRV_VERSION "7.3.21-k8-NAPI"
 const char e1000_driver_version[] = DRV_VERSION;
 static const char e1000_copyright[] = "Copyright (c) 1999-2006 Intel Corporation.";
 
@@ -485,9 +485,6 @@ void e1000_down(struct e1000_adapter *adapter)
         struct net_device *netdev = adapter->netdev;
         u32 rctl, tctl;
 
-        /* signal that we're down so the interrupt handler does not
-         * reschedule our watchdog timer */
-        set_bit(__E1000_DOWN, &adapter->flags);
 
         /* disable receives in the hardware */
         rctl = er32(RCTL);
@@ -508,6 +505,13 @@ void e1000_down(struct e1000_adapter *adapter)
 
         e1000_irq_disable(adapter);
 
+        /*
+         * Setting DOWN must be after irq_disable to prevent
+         * a screaming interrupt.  Setting DOWN also prevents
+         * timers and tasks from rescheduling.
+         */
+        set_bit(__E1000_DOWN, &adapter->flags);
+
         del_timer_sync(&adapter->tx_fifo_stall_timer);
         del_timer_sync(&adapter->watchdog_timer);
         del_timer_sync(&adapter->phy_info_timer);
diff --git a/drivers/net/gianfar.c b/drivers/net/gianfar.c
index 49e4ce1246a7..d1bec6269173 100644
--- a/drivers/net/gianfar.c
+++ b/drivers/net/gianfar.c
@@ -577,11 +577,10 @@ static int gfar_parse_group(struct device_node *np,
                         irq_of_parse_and_map(np, 1);
                 priv->gfargrp[priv->num_grps].interruptError =
                         irq_of_parse_and_map(np,2);
-                if (priv->gfargrp[priv->num_grps].interruptTransmit < 0 ||
-                        priv->gfargrp[priv->num_grps].interruptReceive < 0 ||
-                        priv->gfargrp[priv->num_grps].interruptError < 0) {
+                if (priv->gfargrp[priv->num_grps].interruptTransmit == NO_IRQ ||
+                    priv->gfargrp[priv->num_grps].interruptReceive  == NO_IRQ ||
+                    priv->gfargrp[priv->num_grps].interruptError    == NO_IRQ)
                         return -EINVAL;
-                }
         }
 
         priv->gfargrp[priv->num_grps].grp_id = priv->num_grps;
diff --git a/drivers/net/ipg.c b/drivers/net/ipg.c
index dc0198092343..aa93655c3aa7 100644
--- a/drivers/net/ipg.c
+++ b/drivers/net/ipg.c
@@ -88,16 +88,14 @@ static const char *ipg_brand_name[] = {
         "IC PLUS IP1000 1000/100/10 based NIC",
         "Sundance Technology ST2021 based NIC",
         "Tamarack Microelectronics TC9020/9021 based NIC",
-        "Tamarack Microelectronics TC9020/9021 based NIC",
         "D-Link NIC IP1000A"
 };
 
 static DEFINE_PCI_DEVICE_TABLE(ipg_pci_tbl) = {
         { PCI_VDEVICE(SUNDANCE, 0x1023), 0 },
         { PCI_VDEVICE(SUNDANCE, 0x2021), 1 },
-        { PCI_VDEVICE(SUNDANCE, 0x1021), 2 },
-        { PCI_VDEVICE(DLINK,    0x9021), 3 },
-        { PCI_VDEVICE(DLINK,    0x4020), 4 },
+        { PCI_VDEVICE(DLINK,    0x9021), 2 },
+        { PCI_VDEVICE(DLINK,    0x4020), 3 },
         { 0, }
 };
 
diff --git a/drivers/net/irda/sh_sir.c b/drivers/net/irda/sh_sir.c
index 00b38bccd6d0..52a7c86af663 100644
--- a/drivers/net/irda/sh_sir.c
+++ b/drivers/net/irda/sh_sir.c
@@ -258,7 +258,7 @@ static int sh_sir_set_baudrate(struct sh_sir_self *self, u32 baudrate)
 
         /* Baud Rate Error Correction x 10000 */
         u32 rate_err_array[] = {
-                0000, 0625, 1250, 1875,
+                   0,  625, 1250, 1875,
                 2500, 3125, 3750, 4375,
                 5000, 5625, 6250, 6875,
                 7500, 8125, 8750, 9375,
diff --git a/drivers/net/phy/marvell.c b/drivers/net/phy/marvell.c
index f0bd1a1aba3a..e8b9c53c304b 100644
--- a/drivers/net/phy/marvell.c
+++ b/drivers/net/phy/marvell.c
@@ -30,11 +30,14 @@
 #include <linux/ethtool.h>
 #include <linux/phy.h>
 #include <linux/marvell_phy.h>
+#include <linux/of.h>
 
 #include <asm/io.h>
 #include <asm/irq.h>
 #include <asm/uaccess.h>
 
+#define MII_MARVELL_PHY_PAGE            22
+
 #define MII_M1011_IEVENT                0x13
 #define MII_M1011_IEVENT_CLEAR          0x0000
 
@@ -80,7 +83,6 @@
 #define MII_88E1121_PHY_LED_CTRL        16
 #define MII_88E1121_PHY_LED_PAGE        3
 #define MII_88E1121_PHY_LED_DEF         0x0030
-#define MII_88E1121_PHY_PAGE            22
 
 #define MII_M1011_PHY_STATUS            0x11
 #define MII_M1011_PHY_STATUS_1000       0x8000
@@ -186,13 +188,94 @@ static int marvell_config_aneg(struct phy_device *phydev)
         return 0;
 }
 
+#ifdef CONFIG_OF_MDIO
+/*
+ * Set and/or override some configuration registers based on the
+ * marvell,reg-init property stored in the of_node for the phydev.
+ *
+ * marvell,reg-init = <reg-page reg mask value>,...;
+ *
+ * There may be one or more sets of <reg-page reg mask value>:
+ *
+ * reg-page: which register bank to use.
+ * reg: the register.
+ * mask: if non-zero, ANDed with existing register value.
+ * value: ORed with the masked value and written to the regiser.
+ *
+ */
+static int marvell_of_reg_init(struct phy_device *phydev)
+{
+        const __be32 *paddr;
+        int len, i, saved_page, current_page, page_changed, ret;
+
+        if (!phydev->dev.of_node)
+                return 0;
+
+        paddr = of_get_property(phydev->dev.of_node, "marvell,reg-init", &len);
+        if (!paddr || len < (4 * sizeof(*paddr)))
+                return 0;
+
+        saved_page = phy_read(phydev, MII_MARVELL_PHY_PAGE);
+        if (saved_page < 0)
+                return saved_page;
+        page_changed = 0;
+        current_page = saved_page;
+
+        ret = 0;
+        len /= sizeof(*paddr);
+        for (i = 0; i < len - 3; i += 4) {
+                u16 reg_page = be32_to_cpup(paddr + i);
+                u16 reg = be32_to_cpup(paddr + i + 1);
+                u16 mask = be32_to_cpup(paddr + i + 2);
+                u16 val_bits = be32_to_cpup(paddr + i + 3);
+                int val;
+
+                if (reg_page != current_page) {
+                        current_page = reg_page;
+                        page_changed = 1;
+                        ret = phy_write(phydev, MII_MARVELL_PHY_PAGE, reg_page);
+                        if (ret < 0)
+                                goto err;
+                }
+
+                val = 0;
+                if (mask) {
+                        val = phy_read(phydev, reg);
+                        if (val < 0) {
+                                ret = val;
+                                goto err;
+                        }
+                        val &= mask;
+                }
+                val |= val_bits;
+
+                ret = phy_write(phydev, reg, val);
+                if (ret < 0)
+                        goto err;
+
+        }
+err:
+        if (page_changed) {
+                i = phy_write(phydev, MII_MARVELL_PHY_PAGE, saved_page);
+                if (ret == 0)
+                        ret = i;
+        }
+        return ret;
+}
+#else
+static int marvell_of_reg_init(struct phy_device *phydev)
+{
+        return 0;
+}
+#endif /* CONFIG_OF_MDIO */
+
 static int m88e1121_config_aneg(struct phy_device *phydev)
 {
         int err, oldpage, mscr;
 
-        oldpage = phy_read(phydev, MII_88E1121_PHY_PAGE);
+        oldpage = phy_read(phydev, MII_MARVELL_PHY_PAGE);
 
-        err = phy_write(phydev, MII_88E1121_PHY_PAGE,
+        err = phy_write(phydev, MII_MARVELL_PHY_PAGE,
                         MII_88E1121_PHY_MSCR_PAGE);
         if (err < 0)
                 return err;
@@ -218,7 +301,7 @@ static int m88e1121_config_aneg(struct phy_device *phydev)
                         return err;
         }
 
-        phy_write(phydev, MII_88E1121_PHY_PAGE, oldpage);
+        phy_write(phydev, MII_MARVELL_PHY_PAGE, oldpage);
 
         err = phy_write(phydev, MII_BMCR, BMCR_RESET);
         if (err < 0)
@@ -229,11 +312,11 @@ static int m88e1121_config_aneg(struct phy_device *phydev)
         if (err < 0)
                 return err;
 
-        oldpage = phy_read(phydev, MII_88E1121_PHY_PAGE);
+        oldpage = phy_read(phydev, MII_MARVELL_PHY_PAGE);
 
-        phy_write(phydev, MII_88E1121_PHY_PAGE, MII_88E1121_PHY_LED_PAGE);
+        phy_write(phydev, MII_MARVELL_PHY_PAGE, MII_88E1121_PHY_LED_PAGE);
         phy_write(phydev, MII_88E1121_PHY_LED_CTRL, MII_88E1121_PHY_LED_DEF);
-        phy_write(phydev, MII_88E1121_PHY_PAGE, oldpage);
+        phy_write(phydev, MII_MARVELL_PHY_PAGE, oldpage);
 
         err = genphy_config_aneg(phydev);
 
@@ -244,9 +327,9 @@ static int m88e1318_config_aneg(struct phy_device *phydev)
 {
         int err, oldpage, mscr;
 
-        oldpage = phy_read(phydev, MII_88E1121_PHY_PAGE);
+        oldpage = phy_read(phydev, MII_MARVELL_PHY_PAGE);
 
-        err = phy_write(phydev, MII_88E1121_PHY_PAGE,
+        err = phy_write(phydev, MII_MARVELL_PHY_PAGE,
                         MII_88E1121_PHY_MSCR_PAGE);
         if (err < 0)
                 return err;
@@ -258,7 +341,7 @@ static int m88e1318_config_aneg(struct phy_device *phydev)
         if (err < 0)
                 return err;
 
-        err = phy_write(phydev, MII_88E1121_PHY_PAGE, oldpage);
+        err = phy_write(phydev, MII_MARVELL_PHY_PAGE, oldpage);
         if (err < 0)
                 return err;
 
@@ -368,6 +451,9 @@ static int m88e1111_config_init(struct phy_device *phydev)
                         return err;
         }
 
+        err = marvell_of_reg_init(phydev);
+        if (err < 0)
+                return err;
 
         err = phy_write(phydev, MII_BMCR, BMCR_RESET);
         if (err < 0)
@@ -398,7 +484,7 @@ static int m88e1118_config_init(struct phy_device *phydev)
         int err;
 
         /* Change address */
-        err = phy_write(phydev, 0x16, 0x0002);
+        err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 0x0002);
         if (err < 0)
                 return err;
 
@@ -408,7 +494,7 @@ static int m88e1118_config_init(struct phy_device *phydev)
                 return err;
 
         /* Change address */
-        err = phy_write(phydev, 0x16, 0x0003);
+        err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 0x0003);
         if (err < 0)
                 return err;
 
@@ -420,8 +506,42 @@ static int m88e1118_config_init(struct phy_device *phydev)
         if (err < 0)
                 return err;
 
+        err = marvell_of_reg_init(phydev);
+        if (err < 0)
+                return err;
+
         /* Reset address */
-        err = phy_write(phydev, 0x16, 0x0);
+        err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 0x0);
+        if (err < 0)
+                return err;
+
+        err = phy_write(phydev, MII_BMCR, BMCR_RESET);
+        if (err < 0)
+                return err;
+
+        return 0;
+}
+
+static int m88e1149_config_init(struct phy_device *phydev)
+{
+        int err;
+
+        /* Change address */
+        err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 0x0002);
+        if (err < 0)
+                return err;
+
+        /* Enable 1000 Mbit */
+        err = phy_write(phydev, 0x15, 0x1048);
+        if (err < 0)
+                return err;
+
+        err = marvell_of_reg_init(phydev);
+        if (err < 0)
+                return err;
+
+        /* Reset address */
+        err = phy_write(phydev, MII_MARVELL_PHY_PAGE, 0x0);
         if (err < 0)
                 return err;
 
@@ -491,6 +611,10 @@ static int m88e1145_config_init(struct phy_device *phydev)
                 }
         }
 
+        err = marvell_of_reg_init(phydev);
+        if (err < 0)
+                return err;
+
         return 0;
 }
 
@@ -685,6 +809,19 @@ static struct phy_driver marvell_drivers[] = {
                 .driver = { .owner = THIS_MODULE },
         },
         {
+                .phy_id = MARVELL_PHY_ID_88E1149R,
+                .phy_id_mask = MARVELL_PHY_ID_MASK,
+                .name = "Marvell 88E1149R",
+                .features = PHY_GBIT_FEATURES,
+                .flags = PHY_HAS_INTERRUPT,
+                .config_init = &m88e1149_config_init,
+                .config_aneg = &m88e1118_config_aneg,
+                .read_status = &genphy_read_status,
+                .ack_interrupt = &marvell_ack_interrupt,
+                .config_intr = &marvell_config_intr,
+                .driver = { .owner = THIS_MODULE },
+        },
+        {
                 .phy_id = MARVELL_PHY_ID_88E1240,
                 .phy_id_mask = MARVELL_PHY_ID_MASK,
                 .name = "Marvell 88E1240",
@@ -735,6 +872,7 @@ static struct mdio_device_id __maybe_unused marvell_tbl[] = {
         { 0x01410e10, 0xfffffff0 },
         { 0x01410cb0, 0xfffffff0 },
         { 0x01410cd0, 0xfffffff0 },
+        { 0x01410e50, 0xfffffff0 },
         { 0x01410e30, 0xfffffff0 },
         { 0x01410e90, 0xfffffff0 },
         { }
diff --git a/drivers/net/qlge/qlge_main.c b/drivers/net/qlge/qlge_main.c
index c30e0fe55a31..528eaef5308f 100644
--- a/drivers/net/qlge/qlge_main.c
+++ b/drivers/net/qlge/qlge_main.c
@@ -62,15 +62,15 @@ static const u32 default_msg =
 /* NETIF_MSG_PKTDATA | */
     NETIF_MSG_HW | NETIF_MSG_WOL | 0;
 
-static int debug = 0x00007fff;  /* defaults above */
-module_param(debug, int, 0);
+static int debug = -1;  /* defaults above */
+module_param(debug, int, 0664);
 MODULE_PARM_DESC(debug, "Debug level (0=none,...,16=all)");
 
 #define MSIX_IRQ 0
 #define MSI_IRQ 1
 #define LEG_IRQ 2
 static int qlge_irq_type = MSIX_IRQ;
-module_param(qlge_irq_type, int, MSIX_IRQ);
+module_param(qlge_irq_type, int, 0664);
 MODULE_PARM_DESC(qlge_irq_type, "0 = MSI-X, 1 = MSI, 2 = Legacy.");
 
 static int qlge_mpi_coredump;
diff --git a/drivers/net/r8169.c b/drivers/net/r8169.c
index 4c4d16905efb..7d33ef4bcb4a 100644
--- a/drivers/net/r8169.c
+++ b/drivers/net/r8169.c
@@ -4440,8 +4440,7 @@ static inline void rtl8169_rx_csum(struct sk_buff *skb, u32 opts1)
         u32 status = opts1 & RxProtoMask;
 
         if (((status == RxProtoTCP) && !(opts1 & TCPFail)) ||
-            ((status == RxProtoUDP) && !(opts1 & UDPFail)) ||
-            ((status == RxProtoIP) && !(opts1 & IPFail)))
+            ((status == RxProtoUDP) && !(opts1 & UDPFail)))
                 skb->ip_summed = CHECKSUM_UNNECESSARY;
         else
                 skb_checksum_none_assert(skb);
diff --git a/drivers/net/tile/Makefile b/drivers/net/tile/Makefile
new file mode 100644
index 000000000000..f634f142cab4
--- /dev/null
+++ b/drivers/net/tile/Makefile
@@ -0,0 +1,10 @@
+#
+# Makefile for the TILE on-chip networking support.
+#
+
+obj-$(CONFIG_TILE_NET) += tile_net.o
+ifdef CONFIG_TILEGX
+tile_net-objs := tilegx.o mpipe.o iorpc_mpipe.o dma_queue.o
+else
+tile_net-objs := tilepro.o
+endif
diff --git a/drivers/net/tile/tilepro.c b/drivers/net/tile/tilepro.c
new file mode 100644
index 000000000000..0e6bac5ec65b
--- /dev/null
+++ b/drivers/net/tile/tilepro.c
@@ -0,0 +1,2406 @@
+/*
+ * Copyright 2010 Tilera Corporation. All Rights Reserved.
+ *
+ *   This program is free software; you can redistribute it and/or
+ *   modify it under the terms of the GNU General Public License
+ *   as published by the Free Software Foundation, version 2.
+ *
+ *   This program is distributed in the hope that it will be useful, but
+ *   WITHOUT ANY WARRANTY; without even the implied warranty of
+ *   MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
+ *   NON INFRINGEMENT.  See the GNU General Public License for
+ *   more details.
+ */
+
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/moduleparam.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>      /* printk() */
+#include <linux/slab.h>        /* kmalloc() */
+#include <linux/errno.h>       /* error codes */
+#include <linux/types.h>       /* size_t */
+#include <linux/interrupt.h>
+#include <linux/in.h>
+#include <linux/netdevice.h>   /* struct device, and other headers */
+#include <linux/etherdevice.h> /* eth_type_trans */
+#include <linux/skbuff.h>
+#include <linux/ioctl.h>
+#include <linux/cdev.h>
+#include <linux/hugetlb.h>
+#include <linux/in6.h>
+#include <linux/timer.h>
+#include <linux/io.h>
+#include <asm/checksum.h>
+#include <asm/homecache.h>
+
+#include <hv/drv_xgbe_intf.h>
+#include <hv/drv_xgbe_impl.h>
+#include <hv/hypervisor.h>
+#include <hv/netio_intf.h>
+
+/* For TSO */
+#include <linux/ip.h>
+#include <linux/tcp.h>
+
+
+/* There is no singlethread_cpu, so schedule work on the current cpu. */
+#define singlethread_cpu -1
+
+
+/*
+ * First, "tile_net_init_module()" initializes all four "devices" which
+ * can be used by linux.
+ *
+ * Then, "ifconfig DEVICE up" calls "tile_net_open()", which analyzes
+ * the network cpus, then uses "tile_net_open_aux()" to initialize
+ * LIPP/LEPP, and then uses "tile_net_open_inner()" to register all
+ * the tiles, provide buffers to LIPP, allow ingress to start, and
+ * turn on hypervisor interrupt handling (and NAPI) on all tiles.
+ *
+ * If registration fails due to the link being down, then "retry_work"
+ * is used to keep calling "tile_net_open_inner()" until it succeeds.
+ *
+ * If "ifconfig DEVICE down" is called, it uses "tile_net_stop()" to
+ * stop egress, drain the LIPP buffers, unregister all the tiles, stop
+ * LIPP/LEPP, and wipe the LEPP queue.
+ *
+ * We start out with the ingress interrupt enabled on each CPU.  When
+ * this interrupt fires, we disable it, and call "napi_schedule()".
+ * This will cause "tile_net_poll()" to be called, which will pull
+ * packets from the netio queue, filtering them out, or passing them
+ * to "netif_receive_skb()".  If our budget is exhausted, we will
+ * return, knowing we will be called again later.  Otherwise, we
+ * reenable the ingress interrupt, and call "napi_complete()".
+ *
+ *
+ * NOTE: The use of "native_driver" ensures that EPP exists, and that
+ * "epp_sendv" is legal, and that "LIPP" is being used.
+ *
+ * NOTE: Failing to free completions for an arbitrarily long time
+ * (which is defined to be illegal) does in fact cause bizarre
+ * problems.  The "egress_timer" helps prevent this from happening.
+ *
+ * NOTE: The egress code can be interrupted by the interrupt handler.
+ */
+
+
+/* HACK: Allow use of "jumbo" packets. */
+/* This should be 1500 if "jumbo" is not set in LIPP. */
+/* This should be at most 10226 (10240 - 14) if "jumbo" is set in LIPP. */
+/* ISSUE: This has not been thoroughly tested (except at 1500). */
+#define TILE_NET_MTU 1500
+
+/* HACK: Define to support GSO. */
+/* ISSUE: This may actually hurt performance of the TCP blaster. */
+/* #define TILE_NET_GSO */
+
+/* Define this to collapse "duplicate" acks. */
+/* #define IGNORE_DUP_ACKS */
+
+/* HACK: Define this to verify incoming packets. */
+/* #define TILE_NET_VERIFY_INGRESS */
+
+/* Use 3000 to enable the Linux Traffic Control (QoS) layer, else 0. */
+#define TILE_NET_TX_QUEUE_LEN 0
+
+/* Define to dump packets (prints out the whole packet on tx and rx). */
+/* #define TILE_NET_DUMP_PACKETS */
+
+/* Define to enable debug spew (all PDEBUG's are enabled). */
+/* #define TILE_NET_DEBUG */
+
+
+/* Define to activate paranoia checks. */
+/* #define TILE_NET_PARANOIA */
+
+/* Default transmit lockup timeout period, in jiffies. */
+#define TILE_NET_TIMEOUT (5 * HZ)
+
+/* Default retry interval for bringing up the NetIO interface, in jiffies. */
+#define TILE_NET_RETRY_INTERVAL (5 * HZ)
+
+/* Number of ports (xgbe0, xgbe1, gbe0, gbe1). */
+#define TILE_NET_DEVS 4
+
+
+
+/* Paranoia. */
+#if NET_IP_ALIGN != LIPP_PACKET_PADDING
+#error "NET_IP_ALIGN must match LIPP_PACKET_PADDING."
+#endif
+
+
+/* Debug print. */
+#ifdef TILE_NET_DEBUG
+#define PDEBUG(fmt, args...) net_printk(fmt, ## args)
+#else
+#define PDEBUG(fmt, args...)
+#endif
+
+
+MODULE_AUTHOR("Tilera");
+MODULE_LICENSE("GPL");
+
+
+#define IS_MULTICAST(mac_addr) \
+        (((u8 *)(mac_addr))[0] & 0x01)
+
+#define IS_BROADCAST(mac_addr) \
+        (((u16 *)(mac_addr))[0] == 0xffff)
+
+
+/*
+ * Queue of incoming packets for a specific cpu and device.
+ *
+ * Includes a pointer to the "system" data, and the actual "user" data.
+ */
+struct tile_netio_queue {
+        netio_queue_impl_t *__system_part;
+        netio_queue_user_impl_t __user_part;
+
+};
+
+
+/*
+ * Statistics counters for a specific cpu and device.
+ */
+struct tile_net_stats_t {
+        u32 rx_packets;
+        u32 rx_bytes;
+        u32 tx_packets;
+        u32 tx_bytes;
+};
+
+
+/*
+ * Info for a specific cpu and device.
+ *
+ * ISSUE: There is a "dev" pointer in "napi" as well.
+ */
+struct tile_net_cpu {
+        /* The NAPI struct. */
+        struct napi_struct napi;
+        /* Packet queue. */
+        struct tile_netio_queue queue;
+        /* Statistics. */
+        struct tile_net_stats_t stats;
+        /* ISSUE: Is this needed? */
+        bool napi_enabled;
+        /* True if this tile has succcessfully registered with the IPP. */
+        bool registered;
+        /* True if the link was down last time we tried to register. */
+        bool link_down;
+        /* True if "egress_timer" is scheduled. */
+        bool egress_timer_scheduled;
+        /* Number of small sk_buffs which must still be provided. */
+        unsigned int num_needed_small_buffers;
+        /* Number of large sk_buffs which must still be provided. */
+        unsigned int num_needed_large_buffers;
+        /* A timer for handling egress completions. */
+        struct timer_list egress_timer;
+};
+
+
+/*
+ * Info for a specific device.
+ */
+struct tile_net_priv {
+        /* Our network device. */
+        struct net_device *dev;
+        /* The actual egress queue. */
+        lepp_queue_t *epp_queue;
+        /* Protects "epp_queue->cmd_tail" and "epp_queue->comp_tail" */
+        spinlock_t cmd_lock;
+        /* Protects "epp_queue->comp_head". */
+        spinlock_t comp_lock;
+        /* The hypervisor handle for this interface. */
+        int hv_devhdl;
+        /* The intr bit mask that IDs this device. */
+        u32 intr_id;
+        /* True iff "tile_net_open_aux()" has succeeded. */
+        int partly_opened;
+        /* True iff "tile_net_open_inner()" has succeeded. */
+        int fully_opened;
+        /* Effective network cpus. */
+        struct cpumask network_cpus_map;
+        /* Number of network cpus. */
+        int network_cpus_count;
+        /* Credits per network cpu. */
+        int network_cpus_credits;
+        /* Network stats. */
+        struct net_device_stats stats;
+        /* For NetIO bringup retries. */
+        struct delayed_work retry_work;
+        /* Quick access to per cpu data. */
+        struct tile_net_cpu *cpu[NR_CPUS];
+};
+
+
+/*
+ * The actual devices (xgbe0, xgbe1, gbe0, gbe1).
+ */
+static struct net_device *tile_net_devs[TILE_NET_DEVS];
+
+/*
+ * The "tile_net_cpu" structures for each device.
+ */
+static DEFINE_PER_CPU(struct tile_net_cpu, hv_xgbe0);
+static DEFINE_PER_CPU(struct tile_net_cpu, hv_xgbe1);
+static DEFINE_PER_CPU(struct tile_net_cpu, hv_gbe0);
+static DEFINE_PER_CPU(struct tile_net_cpu, hv_gbe1);
+
+
+/*
+ * True if "network_cpus" was specified.
+ */
+static bool network_cpus_used;
+
+/*
+ * The actual cpus in "network_cpus".
+ */
+static struct cpumask network_cpus_map;
+
+
+
+#ifdef TILE_NET_DEBUG
+/*
+ * printk with extra stuff.
+ *
+ * We print the CPU we're running in brackets.
+ */
+static void net_printk(char *fmt, ...)
+{
+        int i;
+        int len;
+        va_list args;
+        static char buf[256];
+
+        len = sprintf(buf, "tile_net[%2.2d]: ", smp_processor_id());
+        va_start(args, fmt);
+        i = vscnprintf(buf + len, sizeof(buf) - len - 1, fmt, args);
+        va_end(args);
+        buf[255] = '\0';
+        pr_notice(buf);
+}
+#endif
+
+
+#ifdef TILE_NET_DUMP_PACKETS
+/*
+ * Dump a packet.
+ */
+static void dump_packet(unsigned char *data, unsigned long length, char *s)
+{
+        unsigned long i;
+        static unsigned int count;
+
+        pr_info("dump_packet(data %p, length 0x%lx s %s count 0x%x)\n",
+               data, length, s, count++);
+
+        pr_info("\n");
+
+        for (i = 0; i < length; i++) {
+                if ((i & 0xf) == 0)
+                        sprintf(buf, "%8.8lx:", i);
+                sprintf(buf + strlen(buf), " %2.2x", data[i]);
+                if ((i & 0xf) == 0xf || i == length - 1)
+                        pr_info("%s\n", buf);
+        }
+}
+#endif
+
+
+/*
+ * Provide support for the __netio_fastio1() swint
+ * (see <hv/drv_xgbe_intf.h> for how it is used).
+ *
+ * The fastio swint2 call may clobber all the caller-saved registers.
+ * It rarely clobbers memory, but we allow for the possibility in
+ * the signature just to be on the safe side.
+ *
+ * Also, gcc doesn't seem to allow an input operand to be
+ * clobbered, so we fake it with dummy outputs.
+ *
+ * This function can't be static because of the way it is declared
+ * in the netio header.
+ */
+inline int __netio_fastio1(u32 fastio_index, u32 arg0)
+{
+        long result, clobber_r1, clobber_r10;
+        asm volatile("swint2"
+                     : "=R00" (result),
+                       "=R01" (clobber_r1), "=R10" (clobber_r10)
+                     : "R10" (fastio_index), "R01" (arg0)
+                     : "memory", "r2", "r3", "r4",
+                       "r5", "r6", "r7", "r8", "r9",
+                       "r11", "r12", "r13", "r14",
+                       "r15", "r16", "r17", "r18", "r19",
+                       "r20", "r21", "r22", "r23", "r24",
+                       "r25", "r26", "r27", "r28", "r29");
+        return result;
+}
+
+
+/*
+ * Provide a linux buffer to LIPP.
+ */
+static void tile_net_provide_linux_buffer(struct tile_net_cpu *info,
+                                          void *va, bool small)
+{
+        struct tile_netio_queue *queue = &info->queue;
+
+        /* Convert "va" and "small" to "linux_buffer_t". */
+        unsigned int buffer = ((unsigned int)(__pa(va) >> 7) << 1) + small;
+
+        __netio_fastio_free_buffer(queue->__user_part.__fastio_index, buffer);
+}
+
+
+/*
+ * Provide a linux buffer for LIPP.
+ */
+static bool tile_net_provide_needed_buffer(struct tile_net_cpu *info,
+                                           bool small)
+{
+        /* ISSUE: What should we use here? */
+        unsigned int large_size = NET_IP_ALIGN + TILE_NET_MTU + 100;
+
+        /* Round up to ensure to avoid "false sharing" with last cache line. */
+        unsigned int buffer_size =
+                 (((small ? LIPP_SMALL_PACKET_SIZE : large_size) +
+                   CHIP_L2_LINE_SIZE() - 1) & -CHIP_L2_LINE_SIZE());
+
+        /*
+         * ISSUE: Since CPAs are 38 bits, and we can only encode the
+         * high 31 bits in a "linux_buffer_t", the low 7 bits must be
+         * zero, and thus, we must align the actual "va" mod 128.
+         */
+        const unsigned long align = 128;
+
+        struct sk_buff *skb;
+        void *va;
+
+        struct sk_buff **skb_ptr;
+
+        /* Note that "dev_alloc_skb()" adds NET_SKB_PAD more bytes, */
+        /* and also "reserves" that many bytes. */
+        /* ISSUE: Can we "share" the NET_SKB_PAD bytes with "skb_ptr"? */
+        int len = sizeof(*skb_ptr) + align + buffer_size;
+
+        while (1) {
+
+                /* Allocate (or fail). */
+                skb = dev_alloc_skb(len);
+                if (skb == NULL)
+                        return false;
+
+                /* Make room for a back-pointer to 'skb'. */
+                skb_reserve(skb, sizeof(*skb_ptr));
+
+                /* Make sure we are aligned. */
+                skb_reserve(skb, -(long)skb->data & (align - 1));
+
+                /* This address is given to IPP. */
+                va = skb->data;
+
+                if (small)
+                        break;
+
+                /* ISSUE: This has never been observed! */
+                /* Large buffers must not span a huge page. */
+                if (((((long)va & ~HPAGE_MASK) + 1535) & HPAGE_MASK) == 0)
+                        break;
+                pr_err("Leaking unaligned linux buffer at %p.\n", va);
+        }
+
+        /* Skip two bytes to satisfy LIPP assumptions. */
+        /* Note that this aligns IP on a 16 byte boundary. */
+        /* ISSUE: Do this when the packet arrives? */
+        skb_reserve(skb, NET_IP_ALIGN);
+
+        /* Save a back-pointer to 'skb'. */
+        skb_ptr = va - sizeof(*skb_ptr);
+        *skb_ptr = skb;
+
+        /* Invalidate the packet buffer. */
+        if (!hash_default)
+                __inv_buffer(skb->data, buffer_size);
+
+        /* Make sure "skb_ptr" has been flushed. */
+        __insn_mf();
+
+#ifdef TILE_NET_PARANOIA
+#if CHIP_HAS_CBOX_HOME_MAP()
+        if (hash_default) {
+                HV_PTE pte = *virt_to_pte(current->mm, (unsigned long)va);
+                if (hv_pte_get_mode(pte) != HV_PTE_MODE_CACHE_HASH_L3)
+                        panic("Non-coherent ingress buffer!");
+        }
+#endif
+#endif
+
+        /* Provide the new buffer. */
+        tile_net_provide_linux_buffer(info, va, small);
+
+        return true;
+}
+
+
+/*
+ * Provide linux buffers for LIPP.
+ */
+static void tile_net_provide_needed_buffers(struct tile_net_cpu *info)
+{
+        while (info->num_needed_small_buffers != 0) {
+                if (!tile_net_provide_needed_buffer(info, true))
+                        goto oops;
+                info->num_needed_small_buffers--;
+        }
+
+        while (info->num_needed_large_buffers != 0) {
+                if (!tile_net_provide_needed_buffer(info, false))
+                        goto oops;
+                info->num_needed_large_buffers--;
+        }
+
+        return;
+
+oops:
+
+        /* Add a description to the page allocation failure dump. */
+        pr_notice("Could not provide a linux buffer to LIPP.\n");
+}
+
+
+/*
+ * Grab some LEPP completions, and store them in "comps", of size
+ * "comps_size", and return the number of completions which were
+ * stored, so the caller can free them.
+ *
+ * If "pending" is not NULL, it will be set to true if there might
+ * still be some pending completions caused by this tile, else false.
+ */
+static unsigned int tile_net_lepp_grab_comps(struct net_device *dev,
+                                             struct sk_buff *comps[],
+                                             unsigned int comps_size,
+                                             bool *pending)
+{
+        struct tile_net_priv *priv = netdev_priv(dev);
+
+        lepp_queue_t *eq = priv->epp_queue;
+
+        unsigned int n = 0;
+
+        unsigned int comp_head;
+        unsigned int comp_busy;
+        unsigned int comp_tail;
+
+        spin_lock(&priv->comp_lock);
+
+        comp_head = eq->comp_head;
+        comp_busy = eq->comp_busy;
+        comp_tail = eq->comp_tail;
+
+        while (comp_head != comp_busy && n < comps_size) {
+                comps[n++] = eq->comps[comp_head];
+                LEPP_QINC(comp_head);
+        }
+
+        if (pending != NULL)
+                *pending = (comp_head != comp_tail);
+
+        eq->comp_head = comp_head;
+
+        spin_unlock(&priv->comp_lock);
+
+        return n;
+}
+
+
+/*
+ * Make sure the egress timer is scheduled.
+ *
+ * Note that we use "schedule if not scheduled" logic instead of the more
+ * obvious "reschedule" logic, because "reschedule" is fairly expensive.
+ */
+static void tile_net_schedule_egress_timer(struct tile_net_cpu *info)
+{
+        if (!info->egress_timer_scheduled) {
+                mod_timer_pinned(&info->egress_timer, jiffies + 1);
+                info->egress_timer_scheduled = true;
+        }
+}
+
+
+/*
+ * The "function" for "info->egress_timer".
+ *
+ * This timer will reschedule itself as long as there are any pending
+ * completions expected (on behalf of any tile).
+ *
+ * ISSUE: Realistically, will the timer ever stop scheduling itself?
+ *
+ * ISSUE: This timer is almost never actually needed, so just use a global
+ * timer that can run on any tile.
+ *
+ * ISSUE: Maybe instead track number of expected completions, and free
+ * only that many, resetting to zero if "pending" is ever false.
+ */
+static void tile_net_handle_egress_timer(unsigned long arg)
+{
+        struct tile_net_cpu *info = (struct tile_net_cpu *)arg;
+        struct net_device *dev = info->napi.dev;
+
+        struct sk_buff *olds[32];
+        unsigned int wanted = 32;
+        unsigned int i, nolds = 0;
+        bool pending;
+
+        /* The timer is no longer scheduled. */
+        info->egress_timer_scheduled = false;
+
+        nolds = tile_net_lepp_grab_comps(dev, olds, wanted, &pending);
+
+        for (i = 0; i < nolds; i++)
+                kfree_skb(olds[i]);
+
+        /* Reschedule timer if needed. */
+        if (pending)
+                tile_net_schedule_egress_timer(info);
+}
+
+
+#ifdef IGNORE_DUP_ACKS
+
+/*
+ * Help detect "duplicate" ACKs.  These are sequential packets (for a
+ * given flow) which are exactly 66 bytes long, sharing everything but
+ * ID=2@0x12, Hsum=2@0x18, Ack=4@0x2a, WinSize=2@0x30, Csum=2@0x32,
+ * Tstamps=10@0x38.  The ID's are +1, the Hsum's are -1, the Ack's are
+ * +N, and the Tstamps are usually identical.
+ *
+ * NOTE: Apparently truly duplicate acks (with identical "ack" values),
+ * should not be collapsed, as they are used for some kind of flow control.
+ */
+static bool is_dup_ack(char *s1, char *s2, unsigned int len)
+{
+        int i;
+
+        unsigned long long ignorable = 0;
+
+        /* Identification. */
+        ignorable |= (1ULL << 0x12);
+        ignorable |= (1ULL << 0x13);
+
+        /* Header checksum. */
+        ignorable |= (1ULL << 0x18);
+        ignorable |= (1ULL << 0x19);
+
+        /* ACK. */
+        ignorable |= (1ULL << 0x2a);
+        ignorable |= (1ULL << 0x2b);
+        ignorable |= (1ULL << 0x2c);
+        ignorable |= (1ULL << 0x2d);
+
+        /* WinSize. */
+        ignorable |= (1ULL << 0x30);
+        ignorable |= (1ULL << 0x31);
+
+        /* Checksum. */
+        ignorable |= (1ULL << 0x32);
+        ignorable |= (1ULL << 0x33);
+
+        for (i = 0; i < len; i++, ignorable >>= 1) {
+
+                if ((ignorable & 1) || (s1[i] == s2[i]))
+                        continue;
+
+#ifdef TILE_NET_DEBUG
+                /* HACK: Mention non-timestamp diffs. */
+                if (i < 0x38 && i != 0x2f &&
+                    net_ratelimit())
+                        pr_info("Diff at 0x%x\n", i);
+#endif
+
+                return false;
+        }
+
+#ifdef TILE_NET_NO_SUPPRESS_DUP_ACKS
+        /* HACK: Do not suppress truly duplicate ACKs. */
+        /* ISSUE: Is this actually necessary or helpful? */
+        if (s1[0x2a] == s2[0x2a] &&
+            s1[0x2b] == s2[0x2b] &&
+            s1[0x2c] == s2[0x2c] &&
+            s1[0x2d] == s2[0x2d]) {
+                return false;
+        }
+#endif
+
+        return true;
+}
+
+#endif
+
+
+
+/*
+ * Like "tile_net_handle_packets()", but just discard packets.
+ */
+static void tile_net_discard_packets(struct net_device *dev)
+{
+        struct tile_net_priv *priv = netdev_priv(dev);
+        int my_cpu = smp_processor_id();
+        struct tile_net_cpu *info = priv->cpu[my_cpu];
+        struct tile_netio_queue *queue = &info->queue;
+        netio_queue_impl_t *qsp = queue->__system_part;
+        netio_queue_user_impl_t *qup = &queue->__user_part;
+
+        while (qup->__packet_receive_read !=
+               qsp->__packet_receive_queue.__packet_write) {
+
+                int index = qup->__packet_receive_read;
+
+                int index2_aux = index + sizeof(netio_pkt_t);
+                int index2 =
+                        ((index2_aux ==
+                          qsp->__packet_receive_queue.__last_packet_plus_one) ?
+                         0 : index2_aux);
+
+                netio_pkt_t *pkt = (netio_pkt_t *)
+                        ((unsigned long) &qsp[1] + index);
+
+                /* Extract the "linux_buffer_t". */
+                unsigned int buffer = pkt->__packet.word;
+
+                /* Convert "linux_buffer_t" to "va". */
+                void *va = __va((phys_addr_t)(buffer >> 1) << 7);
+
+                /* Acquire the associated "skb". */
+                struct sk_buff **skb_ptr = va - sizeof(*skb_ptr);
+                struct sk_buff *skb = *skb_ptr;
+
+                kfree_skb(skb);
+
+                /* Consume this packet. */
+                qup->__packet_receive_read = index2;
+        }
+}
+
+
+/*
+ * Handle the next packet.  Return true if "processed", false if "filtered".
+ */
+static bool tile_net_poll_aux(struct tile_net_cpu *info, int index)
+{
+        struct net_device *dev = info->napi.dev;
+
+        struct tile_netio_queue *queue = &info->queue;
+        netio_queue_impl_t *qsp = queue->__system_part;
+        netio_queue_user_impl_t *qup = &queue->__user_part;
+        struct tile_net_stats_t *stats = &info->stats;
+
+        int filter;
+
+        int index2_aux = index + sizeof(netio_pkt_t);
+        int index2 =
+                ((index2_aux ==
+                  qsp->__packet_receive_queue.__last_packet_plus_one) ?
+                 0 : index2_aux);
+
+        netio_pkt_t *pkt = (netio_pkt_t *)((unsigned long) &qsp[1] + index);
+
+        netio_pkt_metadata_t *metadata = NETIO_PKT_METADATA(pkt);
+
+        /* Extract the packet size. */
+        unsigned long len =
+                (NETIO_PKT_CUSTOM_LENGTH(pkt) +
+                 NET_IP_ALIGN - NETIO_PACKET_PADDING);
+
+        /* Extract the "linux_buffer_t". */
+        unsigned int buffer = pkt->__packet.word;
+
+        /* Extract "small" (vs "large"). */
+        bool small = ((buffer & 1) != 0);
+
+        /* Convert "linux_buffer_t" to "va". */
+        void *va = __va((phys_addr_t)(buffer >> 1) << 7);
+
+        /* Extract the packet data pointer. */
+        /* Compare to "NETIO_PKT_CUSTOM_DATA(pkt)". */
+        unsigned char *buf = va + NET_IP_ALIGN;
+
+#ifdef IGNORE_DUP_ACKS
+
+        static int other;
+        static int final;
+        static int keep;
+        static int skip;
+
+#endif
+
+        /* Invalidate the packet buffer. */
+        if (!hash_default)
+                __inv_buffer(buf, len);
+
+        /* ISSUE: Is this needed? */
+        dev->last_rx = jiffies;
+
+#ifdef TILE_NET_DUMP_PACKETS
+        dump_packet(buf, len, "rx");
+#endif /* TILE_NET_DUMP_PACKETS */
+
+#ifdef TILE_NET_VERIFY_INGRESS
+        if (!NETIO_PKT_L4_CSUM_CORRECT_M(metadata, pkt) &&
+            NETIO_PKT_L4_CSUM_CALCULATED_M(metadata, pkt)) {
+                /*
+                 * FIXME: This complains about UDP packets
+                 * with a "zero" checksum (bug 6624).
+                 */
+#ifdef TILE_NET_PANIC_ON_BAD
+                dump_packet(buf, len, "rx");
+                panic("Bad L4 checksum.");
+#else
+                pr_warning("Bad L4 checksum on %d byte packet.\n", len);
+#endif
+        }
+        if (!NETIO_PKT_L3_CSUM_CORRECT_M(metadata, pkt) &&
+            NETIO_PKT_L3_CSUM_CALCULATED_M(metadata, pkt)) {
+                dump_packet(buf, len, "rx");
+                panic("Bad L3 checksum.");
+        }
+        switch (NETIO_PKT_STATUS_M(metadata, pkt)) {
+        case NETIO_PKT_STATUS_OVERSIZE:
+                if (len >= 64) {
+                        dump_packet(buf, len, "rx");
+                        panic("Unexpected OVERSIZE.");
+                }
+                break;
+        case NETIO_PKT_STATUS_BAD:
+#ifdef TILE_NET_PANIC_ON_BAD
+                dump_packet(buf, len, "rx");
+                panic("Unexpected BAD packet.");
+#else
+                pr_warning("Unexpected BAD %d byte packet.\n", len);
+#endif
+        }
+#endif
+
+        filter = 0;
+
+        if (!(dev->flags & IFF_UP)) {
+                /* Filter packets received before we're up. */
+                filter = 1;
+        } else if (!(dev->flags & IFF_PROMISC)) {
+                /*
+                 * FIXME: Implement HW multicast filter.
+                 */
+                if (!IS_MULTICAST(buf) && !IS_BROADCAST(buf)) {
+                        /* Filter packets not for our address. */
+                        const u8 *mine = dev->dev_addr;
+                        filter = compare_ether_addr(mine, buf);
+                }
+        }
+
+#ifdef IGNORE_DUP_ACKS
+
+        if (len != 66) {
+                /* FIXME: Must check "is_tcp_ack(buf, len)" somehow. */
+
+                other++;
+
+        } else if (index2 ==
+                   qsp->__packet_receive_queue.__packet_write) {
+
+                final++;
+
+        } else {
+
+                netio_pkt_t *pkt2 = (netio_pkt_t *)
+                        ((unsigned long) &qsp[1] + index2);
+
+                netio_pkt_metadata_t *metadata2 =
+                        NETIO_PKT_METADATA(pkt2);
+
+                /* Extract the packet size. */
+                unsigned long len2 =
+                        (NETIO_PKT_CUSTOM_LENGTH(pkt2) +
+                         NET_IP_ALIGN - NETIO_PACKET_PADDING);
+
+                if (len2 == 66 &&
+                    NETIO_PKT_FLOW_HASH_M(metadata, pkt) ==
+                    NETIO_PKT_FLOW_HASH_M(metadata2, pkt2)) {
+
+                        /* Extract the "linux_buffer_t". */
+                        unsigned int buffer2 = pkt2->__packet.word;
+
+                        /* Convert "linux_buffer_t" to "va". */
+                        void *va2 =
+                                __va((phys_addr_t)(buffer2 >> 1) << 7);
+
+                        /* Extract the packet data pointer. */
+                        /* Compare to "NETIO_PKT_CUSTOM_DATA(pkt)". */
+                        unsigned char *buf2 = va2 + NET_IP_ALIGN;
+
+                        /* Invalidate the packet buffer. */
+                        if (!hash_default)
+                                __inv_buffer(buf2, len2);
+
+                        if (is_dup_ack(buf, buf2, len)) {
+                                skip++;
+                                filter = 1;
+                        } else {
+                                keep++;
+                        }
+                }
+        }
+
+        if (net_ratelimit())
+                pr_info("Other %d Final %d Keep %d Skip %d.\n",
+                        other, final, keep, skip);
+
+#endif
+
+        if (filter) {
+
+                /* ISSUE: Update "drop" statistics? */
+
+                tile_net_provide_linux_buffer(info, va, small);
+
+        } else {
+
+                /* Acquire the associated "skb". */
+                struct sk_buff **skb_ptr = va - sizeof(*skb_ptr);
+                struct sk_buff *skb = *skb_ptr;
+
+                /* Paranoia. */
+                if (skb->data != buf)
+                        panic("Corrupt linux buffer from LIPP! "
+                              "VA=%p, skb=%p, skb->data=%p\n",
+                              va, skb, skb->data);
+
+                /* Encode the actual packet length. */
+                skb_put(skb, len);
+
+                /* NOTE: This call also sets "skb->dev = dev". */
+                skb->protocol = eth_type_trans(skb, dev);
+
+                /* ISSUE: Discard corrupt packets? */
+                /* ISSUE: Discard packets with bad checksums? */
+
+                /* Avoid recomputing TCP/UDP checksums. */
+                if (NETIO_PKT_L4_CSUM_CORRECT_M(metadata, pkt))
+                        skb->ip_summed = CHECKSUM_UNNECESSARY;
+
+                netif_receive_skb(skb);
+
+                stats->rx_packets++;
+                stats->rx_bytes += len;
+
+                if (small)
+                        info->num_needed_small_buffers++;
+                else
+                        info->num_needed_large_buffers++;
+        }
+
+        /* Return four credits after every fourth packet. */
+        if (--qup->__receive_credit_remaining == 0) {
+                u32 interval = qup->__receive_credit_interval;
+                qup->__receive_credit_remaining = interval;
+                __netio_fastio_return_credits(qup->__fastio_index, interval);
+        }
+
+        /* Consume this packet. */
+        qup->__packet_receive_read = index2;
+
+        return !filter;
+}
+
+
+/*
+ * Handle some packets for the given device on the current CPU.
+ *
+ * ISSUE: The "rotting packet" race condition occurs if a packet
+ * arrives after the queue appears to be empty, and before the
+ * hypervisor interrupt is re-enabled.
+ */
+static int tile_net_poll(struct napi_struct *napi, int budget)
+{
+        struct net_device *dev = napi->dev;
+        struct tile_net_priv *priv = netdev_priv(dev);
+        int my_cpu = smp_processor_id();
+        struct tile_net_cpu *info = priv->cpu[my_cpu];
+        struct tile_netio_queue *queue = &info->queue;
+        netio_queue_impl_t *qsp = queue->__system_part;
+        netio_queue_user_impl_t *qup = &queue->__user_part;
+
+        unsigned int work = 0;
+
+        while (1) {
+                int index = qup->__packet_receive_read;
+                if (index == qsp->__packet_receive_queue.__packet_write)
+                        break;
+
+                if (tile_net_poll_aux(info, index)) {
+                        if (++work >= budget)
+                                goto done;
+                }
+        }
+
+        napi_complete(&info->napi);
+
+        /* Re-enable hypervisor interrupts. */
+        enable_percpu_irq(priv->intr_id);
+
+        /* HACK: Avoid the "rotting packet" problem. */
+        if (qup->__packet_receive_read !=
+            qsp->__packet_receive_queue.__packet_write)
+                napi_schedule(&info->napi);
+
+        /* ISSUE: Handle completions? */
+
+done:
+
+        tile_net_provide_needed_buffers(info);
+
+        return work;
+}
+
+
+/*
+ * Handle an ingress interrupt for the given device on the current cpu.
+ */
+static irqreturn_t tile_net_handle_ingress_interrupt(int irq, void *dev_ptr)
+{
+        struct net_device *dev = (struct net_device *)dev_ptr;
+        struct tile_net_priv *priv = netdev_priv(dev);
+        int my_cpu = smp_processor_id();
+        struct tile_net_cpu *info = priv->cpu[my_cpu];
+
+        /* Disable hypervisor interrupt. */
+        disable_percpu_irq(priv->intr_id);
+
+        napi_schedule(&info->napi);
+
+        return IRQ_HANDLED;
+}
+
+
+/*
+ * One time initialization per interface.
+ */
+static int tile_net_open_aux(struct net_device *dev)
+{
+        struct tile_net_priv *priv = netdev_priv(dev);
+
+        int ret;
+        int dummy;
+        unsigned int epp_lotar;
+
+        /*
+         * Find out where EPP memory should be homed.
+         */
+        ret = hv_dev_pread(priv->hv_devhdl, 0,
+                           (HV_VirtAddr)&epp_lotar, sizeof(epp_lotar),
+                           NETIO_EPP_SHM_OFF);
+        if (ret < 0) {
+                pr_err("could not read epp_shm_queue lotar.\n");
+                return -EIO;
+        }
+
+        /*
+         * Home the page on the EPP.
+         */
+        {
+                int epp_home = hv_lotar_to_cpu(epp_lotar);
+                struct page *page = virt_to_page(priv->epp_queue);
+                homecache_change_page_home(page, 0, epp_home);
+        }
+
+        /*
+         * Register the EPP shared memory queue.
+         */
+        {
+                netio_ipp_address_t ea = {
+                        .va = 0,
+                        .pa = __pa(priv->epp_queue),
+                        .pte = hv_pte(0),
+                        .size = PAGE_SIZE,
+                };
+                ea.pte = hv_pte_set_lotar(ea.pte, epp_lotar);
+                ea.pte = hv_pte_set_mode(ea.pte, HV_PTE_MODE_CACHE_TILE_L3);
+                ret = hv_dev_pwrite(priv->hv_devhdl, 0,
+                                    (HV_VirtAddr)&ea,
+                                    sizeof(ea),
+                                    NETIO_EPP_SHM_OFF);
+                if (ret < 0)
+                        return -EIO;
+        }
+
+        /*
+         * Start LIPP/LEPP.
+         */
+        if (hv_dev_pwrite(priv->hv_devhdl, 0, (HV_VirtAddr)&dummy,
+                          sizeof(dummy), NETIO_IPP_START_SHIM_OFF) < 0) {
+                pr_warning("Failed to start LIPP/LEPP.\n");
+                return -EIO;
+        }
+
+        return 0;
+}
+
+
+/*
+ * Register with hypervisor on each CPU.
+ *
+ * Strangely, this function does important things even if it "fails",
+ * which is especially common if the link is not up yet.  Hopefully
+ * these things are all "harmless" if done twice!
+ */
+static void tile_net_register(void *dev_ptr)
+{
+        struct net_device *dev = (struct net_device *)dev_ptr;
+        struct tile_net_priv *priv = netdev_priv(dev);
+        int my_cpu = smp_processor_id();
+        struct tile_net_cpu *info;
+
+        struct tile_netio_queue *queue;
+
+        /* Only network cpus can receive packets. */
+        int queue_id =
+                cpumask_test_cpu(my_cpu, &priv->network_cpus_map) ? 0 : 255;
+
+        netio_input_config_t config = {
+                .flags = 0,
+                .num_receive_packets = priv->network_cpus_credits,
+                .queue_id = queue_id
+        };
+
+        int ret = 0;
+        netio_queue_impl_t *queuep;
+
+        PDEBUG("tile_net_register(queue_id %d)\n", queue_id);
+
+        if (!strcmp(dev->name, "xgbe0"))
+                info = &__get_cpu_var(hv_xgbe0);
+        else if (!strcmp(dev->name, "xgbe1"))
+                info = &__get_cpu_var(hv_xgbe1);
+        else if (!strcmp(dev->name, "gbe0"))
+                info = &__get_cpu_var(hv_gbe0);
+        else if (!strcmp(dev->name, "gbe1"))
+                info = &__get_cpu_var(hv_gbe1);
+        else
+                BUG();
+
+        /* Initialize the egress timer. */
+        init_timer(&info->egress_timer);
+        info->egress_timer.data = (long)info;
+        info->egress_timer.function = tile_net_handle_egress_timer;
+
+        priv->cpu[my_cpu] = info;
+
+        /*
+         * Register ourselves with the IPP.
+         */
+        ret = hv_dev_pwrite(priv->hv_devhdl, 0,
+                            (HV_VirtAddr)&config,
+                            sizeof(netio_input_config_t),
+                            NETIO_IPP_INPUT_REGISTER_OFF);
+        PDEBUG("hv_dev_pwrite(NETIO_IPP_INPUT_REGISTER_OFF) returned %d\n",
+               ret);
+        if (ret < 0) {
+                printk(KERN_DEBUG "hv_dev_pwrite NETIO_IPP_INPUT_REGISTER_OFF"
+                       " failure %d\n", ret);
+                info->link_down = (ret == NETIO_LINK_DOWN);
+                return;
+        }
+
+        /*
+         * Get the pointer to our queue's system part.
+         */
+
+        ret = hv_dev_pread(priv->hv_devhdl, 0,
+                           (HV_VirtAddr)&queuep,
+                           sizeof(netio_queue_impl_t *),
+                           NETIO_IPP_INPUT_REGISTER_OFF);
+        PDEBUG("hv_dev_pread(NETIO_IPP_INPUT_REGISTER_OFF) returned %d\n",
+               ret);
+        PDEBUG("queuep %p\n", queuep);
+        if (ret <= 0) {
+                /* ISSUE: Shouldn't this be a fatal error? */
+                pr_err("hv_dev_pread NETIO_IPP_INPUT_REGISTER_OFF failure\n");
+                return;
+        }
+
+        queue = &info->queue;
+
+        queue->__system_part = queuep;
+
+        memset(&queue->__user_part, 0, sizeof(netio_queue_user_impl_t));
+
+        /* This is traditionally "config.num_receive_packets / 2". */
+        queue->__user_part.__receive_credit_interval = 4;
+        queue->__user_part.__receive_credit_remaining =
+                queue->__user_part.__receive_credit_interval;
+
+        /*
+         * Get a fastio index from the hypervisor.
+         * ISSUE: Shouldn't this check the result?
+         */
+        ret = hv_dev_pread(priv->hv_devhdl, 0,
+                           (HV_VirtAddr)&queue->__user_part.__fastio_index,
+                           sizeof(queue->__user_part.__fastio_index),
+                           NETIO_IPP_GET_FASTIO_OFF);
+        PDEBUG("hv_dev_pread(NETIO_IPP_GET_FASTIO_OFF) returned %d\n", ret);
+
+        netif_napi_add(dev, &info->napi, tile_net_poll, 64);
+
+        /* Now we are registered. */
+        info->registered = true;
+}
+
+
+/*
+ * Unregister with hypervisor on each CPU.
+ */
+static void tile_net_unregister(void *dev_ptr)
+{
+        struct net_device *dev = (struct net_device *)dev_ptr;
+        struct tile_net_priv *priv = netdev_priv(dev);
+        int my_cpu = smp_processor_id();
+        struct tile_net_cpu *info = priv->cpu[my_cpu];
+
+        int ret = 0;
+        int dummy = 0;
+
+        /* Do nothing if never registered. */
+        if (info == NULL)
+                return;
+
+        /* Do nothing if already unregistered. */
+        if (!info->registered)
+                return;
+
+        /*
+         * Unregister ourselves with LIPP.
+         */
+        ret = hv_dev_pwrite(priv->hv_devhdl, 0, (HV_VirtAddr)&dummy,
+                            sizeof(dummy), NETIO_IPP_INPUT_UNREGISTER_OFF);
+        PDEBUG("hv_dev_pwrite(NETIO_IPP_INPUT_UNREGISTER_OFF) returned %d\n",
+               ret);
+        if (ret < 0) {
+                /* FIXME: Just panic? */
+                pr_err("hv_dev_pwrite NETIO_IPP_INPUT_UNREGISTER_OFF"
+                       " failure %d\n", ret);
+        }
+
+        /*
+         * Discard all packets still in our NetIO queue.  Hopefully,
+         * once the unregister call is complete, there will be no
+         * packets still in flight on the IDN.
+         */
+        tile_net_discard_packets(dev);
+
+        /* Reset state. */
+        info->num_needed_small_buffers = 0;
+        info->num_needed_large_buffers = 0;
+
+        /* Cancel egress timer. */
+        del_timer(&info->egress_timer);
+        info->egress_timer_scheduled = false;
+
+        netif_napi_del(&info->napi);
+
+        /* Now we are unregistered. */
+        info->registered = false;
+}
+
+
+/*
+ * Helper function for "tile_net_stop()".
+ *
+ * Also used to handle registration failure in "tile_net_open_inner()",
+ * when "fully_opened" is known to be false, and the various extra
+ * steps in "tile_net_stop()" are not necessary.  ISSUE: It might be
+ * simpler if we could just call "tile_net_stop()" anyway.
+ */
+static void tile_net_stop_aux(struct net_device *dev)
+{
+        struct tile_net_priv *priv = netdev_priv(dev);
+
+        int dummy = 0;
+
+        /* Unregister all tiles, so LIPP will stop delivering packets. */
+        on_each_cpu(tile_net_unregister, (void *)dev, 1);
+
+        /* Stop LIPP/LEPP. */
+        if (hv_dev_pwrite(priv->hv_devhdl, 0, (HV_VirtAddr)&dummy,
+                          sizeof(dummy), NETIO_IPP_STOP_SHIM_OFF) < 0)
+                panic("Failed to stop LIPP/LEPP!\n");
+
+        priv->partly_opened = 0;
+}
+
+
+/*
+ * Disable ingress interrupts for the given device on the current cpu.
+ */
+static void tile_net_disable_intr(void *dev_ptr)
+{
+        struct net_device *dev = (struct net_device *)dev_ptr;
+        struct tile_net_priv *priv = netdev_priv(dev);
+        int my_cpu = smp_processor_id();
+        struct tile_net_cpu *info = priv->cpu[my_cpu];
+
+        /* Disable hypervisor interrupt. */
+        disable_percpu_irq(priv->intr_id);
+
+        /* Disable NAPI if needed. */
+        if (info != NULL && info->napi_enabled) {
+                napi_disable(&info->napi);
+                info->napi_enabled = false;
+        }
+}
+
+
+/*
+ * Enable ingress interrupts for the given device on the current cpu.
+ */
+static void tile_net_enable_intr(void *dev_ptr)
+{
+        struct net_device *dev = (struct net_device *)dev_ptr;
+        struct tile_net_priv *priv = netdev_priv(dev);
+        int my_cpu = smp_processor_id();
+        struct tile_net_cpu *info = priv->cpu[my_cpu];
+
+        /* Enable hypervisor interrupt. */
+        enable_percpu_irq(priv->intr_id);
+
+        /* Enable NAPI. */
+        napi_enable(&info->napi);
+        info->napi_enabled = true;
+}
+
+
+/*
+ * tile_net_open_inner does most of the work of bringing up the interface.
+ * It's called from tile_net_open(), and also from tile_net_retry_open().
+ * The return value is 0 if the interface was brought up, < 0 if
+ * tile_net_open() should return the return value as an error, and > 0 if
+ * tile_net_open() should return success and schedule a work item to
+ * periodically retry the bringup.
+ */
+static int tile_net_open_inner(struct net_device *dev)
+{
+        struct tile_net_priv *priv = netdev_priv(dev);
+        int my_cpu = smp_processor_id();
+        struct tile_net_cpu *info;
+        struct tile_netio_queue *queue;
+        unsigned int irq;
+        int i;
+
+        /*
+         * First try to register just on the local CPU, and handle any
+         * semi-expected "link down" failure specially.  Note that we
+         * do NOT call "tile_net_stop_aux()", unlike below.
+         */
+        tile_net_register(dev);
+        info = priv->cpu[my_cpu];
+        if (!info->registered) {
+                if (info->link_down)
+                        return 1;
+                return -EAGAIN;
+        }
+
+        /*
+         * Now register everywhere else.  If any registration fails,
+         * even for "link down" (which might not be possible), we
+         * clean up using "tile_net_stop_aux()".
+         */
+        smp_call_function(tile_net_register, (void *)dev, 1);
+        for_each_online_cpu(i) {
+                if (!priv->cpu[i]->registered) {
+                        tile_net_stop_aux(dev);
+                        return -EAGAIN;
+                }
+        }
+
+        queue = &info->queue;
+
+        /*
+         * Set the device intr bit mask.
+         * The tile_net_register above sets per tile __intr_id.
+         */
+        priv->intr_id = queue->__system_part->__intr_id;
+        BUG_ON(!priv->intr_id);
+
+        /*
+         * Register the device interrupt handler.
+         * The __ffs() function returns the index into the interrupt handler
+         * table from the interrupt bit mask which should have one bit
+         * and one bit only set.
+         */
+        irq = __ffs(priv->intr_id);
+        tile_irq_activate(irq, TILE_IRQ_PERCPU);
+        BUG_ON(request_irq(irq, tile_net_handle_ingress_interrupt,
+                           0, dev->name, (void *)dev) != 0);
+
+        /* ISSUE: How could "priv->fully_opened" ever be "true" here? */
+
+        if (!priv->fully_opened) {
+
+                int dummy = 0;
+
+                /* Allocate initial buffers. */
+
+                int max_buffers =
+                        priv->network_cpus_count * priv->network_cpus_credits;
+
+                info->num_needed_small_buffers =
+                        min(LIPP_SMALL_BUFFERS, max_buffers);
+
+                info->num_needed_large_buffers =
+                        min(LIPP_LARGE_BUFFERS, max_buffers);
+
+                tile_net_provide_needed_buffers(info);
+
+                if (info->num_needed_small_buffers != 0 ||
+                    info->num_needed_large_buffers != 0)
+                        panic("Insufficient memory for buffer stack!");
+
+                /* Start LIPP/LEPP and activate "ingress" at the shim. */
+                if (hv_dev_pwrite(priv->hv_devhdl, 0, (HV_VirtAddr)&dummy,
+                                  sizeof(dummy), NETIO_IPP_INPUT_INIT_OFF) < 0)
+                        panic("Failed to activate the LIPP Shim!\n");
+
+                priv->fully_opened = 1;
+        }
+
+        /* On each tile, enable the hypervisor to trigger interrupts. */
+        /* ISSUE: Do this before starting LIPP/LEPP? */
+        on_each_cpu(tile_net_enable_intr, (void *)dev, 1);
+
+        /* Start our transmit queue. */
+        netif_start_queue(dev);
+
+        return 0;
+}
+
+
+/*
+ * Called periodically to retry bringing up the NetIO interface,
+ * if it doesn't come up cleanly during tile_net_open().
+ */
+static void tile_net_open_retry(struct work_struct *w)
+{
+        struct delayed_work *dw =
+                container_of(w, struct delayed_work, work);
+
+        struct tile_net_priv *priv =
+                container_of(dw, struct tile_net_priv, retry_work);
+
+        /*
+         * Try to bring the NetIO interface up.  If it fails, reschedule
+         * ourselves to try again later; otherwise, tell Linux we now have
+         * a working link.  ISSUE: What if the return value is negative?
+         */
+        if (tile_net_open_inner(priv->dev))
+                schedule_delayed_work_on(singlethread_cpu, &priv->retry_work,
+                                         TILE_NET_RETRY_INTERVAL);
+        else
+                netif_carrier_on(priv->dev);
+}
+
+
+/*
+ * Called when a network interface is made active.
+ *
+ * Returns 0 on success, negative value on failure.
+ *
+ * The open entry point is called when a network interface is made
+ * active by the system (IFF_UP).  At this point all resources needed
+ * for transmit and receive operations are allocated, the interrupt
+ * handler is registered with the OS, the watchdog timer is started,
+ * and the stack is notified that the interface is ready.
+ *
+ * If the actual link is not available yet, then we tell Linux that
+ * we have no carrier, and we keep checking until the link comes up.
+ */
+static int tile_net_open(struct net_device *dev)
+{
+        int ret = 0;
+        struct tile_net_priv *priv = netdev_priv(dev);
+
+        /*
+         * We rely on priv->partly_opened to tell us if this is the
+         * first time this interface is being brought up. If it is
+         * set, the IPP was already initialized and should not be
+         * initialized again.
+         */
+        if (!priv->partly_opened) {
+
+                int count;
+                int credits;
+
+                /* Initialize LIPP/LEPP, and start the Shim. */
+                ret = tile_net_open_aux(dev);
+                if (ret < 0) {
+                        pr_err("tile_net_open_aux failed: %d\n", ret);
+                        return ret;
+                }
+
+                /* Analyze the network cpus. */
+
+                if (network_cpus_used)
+                        cpumask_copy(&priv->network_cpus_map,
+                                     &network_cpus_map);
+                else
+                        cpumask_copy(&priv->network_cpus_map, cpu_online_mask);
+
+
+                count = cpumask_weight(&priv->network_cpus_map);
+
+                /* Limit credits to available buffers, and apply min. */
+                credits = max(16, (LIPP_LARGE_BUFFERS / count) & ~1);
+
+                /* Apply "GBE" max limit. */
+                /* ISSUE: Use higher limit for XGBE? */
+                credits = min(NETIO_MAX_RECEIVE_PKTS, credits);
+
+                priv->network_cpus_count = count;
+                priv->network_cpus_credits = credits;
+
+#ifdef TILE_NET_DEBUG
+                pr_info("Using %d network cpus, with %d credits each\n",
+                       priv->network_cpus_count, priv->network_cpus_credits);
+#endif
+
+                priv->partly_opened = 1;
+        }
+
+        /*
+         * Attempt to bring up the link.
+         */
+        ret = tile_net_open_inner(dev);
+        if (ret <= 0) {
+                if (ret == 0)
+                        netif_carrier_on(dev);
+                return ret;
+        }
+
+        /*
+         * We were unable to bring up the NetIO interface, but we want to
+         * try again in a little bit.  Tell Linux that we have no carrier
+         * so it doesn't try to use the interface before the link comes up
+         * and then remember to try again later.
+         */
+        netif_carrier_off(dev);
+        schedule_delayed_work_on(singlethread_cpu, &priv->retry_work,
+                                 TILE_NET_RETRY_INTERVAL);
+
+        return 0;
+}
+
+
+/*
+ * Disables a network interface.
+ *
+ * Returns 0, this is not allowed to fail.
+ *
+ * The close entry point is called when an interface is de-activated
+ * by the OS.  The hardware is still under the drivers control, but
+ * needs to be disabled.  A global MAC reset is issued to stop the
+ * hardware, and all transmit and receive resources are freed.
+ *
+ * ISSUE: Can this can be called while "tile_net_poll()" is running?
+ */
+static int tile_net_stop(struct net_device *dev)
+{
+        struct tile_net_priv *priv = netdev_priv(dev);
+
+        bool pending = true;
+
+        PDEBUG("tile_net_stop()\n");
+
+        /* ISSUE: Only needed if not yet fully open. */
+        cancel_delayed_work_sync(&priv->retry_work);
+
+        /* Can't transmit any more. */
+        netif_stop_queue(dev);
+
+        /*
+         * Disable hypervisor interrupts on each tile.
+         */
+        on_each_cpu(tile_net_disable_intr, (void *)dev, 1);
+
+        /*
+         * Unregister the interrupt handler.
+         * The __ffs() function returns the index into the interrupt handler
+         * table from the interrupt bit mask which should have one bit
+         * and one bit only set.
+         */
+        if (priv->intr_id)
+                free_irq(__ffs(priv->intr_id), dev);
+
+        /*
+         * Drain all the LIPP buffers.
+         */
+
+        while (true) {
+                int buffer;
+
+                /* NOTE: This should never fail. */
+                if (hv_dev_pread(priv->hv_devhdl, 0, (HV_VirtAddr)&buffer,
+                                 sizeof(buffer), NETIO_IPP_DRAIN_OFF) < 0)
+                        break;
+
+                /* Stop when done. */
+                if (buffer == 0)
+                        break;
+
+                {
+                        /* Convert "linux_buffer_t" to "va". */
+                        void *va = __va((phys_addr_t)(buffer >> 1) << 7);
+
+                        /* Acquire the associated "skb". */
+                        struct sk_buff **skb_ptr = va - sizeof(*skb_ptr);
+                        struct sk_buff *skb = *skb_ptr;
+
+                        kfree_skb(skb);
+                }
+        }
+
+        /* Stop LIPP/LEPP. */
+        tile_net_stop_aux(dev);
+
+
+        priv->fully_opened = 0;
+
+
+        /*
+         * XXX: ISSUE: It appears that, in practice anyway, by the
+         * time we get here, there are no pending completions.
+         */
+        while (pending) {
+
+                struct sk_buff *olds[32];
+                unsigned int wanted = 32;
+                unsigned int i, nolds = 0;
+
+                nolds = tile_net_lepp_grab_comps(dev, olds,
+                                                 wanted, &pending);
+
+                /* ISSUE: We have never actually seen this debug spew. */
+                if (nolds != 0)
+                        pr_info("During tile_net_stop(), grabbed %d comps.\n",
+                               nolds);
+
+                for (i = 0; i < nolds; i++)
+                        kfree_skb(olds[i]);
+        }
+
+
+        /* Wipe the EPP queue. */
+        memset(priv->epp_queue, 0, sizeof(lepp_queue_t));
+
+        /* Evict the EPP queue. */
+        finv_buffer(priv->epp_queue, PAGE_SIZE);
+
+        return 0;
+}
+
+
+/*
+ * Prepare the "frags" info for the resulting LEPP command.
+ *
+ * If needed, flush the memory used by the frags.
+ */
+static unsigned int tile_net_tx_frags(lepp_frag_t *frags,
+                                      struct sk_buff *skb,
+                                      void *b_data, unsigned int b_len)
+{
+        unsigned int i, n = 0;
+
+        struct skb_shared_info *sh = skb_shinfo(skb);
+
+        phys_addr_t cpa;
+
+        if (b_len != 0) {
+
+                if (!hash_default)
+                        finv_buffer_remote(b_data, b_len);
+
+                cpa = __pa(b_data);
+                frags[n].cpa_lo = cpa;
+                frags[n].cpa_hi = cpa >> 32;
+                frags[n].length = b_len;
+                frags[n].hash_for_home = hash_default;
+                n++;
+        }
+
+        for (i = 0; i < sh->nr_frags; i++) {
+
+                skb_frag_t *f = &sh->frags[i];
+                unsigned long pfn = page_to_pfn(f->page);
+
+                /* FIXME: Compute "hash_for_home" properly. */
+                /* ISSUE: The hypervisor checks CHIP_HAS_REV1_DMA_PACKETS(). */
+                int hash_for_home = hash_default;
+
+                /* FIXME: Hmmm. */
+                if (!hash_default) {
+                        void *va = pfn_to_kaddr(pfn) + f->page_offset;
+                        BUG_ON(PageHighMem(f->page));
+                        finv_buffer_remote(va, f->size);
+                }
+
+                cpa = ((phys_addr_t)pfn << PAGE_SHIFT) + f->page_offset;
+                frags[n].cpa_lo = cpa;
+                frags[n].cpa_hi = cpa >> 32;
+                frags[n].length = f->size;
+                frags[n].hash_for_home = hash_for_home;
+                n++;
+        }
+
+        return n;
+}
+
+
+/*
+ * This function takes "skb", consisting of a header template and a
+ * payload, and hands it to LEPP, to emit as one or more segments,
+ * each consisting of a possibly modified header, plus a piece of the
+ * payload, via a process known as "tcp segmentation offload".
+ *
+ * Usually, "data" will contain the header template, of size "sh_len",
+ * and "sh->frags" will contain "skb->data_len" bytes of payload, and
+ * there will be "sh->gso_segs" segments.
+ *
+ * Sometimes, if "sendfile()" requires copying, we will be called with
+ * "data" containing the header and payload, with "frags" being empty.
+ *
+ * In theory, "sh->nr_frags" could be 3, but in practice, it seems
+ * that this will never actually happen.
+ *
+ * See "emulate_large_send_offload()" for some reference code, which
+ * does not handle checksumming.
+ *
+ * ISSUE: How do we make sure that high memory DMA does not migrate?
+ */
+static int tile_net_tx_tso(struct sk_buff *skb, struct net_device *dev)
+{
+        struct tile_net_priv *priv = netdev_priv(dev);
+        int my_cpu = smp_processor_id();
+        struct tile_net_cpu *info = priv->cpu[my_cpu];
+        struct tile_net_stats_t *stats = &info->stats;
+
+        struct skb_shared_info *sh = skb_shinfo(skb);
+
+        unsigned char *data = skb->data;
+
+        /* The ip header follows the ethernet header. */
+        struct iphdr *ih = ip_hdr(skb);
+        unsigned int ih_len = ih->ihl * 4;
+
+        /* Note that "nh == ih", by definition. */
+        unsigned char *nh = skb_network_header(skb);
+        unsigned int eh_len = nh - data;
+
+        /* The tcp header follows the ip header. */
+        struct tcphdr *th = (struct tcphdr *)(nh + ih_len);
+        unsigned int th_len = th->doff * 4;
+
+        /* The total number of header bytes. */
+        /* NOTE: This may be less than skb_headlen(skb). */
+        unsigned int sh_len = eh_len + ih_len + th_len;
+
+        /* The number of payload bytes at "skb->data + sh_len". */
+        /* This is non-zero for sendfile() without HIGHDMA. */
+        unsigned int b_len = skb_headlen(skb) - sh_len;
+
+        /* The total number of payload bytes. */
+        unsigned int d_len = b_len + skb->data_len;
+
+        /* The maximum payload size. */
+        unsigned int p_len = sh->gso_size;
+
+        /* The total number of segments. */
+        unsigned int num_segs = sh->gso_segs;
+
+        /* The temporary copy of the command. */
+        u32 cmd_body[(LEPP_MAX_CMD_SIZE + 3) / 4];
+        lepp_tso_cmd_t *cmd = (lepp_tso_cmd_t *)cmd_body;
+
+        /* Analyze the "frags". */
+        unsigned int num_frags =
+                tile_net_tx_frags(cmd->frags, skb, data + sh_len, b_len);
+
+        /* The size of the command, including frags and header. */
+        size_t cmd_size = LEPP_TSO_CMD_SIZE(num_frags, sh_len);
+
+        /* The command header. */
+        lepp_tso_cmd_t cmd_init = {
+                .tso = true,
+                .header_size = sh_len,
+                .ip_offset = eh_len,
+                .tcp_offset = eh_len + ih_len,
+                .payload_size = p_len,
+                .num_frags = num_frags,
+        };
+
+        unsigned long irqflags;
+
+        lepp_queue_t *eq = priv->epp_queue;
+
+        struct sk_buff *olds[4];
+        unsigned int wanted = 4;
+        unsigned int i, nolds = 0;
+
+        unsigned int cmd_head, cmd_tail, cmd_next;
+        unsigned int comp_tail;
+
+        unsigned int free_slots;
+
+
+        /* Paranoia. */
+        BUG_ON(skb->protocol != htons(ETH_P_IP));
+        BUG_ON(ih->protocol != IPPROTO_TCP);
+        BUG_ON(skb->ip_summed != CHECKSUM_PARTIAL);
+        BUG_ON(num_frags > LEPP_MAX_FRAGS);
+        /*--BUG_ON(num_segs != (d_len + (p_len - 1)) / p_len); */
+        BUG_ON(num_segs <= 1);
+
+
+        /* Finish preparing the command. */
+
+        /* Copy the command header. */
+        *cmd = cmd_init;
+
+        /* Copy the "header". */
+        memcpy(&cmd->frags[num_frags], data, sh_len);
+
+
+        /* Prefetch and wait, to minimize time spent holding the spinlock. */
+        prefetch_L1(&eq->comp_tail);
+        prefetch_L1(&eq->cmd_tail);
+        mb();
+
+
+        /* Enqueue the command. */
+
+        spin_lock_irqsave(&priv->cmd_lock, irqflags);
+
+        /*
+         * Handle completions if needed to make room.
+         * HACK: Spin until there is sufficient room.
+         */
+        free_slots = lepp_num_free_comp_slots(eq);
+        if (free_slots < 1) {
+spin:
+                nolds += tile_net_lepp_grab_comps(dev, olds + nolds,
+                                                  wanted - nolds, NULL);
+                if (lepp_num_free_comp_slots(eq) < 1)
+                        goto spin;
+        }
+
+        cmd_head = eq->cmd_head;
+        cmd_tail = eq->cmd_tail;
+
+        /* NOTE: The "gotos" below are untested. */
+
+        /* Prepare to advance, detecting full queue. */
+        cmd_next = cmd_tail + cmd_size;
+        if (cmd_tail < cmd_head && cmd_next >= cmd_head)
+                goto spin;
+        if (cmd_next > LEPP_CMD_LIMIT) {
+                cmd_next = 0;
+                if (cmd_next == cmd_head)
+                        goto spin;
+        }
+
+        /* Copy the command. */
+        memcpy(&eq->cmds[cmd_tail], cmd, cmd_size);
+
+        /* Advance. */
+        cmd_tail = cmd_next;
+
+        /* Record "skb" for eventual freeing. */
+        comp_tail = eq->comp_tail;
+        eq->comps[comp_tail] = skb;
+        LEPP_QINC(comp_tail);
+        eq->comp_tail = comp_tail;
+
+        /* Flush before allowing LEPP to handle the command. */
+        __insn_mf();
+
+        eq->cmd_tail = cmd_tail;
+
+        spin_unlock_irqrestore(&priv->cmd_lock, irqflags);
+
+        if (nolds == 0)
+                nolds = tile_net_lepp_grab_comps(dev, olds, wanted, NULL);
+
+        /* Handle completions. */
+        for (i = 0; i < nolds; i++)
+                kfree_skb(olds[i]);
+
+        /* Update stats. */
+        stats->tx_packets += num_segs;
+        stats->tx_bytes += (num_segs * sh_len) + d_len;
+
+        /* Make sure the egress timer is scheduled. */
+        tile_net_schedule_egress_timer(info);
+
+        return NETDEV_TX_OK;
+}
+
+
+/*
+ * Transmit a packet (called by the kernel via "hard_start_xmit" hook).
+ */
+static int tile_net_tx(struct sk_buff *skb, struct net_device *dev)
+{
+        struct tile_net_priv *priv = netdev_priv(dev);
+        int my_cpu = smp_processor_id();
+        struct tile_net_cpu *info = priv->cpu[my_cpu];
+        struct tile_net_stats_t *stats = &info->stats;
+
+        unsigned long irqflags;
+
+        struct skb_shared_info *sh = skb_shinfo(skb);
+
+        unsigned int len = skb->len;
+        unsigned char *data = skb->data;
+
+        unsigned int csum_start = skb->csum_start - skb_headroom(skb);
+
+        lepp_frag_t frags[LEPP_MAX_FRAGS];
+
+        unsigned int num_frags;
+
+        lepp_queue_t *eq = priv->epp_queue;
+
+        struct sk_buff *olds[4];
+        unsigned int wanted = 4;
+        unsigned int i, nolds = 0;
+
+        unsigned int cmd_size = sizeof(lepp_cmd_t);
+
+        unsigned int cmd_head, cmd_tail, cmd_next;
+        unsigned int comp_tail;
+
+        lepp_cmd_t cmds[LEPP_MAX_FRAGS];
+
+        unsigned int free_slots;
+
+
+        /*
+         * This is paranoia, since we think that if the link doesn't come
+         * up, telling Linux we have no carrier will keep it from trying
+         * to transmit.  If it does, though, we can't execute this routine,
+         * since data structures we depend on aren't set up yet.
+         */
+        if (!info->registered)
+                return NETDEV_TX_BUSY;
+
+
+        /* Save the timestamp. */
+        dev->trans_start = jiffies;
+
+
+#ifdef TILE_NET_PARANOIA
+#if CHIP_HAS_CBOX_HOME_MAP()
+        if (hash_default) {
+                HV_PTE pte = *virt_to_pte(current->mm, (unsigned long)data);
+                if (hv_pte_get_mode(pte) != HV_PTE_MODE_CACHE_HASH_L3)
+                        panic("Non-coherent egress buffer!");
+        }
+#endif
+#endif
+
+
+#ifdef TILE_NET_DUMP_PACKETS
+        /* ISSUE: Does not dump the "frags". */
+        dump_packet(data, skb_headlen(skb), "tx");
+#endif /* TILE_NET_DUMP_PACKETS */
+
+
+        if (sh->gso_size != 0)
+                return tile_net_tx_tso(skb, dev);
+
+
+        /* Prepare the commands. */
+
+        num_frags = tile_net_tx_frags(frags, skb, data, skb_headlen(skb));
+
+        for (i = 0; i < num_frags; i++) {
+
+                bool final = (i == num_frags - 1);
+
+                lepp_cmd_t cmd = {
+                        .cpa_lo = frags[i].cpa_lo,
+                        .cpa_hi = frags[i].cpa_hi,
+                        .length = frags[i].length,
+                        .hash_for_home = frags[i].hash_for_home,
+                        .send_completion = final,
+                        .end_of_packet = final
+                };
+
+                if (i == 0 && skb->ip_summed == CHECKSUM_PARTIAL) {
+                        cmd.compute_checksum = 1;
+                        cmd.checksum_data.bits.start_byte = csum_start;
+                        cmd.checksum_data.bits.count = len - csum_start;
+                        cmd.checksum_data.bits.destination_byte =
+                                csum_start + skb->csum_offset;
+                }
+
+                cmds[i] = cmd;
+        }
+
+
+        /* Prefetch and wait, to minimize time spent holding the spinlock. */
+        prefetch_L1(&eq->comp_tail);
+        prefetch_L1(&eq->cmd_tail);
+        mb();
+
+
+        /* Enqueue the commands. */
+
+        spin_lock_irqsave(&priv->cmd_lock, irqflags);
+
+        /*
+         * Handle completions if needed to make room.
+         * HACK: Spin until there is sufficient room.
+         */
+        free_slots = lepp_num_free_comp_slots(eq);
+        if (free_slots < 1) {
+spin:
+                nolds += tile_net_lepp_grab_comps(dev, olds + nolds,
+                                                  wanted - nolds, NULL);
+                if (lepp_num_free_comp_slots(eq) < 1)
+                        goto spin;
+        }
+
+        cmd_head = eq->cmd_head;
+        cmd_tail = eq->cmd_tail;
+
+        /* NOTE: The "gotos" below are untested. */
+
+        /* Copy the commands, or fail. */
+        for (i = 0; i < num_frags; i++) {
+
+                /* Prepare to advance, detecting full queue. */
+                cmd_next = cmd_tail + cmd_size;
+                if (cmd_tail < cmd_head && cmd_next >= cmd_head)
+                        goto spin;
+                if (cmd_next > LEPP_CMD_LIMIT) {
+                        cmd_next = 0;
+                        if (cmd_next == cmd_head)
+                                goto spin;
+                }
+
+                /* Copy the command. */
+                *(lepp_cmd_t *)&eq->cmds[cmd_tail] = cmds[i];
+
+                /* Advance. */
+                cmd_tail = cmd_next;
+        }
+
+        /* Record "skb" for eventual freeing. */
+        comp_tail = eq->comp_tail;
+        eq->comps[comp_tail] = skb;
+        LEPP_QINC(comp_tail);
+        eq->comp_tail = comp_tail;
+
+        /* Flush before allowing LEPP to handle the command. */
+        __insn_mf();
+
+        eq->cmd_tail = cmd_tail;
+
+        spin_unlock_irqrestore(&priv->cmd_lock, irqflags);
+
+        if (nolds == 0)
+                nolds = tile_net_lepp_grab_comps(dev, olds, wanted, NULL);
+
+        /* Handle completions. */
+        for (i = 0; i < nolds; i++)
+                kfree_skb(olds[i]);
+
+        /* HACK: Track "expanded" size for short packets (e.g. 42 < 60). */
+        stats->tx_packets++;
+        stats->tx_bytes += ((len >= ETH_ZLEN) ? len : ETH_ZLEN);
+
+        /* Make sure the egress timer is scheduled. */
+        tile_net_schedule_egress_timer(info);
+
+        return NETDEV_TX_OK;
+}
+
+
+/*
+ * Deal with a transmit timeout.
+ */
+static void tile_net_tx_timeout(struct net_device *dev)
+{
+        PDEBUG("tile_net_tx_timeout()\n");
+        PDEBUG("Transmit timeout at %ld, latency %ld\n", jiffies,
+               jiffies - dev->trans_start);
+
+        /* XXX: ISSUE: This doesn't seem useful for us. */
+        netif_wake_queue(dev);
+}
+
+
+/*
+ * Ioctl commands.
+ */
+static int tile_net_ioctl(struct net_device *dev, struct ifreq *rq, int cmd)
+{
+        return -EOPNOTSUPP;
+}
+
+
+/*
+ * Get System Network Statistics.
+ *
+ * Returns the address of the device statistics structure.
+ */
+static struct net_device_stats *tile_net_get_stats(struct net_device *dev)
+{
+        struct tile_net_priv *priv = netdev_priv(dev);
+        u32 rx_packets = 0;
+        u32 tx_packets = 0;
+        u32 rx_bytes = 0;
+        u32 tx_bytes = 0;
+        int i;
+
+        for_each_online_cpu(i) {
+                if (priv->cpu[i]) {
+                        rx_packets += priv->cpu[i]->stats.rx_packets;
+                        rx_bytes += priv->cpu[i]->stats.rx_bytes;
+                        tx_packets += priv->cpu[i]->stats.tx_packets;
+                        tx_bytes += priv->cpu[i]->stats.tx_bytes;
+                }
+        }
+
+        priv->stats.rx_packets = rx_packets;
+        priv->stats.rx_bytes = rx_bytes;
+        priv->stats.tx_packets = tx_packets;
+        priv->stats.tx_bytes = tx_bytes;
+
+        return &priv->stats;
+}
+
+
+/*
+ * Change the "mtu".
+ *
+ * The "change_mtu" method is usually not needed.
+ * If you need it, it must be like this.
+ */
+static int tile_net_change_mtu(struct net_device *dev, int new_mtu)
+{
+        PDEBUG("tile_net_change_mtu()\n");
+
+        /* Check ranges. */
+        if ((new_mtu < 68) || (new_mtu > 1500))
+                return -EINVAL;
+
+        /* Accept the value. */
+        dev->mtu = new_mtu;
+
+        return 0;
+}
+
+
+/*
+ * Change the Ethernet Address of the NIC.
+ *
+ * The hypervisor driver does not support changing MAC address.  However,
+ * the IPP does not do anything with the MAC address, so the address which
+ * gets used on outgoing packets, and which is accepted on incoming packets,
+ * is completely up to the NetIO program or kernel driver which is actually
+ * handling them.
+ *
+ * Returns 0 on success, negative on failure.
+ */
+static int tile_net_set_mac_address(struct net_device *dev, void *p)
+{
+        struct sockaddr *addr = p;
+
+        if (!is_valid_ether_addr(addr->sa_data))
+                return -EINVAL;
+
+        /* ISSUE: Note that "dev_addr" is now a pointer. */
+        memcpy(dev->dev_addr, addr->sa_data, dev->addr_len);
+
+        return 0;
+}
+
+
+/*
+ * Obtain the MAC address from the hypervisor.
+ * This must be done before opening the device.
+ */
+static int tile_net_get_mac(struct net_device *dev)
+{
+        struct tile_net_priv *priv = netdev_priv(dev);
+
+        char hv_dev_name[32];
+        int len;
+
+        __netio_getset_offset_t offset = { .word = NETIO_IPP_PARAM_OFF };
+
+        int ret;
+
+        /* For example, "xgbe0". */
+        strcpy(hv_dev_name, dev->name);
+        len = strlen(hv_dev_name);
+
+        /* For example, "xgbe/0". */
+        hv_dev_name[len] = hv_dev_name[len - 1];
+        hv_dev_name[len - 1] = '/';
+        len++;
+
+        /* For example, "xgbe/0/native_hash". */
+        strcpy(hv_dev_name + len, hash_default ? "/native_hash" : "/native");
+
+        /* Get the hypervisor handle for this device. */
+        priv->hv_devhdl = hv_dev_open((HV_VirtAddr)hv_dev_name, 0);
+        PDEBUG("hv_dev_open(%s) returned %d %p\n",
+               hv_dev_name, priv->hv_devhdl, &priv->hv_devhdl);
+        if (priv->hv_devhdl < 0) {
+                if (priv->hv_devhdl == HV_ENODEV)
+                        printk(KERN_DEBUG "Ignoring unconfigured device %s\n",
+                                 hv_dev_name);
+                else
+                        printk(KERN_DEBUG "hv_dev_open(%s) returned %d\n",
+                                 hv_dev_name, priv->hv_devhdl);
+                return -1;
+        }
+
+        /*
+         * Read the hardware address from the hypervisor.
+         * ISSUE: Note that "dev_addr" is now a pointer.
+         */
+        offset.bits.class = NETIO_PARAM;
+        offset.bits.addr = NETIO_PARAM_MAC;
+        ret = hv_dev_pread(priv->hv_devhdl, 0,
+                           (HV_VirtAddr)dev->dev_addr, dev->addr_len,
+                           offset.word);
+        PDEBUG("hv_dev_pread(NETIO_PARAM_MAC) returned %d\n", ret);
+        if (ret <= 0) {
+                printk(KERN_DEBUG "hv_dev_pread(NETIO_PARAM_MAC) %s failed\n",
+                       dev->name);
+                /*
+                 * Since the device is configured by the hypervisor but we
+                 * can't get its MAC address, we are most likely running
+                 * the simulator, so let's generate a random MAC address.
+                 */
+                random_ether_addr(dev->dev_addr);
+        }
+
+        return 0;
+}
+
+
+static struct net_device_ops tile_net_ops = {
+        .ndo_open = tile_net_open,
+        .ndo_stop = tile_net_stop,
+        .ndo_start_xmit = tile_net_tx,
+        .ndo_do_ioctl = tile_net_ioctl,
+        .ndo_get_stats = tile_net_get_stats,
+        .ndo_change_mtu = tile_net_change_mtu,
+        .ndo_tx_timeout = tile_net_tx_timeout,
+        .ndo_set_mac_address = tile_net_set_mac_address
+};
+
+
+/*
+ * The setup function.
+ *
+ * This uses ether_setup() to assign various fields in dev, including
+ * setting IFF_BROADCAST and IFF_MULTICAST, then sets some extra fields.
+ */
+static void tile_net_setup(struct net_device *dev)
+{
+        PDEBUG("tile_net_setup()\n");
+
+        ether_setup(dev);
+
+        dev->netdev_ops = &tile_net_ops;
+
+        dev->watchdog_timeo = TILE_NET_TIMEOUT;
+
+        /* We want lockless xmit. */
+        dev->features |= NETIF_F_LLTX;
+
+        /* We support hardware tx checksums. */
+        dev->features |= NETIF_F_HW_CSUM;
+
+        /* We support scatter/gather. */
+        dev->features |= NETIF_F_SG;
+
+        /* We support TSO. */
+        dev->features |= NETIF_F_TSO;
+
+#ifdef TILE_NET_GSO
+        /* We support GSO. */
+        dev->features |= NETIF_F_GSO;
+#endif
+
+        if (hash_default)
+                dev->features |= NETIF_F_HIGHDMA;
+
+        /* ISSUE: We should support NETIF_F_UFO. */
+
+        dev->tx_queue_len = TILE_NET_TX_QUEUE_LEN;
+
+        dev->mtu = TILE_NET_MTU;
+}
+
+
+/*
+ * Allocate the device structure, register the device, and obtain the
+ * MAC address from the hypervisor.
+ */
+static struct net_device *tile_net_dev_init(const char *name)
+{
+        int ret;
+        struct net_device *dev;
+        struct tile_net_priv *priv;
+        struct page *page;
+
+        /*
+         * Allocate the device structure.  This allocates "priv", calls
+         * tile_net_setup(), and saves "name".  Normally, "name" is a
+         * template, instantiated by register_netdev(), but not for us.
+         */
+        dev = alloc_netdev(sizeof(*priv), name, tile_net_setup);
+        if (!dev) {
+                pr_err("alloc_netdev(%s) failed\n", name);
+                return NULL;
+        }
+
+        priv = netdev_priv(dev);
+
+        /* Initialize "priv". */
+
+        memset(priv, 0, sizeof(*priv));
+
+        /* Save "dev" for "tile_net_open_retry()". */
+        priv->dev = dev;
+
+        INIT_DELAYED_WORK(&priv->retry_work, tile_net_open_retry);
+
+        spin_lock_init(&priv->cmd_lock);
+        spin_lock_init(&priv->comp_lock);
+
+        /* Allocate "epp_queue". */
+        BUG_ON(get_order(sizeof(lepp_queue_t)) != 0);
+        page = alloc_pages(GFP_KERNEL | __GFP_ZERO, 0);
+        if (!page) {
+                free_netdev(dev);
+                return NULL;
+        }
+        priv->epp_queue = page_address(page);
+
+        /* Register the network device. */
+        ret = register_netdev(dev);
+        if (ret) {
+                pr_err("register_netdev %s failed %d\n", dev->name, ret);
+                free_page((unsigned long)priv->epp_queue);
+                free_netdev(dev);
+                return NULL;
+        }
+
+        /* Get the MAC address. */
+        ret = tile_net_get_mac(dev);
+        if (ret < 0) {
+                unregister_netdev(dev);
+                free_page((unsigned long)priv->epp_queue);
+                free_netdev(dev);
+                return NULL;
+        }
+
+        return dev;
+}
+
+
+/*
+ * Module cleanup.
+ */
+static void tile_net_cleanup(void)
+{
+        int i;
+
+        for (i = 0; i < TILE_NET_DEVS; i++) {
+                if (tile_net_devs[i]) {
+                        struct net_device *dev = tile_net_devs[i];
+                        struct tile_net_priv *priv = netdev_priv(dev);
+                        unregister_netdev(dev);
+                        finv_buffer(priv->epp_queue, PAGE_SIZE);
+                        free_page((unsigned long)priv->epp_queue);
+                        free_netdev(dev);
+                }
+        }
+}
+
+
+/*
+ * Module initialization.
+ */
+static int tile_net_init_module(void)
+{
+        pr_info("Tilera IPP Net Driver\n");
+
+        tile_net_devs[0] = tile_net_dev_init("xgbe0");
+        tile_net_devs[1] = tile_net_dev_init("xgbe1");
+        tile_net_devs[2] = tile_net_dev_init("gbe0");
+        tile_net_devs[3] = tile_net_dev_init("gbe1");
+
+        return 0;
+}
+
+
+#ifndef MODULE
+/*
+ * The "network_cpus" boot argument specifies the cpus that are dedicated
+ * to handle ingress packets.
+ *
+ * The parameter should be in the form "network_cpus=m-n[,x-y]", where
+ * m, n, x, y are integer numbers that represent the cpus that can be
+ * neither a dedicated cpu nor a dataplane cpu.
+ */
+static int __init network_cpus_setup(char *str)
+{
+        int rc = cpulist_parse_crop(str, &network_cpus_map);
+        if (rc != 0) {
+                pr_warning("network_cpus=%s: malformed cpu list\n",
+                       str);
+        } else {
+
+                /* Remove dedicated cpus. */
+                cpumask_and(&network_cpus_map, &network_cpus_map,
+                            cpu_possible_mask);
+
+
+                if (cpumask_empty(&network_cpus_map)) {
+                        pr_warning("Ignoring network_cpus='%s'.\n",
+                               str);
+                } else {
+                        char buf[1024];
+                        cpulist_scnprintf(buf, sizeof(buf), &network_cpus_map);
+                        pr_info("Linux network CPUs: %s\n", buf);
+                        network_cpus_used = true;
+                }
+        }
+
+        return 0;
+}
+__setup("network_cpus=", network_cpus_setup);
+#endif
+
+
+module_init(tile_net_init_module);
+module_exit(tile_net_cleanup);
diff --git a/drivers/net/wireless/ath/ath9k/eeprom_9287.c b/drivers/net/wireless/ath/ath9k/eeprom_9287.c
index 966b9496a9dd..195406db3bd8 100644
--- a/drivers/net/wireless/ath/ath9k/eeprom_9287.c
+++ b/drivers/net/wireless/ath/ath9k/eeprom_9287.c
@@ -37,7 +37,7 @@ static bool ath9k_hw_ar9287_fill_eeprom(struct ath_hw *ah)
         int addr, eep_start_loc;
         eep_data = (u16 *)eep;
 
-        if (ah->hw_version.devid == 0x7015)
+        if (AR9287_HTC_DEVID(ah))
                 eep_start_loc = AR9287_HTC_EEP_START_LOC;
         else
                 eep_start_loc = AR9287_EEP_START_LOC;
diff --git a/drivers/net/wireless/ath/ath9k/hif_usb.c b/drivers/net/wireless/ath/ath9k/hif_usb.c
index f7ec31b4ddd3..dfb6560dab92 100644
--- a/drivers/net/wireless/ath/ath9k/hif_usb.c
+++ b/drivers/net/wireless/ath/ath9k/hif_usb.c
@@ -36,8 +36,13 @@ static struct usb_device_id ath9k_hif_usb_ids[] = {
         { USB_DEVICE(0x13D3, 0x3327) }, /* Azurewave */
         { USB_DEVICE(0x13D3, 0x3328) }, /* Azurewave */
         { USB_DEVICE(0x13D3, 0x3346) }, /* IMC Networks */
+        { USB_DEVICE(0x13D3, 0x3348) }, /* Azurewave */
+        { USB_DEVICE(0x13D3, 0x3349) }, /* Azurewave */
+        { USB_DEVICE(0x13D3, 0x3350) }, /* Azurewave */
         { USB_DEVICE(0x04CA, 0x4605) }, /* Liteon */
         { USB_DEVICE(0x083A, 0xA704) }, /* SMC Networks */
+        { USB_DEVICE(0x040D, 0x3801) }, /* VIA */
+        { USB_DEVICE(0x1668, 0x1200) }, /* Verizon */
         { },
 };
 
@@ -806,6 +811,8 @@ static int ath9k_hif_usb_download_fw(struct hif_device_usb *hif_dev)
         case 0x7010:
         case 0x7015:
         case 0x9018:
+        case 0xA704:
+        case 0x1200:
                 firm_offset = AR7010_FIRMWARE_TEXT;
                 break;
         default:
@@ -928,6 +935,8 @@ static int ath9k_hif_usb_probe(struct usb_interface *interface,
         case 0x7010:
         case 0x7015:
         case 0x9018:
+        case 0xA704:
+        case 0x1200:
                 if (le16_to_cpu(udev->descriptor.bcdDevice) == 0x0202)
                         hif_dev->fw_name = FIRMWARE_AR7010_1_1;
                 else
diff --git a/drivers/net/wireless/ath/ath9k/htc_drv_init.c b/drivers/net/wireless/ath/ath9k/htc_drv_init.c
index 3d7b97f1b3ae..7c8a38d04561 100644
--- a/drivers/net/wireless/ath/ath9k/htc_drv_init.c
+++ b/drivers/net/wireless/ath/ath9k/htc_drv_init.c
@@ -249,6 +249,8 @@ static int ath9k_init_htc_services(struct ath9k_htc_priv *priv, u16 devid)
         case 0x7010:
         case 0x7015:
         case 0x9018:
+        case 0xA704:
+        case 0x1200:
                 priv->htc->credits = 45;
                 break;
         default:
diff --git a/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c b/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c
index 3d19b5bc937f..29d80ca78393 100644
--- a/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c
+++ b/drivers/net/wireless/ath/ath9k/htc_drv_txrx.c
@@ -121,7 +121,7 @@ int ath9k_htc_tx_start(struct ath9k_htc_priv *priv, struct sk_buff *skb)
                         tx_hdr.data_type = ATH9K_HTC_NORMAL;
                 }
 
-                if (ieee80211_is_data(fc)) {
+                if (ieee80211_is_data_qos(fc)) {
                         qc = ieee80211_get_qos_ctl(hdr);
                         tx_hdr.tidno = qc[0] & IEEE80211_QOS_CTL_TID_MASK;
                 }
diff --git a/drivers/net/wireless/ath/ath9k/init.c b/drivers/net/wireless/ath/ath9k/init.c
index 6a0d99eff404..92bc5c5f4876 100644
--- a/drivers/net/wireless/ath/ath9k/init.c
+++ b/drivers/net/wireless/ath/ath9k/init.c
@@ -817,8 +817,6 @@ void ath9k_deinit_device(struct ath_softc *sc)
 
         ath9k_ps_wakeup(sc);
 
-        pm_qos_remove_request(&ath9k_pm_qos_req);
-
         wiphy_rfkill_stop_polling(sc->hw->wiphy);
         ath_deinit_leds(sc);
 
@@ -832,6 +830,7 @@ void ath9k_deinit_device(struct ath_softc *sc)
         }
 
         ieee80211_unregister_hw(hw);
+        pm_qos_remove_request(&ath9k_pm_qos_req);
         ath_rx_cleanup(sc);
         ath_tx_cleanup(sc);
         ath9k_deinit_softc(sc);
diff --git a/drivers/net/wireless/ath/ath9k/reg.h b/drivers/net/wireless/ath/ath9k/reg.h
index fa05b711e5cd..dddf579aacf1 100644
--- a/drivers/net/wireless/ath/ath9k/reg.h
+++ b/drivers/net/wireless/ath/ath9k/reg.h
@@ -866,7 +866,13 @@
 #define AR_DEVID_7010(_ah) \
         (((_ah)->hw_version.devid == 0x7010) || \
          ((_ah)->hw_version.devid == 0x7015) || \
-         ((_ah)->hw_version.devid == 0x9018))
+         ((_ah)->hw_version.devid == 0x9018) || \
+         ((_ah)->hw_version.devid == 0xA704) || \
+         ((_ah)->hw_version.devid == 0x1200))
+
+#define AR9287_HTC_DEVID(_ah) \
+        (((_ah)->hw_version.devid == 0x7015) || \
+         ((_ah)->hw_version.devid == 0x1200))
 
 #define AR_RADIO_SREV_MAJOR                   0xf0
 #define AR_RAD5133_SREV_MAJOR                 0xc0
diff --git a/drivers/net/wireless/ath/carl9170/usb.c b/drivers/net/wireless/ath/carl9170/usb.c
index 3317039cd28f..7504ed14c725 100644
--- a/drivers/net/wireless/ath/carl9170/usb.c
+++ b/drivers/net/wireless/ath/carl9170/usb.c
@@ -553,12 +553,12 @@ static int carl9170_usb_flush(struct ar9170 *ar)
                 usb_free_urb(urb);
         }
 
-        ret = usb_wait_anchor_empty_timeout(&ar->tx_cmd, HZ);
+        ret = usb_wait_anchor_empty_timeout(&ar->tx_cmd, 1000);
         if (ret == 0)
                 err = -ETIMEDOUT;
 
         /* lets wait a while until the tx - queues are dried out */
-        ret = usb_wait_anchor_empty_timeout(&ar->tx_anch, HZ);
+        ret = usb_wait_anchor_empty_timeout(&ar->tx_anch, 1000);
         if (ret == 0)
                 err = -ETIMEDOUT;
 
diff --git a/drivers/net/wireless/orinoco/orinoco_usb.c b/drivers/net/wireless/orinoco/orinoco_usb.c
index a38a7bd25f19..b9aedf18a046 100644
--- a/drivers/net/wireless/orinoco/orinoco_usb.c
+++ b/drivers/net/wireless/orinoco/orinoco_usb.c
@@ -57,7 +57,6 @@
 #include <linux/fcntl.h>
 #include <linux/spinlock.h>
 #include <linux/list.h>
-#include <linux/smp_lock.h>
 #include <linux/usb.h>
 #include <linux/timer.h>
 
diff --git a/drivers/parisc/eisa_eeprom.c b/drivers/parisc/eisa_eeprom.c
index cce00ed81f37..af212c6a6158 100644
--- a/drivers/parisc/eisa_eeprom.c
+++ b/drivers/parisc/eisa_eeprom.c
@@ -24,7 +24,6 @@
 #include <linux/kernel.h>
 #include <linux/miscdevice.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/fs.h>
 #include <asm/io.h>
 #include <asm/uaccess.h>
diff --git a/drivers/pci/Makefile b/drivers/pci/Makefile
index f01e344cf4bd..98e6fdf34d30 100644
--- a/drivers/pci/Makefile
+++ b/drivers/pci/Makefile
@@ -49,6 +49,7 @@ obj-$(CONFIG_MIPS) += setup-bus.o setup-irq.o
 obj-$(CONFIG_X86_VISWS) += setup-irq.o
 obj-$(CONFIG_MN10300) += setup-bus.o
 obj-$(CONFIG_MICROBLAZE) += setup-bus.o
+obj-$(CONFIG_TILE) += setup-bus.o setup-irq.o
 
 #
 # ACPI Related PCI FW Functions
diff --git a/drivers/pci/bus.c b/drivers/pci/bus.c
index 5624db8c9ad0..003170ea2e39 100644
--- a/drivers/pci/bus.c
+++ b/drivers/pci/bus.c
@@ -64,17 +64,57 @@ void pci_bus_remove_resources(struct pci_bus *bus)
         }
 }
 
+static bool pci_bus_resource_better(struct resource *res1, bool pos1,
+                                    struct resource *res2, bool pos2)
+{
+        /* If exactly one is positive decode, always prefer that one */
+        if (pos1 != pos2)
+                return pos1 ? true : false;
+
+        /* Prefer the one that contains the highest address */
+        if (res1->end != res2->end)
+                return (res1->end > res2->end) ? true : false;
+
+        /* Otherwise, prefer the one with highest "center of gravity" */
+        if (res1->start != res2->start)
+                return (res1->start > res2->start) ? true : false;
+
+        /* Otherwise, choose one arbitrarily (but consistently) */
+        return (res1 > res2) ? true : false;
+}
+
+static bool pci_bus_resource_positive(struct pci_bus *bus, struct resource *res)
+{
+        struct pci_bus_resource *bus_res;
+
+        /*
+         * This relies on the fact that pci_bus.resource[] refers to P2P or
+         * CardBus bridge base/limit registers, which are always positively
+         * decoded.  The pci_bus.resources list contains host bridge or
+         * subtractively decoded resources.
+         */
+        list_for_each_entry(bus_res, &bus->resources, list) {
+                if (bus_res->res == res)
+                        return (bus_res->flags & PCI_SUBTRACTIVE_DECODE) ?
+                                false : true;
+        }
+        return true;
+}
+
 /*
- * Find the highest-address bus resource below the cursor "res".  If the
- * cursor is NULL, return the highest resource.
+ * Find the next-best bus resource after the cursor "res".  If the cursor is
+ * NULL, return the best resource.  "Best" means that we prefer positive
+ * decode regions over subtractive decode, then those at higher addresses.
  */
 static struct resource *pci_bus_find_resource_prev(struct pci_bus *bus,
                                                    unsigned int type,
                                                    struct resource *res)
 {
+        bool res_pos, r_pos, prev_pos = false;
         struct resource *r, *prev = NULL;
         int i;
 
+        res_pos = pci_bus_resource_positive(bus, res);
         pci_bus_for_each_resource(bus, r, i) {
                 if (!r)
                         continue;
@@ -82,26 +122,14 @@ static struct resource *pci_bus_find_resource_prev(struct pci_bus *bus,
                 if ((r->flags & IORESOURCE_TYPE_BITS) != type)
                         continue;
 
-                /* If this resource is at or past the cursor, skip it */
-                if (res) {
-                        if (r == res)
-                                continue;
-                        if (r->end > res->end)
-                                continue;
-                        if (r->end == res->end && r->start > res->start)
-                                continue;
+                r_pos = pci_bus_resource_positive(bus, r);
+                if (!res || pci_bus_resource_better(res, res_pos, r, r_pos)) {
+                        if (!prev || pci_bus_resource_better(r, r_pos,
+                                                             prev, prev_pos)) {
+                                prev = r;
+                                prev_pos = r_pos;
+                        }
                 }
-
-                if (!prev)
-                        prev = r;
-
-                /*
-                 * A small resource is higher than a large one that ends at
-                 * the same address.
-                 */
-                if (r->end > prev->end ||
-                    (r->end == prev->end && r->start > prev->start))
-                        prev = r;
         }
 
         return prev;
diff --git a/drivers/pci/hotplug/ibmphp_ebda.c b/drivers/pci/hotplug/ibmphp_ebda.c
index 5becbdee4027..2850e64dedae 100644
--- a/drivers/pci/hotplug/ibmphp_ebda.c
+++ b/drivers/pci/hotplug/ibmphp_ebda.c
@@ -276,6 +276,12 @@ int __init ibmphp_access_ebda (void)
 
         for (;;) {
                 offset = next_offset;
+
+                /* Make sure what we read is still in the mapped section */
+                if (WARN(offset > (ebda_sz * 1024 - 4),
+                         "ibmphp_ebda: next read is beyond ebda_sz\n"))
+                        break;
+
                 next_offset = readw (io_mem + offset);  /* offset of next blk */
 
                 offset += 2;
diff --git a/drivers/pci/pci-sysfs.c b/drivers/pci/pci-sysfs.c
index b5a7d9bfcb24..63d5042f2079 100644
--- a/drivers/pci/pci-sysfs.c
+++ b/drivers/pci/pci-sysfs.c
@@ -705,17 +705,21 @@ void pci_remove_legacy_files(struct pci_bus *b)
 
 #ifdef HAVE_PCI_MMAP
 
-int pci_mmap_fits(struct pci_dev *pdev, int resno, struct vm_area_struct *vma)
+int pci_mmap_fits(struct pci_dev *pdev, int resno, struct vm_area_struct *vma,
+                  enum pci_mmap_api mmap_api)
 {
-        unsigned long nr, start, size;
+        unsigned long nr, start, size, pci_start;
 
+        if (pci_resource_len(pdev, resno) == 0)
+                return 0;
         nr = (vma->vm_end - vma->vm_start) >> PAGE_SHIFT;
         start = vma->vm_pgoff;
         size = ((pci_resource_len(pdev, resno) - 1) >> PAGE_SHIFT) + 1;
-        if (start < size && size - start >= nr)
+        pci_start = (mmap_api == PCI_MMAP_PROCFS) ?
+                        pci_resource_start(pdev, resno) >> PAGE_SHIFT : 0;
+        if (start >= pci_start && start < pci_start + size &&
+                        start + nr <= pci_start + size)
                 return 1;
-        WARN(1, "process \"%s\" tried to map 0x%08lx-0x%08lx on %s BAR %d (size 0x%08lx)\n",
-                current->comm, start, start+nr, pci_name(pdev), resno, size);
         return 0;
 }
 
@@ -745,8 +749,15 @@ pci_mmap_resource(struct kobject *kobj, struct bin_attribute *attr,
         if (i >= PCI_ROM_RESOURCE)
                 return -ENODEV;
 
-        if (!pci_mmap_fits(pdev, i, vma))
+        if (!pci_mmap_fits(pdev, i, vma, PCI_MMAP_SYSFS)) {
+                WARN(1, "process \"%s\" tried to map 0x%08lx bytes "
+                        "at page 0x%08lx on %s BAR %d (start 0x%16Lx, size 0x%16Lx)\n",
+                        current->comm, vma->vm_end-vma->vm_start, vma->vm_pgoff,
+                        pci_name(pdev), i,
+                        (u64)pci_resource_start(pdev, i),
+                        (u64)pci_resource_len(pdev, i));
                 return -EINVAL;
+        }
 
         /* pci_mmap_page_range() expects the same kind of entry as coming
          * from /proc/bus/pci/ which is a "user visible" value. If this is
diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index e98c8104297b..710c8a29be0d 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -1007,6 +1007,18 @@ static int __pci_enable_device_flags(struct pci_dev *dev,
         int err;
         int i, bars = 0;
 
+        /*
+         * Power state could be unknown at this point, either due to a fresh
+         * boot or a device removal call.  So get the current power state
+         * so that things like MSI message writing will behave as expected
+         * (e.g. if the device really is in D0 at enable time).
+         */
+        if (dev->pm_cap) {
+                u16 pmcsr;
+                pci_read_config_word(dev, dev->pm_cap + PCI_PM_CTRL, &pmcsr);
+                dev->current_state = (pmcsr & PCI_PM_CTRL_STATE_MASK);
+        }
+
         if (atomic_add_return(1, &dev->enable_cnt) > 1)
                 return 0;               /* already enabled */
 
diff --git a/drivers/pci/pci.h b/drivers/pci/pci.h
index f5c7c382765f..7d33f6673868 100644
--- a/drivers/pci/pci.h
+++ b/drivers/pci/pci.h
@@ -22,8 +22,13 @@ extern void pci_remove_firmware_label_files(struct pci_dev *pdev);
 #endif
 extern void pci_cleanup_rom(struct pci_dev *dev);
 #ifdef HAVE_PCI_MMAP
+enum pci_mmap_api {
+        PCI_MMAP_SYSFS, /* mmap on /sys/bus/pci/devices/<BDF>/resource<N> */
+        PCI_MMAP_PROCFS /* mmap on /proc/bus/pci/<BDF> */
+};
 extern int pci_mmap_fits(struct pci_dev *pdev, int resno,
-                         struct vm_area_struct *vma);
+                         struct vm_area_struct *vmai,
+                         enum pci_mmap_api mmap_api);
 #endif
 int pci_probe_reset_function(struct pci_dev *dev);
 
diff --git a/drivers/pci/proc.c b/drivers/pci/proc.c
index 297b72c880a1..27911b55c2a5 100644
--- a/drivers/pci/proc.c
+++ b/drivers/pci/proc.c
@@ -10,7 +10,6 @@
 #include <linux/module.h>
 #include <linux/proc_fs.h>
 #include <linux/seq_file.h>
-#include <linux/smp_lock.h>
 #include <linux/capability.h>
 #include <asm/uaccess.h>
 #include <asm/byteorder.h>
@@ -257,7 +256,7 @@ static int proc_bus_pci_mmap(struct file *file, struct vm_area_struct *vma)
 
         /* Make sure the caller is mapping a real resource for this device */
         for (i = 0; i < PCI_ROM_RESOURCE; i++) {
-                if (pci_mmap_fits(dev, i, vma))
+                if (pci_mmap_fits(dev, i, vma,  PCI_MMAP_PROCFS))
                         break;
         }
 
diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
index f5c63fe9db5c..6f9350cabbd5 100644
--- a/drivers/pci/quirks.c
+++ b/drivers/pci/quirks.c
@@ -2136,6 +2136,24 @@ DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82865_HB,
 DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_82875_HB,
                         quirk_unhide_mch_dev6);
 
+#ifdef CONFIG_TILE
+/*
+ * The Tilera TILEmpower platform needs to set the link speed
+ * to 2.5GT(Giga-Transfers)/s (Gen 1). The default link speed
+ * setting is 5GT/s (Gen 2). 0x98 is the Link Control2 PCIe
+ * capability register of the PEX8624 PCIe switch. The switch
+ * supports link speed auto negotiation, but falsely sets
+ * the link speed to 5GT/s.
+ */
+static void __devinit quirk_tile_plx_gen1(struct pci_dev *dev)
+{
+        if (tile_plx_gen1) {
+                pci_write_config_dword(dev, 0x98, 0x1);
+                mdelay(50);
+        }
+}
+DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_PLX, 0x8624, quirk_tile_plx_gen1);
+#endif /* CONFIG_TILE */
 
 #ifdef CONFIG_PCI_MSI
 /* Some chipsets do not support MSI. We cannot easily rely on setting
diff --git a/drivers/pnp/isapnp/proc.c b/drivers/pnp/isapnp/proc.c
index e73ebefdf3e0..315b3112aca8 100644
--- a/drivers/pnp/isapnp/proc.c
+++ b/drivers/pnp/isapnp/proc.c
@@ -21,7 +21,6 @@
 #include <linux/isapnp.h>
 #include <linux/proc_fs.h>
 #include <linux/init.h>
-#include <linux/smp_lock.h>
 #include <asm/uaccess.h>
 
 extern struct pnp_protocol isapnp_protocol;
diff --git a/drivers/rtc/rtc-sh.c b/drivers/rtc/rtc-sh.c
index 5efbd5990ff8..06e41ed93230 100644
--- a/drivers/rtc/rtc-sh.c
+++ b/drivers/rtc/rtc-sh.c
@@ -761,7 +761,7 @@ err_unmap:
         clk_put(rtc->clk);
         iounmap(rtc->regbase);
 err_badmap:
-        release_resource(rtc->res);
+        release_mem_region(rtc->res->start, rtc->regsize);
 err_badres:
         kfree(rtc);
 
@@ -786,7 +786,7 @@ static int __exit sh_rtc_remove(struct platform_device *pdev)
         }
 
         iounmap(rtc->regbase);
-        release_resource(rtc->res);
+        release_mem_region(rtc->res->start, rtc->regsize);
 
         clk_disable(rtc->clk);
         clk_put(rtc->clk);
diff --git a/drivers/s390/block/dasd_eer.c b/drivers/s390/block/dasd_eer.c
index c71d89dba302..83b4615a3b62 100644
--- a/drivers/s390/block/dasd_eer.c
+++ b/drivers/s390/block/dasd_eer.c
@@ -17,7 +17,6 @@
 #include <linux/device.h>
 #include <linux/poll.h>
 #include <linux/mutex.h>
-#include <linux/smp_lock.h>
 #include <linux/err.h>
 #include <linux/slab.h>
 
diff --git a/drivers/s390/char/fs3270.c b/drivers/s390/char/fs3270.c
index eb28fb01a38a..f6489eb7e976 100644
--- a/drivers/s390/char/fs3270.c
+++ b/drivers/s390/char/fs3270.c
@@ -14,7 +14,6 @@
 #include <linux/list.h>
 #include <linux/slab.h>
 #include <linux/types.h>
-#include <linux/smp_lock.h>
 
 #include <asm/compat.h>
 #include <asm/ccwdev.h>
diff --git a/drivers/s390/char/tape_char.c b/drivers/s390/char/tape_char.c
index 883e2db02bd3..e090a307fdee 100644
--- a/drivers/s390/char/tape_char.c
+++ b/drivers/s390/char/tape_char.c
@@ -17,7 +17,6 @@
 #include <linux/types.h>
 #include <linux/proc_fs.h>
 #include <linux/mtio.h>
-#include <linux/smp_lock.h>
 #include <linux/compat.h>
 
 #include <asm/uaccess.h>
diff --git a/drivers/s390/char/tape_core.c b/drivers/s390/char/tape_core.c
index 6c408670e08d..b3a3e8e8656e 100644
--- a/drivers/s390/char/tape_core.c
+++ b/drivers/s390/char/tape_core.c
@@ -209,29 +209,79 @@ tape_state_set(struct tape_device *device, enum tape_state newstate)
         wake_up(&device->state_change_wq);
 }
 
+struct tape_med_state_work_data {
+        struct tape_device *device;
+        enum tape_medium_state state;
+        struct work_struct  work;
+};
+
+static void
+tape_med_state_work_handler(struct work_struct *work)
+{
+        static char env_state_loaded[] = "MEDIUM_STATE=LOADED";
+        static char env_state_unloaded[] = "MEDIUM_STATE=UNLOADED";
+        struct tape_med_state_work_data *p =
+                container_of(work, struct tape_med_state_work_data, work);
+        struct tape_device *device = p->device;
+        char *envp[] = { NULL, NULL };
+
+        switch (p->state) {
+        case MS_UNLOADED:
+                pr_info("%s: The tape cartridge has been successfully "
+                        "unloaded\n", dev_name(&device->cdev->dev));
+                envp[0] = env_state_unloaded;
+                kobject_uevent_env(&device->cdev->dev.kobj, KOBJ_CHANGE, envp);
+                break;
+        case MS_LOADED:
+                pr_info("%s: A tape cartridge has been mounted\n",
+                        dev_name(&device->cdev->dev));
+                envp[0] = env_state_loaded;
+                kobject_uevent_env(&device->cdev->dev.kobj, KOBJ_CHANGE, envp);
+                break;
+        default:
+                break;
+        }
+        tape_put_device(device);
+        kfree(p);
+}
+
+static void
+tape_med_state_work(struct tape_device *device, enum tape_medium_state state)
+{
+        struct tape_med_state_work_data *p;
+
+        p = kzalloc(sizeof(*p), GFP_ATOMIC);
+        if (p) {
+                INIT_WORK(&p->work, tape_med_state_work_handler);
+                p->device = tape_get_device(device);
+                p->state = state;
+                schedule_work(&p->work);
+        }
+}
+
 void
 tape_med_state_set(struct tape_device *device, enum tape_medium_state newstate)
 {
-        if (device->medium_state == newstate)
+        enum tape_medium_state oldstate;
+
+        oldstate = device->medium_state;
+        if (oldstate == newstate)
                 return;
+        device->medium_state = newstate;
         switch(newstate){
         case MS_UNLOADED:
                 device->tape_generic_status |= GMT_DR_OPEN(~0);
-                if (device->medium_state == MS_LOADED)
-                        pr_info("%s: The tape cartridge has been successfully "
-                                "unloaded\n", dev_name(&device->cdev->dev));
+                if (oldstate == MS_LOADED)
+                        tape_med_state_work(device, MS_UNLOADED);
                 break;
         case MS_LOADED:
                 device->tape_generic_status &= ~GMT_DR_OPEN(~0);
-                if (device->medium_state == MS_UNLOADED)
-                        pr_info("%s: A tape cartridge has been mounted\n",
-                                dev_name(&device->cdev->dev));
+                if (oldstate == MS_UNLOADED)
+                        tape_med_state_work(device, MS_LOADED);
                 break;
         default:
-                // print nothing
                 break;
         }
-        device->medium_state = newstate;
         wake_up(&device->state_change_wq);
 }
 
diff --git a/drivers/s390/char/vmlogrdr.c b/drivers/s390/char/vmlogrdr.c
index 9f661426e4a1..c837d7419a6a 100644
--- a/drivers/s390/char/vmlogrdr.c
+++ b/drivers/s390/char/vmlogrdr.c
@@ -30,7 +30,6 @@
 #include <linux/kmod.h>
 #include <linux/cdev.h>
 #include <linux/device.h>
-#include <linux/smp_lock.h>
 #include <linux/string.h>
 
 MODULE_AUTHOR
@@ -249,27 +248,25 @@ static int vmlogrdr_recording(struct vmlogrdr_priv_t * logptr,
         char cp_command[80];
         char cp_response[160];
         char *onoff, *qid_string;
+        int rc;
 
-        memset(cp_command, 0x00, sizeof(cp_command));
-        memset(cp_response, 0x00, sizeof(cp_response));
-
-        onoff = ((action == 1) ? "ON" : "OFF");
+        onoff = ((action == 1) ? "ON" : "OFF");
         qid_string = ((recording_class_AB == 1) ? " QID * " : "");
 
-        /*
+        /*
          * The recording commands needs to be called with option QID
          * for guests that have previlege classes A or B.
          * Purging has to be done as separate step, because recording
          * can't be switched on as long as records are on the queue.
          * Doing both at the same time doesn't work.
          */
-
-        if (purge) {
+        if (purge && (action == 1)) {
+                memset(cp_command, 0x00, sizeof(cp_command));
+                memset(cp_response, 0x00, sizeof(cp_response));
                 snprintf(cp_command, sizeof(cp_command),
                          "RECORDING %s PURGE %s",
                          logptr->recording_name,
                          qid_string);
-
                 cpcmd(cp_command, cp_response, sizeof(cp_response), NULL);
         }
 
@@ -279,19 +276,33 @@ static int vmlogrdr_recording(struct vmlogrdr_priv_t * logptr,
                 logptr->recording_name,
                 onoff,
                 qid_string);
-
         cpcmd(cp_command, cp_response, sizeof(cp_response), NULL);
         /* The recording command will usually answer with 'Command complete'
          * on success, but when the specific service was never connected
          * before then there might be an additional informational message
          * 'HCPCRC8072I Recording entry not found' before the
-         * 'Command complete'. So I use strstr rather then the strncmp.
+         * 'Command complete'. So I use strstr rather then the strncmp.
          */
         if (strstr(cp_response,"Command complete"))
-                return 0;
+                rc = 0;
         else
-                return -EIO;
+                rc = -EIO;
+        /*
+         * If we turn recording off, we have to purge any remaining records
+         * afterwards, as a large number of queued records may impact z/VM
+         * performance.
+         */
+        if (purge && (action == 0)) {
+                memset(cp_command, 0x00, sizeof(cp_command));
+                memset(cp_response, 0x00, sizeof(cp_response));
+                snprintf(cp_command, sizeof(cp_command),
+                         "RECORDING %s PURGE %s",
+                         logptr->recording_name,
+                         qid_string);
+                cpcmd(cp_command, cp_response, sizeof(cp_response), NULL);
+        }
 
+        return rc;
 }
 
 
diff --git a/drivers/s390/char/vmur.c b/drivers/s390/char/vmur.c
index 1de672f21037..f7e4ae6bf15a 100644
--- a/drivers/s390/char/vmur.c
+++ b/drivers/s390/char/vmur.c
@@ -13,7 +13,6 @@
 
 #include <linux/cdev.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 
 #include <asm/uaccess.h>
 #include <asm/cio.h>
diff --git a/drivers/s390/cio/device.c b/drivers/s390/cio/device.c
index 2ff8a22d4257..e8391b89eff4 100644
--- a/drivers/s390/cio/device.c
+++ b/drivers/s390/cio/device.c
@@ -1455,7 +1455,16 @@ static int io_subchannel_sch_event(struct subchannel *sch, int process)
                 break;
         case IO_SCH_UNREG_ATTACH:
         case IO_SCH_UNREG:
-                if (cdev)
+                if (!cdev)
+                        break;
+                if (cdev->private->state == DEV_STATE_SENSE_ID) {
+                        /*
+                         * Note: delayed work triggered by this event
+                         * and repeated calls to sch_event are synchronized
+                         * by the above check for work_pending(cdev).
+                         */
+                        dev_fsm_event(cdev, DEV_EVENT_NOTOPER);
+                } else
                         ccw_device_set_notoper(cdev);
                 break;
         case IO_SCH_NOP:
diff --git a/drivers/s390/crypto/zcrypt_api.c b/drivers/s390/crypto/zcrypt_api.c
index f5221749d180..7fca9c10ffcf 100644
--- a/drivers/s390/crypto/zcrypt_api.c
+++ b/drivers/s390/crypto/zcrypt_api.c
@@ -35,7 +35,6 @@
 #include <linux/proc_fs.h>
 #include <linux/seq_file.h>
 #include <linux/compat.h>
-#include <linux/smp_lock.h>
 #include <linux/slab.h>
 #include <asm/atomic.h>
 #include <asm/uaccess.h>
diff --git a/drivers/s390/scsi/zfcp_scsi.c b/drivers/s390/scsi/zfcp_scsi.c
index 50286d8707f3..6bd2dbc4c316 100644
--- a/drivers/s390/scsi/zfcp_scsi.c
+++ b/drivers/s390/scsi/zfcp_scsi.c
@@ -76,7 +76,7 @@ static void zfcp_scsi_command_fail(struct scsi_cmnd *scpnt, int result)
         scpnt->scsi_done(scpnt);
 }
 
-static int zfcp_scsi_queuecommand(struct scsi_cmnd *scpnt,
+static int zfcp_scsi_queuecommand_lck(struct scsi_cmnd *scpnt,
                                   void (*done) (struct scsi_cmnd *))
 {
         struct zfcp_scsi_dev *zfcp_sdev = sdev_to_zfcp(scpnt->device);
@@ -127,6 +127,8 @@ static int zfcp_scsi_queuecommand(struct scsi_cmnd *scpnt,
         return ret;
 }
 
+static DEF_SCSI_QCMD(zfcp_scsi_queuecommand)
+
 static int zfcp_scsi_slave_alloc(struct scsi_device *sdev)
 {
         struct fc_rport *rport = starget_to_rport(scsi_target(sdev));
diff --git a/drivers/scsi/3w-9xxx.c b/drivers/scsi/3w-9xxx.c
index fcf08b3f52c1..b7bd5b0cc7aa 100644
--- a/drivers/scsi/3w-9xxx.c
+++ b/drivers/scsi/3w-9xxx.c
@@ -1765,7 +1765,7 @@ out:
 } /* End twa_scsi_eh_reset() */
 
 /* This is the main scsi queue function to handle scsi opcodes */
-static int twa_scsi_queue(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
+static int twa_scsi_queue_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
 {
         int request_id, retval;
         TW_Device_Extension *tw_dev = (TW_Device_Extension *)SCpnt->device->host->hostdata;
@@ -1812,6 +1812,8 @@ out:
         return retval;
 } /* End twa_scsi_queue() */
 
+static DEF_SCSI_QCMD(twa_scsi_queue)
+
 /* This function hands scsi cdb's to the firmware */
 static int twa_scsiop_execute_scsi(TW_Device_Extension *tw_dev, int request_id, char *cdb, int use_sg, TW_SG_Entry *sglistarg)
 {
diff --git a/drivers/scsi/3w-sas.c b/drivers/scsi/3w-sas.c
index 6a95d111d207..13e39e1fdfe2 100644
--- a/drivers/scsi/3w-sas.c
+++ b/drivers/scsi/3w-sas.c
@@ -1501,7 +1501,7 @@ out:
 } /* End twl_scsi_eh_reset() */
 
 /* This is the main scsi queue function to handle scsi opcodes */
-static int twl_scsi_queue(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
+static int twl_scsi_queue_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
 {
         int request_id, retval;
         TW_Device_Extension *tw_dev = (TW_Device_Extension *)SCpnt->device->host->hostdata;
@@ -1536,6 +1536,8 @@ out:
         return retval;
 } /* End twl_scsi_queue() */
 
+static DEF_SCSI_QCMD(twl_scsi_queue)
+
 /* This function tells the controller to shut down */
 static void __twl_shutdown(TW_Device_Extension *tw_dev)
 {
diff --git a/drivers/scsi/3w-xxxx.c b/drivers/scsi/3w-xxxx.c
index b1125341f4c8..7fe96ff60c58 100644
--- a/drivers/scsi/3w-xxxx.c
+++ b/drivers/scsi/3w-xxxx.c
@@ -1947,7 +1947,7 @@ static int tw_scsiop_test_unit_ready_complete(TW_Device_Extension *tw_dev, int r
 } /* End tw_scsiop_test_unit_ready_complete() */
 
 /* This is the main scsi queue function to handle scsi opcodes */
-static int tw_scsi_queue(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *)) 
+static int tw_scsi_queue_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
 {
         unsigned char *command = SCpnt->cmnd;
         int request_id = 0;
@@ -2023,6 +2023,8 @@ static int tw_scsi_queue(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd
         return retval;
 } /* End tw_scsi_queue() */
 
+static DEF_SCSI_QCMD(tw_scsi_queue)
+
 /* This function is the interrupt service routine */
 static irqreturn_t tw_interrupt(int irq, void *dev_instance) 
 {
diff --git a/drivers/scsi/53c700.c b/drivers/scsi/53c700.c
index 89fc1c8af86b..f672491774eb 100644
--- a/drivers/scsi/53c700.c
+++ b/drivers/scsi/53c700.c
@@ -167,7 +167,7 @@ MODULE_LICENSE("GPL");
 #include "53c700_d.h"
 
 
-STATIC int NCR_700_queuecommand(struct scsi_cmnd *, void (*done)(struct scsi_cmnd *));
+STATIC int NCR_700_queuecommand(struct Scsi_Host *h, struct scsi_cmnd *);
 STATIC int NCR_700_abort(struct scsi_cmnd * SCpnt);
 STATIC int NCR_700_bus_reset(struct scsi_cmnd * SCpnt);
 STATIC int NCR_700_host_reset(struct scsi_cmnd * SCpnt);
@@ -1749,8 +1749,8 @@ NCR_700_intr(int irq, void *dev_id)
         return IRQ_RETVAL(handled);
 }
 
-STATIC int
-NCR_700_queuecommand(struct scsi_cmnd *SCp, void (*done)(struct scsi_cmnd *))
+static int
+NCR_700_queuecommand_lck(struct scsi_cmnd *SCp, void (*done)(struct scsi_cmnd *))
 {
         struct NCR_700_Host_Parameters *hostdata = 
                 (struct NCR_700_Host_Parameters *)SCp->device->host->hostdata[0];
@@ -1904,6 +1904,8 @@ NCR_700_queuecommand(struct scsi_cmnd *SCp, void (*done)(struct scsi_cmnd *))
         return 0;
 }
 
+STATIC DEF_SCSI_QCMD(NCR_700_queuecommand)
+
 STATIC int
 NCR_700_abort(struct scsi_cmnd * SCp)
 {
diff --git a/drivers/scsi/BusLogic.c b/drivers/scsi/BusLogic.c
index fc0b4b81d552..f66c33b9ab41 100644
--- a/drivers/scsi/BusLogic.c
+++ b/drivers/scsi/BusLogic.c
@@ -2807,7 +2807,7 @@ static int BusLogic_host_reset(struct scsi_cmnd * SCpnt)
   Outgoing Mailbox for execution by the associated Host Adapter.
 */
 
-static int BusLogic_QueueCommand(struct scsi_cmnd *Command, void (*CompletionRoutine) (struct scsi_cmnd *))
+static int BusLogic_QueueCommand_lck(struct scsi_cmnd *Command, void (*CompletionRoutine) (struct scsi_cmnd *))
 {
         struct BusLogic_HostAdapter *HostAdapter = (struct BusLogic_HostAdapter *) Command->device->host->hostdata;
         struct BusLogic_TargetFlags *TargetFlags = &HostAdapter->TargetFlags[Command->device->id];
@@ -2994,6 +2994,7 @@ static int BusLogic_QueueCommand(struct scsi_cmnd *Command, void (*CompletionRou
         return 0;
 }
 
+static DEF_SCSI_QCMD(BusLogic_QueueCommand)
 
 #if 0
 /*
diff --git a/drivers/scsi/BusLogic.h b/drivers/scsi/BusLogic.h
index 73f237a1ed94..649fcb31f26d 100644
--- a/drivers/scsi/BusLogic.h
+++ b/drivers/scsi/BusLogic.h
@@ -1319,7 +1319,7 @@ static inline void BusLogic_IncrementSizeBucket(BusLogic_CommandSizeBuckets_T Co
 */
 
 static const char *BusLogic_DriverInfo(struct Scsi_Host *);
-static int BusLogic_QueueCommand(struct scsi_cmnd *, void (*CompletionRoutine) (struct scsi_cmnd *));
+static int BusLogic_QueueCommand(struct Scsi_Host *h, struct scsi_cmnd *);
 static int BusLogic_BIOSDiskParameters(struct scsi_device *, struct block_device *, sector_t, int *);
 static int BusLogic_ProcDirectoryInfo(struct Scsi_Host *, char *, char **, off_t, int, int);
 static int BusLogic_SlaveConfigure(struct scsi_device *);
diff --git a/drivers/scsi/NCR5380.c b/drivers/scsi/NCR5380.c
index 5d2f148889ad..9a5629f94f95 100644
--- a/drivers/scsi/NCR5380.c
+++ b/drivers/scsi/NCR5380.c
@@ -952,7 +952,7 @@ static void NCR5380_exit(struct Scsi_Host *instance)
  *      Locks: host lock taken by caller
  */
 
-static int NCR5380_queue_command(Scsi_Cmnd * cmd, void (*done) (Scsi_Cmnd *)) 
+static int NCR5380_queue_command_lck(Scsi_Cmnd * cmd, void (*done) (Scsi_Cmnd *))
 {
         struct Scsi_Host *instance = cmd->device->host;
         struct NCR5380_hostdata *hostdata = (struct NCR5380_hostdata *) instance->hostdata;
@@ -1021,6 +1021,7 @@ static int NCR5380_queue_command(Scsi_Cmnd * cmd, void (*done) (Scsi_Cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(NCR5380_queue_command)
 
 /**
  *      NCR5380_main    -       NCR state machines
diff --git a/drivers/scsi/NCR5380.h b/drivers/scsi/NCR5380.h
index bdc468c9e1d9..fd40a32b1f6f 100644
--- a/drivers/scsi/NCR5380.h
+++ b/drivers/scsi/NCR5380.h
@@ -313,7 +313,7 @@ static void NCR5380_print(struct Scsi_Host *instance);
 #endif
 static int NCR5380_abort(Scsi_Cmnd * cmd);
 static int NCR5380_bus_reset(Scsi_Cmnd * cmd);
-static int NCR5380_queue_command(Scsi_Cmnd * cmd, void (*done) (Scsi_Cmnd *));
+static int NCR5380_queue_command(struct Scsi_Host *, struct scsi_cmnd *);
 static int __maybe_unused NCR5380_proc_info(struct Scsi_Host *instance,
         char *buffer, char **start, off_t offset, int length, int inout);
 
diff --git a/drivers/scsi/NCR53c406a.c b/drivers/scsi/NCR53c406a.c
index 6961f78742ae..c91888a0a23c 100644
--- a/drivers/scsi/NCR53c406a.c
+++ b/drivers/scsi/NCR53c406a.c
@@ -693,7 +693,7 @@ static void wait_intr(void)
 }
 #endif
 
-static int NCR53c406a_queue(Scsi_Cmnd * SCpnt, void (*done) (Scsi_Cmnd *))
+static int NCR53c406a_queue_lck(Scsi_Cmnd * SCpnt, void (*done) (Scsi_Cmnd *))
 {
         int i;
 
@@ -726,6 +726,8 @@ static int NCR53c406a_queue(Scsi_Cmnd * SCpnt, void (*done) (Scsi_Cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(NCR53c406a_queue)
+
 static int NCR53c406a_host_reset(Scsi_Cmnd * SCpnt)
 {
         DEB(printk("NCR53c406a_reset called\n"));
diff --git a/drivers/scsi/a100u2w.c b/drivers/scsi/a100u2w.c
index dbbc601948e5..dc5ac6e528c4 100644
--- a/drivers/scsi/a100u2w.c
+++ b/drivers/scsi/a100u2w.c
@@ -911,7 +911,7 @@ static int inia100_build_scb(struct orc_host * host, struct orc_scb * scb, struc
  *      queue the command down to the controller
  */
 
-static int inia100_queue(struct scsi_cmnd * cmd, void (*done) (struct scsi_cmnd *))
+static int inia100_queue_lck(struct scsi_cmnd * cmd, void (*done) (struct scsi_cmnd *))
 {
         struct orc_scb *scb;
         struct orc_host *host;          /* Point to Host adapter control block */
@@ -930,6 +930,8 @@ static int inia100_queue(struct scsi_cmnd * cmd, void (*done) (struct scsi_cmnd
         return 0;
 }
 
+static DEF_SCSI_QCMD(inia100_queue)
+
 /*****************************************************************************
  Function name  : inia100_abort
  Description    : Abort a queued command.
diff --git a/drivers/scsi/aacraid/linit.c b/drivers/scsi/aacraid/linit.c
index 29c0ed1cf507..2c93d9496d62 100644
--- a/drivers/scsi/aacraid/linit.c
+++ b/drivers/scsi/aacraid/linit.c
@@ -248,7 +248,7 @@ static struct aac_driver_ident aac_drivers[] = {
  *      TODO: unify with aac_scsi_cmd().
  */
 
-static int aac_queuecommand(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
+static int aac_queuecommand_lck(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 {
         struct Scsi_Host *host = cmd->device->host;
         struct aac_dev *dev = (struct aac_dev *)host->hostdata;
@@ -267,6 +267,8 @@ static int aac_queuecommand(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd
         return (aac_scsi_cmd(cmd) ? FAILED : 0);
 }
 
+static DEF_SCSI_QCMD(aac_queuecommand)
+
 /**
  *      aac_info                -       Returns the host adapter name
  *      @shost:         Scsi host to report on
diff --git a/drivers/scsi/advansys.c b/drivers/scsi/advansys.c
index 0ec3da6f3e12..081c6de92bc5 100644
--- a/drivers/scsi/advansys.c
+++ b/drivers/scsi/advansys.c
@@ -9500,7 +9500,7 @@ static int asc_execute_scsi_cmnd(struct scsi_cmnd *scp)
  * in the 'scp' result field.
  */
 static int
-advansys_queuecommand(struct scsi_cmnd *scp, void (*done)(struct scsi_cmnd *))
+advansys_queuecommand_lck(struct scsi_cmnd *scp, void (*done)(struct scsi_cmnd *))
 {
         struct Scsi_Host *shost = scp->device->host;
         int asc_res, result = 0;
@@ -9525,6 +9525,8 @@ advansys_queuecommand(struct scsi_cmnd *scp, void (*done)(struct scsi_cmnd *))
         return result;
 }
 
+static DEF_SCSI_QCMD(advansys_queuecommand)
+
 static ushort __devinit AscGetEisaChipCfg(PortAddr iop_base)
 {
         PortAddr eisa_cfg_iop = (PortAddr) ASC_GET_EISA_SLOT(iop_base) |
diff --git a/drivers/scsi/aha152x.c b/drivers/scsi/aha152x.c
index 8eab8587ff21..c5169f01c1cd 100644
--- a/drivers/scsi/aha152x.c
+++ b/drivers/scsi/aha152x.c
@@ -1056,7 +1056,7 @@ static int aha152x_internal_queue(Scsi_Cmnd *SCpnt, struct completion *complete,
  *  queue a command
  *
  */
-static int aha152x_queue(Scsi_Cmnd *SCpnt, void (*done)(Scsi_Cmnd *))
+static int aha152x_queue_lck(Scsi_Cmnd *SCpnt, void (*done)(Scsi_Cmnd *))
 {
 #if 0
         if(*SCpnt->cmnd == REQUEST_SENSE) {
@@ -1070,6 +1070,8 @@ static int aha152x_queue(Scsi_Cmnd *SCpnt, void (*done)(Scsi_Cmnd *))
         return aha152x_internal_queue(SCpnt, NULL, 0, done);
 }
 
+static DEF_SCSI_QCMD(aha152x_queue)
+
 
 /*
  *  
diff --git a/drivers/scsi/aha1542.c b/drivers/scsi/aha1542.c
index 4f785f254c1f..195823a51aab 100644
--- a/drivers/scsi/aha1542.c
+++ b/drivers/scsi/aha1542.c
@@ -558,7 +558,7 @@ static void aha1542_intr_handle(struct Scsi_Host *shost)
         };
 }
 
-static int aha1542_queuecommand(Scsi_Cmnd * SCpnt, void (*done) (Scsi_Cmnd *))
+static int aha1542_queuecommand_lck(Scsi_Cmnd * SCpnt, void (*done) (Scsi_Cmnd *))
 {
         unchar ahacmd = CMD_START_SCSI;
         unchar direction;
@@ -718,6 +718,8 @@ static int aha1542_queuecommand(Scsi_Cmnd * SCpnt, void (*done) (Scsi_Cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(aha1542_queuecommand)
+
 /* Initialize mailboxes */
 static void setup_mailboxes(int bse, struct Scsi_Host *shpnt)
 {
diff --git a/drivers/scsi/aha1542.h b/drivers/scsi/aha1542.h
index 1db538552d56..b871d2b57f93 100644
--- a/drivers/scsi/aha1542.h
+++ b/drivers/scsi/aha1542.h
@@ -132,7 +132,7 @@ struct ccb {			/* Command Control Block 5.3 */
 };
 
 static int aha1542_detect(struct scsi_host_template *);
-static int aha1542_queuecommand(Scsi_Cmnd *, void (*done)(Scsi_Cmnd *));
+static int aha1542_queuecommand(struct Scsi_Host *, struct scsi_cmnd *);
 static int aha1542_bus_reset(Scsi_Cmnd * SCpnt);
 static int aha1542_dev_reset(Scsi_Cmnd * SCpnt);
 static int aha1542_host_reset(Scsi_Cmnd * SCpnt);
diff --git a/drivers/scsi/aha1740.c b/drivers/scsi/aha1740.c
index 0107a4cc3331..d058f1ab82b5 100644
--- a/drivers/scsi/aha1740.c
+++ b/drivers/scsi/aha1740.c
@@ -331,7 +331,7 @@ static irqreturn_t aha1740_intr_handle(int irq, void *dev_id)
         return IRQ_RETVAL(handled);
 }
 
-static int aha1740_queuecommand(Scsi_Cmnd * SCpnt, void (*done)(Scsi_Cmnd *))
+static int aha1740_queuecommand_lck(Scsi_Cmnd * SCpnt, void (*done)(Scsi_Cmnd *))
 {
         unchar direction;
         unchar *cmd = (unchar *) SCpnt->cmnd;
@@ -503,6 +503,8 @@ static int aha1740_queuecommand(Scsi_Cmnd * SCpnt, void (*done)(Scsi_Cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(aha1740_queuecommand)
+
 /* Query the board for its irq_level and irq_type.  Nothing else matters
    in enhanced mode on an EISA bus. */
 
diff --git a/drivers/scsi/aic7xxx/aic79xx_osm.c b/drivers/scsi/aic7xxx/aic79xx_osm.c
index 88ad8482ef59..25d066624476 100644
--- a/drivers/scsi/aic7xxx/aic79xx_osm.c
+++ b/drivers/scsi/aic7xxx/aic79xx_osm.c
@@ -573,7 +573,7 @@ ahd_linux_info(struct Scsi_Host *host)
  * Queue an SCB to the controller.
  */
 static int
-ahd_linux_queue(struct scsi_cmnd * cmd, void (*scsi_done) (struct scsi_cmnd *))
+ahd_linux_queue_lck(struct scsi_cmnd * cmd, void (*scsi_done) (struct scsi_cmnd *))
 {
         struct   ahd_softc *ahd;
         struct   ahd_linux_device *dev = scsi_transport_device_data(cmd->device);
@@ -588,6 +588,8 @@ ahd_linux_queue(struct scsi_cmnd * cmd, void (*scsi_done) (struct scsi_cmnd *))
         return rtn;
 }
 
+static DEF_SCSI_QCMD(ahd_linux_queue)
+
 static struct scsi_target **
 ahd_linux_target_in_softc(struct scsi_target *starget)
 {
diff --git a/drivers/scsi/aic7xxx/aic7xxx_osm.c b/drivers/scsi/aic7xxx/aic7xxx_osm.c
index aeea7a61478e..4a359bb307c6 100644
--- a/drivers/scsi/aic7xxx/aic7xxx_osm.c
+++ b/drivers/scsi/aic7xxx/aic7xxx_osm.c
@@ -528,7 +528,7 @@ ahc_linux_info(struct Scsi_Host *host)
  * Queue an SCB to the controller.
  */
 static int
-ahc_linux_queue(struct scsi_cmnd * cmd, void (*scsi_done) (struct scsi_cmnd *))
+ahc_linux_queue_lck(struct scsi_cmnd * cmd, void (*scsi_done) (struct scsi_cmnd *))
 {
         struct   ahc_softc *ahc;
         struct   ahc_linux_device *dev = scsi_transport_device_data(cmd->device);
@@ -548,6 +548,8 @@ ahc_linux_queue(struct scsi_cmnd * cmd, void (*scsi_done) (struct scsi_cmnd *))
         return rtn;
 }
 
+static DEF_SCSI_QCMD(ahc_linux_queue)
+
 static inline struct scsi_target **
 ahc_linux_target_in_softc(struct scsi_target *starget)
 {
diff --git a/drivers/scsi/aic7xxx_old.c b/drivers/scsi/aic7xxx_old.c
index aee73fafccc8..4ff60a08df0f 100644
--- a/drivers/scsi/aic7xxx_old.c
+++ b/drivers/scsi/aic7xxx_old.c
@@ -10234,7 +10234,7 @@ static void aic7xxx_buildscb(struct aic7xxx_host *p, struct scsi_cmnd *cmd,
  * Description:
  *   Queue a SCB to the controller.
  *-F*************************************************************************/
-static int aic7xxx_queue(struct scsi_cmnd *cmd, void (*fn)(struct scsi_cmnd *))
+static int aic7xxx_queue_lck(struct scsi_cmnd *cmd, void (*fn)(struct scsi_cmnd *))
 {
   struct aic7xxx_host *p;
   struct aic7xxx_scb *scb;
@@ -10292,6 +10292,8 @@ static int aic7xxx_queue(struct scsi_cmnd *cmd, void (*fn)(struct scsi_cmnd *))
   return (0);
 }
 
+static DEF_SCSI_QCMD(aic7xxx_queue)
+
 /*+F*************************************************************************
  * Function:
  *   aic7xxx_bus_device_reset
diff --git a/drivers/scsi/arcmsr/arcmsr_hba.c b/drivers/scsi/arcmsr/arcmsr_hba.c
index 05a78e515a24..17e3df4f016f 100644
--- a/drivers/scsi/arcmsr/arcmsr_hba.c
+++ b/drivers/scsi/arcmsr/arcmsr_hba.c
@@ -85,8 +85,7 @@ static int arcmsr_abort(struct scsi_cmnd *);
 static int arcmsr_bus_reset(struct scsi_cmnd *);
 static int arcmsr_bios_param(struct scsi_device *sdev,
                 struct block_device *bdev, sector_t capacity, int *info);
-static int arcmsr_queue_command(struct scsi_cmnd *cmd,
-                                        void (*done) (struct scsi_cmnd *));
+static int arcmsr_queue_command(struct Scsi_Host *h, struct scsi_cmnd *cmd);
 static int arcmsr_probe(struct pci_dev *pdev,
                                 const struct pci_device_id *id);
 static void arcmsr_remove(struct pci_dev *pdev);
@@ -2081,7 +2080,7 @@ static void arcmsr_handle_virtual_command(struct AdapterControlBlock *acb,
         }
 }
 
-static int arcmsr_queue_command(struct scsi_cmnd *cmd,
+static int arcmsr_queue_command_lck(struct scsi_cmnd *cmd,
         void (* done)(struct scsi_cmnd *))
 {
         struct Scsi_Host *host = cmd->device->host;
@@ -2124,6 +2123,8 @@ static int arcmsr_queue_command(struct scsi_cmnd *cmd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(arcmsr_queue_command)
+
 static bool arcmsr_get_hba_config(struct AdapterControlBlock *acb)
 {
         struct MessageUnit_A __iomem *reg = acb->pmuA;
diff --git a/drivers/scsi/arm/acornscsi.c b/drivers/scsi/arm/acornscsi.c
index 918ccf818757..ec166726b314 100644
--- a/drivers/scsi/arm/acornscsi.c
+++ b/drivers/scsi/arm/acornscsi.c
@@ -2511,7 +2511,7 @@ acornscsi_intr(int irq, void *dev_id)
  *            done - function called on completion, with pointer to command descriptor
  * Returns  : 0, or < 0 on error.
  */
-int acornscsi_queuecmd(struct scsi_cmnd *SCpnt,
+static int acornscsi_queuecmd_lck(struct scsi_cmnd *SCpnt,
                        void (*done)(struct scsi_cmnd *))
 {
     AS_Host *host = (AS_Host *)SCpnt->device->host->hostdata;
@@ -2561,6 +2561,8 @@ int acornscsi_queuecmd(struct scsi_cmnd *SCpnt,
     return 0;
 }
 
+DEF_SCSI_QCMD(acornscsi_queuecmd)
+
 /*
  * Prototype: void acornscsi_reportstatus(struct scsi_cmnd **SCpntp1, struct scsi_cmnd **SCpntp2, int result)
  * Purpose  : pass a result to *SCpntp1, and check if *SCpntp1 = *SCpntp2
diff --git a/drivers/scsi/arm/fas216.c b/drivers/scsi/arm/fas216.c
index 9e71ac611146..2b2ce21e227e 100644
--- a/drivers/scsi/arm/fas216.c
+++ b/drivers/scsi/arm/fas216.c
@@ -2198,7 +2198,7 @@ no_command:
  * Returns: 0 on success, else error.
  * Notes: io_request_lock is held, interrupts are disabled.
  */
-int fas216_queue_command(struct scsi_cmnd *SCpnt,
+static int fas216_queue_command_lck(struct scsi_cmnd *SCpnt,
                          void (*done)(struct scsi_cmnd *))
 {
         FAS216_Info *info = (FAS216_Info *)SCpnt->device->host->hostdata;
@@ -2240,6 +2240,8 @@ int fas216_queue_command(struct scsi_cmnd *SCpnt,
         return result;
 }
 
+DEF_SCSI_QCMD(fas216_queue_command)
+
 /**
  * fas216_internal_done - trigger restart of a waiting thread in fas216_noqueue_command
  * @SCpnt: Command to wake
@@ -2263,7 +2265,7 @@ static void fas216_internal_done(struct scsi_cmnd *SCpnt)
  * Returns: scsi result code.
  * Notes: io_request_lock is held, interrupts are disabled.
  */
-int fas216_noqueue_command(struct scsi_cmnd *SCpnt,
+static int fas216_noqueue_command_lck(struct scsi_cmnd *SCpnt,
                            void (*done)(struct scsi_cmnd *))
 {
         FAS216_Info *info = (FAS216_Info *)SCpnt->device->host->hostdata;
@@ -2277,7 +2279,7 @@ int fas216_noqueue_command(struct scsi_cmnd *SCpnt,
         BUG_ON(info->scsi.irq != NO_IRQ);
 
         info->internal_done = 0;
-        fas216_queue_command(SCpnt, fas216_internal_done);
+        fas216_queue_command_lck(SCpnt, fas216_internal_done);
 
         /*
          * This wastes time, since we can't return until the command is
@@ -2310,6 +2312,8 @@ int fas216_noqueue_command(struct scsi_cmnd *SCpnt,
         return 0;
 }
 
+DEF_SCSI_QCMD(fas216_noqueue_command)
+
 /*
  * Error handler timeout function.  Indicate that we timed out,
  * and wake up any error handler process so it can continue.
diff --git a/drivers/scsi/arm/fas216.h b/drivers/scsi/arm/fas216.h
index b65f4cf0eec9..f30f8d659dc4 100644
--- a/drivers/scsi/arm/fas216.h
+++ b/drivers/scsi/arm/fas216.h
@@ -331,23 +331,21 @@ extern int fas216_init (struct Scsi_Host *instance);
  */
 extern int fas216_add (struct Scsi_Host *instance, struct device *dev);
 
-/* Function: int fas216_queue_command(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
+/* Function: int fas216_queue_command(struct Scsi_Host *h, struct scsi_cmnd *SCpnt)
  * Purpose : queue a command for adapter to process.
- * Params  : SCpnt - Command to queue
- *           done  - done function to call once command is complete
+ * Params  : h - host adapter
+ *         : SCpnt - Command to queue
  * Returns : 0 - success, else error
  */
-extern int fas216_queue_command(struct scsi_cmnd *,
-                                void (*done)(struct scsi_cmnd *));
+extern int fas216_queue_command(struct Scsi_Host *h, struct scsi_cmnd *SCpnt);
 
-/* Function: int fas216_noqueue_command(istruct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
+/* Function: int fas216_noqueue_command(struct Scsi_Host *h, struct scsi_cmnd *SCpnt)
  * Purpose : queue a command for adapter to process, and process it to completion.
- * Params  : SCpnt - Command to queue
- *           done  - done function to call once command is complete
+ * Params  : h - host adapter
+ *         : SCpnt - Command to queue
  * Returns : 0 - success, else error
  */
-extern int fas216_noqueue_command(struct scsi_cmnd *,
-                                  void (*done)(struct scsi_cmnd *));
+extern int fas216_noqueue_command(struct Scsi_Host *, struct scsi_cmnd *);
 
 /* Function: irqreturn_t fas216_intr (FAS216_Info *info)
  * Purpose : handle interrupts from the interface to progress a command
diff --git a/drivers/scsi/atari_NCR5380.c b/drivers/scsi/atari_NCR5380.c
index 158ebc3644d8..88b2928b4d3b 100644
--- a/drivers/scsi/atari_NCR5380.c
+++ b/drivers/scsi/atari_NCR5380.c
@@ -910,7 +910,7 @@ static int __init NCR5380_init(struct Scsi_Host *instance, int flags)
  *
  */
 
-static int NCR5380_queue_command(Scsi_Cmnd *cmd, void (*done)(Scsi_Cmnd *))
+static int NCR5380_queue_command_lck(Scsi_Cmnd *cmd, void (*done)(Scsi_Cmnd *))
 {
         SETUP_HOSTDATA(cmd->device->host);
         Scsi_Cmnd *tmp;
@@ -1022,6 +1022,8 @@ static int NCR5380_queue_command(Scsi_Cmnd *cmd, void (*done)(Scsi_Cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(NCR5380_queue_command)
+
 /*
  * Function : NCR5380_main (void)
  *
diff --git a/drivers/scsi/atari_scsi.c b/drivers/scsi/atari_scsi.c
index ad7a23aef0ec..3e8658e2f154 100644
--- a/drivers/scsi/atari_scsi.c
+++ b/drivers/scsi/atari_scsi.c
@@ -572,23 +572,6 @@ static void falcon_get_lock(void)
 }
 
 
-/* This is the wrapper function for NCR5380_queue_command(). It just
- * tries to get the lock on the ST-DMA (see above) and then calls the
- * original function.
- */
-
-#if 0
-int atari_queue_command(Scsi_Cmnd *cmd, void (*done)(Scsi_Cmnd *))
-{
-        /* falcon_get_lock();
-         * ++guenther: moved to NCR5380_queue_command() to prevent
-         * race condition, see there for an explanation.
-         */
-        return NCR5380_queue_command(cmd, done);
-}
-#endif
-
-
 int __init atari_scsi_detect(struct scsi_host_template *host)
 {
         static int called = 0;
diff --git a/drivers/scsi/atp870u.c b/drivers/scsi/atp870u.c
index ab5bdda6903e..76029d570beb 100644
--- a/drivers/scsi/atp870u.c
+++ b/drivers/scsi/atp870u.c
@@ -605,7 +605,7 @@ handled:
  *
  *      Queue a command to the ATP queue. Called with the host lock held.
  */
-static int atp870u_queuecommand(struct scsi_cmnd * req_p, 
+static int atp870u_queuecommand_lck(struct scsi_cmnd *req_p,
                          void (*done) (struct scsi_cmnd *))
 {
         unsigned char c;
@@ -694,6 +694,8 @@ static int atp870u_queuecommand(struct scsi_cmnd * req_p,
         return 0;
 }
 
+static DEF_SCSI_QCMD(atp870u_queuecommand)
+
 /**
  *      send_s870       -       send a command to the controller
  *      @host: host
diff --git a/drivers/scsi/bfa/bfad_im.c b/drivers/scsi/bfa/bfad_im.c
index 8daa716739d1..8ca967dee66d 100644
--- a/drivers/scsi/bfa/bfad_im.c
+++ b/drivers/scsi/bfa/bfad_im.c
@@ -30,8 +30,7 @@ DEFINE_IDR(bfad_im_port_index);
 struct scsi_transport_template *bfad_im_scsi_transport_template;
 struct scsi_transport_template *bfad_im_scsi_vport_transport_template;
 static void bfad_im_itnim_work_handler(struct work_struct *work);
-static int bfad_im_queuecommand(struct scsi_cmnd *cmnd,
-                                void (*done)(struct scsi_cmnd *));
+static int bfad_im_queuecommand(struct Scsi_Host *h, struct scsi_cmnd *cmnd);
 static int bfad_im_slave_alloc(struct scsi_device *sdev);
 static void bfad_im_fc_rport_add(struct bfad_im_port_s  *im_port,
                                 struct bfad_itnim_s *itnim);
@@ -1120,7 +1119,7 @@ bfad_im_itnim_work_handler(struct work_struct *work)
  * Scsi_Host template entry, queue a SCSI command to the BFAD.
  */
 static int
-bfad_im_queuecommand(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
+bfad_im_queuecommand_lck(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
 {
         struct bfad_im_port_s *im_port =
                 (struct bfad_im_port_s *) cmnd->device->host->hostdata[0];
@@ -1187,6 +1186,8 @@ out_fail_cmd:
         return 0;
 }
 
+static DEF_SCSI_QCMD(bfad_im_queuecommand)
+
 void
 bfad_os_rport_online_wait(struct bfad_s *bfad)
 {
diff --git a/drivers/scsi/dc395x.c b/drivers/scsi/dc395x.c
index 54f50b07dac7..8f1b5c8bf903 100644
--- a/drivers/scsi/dc395x.c
+++ b/drivers/scsi/dc395x.c
@@ -1080,7 +1080,7 @@ static void build_srb(struct scsi_cmnd *cmd, struct DeviceCtlBlk *dcb,
  *        and is expected to be held on return.
  *
  **/
-static int dc395x_queue_command(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
+static int dc395x_queue_command_lck(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 {
         struct DeviceCtlBlk *dcb;
         struct ScsiReqBlk *srb;
@@ -1154,6 +1154,7 @@ complete:
         return 0;
 }
 
+static DEF_SCSI_QCMD(dc395x_queue_command)
 
 /*
  * Return the disk geometry for the given SCSI device.
diff --git a/drivers/scsi/dpt_i2o.c b/drivers/scsi/dpt_i2o.c
index 23dec0063385..cffcb108ac96 100644
--- a/drivers/scsi/dpt_i2o.c
+++ b/drivers/scsi/dpt_i2o.c
@@ -423,7 +423,7 @@ static int adpt_slave_configure(struct scsi_device * device)
         return 0;
 }
 
-static int adpt_queue(struct scsi_cmnd * cmd, void (*done) (struct scsi_cmnd *))
+static int adpt_queue_lck(struct scsi_cmnd * cmd, void (*done) (struct scsi_cmnd *))
 {
         adpt_hba* pHba = NULL;
         struct adpt_device* pDev = NULL;        /* dpt per device information */
@@ -491,6 +491,8 @@ static int adpt_queue(struct scsi_cmnd * cmd, void (*done) (struct scsi_cmnd *))
         return adpt_scsi_to_i2o(pHba, cmd, pDev);
 }
 
+static DEF_SCSI_QCMD(adpt_queue)
+
 static int adpt_bios_param(struct scsi_device *sdev, struct block_device *dev,
                 sector_t capacity, int geom[])
 {
diff --git a/drivers/scsi/dpti.h b/drivers/scsi/dpti.h
index 337746d46043..beded716f93f 100644
--- a/drivers/scsi/dpti.h
+++ b/drivers/scsi/dpti.h
@@ -29,7 +29,7 @@
  */
 
 static int adpt_detect(struct scsi_host_template * sht);
-static int adpt_queue(struct scsi_cmnd * cmd, void (*cmdcomplete) (struct scsi_cmnd *));
+static int adpt_queue(struct Scsi_Host *h, struct scsi_cmnd * cmd);
 static int adpt_abort(struct scsi_cmnd * cmd);
 static int adpt_reset(struct scsi_cmnd* cmd);
 static int adpt_release(struct Scsi_Host *host);
diff --git a/drivers/scsi/dtc.h b/drivers/scsi/dtc.h
index 0b205f8c7326..cdc621204b66 100644
--- a/drivers/scsi/dtc.h
+++ b/drivers/scsi/dtc.h
@@ -36,7 +36,7 @@ static int dtc_abort(Scsi_Cmnd *);
 static int dtc_biosparam(struct scsi_device *, struct block_device *,
                          sector_t, int*);
 static int dtc_detect(struct scsi_host_template *);
-static int dtc_queue_command(Scsi_Cmnd *, void (*done)(Scsi_Cmnd *));
+static int dtc_queue_command(struct Scsi_Host *, struct scsi_cmnd *);
 static int dtc_bus_reset(Scsi_Cmnd *);
 
 #ifndef CMD_PER_LUN
diff --git a/drivers/scsi/eata.c b/drivers/scsi/eata.c
index d1c31378f6da..53925ac178fd 100644
--- a/drivers/scsi/eata.c
+++ b/drivers/scsi/eata.c
@@ -505,8 +505,7 @@
 
 static int eata2x_detect(struct scsi_host_template *);
 static int eata2x_release(struct Scsi_Host *);
-static int eata2x_queuecommand(struct scsi_cmnd *,
-                               void (*done) (struct scsi_cmnd *));
+static int eata2x_queuecommand(struct Scsi_Host *, struct scsi_cmnd *);
 static int eata2x_eh_abort(struct scsi_cmnd *);
 static int eata2x_eh_host_reset(struct scsi_cmnd *);
 static int eata2x_bios_param(struct scsi_device *, struct block_device *,
@@ -1758,7 +1757,7 @@ static void scsi_to_dev_dir(unsigned int i, struct hostdata *ha)
 
 }
 
-static int eata2x_queuecommand(struct scsi_cmnd *SCpnt,
+static int eata2x_queuecommand_lck(struct scsi_cmnd *SCpnt,
                                void (*done) (struct scsi_cmnd *))
 {
         struct Scsi_Host *shost = SCpnt->device->host;
@@ -1843,6 +1842,8 @@ static int eata2x_queuecommand(struct scsi_cmnd *SCpnt,
         return 0;
 }
 
+static DEF_SCSI_QCMD(eata2x_queuecommand)
+
 static int eata2x_eh_abort(struct scsi_cmnd *SCarg)
 {
         struct Scsi_Host *shost = SCarg->device->host;
diff --git a/drivers/scsi/eata_pio.c b/drivers/scsi/eata_pio.c
index 60886c19065e..4a9641e69f54 100644
--- a/drivers/scsi/eata_pio.c
+++ b/drivers/scsi/eata_pio.c
@@ -335,7 +335,7 @@ static inline unsigned int eata_pio_send_command(unsigned long base, unsigned ch
         return 0;
 }
 
-static int eata_pio_queue(struct scsi_cmnd *cmd,
+static int eata_pio_queue_lck(struct scsi_cmnd *cmd,
                 void (*done)(struct scsi_cmnd *))
 {
         unsigned int x, y;
@@ -438,6 +438,8 @@ static int eata_pio_queue(struct scsi_cmnd *cmd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(eata_pio_queue)
+
 static int eata_pio_abort(struct scsi_cmnd *cmd)
 {
         unsigned int loop = 100;
diff --git a/drivers/scsi/esp_scsi.c b/drivers/scsi/esp_scsi.c
index e2bc779f86c1..57558523c1b8 100644
--- a/drivers/scsi/esp_scsi.c
+++ b/drivers/scsi/esp_scsi.c
@@ -916,7 +916,7 @@ static void esp_event_queue_full(struct esp *esp, struct esp_cmd_entry *ent)
         scsi_track_queue_full(dev, lp->num_tagged - 1);
 }
 
-static int esp_queuecommand(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
+static int esp_queuecommand_lck(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 {
         struct scsi_device *dev = cmd->device;
         struct esp *esp = shost_priv(dev->host);
@@ -941,6 +941,8 @@ static int esp_queuecommand(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd
         return 0;
 }
 
+static DEF_SCSI_QCMD(esp_queuecommand)
+
 static int esp_check_gross_error(struct esp *esp)
 {
         if (esp->sreg & ESP_STAT_SPAM) {
diff --git a/drivers/scsi/fd_mcs.c b/drivers/scsi/fd_mcs.c
index 2ad95aa8f585..a2c6135d337e 100644
--- a/drivers/scsi/fd_mcs.c
+++ b/drivers/scsi/fd_mcs.c
@@ -1072,7 +1072,7 @@ static int fd_mcs_release(struct Scsi_Host *shpnt)
         return 0;
 }
 
-static int fd_mcs_queue(Scsi_Cmnd * SCpnt, void (*done) (Scsi_Cmnd *))
+static int fd_mcs_queue_lck(Scsi_Cmnd * SCpnt, void (*done) (Scsi_Cmnd *))
 {
         struct Scsi_Host *shpnt = SCpnt->device->host;
 
@@ -1122,6 +1122,8 @@ static int fd_mcs_queue(Scsi_Cmnd * SCpnt, void (*done) (Scsi_Cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(fd_mcs_queue)
+
 #if DEBUG_ABORT || DEBUG_RESET
 static void fd_mcs_print_info(Scsi_Cmnd * SCpnt)
 {
diff --git a/drivers/scsi/fdomain.c b/drivers/scsi/fdomain.c
index e296bcc57d5c..69b7aa54f43f 100644
--- a/drivers/scsi/fdomain.c
+++ b/drivers/scsi/fdomain.c
@@ -1419,7 +1419,7 @@ static irqreturn_t do_fdomain_16x0_intr(int irq, void *dev_id)
    return IRQ_HANDLED;
 }
 
-static int fdomain_16x0_queue(struct scsi_cmnd *SCpnt,
+static int fdomain_16x0_queue_lck(struct scsi_cmnd *SCpnt,
                 void (*done)(struct scsi_cmnd *))
 {
    if (in_command) {
@@ -1469,6 +1469,8 @@ static int fdomain_16x0_queue(struct scsi_cmnd *SCpnt,
    return 0;
 }
 
+static DEF_SCSI_QCMD(fdomain_16x0_queue)
+
 #if DEBUG_ABORT
 static void print_info(struct scsi_cmnd *SCpnt)
 {
diff --git a/drivers/scsi/fnic/fnic.h b/drivers/scsi/fnic/fnic.h
index cbb20b13b228..92f185081e62 100644
--- a/drivers/scsi/fnic/fnic.h
+++ b/drivers/scsi/fnic/fnic.h
@@ -246,7 +246,7 @@ void fnic_set_port_id(struct fc_lport *, u32, struct fc_frame *);
 void fnic_update_mac(struct fc_lport *, u8 *new);
 void fnic_update_mac_locked(struct fnic *, u8 *new);
 
-int fnic_queuecommand(struct scsi_cmnd *, void (*done)(struct scsi_cmnd *));
+int fnic_queuecommand(struct Scsi_Host *, struct scsi_cmnd *);
 int fnic_abort_cmd(struct scsi_cmnd *);
 int fnic_device_reset(struct scsi_cmnd *);
 int fnic_host_reset(struct scsi_cmnd *);
diff --git a/drivers/scsi/fnic/fnic_scsi.c b/drivers/scsi/fnic/fnic_scsi.c
index 198cbab3e894..22d02404d15f 100644
--- a/drivers/scsi/fnic/fnic_scsi.c
+++ b/drivers/scsi/fnic/fnic_scsi.c
@@ -349,7 +349,7 @@ static inline int fnic_queue_wq_copy_desc(struct fnic *fnic,
  * Routine to send a scsi cdb
  * Called with host_lock held and interrupts disabled.
  */
-int fnic_queuecommand(struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *))
+static int fnic_queuecommand_lck(struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *))
 {
         struct fc_lport *lp;
         struct fc_rport *rport;
@@ -457,6 +457,8 @@ out:
         return ret;
 }
 
+DEF_SCSI_QCMD(fnic_queuecommand)
+
 /*
  * fnic_fcpio_fw_reset_cmpl_handler
  * Routine to handle fw reset completion
diff --git a/drivers/scsi/g_NCR5380.h b/drivers/scsi/g_NCR5380.h
index 921764c9ab24..1bcdb7beb77b 100644
--- a/drivers/scsi/g_NCR5380.h
+++ b/drivers/scsi/g_NCR5380.h
@@ -46,7 +46,7 @@
 static int generic_NCR5380_abort(Scsi_Cmnd *);
 static int generic_NCR5380_detect(struct scsi_host_template *);
 static int generic_NCR5380_release_resources(struct Scsi_Host *);
-static int generic_NCR5380_queue_command(Scsi_Cmnd *, void (*done)(Scsi_Cmnd *));
+static int generic_NCR5380_queue_command(struct Scsi_Host *, struct scsi_cmnd *);
 static int generic_NCR5380_bus_reset(Scsi_Cmnd *);
 static const char* generic_NCR5380_info(struct Scsi_Host *);
 
diff --git a/drivers/scsi/gdth.c b/drivers/scsi/gdth.c
index 841101846b88..76365700e2d5 100644
--- a/drivers/scsi/gdth.c
+++ b/drivers/scsi/gdth.c
@@ -185,7 +185,7 @@ static long gdth_unlocked_ioctl(struct file *filep, unsigned int cmd,
                                 unsigned long arg);
 
 static void gdth_flush(gdth_ha_str *ha);
-static int gdth_queuecommand(Scsi_Cmnd *scp,void (*done)(Scsi_Cmnd *));
+static int gdth_queuecommand(struct Scsi_Host *h, struct scsi_cmnd *cmd);
 static int __gdth_queuecommand(gdth_ha_str *ha, struct scsi_cmnd *scp,
                                 struct gdth_cmndinfo *cmndinfo);
 static void gdth_scsi_done(struct scsi_cmnd *scp);
@@ -4004,7 +4004,7 @@ static int gdth_bios_param(struct scsi_device *sdev,struct block_device *bdev,se
 }
 
 
-static int gdth_queuecommand(struct scsi_cmnd *scp,
+static int gdth_queuecommand_lck(struct scsi_cmnd *scp,
                                 void (*done)(struct scsi_cmnd *))
 {
     gdth_ha_str *ha = shost_priv(scp->device->host);
@@ -4022,6 +4022,8 @@ static int gdth_queuecommand(struct scsi_cmnd *scp,
     return __gdth_queuecommand(ha, scp, cmndinfo);
 }
 
+static DEF_SCSI_QCMD(gdth_queuecommand)
+
 static int __gdth_queuecommand(gdth_ha_str *ha, struct scsi_cmnd *scp,
                                 struct gdth_cmndinfo *cmndinfo)
 {
diff --git a/drivers/scsi/hpsa.c b/drivers/scsi/hpsa.c
index c5d0606ad097..b2fb2b2a6e70 100644
--- a/drivers/scsi/hpsa.c
+++ b/drivers/scsi/hpsa.c
@@ -31,7 +31,6 @@
 #include <linux/seq_file.h>
 #include <linux/init.h>
 #include <linux/spinlock.h>
-#include <linux/smp_lock.h>
 #include <linux/compat.h>
 #include <linux/blktrace_api.h>
 #include <linux/uaccess.h>
@@ -143,8 +142,7 @@ static void fill_cmd(struct CommandList *c, u8 cmd, struct ctlr_info *h,
         void *buff, size_t size, u8 page_code, unsigned char *scsi3addr,
         int cmd_type);
 
-static int hpsa_scsi_queue_command(struct scsi_cmnd *cmd,
-                void (*done)(struct scsi_cmnd *));
+static int hpsa_scsi_queue_command(struct Scsi_Host *h, struct scsi_cmnd *cmd);
 static void hpsa_scan_start(struct Scsi_Host *);
 static int hpsa_scan_finished(struct Scsi_Host *sh,
         unsigned long elapsed_time);
@@ -1926,7 +1924,7 @@ sglist_finished:
 }
 
 
-static int hpsa_scsi_queue_command(struct scsi_cmnd *cmd,
+static int hpsa_scsi_queue_command_lck(struct scsi_cmnd *cmd,
         void (*done)(struct scsi_cmnd *))
 {
         struct ctlr_info *h;
@@ -2020,6 +2018,8 @@ static int hpsa_scsi_queue_command(struct scsi_cmnd *cmd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(hpsa_scsi_queue_command)
+
 static void hpsa_scan_start(struct Scsi_Host *sh)
 {
         struct ctlr_info *h = shost_to_hba(sh);
diff --git a/drivers/scsi/hptiop.c b/drivers/scsi/hptiop.c
index 0729f150b33a..10b65556937b 100644
--- a/drivers/scsi/hptiop.c
+++ b/drivers/scsi/hptiop.c
@@ -751,7 +751,7 @@ static void hptiop_post_req_mv(struct hptiop_hba *hba,
                 MVIOP_MU_QUEUE_ADDR_HOST_BIT | size_bit, hba);
 }
 
-static int hptiop_queuecommand(struct scsi_cmnd *scp,
+static int hptiop_queuecommand_lck(struct scsi_cmnd *scp,
                                 void (*done)(struct scsi_cmnd *))
 {
         struct Scsi_Host *host = scp->device->host;
@@ -819,6 +819,8 @@ cmd_done:
         return 0;
 }
 
+static DEF_SCSI_QCMD(hptiop_queuecommand)
+
 static const char *hptiop_info(struct Scsi_Host *host)
 {
         return driver_name_long;
diff --git a/drivers/scsi/ibmmca.c b/drivers/scsi/ibmmca.c
index 9a4b69d4f4eb..67fc8ffd52e6 100644
--- a/drivers/scsi/ibmmca.c
+++ b/drivers/scsi/ibmmca.c
@@ -39,7 +39,7 @@
 #include <scsi/scsi_host.h>
 
 /* Common forward declarations for all Linux-versions: */
-static int ibmmca_queuecommand (Scsi_Cmnd *, void (*done) (Scsi_Cmnd *));
+static int ibmmca_queuecommand (struct Scsi_Host *, struct scsi_cmnd *);
 static int ibmmca_abort (Scsi_Cmnd *);
 static int ibmmca_host_reset (Scsi_Cmnd *);
 static int ibmmca_biosparam (struct scsi_device *, struct block_device *, sector_t, int *);
@@ -1691,7 +1691,7 @@ static int __devexit ibmmca_remove(struct device *dev)
 }
 
 /* The following routine is the SCSI command queue for the midlevel driver */
-static int ibmmca_queuecommand(Scsi_Cmnd * cmd, void (*done) (Scsi_Cmnd *))
+static int ibmmca_queuecommand_lck(Scsi_Cmnd * cmd, void (*done) (Scsi_Cmnd *))
 {
         unsigned int ldn;
         unsigned int scsi_cmd;
@@ -1996,6 +1996,8 @@ static int ibmmca_queuecommand(Scsi_Cmnd * cmd, void (*done) (Scsi_Cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(ibmmca_queuecommand)
+
 static int __ibmmca_abort(Scsi_Cmnd * cmd)
 {
         /* Abort does not work, as the adapter never generates an interrupt on
diff --git a/drivers/scsi/ibmvscsi/ibmvfc.c b/drivers/scsi/ibmvscsi/ibmvfc.c
index 00d08b25425f..57cad7e20caa 100644
--- a/drivers/scsi/ibmvscsi/ibmvfc.c
+++ b/drivers/scsi/ibmvscsi/ibmvfc.c
@@ -1606,7 +1606,7 @@ static inline int ibmvfc_host_chkready(struct ibmvfc_host *vhost)
  * Returns:
  *      0 on success / other on failure
  **/
-static int ibmvfc_queuecommand(struct scsi_cmnd *cmnd,
+static int ibmvfc_queuecommand_lck(struct scsi_cmnd *cmnd,
                                void (*done) (struct scsi_cmnd *))
 {
         struct ibmvfc_host *vhost = shost_priv(cmnd->device->host);
@@ -1672,6 +1672,8 @@ static int ibmvfc_queuecommand(struct scsi_cmnd *cmnd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(ibmvfc_queuecommand)
+
 /**
  * ibmvfc_sync_completion - Signal that a synchronous command has completed
  * @evt:        ibmvfc event struct
diff --git a/drivers/scsi/ibmvscsi/ibmvscsi.c b/drivers/scsi/ibmvscsi/ibmvscsi.c
index 67f78a470f5f..041958453e2a 100644
--- a/drivers/scsi/ibmvscsi/ibmvscsi.c
+++ b/drivers/scsi/ibmvscsi/ibmvscsi.c
@@ -713,7 +713,7 @@ static inline u16 lun_from_dev(struct scsi_device *dev)
  * @cmd:        struct scsi_cmnd to be executed
  * @done:       Callback function to be called when cmd is completed
 */
-static int ibmvscsi_queuecommand(struct scsi_cmnd *cmnd,
+static int ibmvscsi_queuecommand_lck(struct scsi_cmnd *cmnd,
                                  void (*done) (struct scsi_cmnd *))
 {
         struct srp_cmd *srp_cmd;
@@ -766,6 +766,8 @@ static int ibmvscsi_queuecommand(struct scsi_cmnd *cmnd,
         return ibmvscsi_send_srp_event(evt_struct, hostdata, 0);
 }
 
+static DEF_SCSI_QCMD(ibmvscsi_queuecommand)
+
 /* ------------------------------------------------------------
  * Routines for driver initialization
  */
diff --git a/drivers/scsi/imm.c b/drivers/scsi/imm.c
index 4734ab0b3ff6..99aa0e5699bc 100644
--- a/drivers/scsi/imm.c
+++ b/drivers/scsi/imm.c
@@ -926,7 +926,7 @@ static int imm_engine(imm_struct *dev, struct scsi_cmnd *cmd)
         return 0;
 }
 
-static int imm_queuecommand(struct scsi_cmnd *cmd,
+static int imm_queuecommand_lck(struct scsi_cmnd *cmd,
                 void (*done)(struct scsi_cmnd *))
 {
         imm_struct *dev = imm_dev(cmd->device->host);
@@ -949,6 +949,8 @@ static int imm_queuecommand(struct scsi_cmnd *cmd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(imm_queuecommand)
+
 /*
  * Apparently the disk->capacity attribute is off by 1 sector 
  * for all disk drives.  We add the one here, but it should really
diff --git a/drivers/scsi/in2000.c b/drivers/scsi/in2000.c
index 52bdc6df6b92..6568aab745a0 100644
--- a/drivers/scsi/in2000.c
+++ b/drivers/scsi/in2000.c
@@ -334,7 +334,7 @@ static uchar calc_sync_xfer(unsigned int period, unsigned int offset)
 
 static void in2000_execute(struct Scsi_Host *instance);
 
-static int in2000_queuecommand(Scsi_Cmnd * cmd, void (*done) (Scsi_Cmnd *))
+static int in2000_queuecommand_lck(Scsi_Cmnd * cmd, void (*done) (Scsi_Cmnd *))
 {
         struct Scsi_Host *instance;
         struct IN2000_hostdata *hostdata;
@@ -431,6 +431,8 @@ static int in2000_queuecommand(Scsi_Cmnd * cmd, void (*done) (Scsi_Cmnd *))
             return 0;
 }
 
+static DEF_SCSI_QCMD(in2000_queuecommand)
+
 
 
 /*
diff --git a/drivers/scsi/in2000.h b/drivers/scsi/in2000.h
index 0fb8b06b8392..5821e1fbce08 100644
--- a/drivers/scsi/in2000.h
+++ b/drivers/scsi/in2000.h
@@ -396,7 +396,7 @@ struct IN2000_hostdata {
                                                            flags)
 
 static int in2000_detect(struct scsi_host_template *) in2000__INIT;
-static int in2000_queuecommand(Scsi_Cmnd *, void (*done)(Scsi_Cmnd *));
+static int in2000_queuecommand(struct Scsi_Host *, struct scsi_cmnd *);
 static int in2000_abort(Scsi_Cmnd *);
 static void in2000_setup(char *, int *) in2000__INIT;
 static int in2000_biosparam(struct scsi_device *, struct block_device *,
diff --git a/drivers/scsi/initio.c b/drivers/scsi/initio.c
index 108797761b95..9627d062e16b 100644
--- a/drivers/scsi/initio.c
+++ b/drivers/scsi/initio.c
@@ -2639,7 +2639,7 @@ static void initio_build_scb(struct initio_host * host, struct scsi_ctrl_blk * c
  *      will cause the mid layer to call us again later with the command)
  */
 
-static int i91u_queuecommand(struct scsi_cmnd *cmd,
+static int i91u_queuecommand_lck(struct scsi_cmnd *cmd,
                 void (*done)(struct scsi_cmnd *))
 {
         struct initio_host *host = (struct initio_host *) cmd->device->host->hostdata;
@@ -2656,6 +2656,8 @@ static int i91u_queuecommand(struct scsi_cmnd *cmd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(i91u_queuecommand)
+
 /**
  *      i91u_bus_reset          -       reset the SCSI bus
  *      @cmnd: Command block we want to trigger the reset for
diff --git a/drivers/scsi/ipr.c b/drivers/scsi/ipr.c
index fa60d7df44be..5bbaee597e88 100644
--- a/drivers/scsi/ipr.c
+++ b/drivers/scsi/ipr.c
@@ -5709,7 +5709,7 @@ static void ipr_scsi_done(struct ipr_cmnd *ipr_cmd)
  *      SCSI_MLQUEUE_DEVICE_BUSY if device is busy
  *      SCSI_MLQUEUE_HOST_BUSY if host is busy
  **/
-static int ipr_queuecommand(struct scsi_cmnd *scsi_cmd,
+static int ipr_queuecommand_lck(struct scsi_cmnd *scsi_cmd,
                             void (*done) (struct scsi_cmnd *))
 {
         struct ipr_ioa_cfg *ioa_cfg;
@@ -5792,6 +5792,8 @@ static int ipr_queuecommand(struct scsi_cmnd *scsi_cmd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(ipr_queuecommand)
+
 /**
  * ipr_ioctl - IOCTL handler
  * @sdev:       scsi device struct
diff --git a/drivers/scsi/ips.c b/drivers/scsi/ips.c
index f83a116955f2..b2511acd39bd 100644
--- a/drivers/scsi/ips.c
+++ b/drivers/scsi/ips.c
@@ -232,7 +232,7 @@ static int ips_detect(struct scsi_host_template *);
 static int ips_release(struct Scsi_Host *);
 static int ips_eh_abort(struct scsi_cmnd *);
 static int ips_eh_reset(struct scsi_cmnd *);
-static int ips_queue(struct scsi_cmnd *, void (*)(struct scsi_cmnd *));
+static int ips_queue(struct Scsi_Host *, struct scsi_cmnd *);
 static const char *ips_info(struct Scsi_Host *);
 static irqreturn_t do_ipsintr(int, void *);
 static int ips_hainit(ips_ha_t *);
@@ -1046,7 +1046,7 @@ static int ips_eh_reset(struct scsi_cmnd *SC)
 /*    Linux obtains io_request_lock before calling this function            */
 /*                                                                          */
 /****************************************************************************/
-static int ips_queue(struct scsi_cmnd *SC, void (*done) (struct scsi_cmnd *))
+static int ips_queue_lck(struct scsi_cmnd *SC, void (*done) (struct scsi_cmnd *))
 {
         ips_ha_t *ha;
         ips_passthru_t *pt;
@@ -1137,6 +1137,8 @@ static int ips_queue(struct scsi_cmnd *SC, void (*done) (struct scsi_cmnd *))
         return (0);
 }
 
+static DEF_SCSI_QCMD(ips_queue)
+
 /****************************************************************************/
 /*                                                                          */
 /* Routine Name: ips_biosparam                                              */
diff --git a/drivers/scsi/libfc/fc_fcp.c b/drivers/scsi/libfc/fc_fcp.c
index e340373b509b..2924363d142b 100644
--- a/drivers/scsi/libfc/fc_fcp.c
+++ b/drivers/scsi/libfc/fc_fcp.c
@@ -1753,7 +1753,7 @@ static inline int fc_fcp_lport_queue_ready(struct fc_lport *lport)
  * This is the i/o strategy routine, called by the SCSI layer. This routine
  * is called with the host_lock held.
  */
-int fc_queuecommand(struct scsi_cmnd *sc_cmd, void (*done)(struct scsi_cmnd *))
+static int fc_queuecommand_lck(struct scsi_cmnd *sc_cmd, void (*done)(struct scsi_cmnd *))
 {
         struct fc_lport *lport;
         struct fc_rport *rport = starget_to_rport(scsi_target(sc_cmd->device));
@@ -1851,6 +1851,8 @@ out:
         spin_lock_irq(lport->host->host_lock);
         return rc;
 }
+
+DEF_SCSI_QCMD(fc_queuecommand)
 EXPORT_SYMBOL(fc_queuecommand);
 
 /**
diff --git a/drivers/scsi/libiscsi.c b/drivers/scsi/libiscsi.c
index 633e09036357..c15fde808c33 100644
--- a/drivers/scsi/libiscsi.c
+++ b/drivers/scsi/libiscsi.c
@@ -1599,7 +1599,7 @@ enum {
         FAILURE_SESSION_NOT_READY,
 };
 
-int iscsi_queuecommand(struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *))
+static int iscsi_queuecommand_lck(struct scsi_cmnd *sc, void (*done)(struct scsi_cmnd *))
 {
         struct iscsi_cls_session *cls_session;
         struct Scsi_Host *host;
@@ -1736,6 +1736,8 @@ fault:
         spin_lock(host->host_lock);
         return 0;
 }
+
+DEF_SCSI_QCMD(iscsi_queuecommand)
 EXPORT_SYMBOL_GPL(iscsi_queuecommand);
 
 int iscsi_change_queue_depth(struct scsi_device *sdev, int depth, int reason)
diff --git a/drivers/scsi/libsas/sas_scsi_host.c b/drivers/scsi/libsas/sas_scsi_host.c
index 55f09e92ab59..29251fabecc6 100644
--- a/drivers/scsi/libsas/sas_scsi_host.c
+++ b/drivers/scsi/libsas/sas_scsi_host.c
@@ -189,7 +189,7 @@ int sas_queue_up(struct sas_task *task)
  * Note: XXX: Remove the host unlock/lock pair when SCSI Core can
  * call us without holding an IRQ spinlock...
  */
-int sas_queuecommand(struct scsi_cmnd *cmd,
+static int sas_queuecommand_lck(struct scsi_cmnd *cmd,
                      void (*scsi_done)(struct scsi_cmnd *))
         __releases(host->host_lock)
         __acquires(dev->sata_dev.ap->lock)
@@ -254,6 +254,8 @@ out:
         return res;
 }
 
+DEF_SCSI_QCMD(sas_queuecommand)
+
 static void sas_eh_finish_cmd(struct scsi_cmnd *cmd)
 {
         struct sas_task *task = TO_SAS_TASK(cmd);
diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c
index f64b65a770b8..581837b3c71a 100644
--- a/drivers/scsi/lpfc/lpfc_scsi.c
+++ b/drivers/scsi/lpfc/lpfc_scsi.c
@@ -2899,7 +2899,7 @@ void lpfc_poll_timeout(unsigned long ptr)
  *   SCSI_MLQUEUE_HOST_BUSY - Block all devices served by this host temporarily.
  **/
 static int
-lpfc_queuecommand(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
+lpfc_queuecommand_lck(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
 {
         struct Scsi_Host  *shost = cmnd->device->host;
         struct lpfc_vport *vport = (struct lpfc_vport *) shost->hostdata;
@@ -3060,6 +3060,8 @@ lpfc_queuecommand(struct scsi_cmnd *cmnd, void (*done) (struct scsi_cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(lpfc_queuecommand)
+
 /**
  * lpfc_abort_handler - scsi_host_template eh_abort_handler entry point
  * @cmnd: Pointer to scsi_cmnd data structure.
diff --git a/drivers/scsi/mac53c94.c b/drivers/scsi/mac53c94.c
index 3ddb4dc62d5d..6c42dff0f4d3 100644
--- a/drivers/scsi/mac53c94.c
+++ b/drivers/scsi/mac53c94.c
@@ -66,7 +66,7 @@ static void cmd_done(struct fsc_state *, int result);
 static void set_dma_cmds(struct fsc_state *, struct scsi_cmnd *);
 
 
-static int mac53c94_queue(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
+static int mac53c94_queue_lck(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 {
         struct fsc_state *state;
 
@@ -99,6 +99,8 @@ static int mac53c94_queue(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *
         return 0;
 }
 
+static DEF_SCSI_QCMD(mac53c94_queue)
+
 static int mac53c94_host_reset(struct scsi_cmnd *cmd)
 {
         struct fsc_state *state = (struct fsc_state *) cmd->device->host->hostdata;
diff --git a/drivers/scsi/megaraid.c b/drivers/scsi/megaraid.c
index 7ceb5cf12c6b..9aa048525eb2 100644
--- a/drivers/scsi/megaraid.c
+++ b/drivers/scsi/megaraid.c
@@ -366,7 +366,7 @@ mega_runpendq(adapter_t *adapter)
  * The command queuing entry point for the mid-layer.
  */
 static int
-megaraid_queue(Scsi_Cmnd *scmd, void (*done)(Scsi_Cmnd *))
+megaraid_queue_lck(Scsi_Cmnd *scmd, void (*done)(Scsi_Cmnd *))
 {
         adapter_t       *adapter;
         scb_t   *scb;
@@ -409,6 +409,8 @@ megaraid_queue(Scsi_Cmnd *scmd, void (*done)(Scsi_Cmnd *))
         return busy;
 }
 
+static DEF_SCSI_QCMD(megaraid_queue)
+
 /**
  * mega_allocate_scb()
  * @adapter - pointer to our soft state
@@ -4456,7 +4458,7 @@ mega_internal_command(adapter_t *adapter, megacmd_t *mc, mega_passthru *pthru)
 
         scb->idx = CMDID_INT_CMDS;
 
-        megaraid_queue(scmd, mega_internal_done);
+        megaraid_queue_lck(scmd, mega_internal_done);
 
         wait_for_completion(&adapter->int_waitq);
 
diff --git a/drivers/scsi/megaraid.h b/drivers/scsi/megaraid.h
index 2b4a048cadf1..f5644745e24e 100644
--- a/drivers/scsi/megaraid.h
+++ b/drivers/scsi/megaraid.h
@@ -987,7 +987,7 @@ static int mega_query_adapter(adapter_t *);
 static int issue_scb(adapter_t *, scb_t *);
 static int mega_setup_mailbox(adapter_t *);
 
-static int megaraid_queue (Scsi_Cmnd *, void (*)(Scsi_Cmnd *));
+static int megaraid_queue (struct Scsi_Host *, struct scsi_cmnd *);
 static scb_t * mega_build_cmd(adapter_t *, Scsi_Cmnd *, int *);
 static void __mega_runpendq(adapter_t *);
 static int issue_scb_block(adapter_t *, u_char *);
diff --git a/drivers/scsi/megaraid/megaraid_mbox.c b/drivers/scsi/megaraid/megaraid_mbox.c
index a7810a106b37..5708cb27d078 100644
--- a/drivers/scsi/megaraid/megaraid_mbox.c
+++ b/drivers/scsi/megaraid/megaraid_mbox.c
@@ -113,8 +113,7 @@ static int megaraid_mbox_fire_sync_cmd(adapter_t *);
 static void megaraid_mbox_display_scb(adapter_t *, scb_t *);
 static void megaraid_mbox_setup_device_map(adapter_t *);
 
-static int megaraid_queue_command(struct scsi_cmnd *,
-                void (*)(struct scsi_cmnd *));
+static int megaraid_queue_command(struct Scsi_Host *, struct scsi_cmnd *);
 static scb_t *megaraid_mbox_build_cmd(adapter_t *, struct scsi_cmnd *, int *);
 static void megaraid_mbox_runpendq(adapter_t *, scb_t *);
 static void megaraid_mbox_prepare_pthru(adapter_t *, scb_t *,
@@ -1484,7 +1483,7 @@ mbox_post_cmd(adapter_t *adapter, scb_t *scb)
  * Queue entry point for mailbox based controllers.
  */
 static int
-megaraid_queue_command(struct scsi_cmnd *scp, void (*done)(struct scsi_cmnd *))
+megaraid_queue_command_lck(struct scsi_cmnd *scp, void (*done)(struct scsi_cmnd *))
 {
         adapter_t       *adapter;
         scb_t           *scb;
@@ -1513,6 +1512,8 @@ megaraid_queue_command(struct scsi_cmnd *scp, void (*done)(struct scsi_cmnd *))
         return if_busy;
 }
 
+static DEF_SCSI_QCMD(megaraid_queue_command)
+
 /**
  * megaraid_mbox_build_cmd - transform the mid-layer scsi commands
  * @adapter     : controller's soft state
diff --git a/drivers/scsi/megaraid/megaraid_sas.c b/drivers/scsi/megaraid/megaraid_sas.c
index eb29d5085131..7451bc096a01 100644
--- a/drivers/scsi/megaraid/megaraid_sas.c
+++ b/drivers/scsi/megaraid/megaraid_sas.c
@@ -1334,7 +1334,7 @@ megasas_dump_pending_frames(struct megasas_instance *instance)
  * @done:                       Callback entry point
  */
 static int
-megasas_queue_command(struct scsi_cmnd *scmd, void (*done) (struct scsi_cmnd *))
+megasas_queue_command_lck(struct scsi_cmnd *scmd, void (*done) (struct scsi_cmnd *))
 {
         u32 frame_count;
         struct megasas_cmd *cmd;
@@ -1417,6 +1417,8 @@ megasas_queue_command(struct scsi_cmnd *scmd, void (*done) (struct scsi_cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(megasas_queue_command)
+
 static struct megasas_instance *megasas_lookup_instance(u16 host_no)
 {
         int i;
diff --git a/drivers/scsi/mesh.c b/drivers/scsi/mesh.c
index 1f784fde2510..197aa1b3f0f3 100644
--- a/drivers/scsi/mesh.c
+++ b/drivers/scsi/mesh.c
@@ -1627,7 +1627,7 @@ static void cmd_complete(struct mesh_state *ms)
  * Called by midlayer with host locked to queue a new
  * request
  */
-static int mesh_queue(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
+static int mesh_queue_lck(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 {
         struct mesh_state *ms;
 
@@ -1648,6 +1648,8 @@ static int mesh_queue(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(mesh_queue)
+
 /*
  * Called to handle interrupts, either call by the interrupt
  * handler (do_mesh_interrupt) or by other functions in
diff --git a/drivers/scsi/mpt2sas/mpt2sas_scsih.c b/drivers/scsi/mpt2sas/mpt2sas_scsih.c
index 16e99b686354..1a96a00418a4 100644
--- a/drivers/scsi/mpt2sas/mpt2sas_scsih.c
+++ b/drivers/scsi/mpt2sas/mpt2sas_scsih.c
@@ -3315,7 +3315,7 @@ _scsih_eedp_error_handling(struct scsi_cmnd *scmd, u16 ioc_status)
  * SCSI_MLQUEUE_HOST_BUSY if the entire host queue is full
  */
 static int
-_scsih_qcmd(struct scsi_cmnd *scmd, void (*done)(struct scsi_cmnd *))
+_scsih_qcmd_lck(struct scsi_cmnd *scmd, void (*done)(struct scsi_cmnd *))
 {
         struct MPT2SAS_ADAPTER *ioc = shost_priv(scmd->device->host);
         struct MPT2SAS_DEVICE *sas_device_priv_data;
@@ -3441,6 +3441,8 @@ _scsih_qcmd(struct scsi_cmnd *scmd, void (*done)(struct scsi_cmnd *))
         return SCSI_MLQUEUE_HOST_BUSY;
 }
 
+static DEF_SCSI_QCMD(_scsih_qcmd)
+
 /**
  * _scsih_normalize_sense - normalize descriptor and fixed format sense data
  * @sense_buffer: sense data returned by target
diff --git a/drivers/scsi/ncr53c8xx.c b/drivers/scsi/ncr53c8xx.c
index d013a2aa2fd5..46cc3825638d 100644
--- a/drivers/scsi/ncr53c8xx.c
+++ b/drivers/scsi/ncr53c8xx.c
@@ -8029,7 +8029,7 @@ static int ncr53c8xx_slave_configure(struct scsi_device *device)
         return 0;
 }
 
-static int ncr53c8xx_queue_command (struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd *))
+static int ncr53c8xx_queue_command_lck (struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 {
      struct ncb *np = ((struct host_data *) cmd->device->host->hostdata)->ncb;
      unsigned long flags;
@@ -8068,6 +8068,8 @@ printk("ncr53c8xx : command successfully queued\n");
      return sts;
 }
 
+static DEF_SCSI_QCMD(ncr53c8xx_queue_command)
+
 irqreturn_t ncr53c8xx_intr(int irq, void *dev_id)
 {
      unsigned long flags;
diff --git a/drivers/scsi/nsp32.c b/drivers/scsi/nsp32.c
index 4c1e54545200..6b8b021400f8 100644
--- a/drivers/scsi/nsp32.c
+++ b/drivers/scsi/nsp32.c
@@ -196,8 +196,7 @@ static void __exit    exit_nsp32  (void);
 static int         nsp32_proc_info   (struct Scsi_Host *, char *, char **, off_t, int, int);
 
 static int         nsp32_detect      (struct pci_dev *pdev);
-static int         nsp32_queuecommand(struct scsi_cmnd *,
-                void (*done)(struct scsi_cmnd *));
+static int         nsp32_queuecommand(struct Scsi_Host *, struct scsi_cmnd *);
 static const char *nsp32_info        (struct Scsi_Host *);
 static int         nsp32_release     (struct Scsi_Host *);
 
@@ -909,7 +908,7 @@ static int nsp32_setup_sg_table(struct scsi_cmnd *SCpnt)
         return TRUE;
 }
 
-static int nsp32_queuecommand(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
+static int nsp32_queuecommand_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
 {
         nsp32_hw_data *data = (nsp32_hw_data *)SCpnt->device->host->hostdata;
         nsp32_target *target;
@@ -1050,6 +1049,8 @@ static int nsp32_queuecommand(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_
         return 0;
 }
 
+static DEF_SCSI_QCMD(nsp32_queuecommand)
+
 /* initialize asic */
 static int nsp32hw_init(nsp32_hw_data *data)
 {
diff --git a/drivers/scsi/pas16.h b/drivers/scsi/pas16.h
index 8dc5b1a5f5da..a04281cace2e 100644
--- a/drivers/scsi/pas16.h
+++ b/drivers/scsi/pas16.h
@@ -118,7 +118,7 @@ static int pas16_abort(Scsi_Cmnd *);
 static int pas16_biosparam(struct scsi_device *, struct block_device *,
                            sector_t, int*);
 static int pas16_detect(struct scsi_host_template *);
-static int pas16_queue_command(Scsi_Cmnd *, void (*done)(Scsi_Cmnd *));
+static int pas16_queue_command(struct Scsi_Host *, struct scsi_cmnd *);
 static int pas16_bus_reset(Scsi_Cmnd *);
 
 #ifndef CMD_PER_LUN
diff --git a/drivers/scsi/pcmcia/nsp_cs.c b/drivers/scsi/pcmcia/nsp_cs.c
index 9326c2c14880..be3f33d31a99 100644
--- a/drivers/scsi/pcmcia/nsp_cs.c
+++ b/drivers/scsi/pcmcia/nsp_cs.c
@@ -184,7 +184,7 @@ static void nsp_scsi_done(struct scsi_cmnd *SCpnt)
         SCpnt->scsi_done(SCpnt);
 }
 
-static int nsp_queuecommand(struct scsi_cmnd *SCpnt,
+static int nsp_queuecommand_lck(struct scsi_cmnd *SCpnt,
                             void (*done)(struct scsi_cmnd *))
 {
 #ifdef NSP_DEBUG
@@ -264,6 +264,8 @@ static int nsp_queuecommand(struct scsi_cmnd *SCpnt,
         return 0;
 }
 
+static DEF_SCSI_QCMD(nsp_queuecommand)
+
 /*
  * setup PIO FIFO transfer mode and enable/disable to data out
  */
diff --git a/drivers/scsi/pcmcia/nsp_cs.h b/drivers/scsi/pcmcia/nsp_cs.h
index d68c9f267c5e..7fc9a9d0a448 100644
--- a/drivers/scsi/pcmcia/nsp_cs.h
+++ b/drivers/scsi/pcmcia/nsp_cs.h
@@ -299,8 +299,7 @@ static        int        nsp_proc_info  (
                                          off_t   offset,
                                          int     length,
                                          int     inout);
-static int nsp_queuecommand(struct scsi_cmnd *SCpnt,
-                            void (* done)(struct scsi_cmnd *SCpnt));
+static int nsp_queuecommand(struct Scsi_Host *h, struct scsi_cmnd *SCpnt);
 
 /* Error handler */
 /*static int nsp_eh_abort       (struct scsi_cmnd *SCpnt);*/
diff --git a/drivers/scsi/pcmcia/sym53c500_cs.c b/drivers/scsi/pcmcia/sym53c500_cs.c
index 0ae27cb5cd6f..8552296edaa1 100644
--- a/drivers/scsi/pcmcia/sym53c500_cs.c
+++ b/drivers/scsi/pcmcia/sym53c500_cs.c
@@ -547,7 +547,7 @@ SYM53C500_info(struct Scsi_Host *SChost)
 }
 
 static int 
-SYM53C500_queue(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
+SYM53C500_queue_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
 {
         int i;
         int port_base = SCpnt->device->host->io_port;
@@ -583,6 +583,8 @@ SYM53C500_queue(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(SYM53C500_queue)
+
 static int 
 SYM53C500_host_reset(struct scsi_cmnd *SCpnt)
 {
diff --git a/drivers/scsi/pm8001/pm8001_sas.h b/drivers/scsi/pm8001/pm8001_sas.h
index 8e38ca8cd101..7f064f9ca828 100644
--- a/drivers/scsi/pm8001/pm8001_sas.h
+++ b/drivers/scsi/pm8001/pm8001_sas.h
@@ -50,7 +50,6 @@
 #include <linux/dma-mapping.h>
 #include <linux/pci.h>
 #include <linux/interrupt.h>
-#include <linux/smp_lock.h>
 #include <scsi/libsas.h>
 #include <scsi/scsi_tcq.h>
 #include <scsi/sas_ata.h>
diff --git a/drivers/scsi/pmcraid.c b/drivers/scsi/pmcraid.c
index cf89091e4c3d..5e76a624cb08 100644
--- a/drivers/scsi/pmcraid.c
+++ b/drivers/scsi/pmcraid.c
@@ -3478,7 +3478,7 @@ static int pmcraid_copy_sglist(
  *        SCSI_MLQUEUE_DEVICE_BUSY if device is busy
  *        SCSI_MLQUEUE_HOST_BUSY if host is busy
  */
-static int pmcraid_queuecommand(
+static int pmcraid_queuecommand_lck(
         struct scsi_cmnd *scsi_cmd,
         void (*done) (struct scsi_cmnd *)
 )
@@ -3584,6 +3584,8 @@ static int pmcraid_queuecommand(
         return rc;
 }
 
+static DEF_SCSI_QCMD(pmcraid_queuecommand)
+
 /**
  * pmcraid_open -char node "open" entry, allowed only users with admin access
  */
diff --git a/drivers/scsi/ppa.c b/drivers/scsi/ppa.c
index 7bc2d796e403..d164c9639361 100644
--- a/drivers/scsi/ppa.c
+++ b/drivers/scsi/ppa.c
@@ -798,7 +798,7 @@ static int ppa_engine(ppa_struct *dev, struct scsi_cmnd *cmd)
         return 0;
 }
 
-static int ppa_queuecommand(struct scsi_cmnd *cmd,
+static int ppa_queuecommand_lck(struct scsi_cmnd *cmd,
                 void (*done) (struct scsi_cmnd *))
 {
         ppa_struct *dev = ppa_dev(cmd->device->host);
@@ -821,6 +821,8 @@ static int ppa_queuecommand(struct scsi_cmnd *cmd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(ppa_queuecommand)
+
 /*
  * Apparently the disk->capacity attribute is off by 1 sector 
  * for all disk drives.  We add the one here, but it should really
diff --git a/drivers/scsi/ps3rom.c b/drivers/scsi/ps3rom.c
index 92ffbb510498..cd178b9e40cd 100644
--- a/drivers/scsi/ps3rom.c
+++ b/drivers/scsi/ps3rom.c
@@ -211,7 +211,7 @@ static int ps3rom_write_request(struct ps3_storage_device *dev,
         return 0;
 }
 
-static int ps3rom_queuecommand(struct scsi_cmnd *cmd,
+static int ps3rom_queuecommand_lck(struct scsi_cmnd *cmd,
                                void (*done)(struct scsi_cmnd *))
 {
         struct ps3rom_private *priv = shost_priv(cmd->device->host);
@@ -260,6 +260,8 @@ static int ps3rom_queuecommand(struct scsi_cmnd *cmd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(ps3rom_queuecommand)
+
 static int decode_lv1_status(u64 status, unsigned char *sense_key,
                              unsigned char *asc, unsigned char *ascq)
 {
diff --git a/drivers/scsi/qla1280.c b/drivers/scsi/qla1280.c
index b8166ecfd0e3..5dec684bf010 100644
--- a/drivers/scsi/qla1280.c
+++ b/drivers/scsi/qla1280.c
@@ -727,7 +727,7 @@ qla1280_info(struct Scsi_Host *host)
  * context which is a big NO! NO!.
  **************************************************************************/
 static int
-qla1280_queuecommand(struct scsi_cmnd *cmd, void (*fn)(struct scsi_cmnd *))
+qla1280_queuecommand_lck(struct scsi_cmnd *cmd, void (*fn)(struct scsi_cmnd *))
 {
         struct Scsi_Host *host = cmd->device->host;
         struct scsi_qla_host *ha = (struct scsi_qla_host *)host->hostdata;
@@ -756,6 +756,8 @@ qla1280_queuecommand(struct scsi_cmnd *cmd, void (*fn)(struct scsi_cmnd *))
         return status;
 }
 
+static DEF_SCSI_QCMD(qla1280_queuecommand)
+
 enum action {
         ABORT_COMMAND,
         DEVICE_RESET,
diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c
index 1830e6e97315..1644eabaafeb 100644
--- a/drivers/scsi/qla2xxx/qla_os.c
+++ b/drivers/scsi/qla2xxx/qla_os.c
@@ -179,8 +179,7 @@ static int qla2xxx_slave_alloc(struct scsi_device *);
 static int qla2xxx_scan_finished(struct Scsi_Host *, unsigned long time);
 static void qla2xxx_scan_start(struct Scsi_Host *);
 static void qla2xxx_slave_destroy(struct scsi_device *);
-static int qla2xxx_queuecommand(struct scsi_cmnd *cmd,
-                void (*fn)(struct scsi_cmnd *));
+static int qla2xxx_queuecommand(struct Scsi_Host *h, struct scsi_cmnd *cmd);
 static int qla2xxx_eh_abort(struct scsi_cmnd *);
 static int qla2xxx_eh_device_reset(struct scsi_cmnd *);
 static int qla2xxx_eh_target_reset(struct scsi_cmnd *);
@@ -535,7 +534,7 @@ qla2x00_get_new_sp(scsi_qla_host_t *vha, fc_port_t *fcport,
 }
 
 static int
-qla2xxx_queuecommand(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
+qla2xxx_queuecommand_lck(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 {
         scsi_qla_host_t *vha = shost_priv(cmd->device->host);
         fc_port_t *fcport = (struct fc_port *) cmd->device->hostdata;
@@ -609,6 +608,8 @@ qc24_fail_command:
         return 0;
 }
 
+static DEF_SCSI_QCMD(qla2xxx_queuecommand)
+
 
 /*
  * qla2x00_eh_wait_on_command
diff --git a/drivers/scsi/qla4xxx/ql4_os.c b/drivers/scsi/qla4xxx/ql4_os.c
index f4cd846abf6d..0d48fb4d1044 100644
--- a/drivers/scsi/qla4xxx/ql4_os.c
+++ b/drivers/scsi/qla4xxx/ql4_os.c
@@ -79,8 +79,7 @@ static enum blk_eh_timer_return qla4xxx_eh_cmd_timed_out(struct scsi_cmnd *sc);
 /*
  * SCSI host template entry points
  */
-static int qla4xxx_queuecommand(struct scsi_cmnd *cmd,
-                                void (*done) (struct scsi_cmnd *));
+static int qla4xxx_queuecommand(struct Scsi_Host *h, struct scsi_cmnd *cmd);
 static int qla4xxx_eh_abort(struct scsi_cmnd *cmd);
 static int qla4xxx_eh_device_reset(struct scsi_cmnd *cmd);
 static int qla4xxx_eh_target_reset(struct scsi_cmnd *cmd);
@@ -464,7 +463,7 @@ void qla4xxx_srb_compl(struct kref *ref)
  * completion handling).   Unfortunely, it sometimes calls the scheduler
  * in interrupt context which is a big NO! NO!.
  **/
-static int qla4xxx_queuecommand(struct scsi_cmnd *cmd,
+static int qla4xxx_queuecommand_lck(struct scsi_cmnd *cmd,
                                 void (*done)(struct scsi_cmnd *))
 {
         struct scsi_qla_host *ha = to_qla_host(cmd->device->host);
@@ -538,6 +537,8 @@ qc_fail_command:
         return 0;
 }
 
+static DEF_SCSI_QCMD(qla4xxx_queuecommand)
+
 /**
  * qla4xxx_mem_free - frees memory allocated to adapter
  * @ha: Pointer to host adapter structure.
diff --git a/drivers/scsi/qlogicfas408.c b/drivers/scsi/qlogicfas408.c
index 1ad51552d6b1..c3a9151ca823 100644
--- a/drivers/scsi/qlogicfas408.c
+++ b/drivers/scsi/qlogicfas408.c
@@ -439,7 +439,7 @@ irqreturn_t qlogicfas408_ihandl(int irq, void *dev_id)
  *      Queued command
  */
 
-int qlogicfas408_queuecommand(struct scsi_cmnd *cmd,
+static int qlogicfas408_queuecommand_lck(struct scsi_cmnd *cmd,
                               void (*done) (struct scsi_cmnd *))
 {
         struct qlogicfas408_priv *priv = get_priv_by_cmd(cmd);
@@ -459,6 +459,8 @@ int qlogicfas408_queuecommand(struct scsi_cmnd *cmd,
         return 0;
 }
 
+DEF_SCSI_QCMD(qlogicfas408_queuecommand)
+
 /* 
  *      Return bios parameters 
  */
diff --git a/drivers/scsi/qlogicfas408.h b/drivers/scsi/qlogicfas408.h
index 260626427a32..2f6c0a166200 100644
--- a/drivers/scsi/qlogicfas408.h
+++ b/drivers/scsi/qlogicfas408.h
@@ -103,8 +103,7 @@ struct qlogicfas408_priv {
 #define get_priv_by_host(x) (struct qlogicfas408_priv *)&((x)->hostdata[0])
 
 irqreturn_t qlogicfas408_ihandl(int irq, void *dev_id);
-int qlogicfas408_queuecommand(struct scsi_cmnd * cmd,
-                              void (*done) (struct scsi_cmnd *));
+int qlogicfas408_queuecommand(struct Scsi_Host *h, struct scsi_cmnd * cmd);
 int qlogicfas408_biosparam(struct scsi_device * disk,
                            struct block_device *dev,
                            sector_t capacity, int ip[]);
diff --git a/drivers/scsi/qlogicpti.c b/drivers/scsi/qlogicpti.c
index f8c561cf751e..664c9572d0c9 100644
--- a/drivers/scsi/qlogicpti.c
+++ b/drivers/scsi/qlogicpti.c
@@ -1003,7 +1003,7 @@ static int qlogicpti_slave_configure(struct scsi_device *sdev)
  *
  * "This code must fly." -davem
  */
-static int qlogicpti_queuecommand(struct scsi_cmnd *Cmnd, void (*done)(struct scsi_cmnd *))
+static int qlogicpti_queuecommand_lck(struct scsi_cmnd *Cmnd, void (*done)(struct scsi_cmnd *))
 {
         struct Scsi_Host *host = Cmnd->device->host;
         struct qlogicpti *qpti = (struct qlogicpti *) host->hostdata;
@@ -1052,6 +1052,8 @@ toss_command:
         return 1;
 }
 
+static DEF_SCSI_QCMD(qlogicpti_queuecommand)
+
 static int qlogicpti_return_status(struct Status_Entry *sts, int id)
 {
         int host_status = DID_ERROR;
diff --git a/drivers/scsi/scsi.c b/drivers/scsi/scsi.c
index 348fba0a8976..2aeb2e9c4d3b 100644
--- a/drivers/scsi/scsi.c
+++ b/drivers/scsi/scsi.c
@@ -634,12 +634,13 @@ void scsi_log_completion(struct scsi_cmnd *cmd, int disposition)
  * Description: a serial number identifies a request for error recovery
  * and debugging purposes.  Protected by the Host_Lock of host.
  */
-static inline void scsi_cmd_get_serial(struct Scsi_Host *host, struct scsi_cmnd *cmd)
+void scsi_cmd_get_serial(struct Scsi_Host *host, struct scsi_cmnd *cmd)
 {
         cmd->serial_number = host->cmd_serial_number++;
         if (cmd->serial_number == 0) 
                 cmd->serial_number = host->cmd_serial_number++;
 }
+EXPORT_SYMBOL(scsi_cmd_get_serial);
 
 /**
  * scsi_dispatch_command - Dispatch a command to the low-level driver.
@@ -651,7 +652,6 @@ static inline void scsi_cmd_get_serial(struct Scsi_Host *host, struct scsi_cmnd
 int scsi_dispatch_cmd(struct scsi_cmnd *cmd)
 {
         struct Scsi_Host *host = cmd->device->host;
-        unsigned long flags = 0;
         unsigned long timeout;
         int rtn = 0;
 
@@ -737,23 +737,15 @@ int scsi_dispatch_cmd(struct scsi_cmnd *cmd)
                 goto out;
         }
 
-        spin_lock_irqsave(host->host_lock, flags);
-        /*
-         * AK: unlikely race here: for some reason the timer could
-         * expire before the serial number is set up below.
-         *
-         * TODO: kill serial or move to blk layer
-         */
-        scsi_cmd_get_serial(host, cmd); 
-
         if (unlikely(host->shost_state == SHOST_DEL)) {
                 cmd->result = (DID_NO_CONNECT << 16);
                 scsi_done(cmd);
         } else {
                 trace_scsi_dispatch_cmd_start(cmd);
-                rtn = host->hostt->queuecommand(cmd, scsi_done);
+                cmd->scsi_done = scsi_done;
+                rtn = host->hostt->queuecommand(host, cmd);
         }
-        spin_unlock_irqrestore(host->host_lock, flags);
+
         if (rtn) {
                 trace_scsi_dispatch_cmd_error(cmd, rtn);
                 if (rtn != SCSI_MLQUEUE_DEVICE_BUSY &&
diff --git a/drivers/scsi/scsi_debug.c b/drivers/scsi/scsi_debug.c
index 2c36bae3bd4b..2f1f9b079b10 100644
--- a/drivers/scsi/scsi_debug.c
+++ b/drivers/scsi/scsi_debug.c
@@ -3538,7 +3538,7 @@ static void sdebug_remove_adapter(void)
 }
 
 static
-int scsi_debug_queuecommand(struct scsi_cmnd *SCpnt, done_funct_t done)
+int scsi_debug_queuecommand_lck(struct scsi_cmnd *SCpnt, done_funct_t done)
 {
         unsigned char *cmd = (unsigned char *) SCpnt->cmnd;
         int len, k;
@@ -3884,6 +3884,8 @@ write:
                              (delay_override ? 0 : scsi_debug_delay));
 }
 
+static DEF_SCSI_QCMD(scsi_debug_queuecommand)
+
 static struct scsi_host_template sdebug_driver_template = {
         .proc_info =            scsi_debug_proc_info,
         .proc_name =            sdebug_proc_name,
diff --git a/drivers/scsi/scsi_error.c b/drivers/scsi/scsi_error.c
index f3cf924a2cd9..824b8fc03ce5 100644
--- a/drivers/scsi/scsi_error.c
+++ b/drivers/scsi/scsi_error.c
@@ -773,17 +773,15 @@ static int scsi_send_eh_cmnd(struct scsi_cmnd *scmd, unsigned char *cmnd,
         struct Scsi_Host *shost = sdev->host;
         DECLARE_COMPLETION_ONSTACK(done);
         unsigned long timeleft;
-        unsigned long flags;
         struct scsi_eh_save ses;
         int rtn;
 
         scsi_eh_prep_cmnd(scmd, &ses, cmnd, cmnd_size, sense_bytes);
         shost->eh_action = &done;
 
-        spin_lock_irqsave(shost->host_lock, flags);
         scsi_log_send(scmd);
-        shost->hostt->queuecommand(scmd, scsi_eh_done);
-        spin_unlock_irqrestore(shost->host_lock, flags);
+        scmd->scsi_done = scsi_eh_done;
+        shost->hostt->queuecommand(shost, scmd);
 
         timeleft = wait_for_completion_timeout(&done, timeout);
 
diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c
index b9ab3a590e4b..956496182c80 100644
--- a/drivers/scsi/sd.c
+++ b/drivers/scsi/sd.c
@@ -46,7 +46,6 @@
 #include <linux/blkdev.h>
 #include <linux/blkpg.h>
 #include <linux/delay.h>
-#include <linux/smp_lock.h>
 #include <linux/mutex.h>
 #include <linux/string_helpers.h>
 #include <linux/async.h>
diff --git a/drivers/scsi/stex.c b/drivers/scsi/stex.c
index 9c73dbda3bbb..606215e54b88 100644
--- a/drivers/scsi/stex.c
+++ b/drivers/scsi/stex.c
@@ -572,7 +572,7 @@ stex_slave_destroy(struct scsi_device *sdev)
 }
 
 static int
-stex_queuecommand(struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd *))
+stex_queuecommand_lck(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 {
         struct st_hba *hba;
         struct Scsi_Host *host;
@@ -698,6 +698,8 @@ stex_queuecommand(struct scsi_cmnd *cmd, void (* done)(struct scsi_cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(stex_queuecommand)
+
 static void stex_scsi_done(struct st_ccb *ccb)
 {
         struct scsi_cmnd *cmd = ccb->cmd;
diff --git a/drivers/scsi/sun3_NCR5380.c b/drivers/scsi/sun3_NCR5380.c
index 713620ed70d9..4f0e5485ffde 100644
--- a/drivers/scsi/sun3_NCR5380.c
+++ b/drivers/scsi/sun3_NCR5380.c
@@ -908,7 +908,7 @@ static int NCR5380_init (struct Scsi_Host *instance, int flags)
  */
 
 /* Only make static if a wrapper function is used */
-static int NCR5380_queue_command(struct scsi_cmnd *cmd,
+static int NCR5380_queue_command_lck(struct scsi_cmnd *cmd,
                                  void (*done)(struct scsi_cmnd *))
 {
     SETUP_HOSTDATA(cmd->device->host);
@@ -1019,6 +1019,8 @@ static int NCR5380_queue_command(struct scsi_cmnd *cmd,
     return 0;
 }
 
+static DEF_SCSI_QCMD(NCR5380_queue_command)
+
 /*
  * Function : NCR5380_main (void) 
  *
diff --git a/drivers/scsi/sun3_scsi.h b/drivers/scsi/sun3_scsi.h
index b29a9d661ca4..bcefd8458e65 100644
--- a/drivers/scsi/sun3_scsi.h
+++ b/drivers/scsi/sun3_scsi.h
@@ -51,8 +51,7 @@ static int sun3scsi_abort(struct scsi_cmnd *);
 static int sun3scsi_detect (struct scsi_host_template *);
 static const char *sun3scsi_info (struct Scsi_Host *);
 static int sun3scsi_bus_reset(struct scsi_cmnd *);
-static int sun3scsi_queue_command(struct scsi_cmnd *,
-                                  void (*done)(struct scsi_cmnd *));
+static int sun3scsi_queue_command(struct Scsi_Host *, struct scsi_cmnd *);
 static int sun3scsi_release (struct Scsi_Host *);
 
 #ifndef CMD_PER_LUN
diff --git a/drivers/scsi/sym53c416.c b/drivers/scsi/sym53c416.c
index e5c369bb568f..190107ae120b 100644
--- a/drivers/scsi/sym53c416.c
+++ b/drivers/scsi/sym53c416.c
@@ -734,7 +734,7 @@ const char *sym53c416_info(struct Scsi_Host *SChost)
         return info;
 }
 
-int sym53c416_queuecommand(Scsi_Cmnd *SCpnt, void (*done)(Scsi_Cmnd *))
+static int sym53c416_queuecommand_lck(Scsi_Cmnd *SCpnt, void (*done)(Scsi_Cmnd *))
 {
         int base;
         unsigned long flags = 0;
@@ -761,6 +761,8 @@ int sym53c416_queuecommand(Scsi_Cmnd *SCpnt, void (*done)(Scsi_Cmnd *))
         return 0;
 }
 
+DEF_SCSI_QCMD(sym53c416_queuecommand)
+
 static int sym53c416_host_reset(Scsi_Cmnd *SCpnt)
 {
         int base;
diff --git a/drivers/scsi/sym53c416.h b/drivers/scsi/sym53c416.h
index 77860d0748ff..387de5d80a70 100644
--- a/drivers/scsi/sym53c416.h
+++ b/drivers/scsi/sym53c416.h
@@ -25,7 +25,7 @@
 static int sym53c416_detect(struct scsi_host_template *);
 static const char *sym53c416_info(struct Scsi_Host *);
 static int sym53c416_release(struct Scsi_Host *);
-static int sym53c416_queuecommand(Scsi_Cmnd *, void (*done)(Scsi_Cmnd *));
+static int sym53c416_queuecommand(struct Scsi_Host *, struct scsi_cmnd *);
 static int sym53c416_host_reset(Scsi_Cmnd *);
 static int sym53c416_bios_param(struct scsi_device *, struct block_device *,
                 sector_t, int *);
diff --git a/drivers/scsi/sym53c8xx_2/sym_glue.c b/drivers/scsi/sym53c8xx_2/sym_glue.c
index 8b955b534a36..6b97ded9d45d 100644
--- a/drivers/scsi/sym53c8xx_2/sym_glue.c
+++ b/drivers/scsi/sym53c8xx_2/sym_glue.c
@@ -505,7 +505,7 @@ void sym_log_bus_error(struct Scsi_Host *shost)
  * queuecommand method.  Entered with the host adapter lock held and
  * interrupts disabled.
  */
-static int sym53c8xx_queue_command(struct scsi_cmnd *cmd,
+static int sym53c8xx_queue_command_lck(struct scsi_cmnd *cmd,
                                         void (*done)(struct scsi_cmnd *))
 {
         struct sym_hcb *np = SYM_SOFTC_PTR(cmd);
@@ -536,6 +536,8 @@ static int sym53c8xx_queue_command(struct scsi_cmnd *cmd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(sym53c8xx_queue_command)
+
 /*
  *  Linux entry point of the interrupt handler.
  */
diff --git a/drivers/scsi/t128.h b/drivers/scsi/t128.h
index 76a069b7ac0b..ada1115079c9 100644
--- a/drivers/scsi/t128.h
+++ b/drivers/scsi/t128.h
@@ -96,8 +96,7 @@ static int t128_abort(struct scsi_cmnd *);
 static int t128_biosparam(struct scsi_device *, struct block_device *,
                           sector_t, int*);
 static int t128_detect(struct scsi_host_template *);
-static int t128_queue_command(struct scsi_cmnd *,
-                              void (*done)(struct scsi_cmnd *));
+static int t128_queue_command(struct Scsi_Host *, struct scsi_cmnd *);
 static int t128_bus_reset(struct scsi_cmnd *);
 
 #ifndef CMD_PER_LUN
diff --git a/drivers/scsi/tmscsim.c b/drivers/scsi/tmscsim.c
index 27866b0adfeb..a124a28f2ccb 100644
--- a/drivers/scsi/tmscsim.c
+++ b/drivers/scsi/tmscsim.c
@@ -1883,7 +1883,7 @@ dc390_ScsiRstDetect( struct dc390_acb* pACB )
     return;
 }
 
-static int DC390_queuecommand(struct scsi_cmnd *cmd,
+static int DC390_queuecommand_lck(struct scsi_cmnd *cmd,
                 void (*done)(struct scsi_cmnd *))
 {
         struct scsi_device *sdev = cmd->device;
@@ -1944,6 +1944,8 @@ static int DC390_queuecommand(struct scsi_cmnd *cmd,
         return SCSI_MLQUEUE_DEVICE_BUSY;
 }
 
+static DEF_SCSI_QCMD(DC390_queuecommand)
+
 static void dc390_dumpinfo (struct dc390_acb* pACB, struct dc390_dcb* pDCB, struct dc390_srb* pSRB)
 {
     struct pci_dev *pdev;
diff --git a/drivers/scsi/u14-34f.c b/drivers/scsi/u14-34f.c
index 5d9fdeeb2315..edfc5da8be4c 100644
--- a/drivers/scsi/u14-34f.c
+++ b/drivers/scsi/u14-34f.c
@@ -433,7 +433,7 @@
 
 static int u14_34f_detect(struct scsi_host_template *);
 static int u14_34f_release(struct Scsi_Host *);
-static int u14_34f_queuecommand(struct scsi_cmnd *, void (*done)(struct scsi_cmnd *));
+static int u14_34f_queuecommand(struct Scsi_Host *, struct scsi_cmnd *);
 static int u14_34f_eh_abort(struct scsi_cmnd *);
 static int u14_34f_eh_host_reset(struct scsi_cmnd *);
 static int u14_34f_bios_param(struct scsi_device *, struct block_device *,
@@ -1248,7 +1248,7 @@ static void scsi_to_dev_dir(unsigned int i, unsigned int j) {
 
 }
 
-static int u14_34f_queuecommand(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *)) {
+static int u14_34f_queuecommand_lck(struct scsi_cmnd *SCpnt, void (*done)(struct scsi_cmnd *)) {
    unsigned int i, j, k;
    struct mscp *cpp;
 
@@ -1329,6 +1329,8 @@ static int u14_34f_queuecommand(struct scsi_cmnd *SCpnt, void (*done)(struct scs
    return 0;
 }
 
+static DEF_SCSI_QCMD(u14_34f_queuecommand)
+
 static int u14_34f_eh_abort(struct scsi_cmnd *SCarg) {
    unsigned int i, j;
 
diff --git a/drivers/scsi/ultrastor.c b/drivers/scsi/ultrastor.c
index 27aa40f3980e..0571ef9639cb 100644
--- a/drivers/scsi/ultrastor.c
+++ b/drivers/scsi/ultrastor.c
@@ -700,7 +700,7 @@ static inline void build_sg_list(struct mscp *mscp, struct scsi_cmnd *SCpnt)
         mscp->transfer_data_length = transfer_length;
 }
 
-static int ultrastor_queuecommand(struct scsi_cmnd *SCpnt,
+static int ultrastor_queuecommand_lck(struct scsi_cmnd *SCpnt,
                                 void (*done) (struct scsi_cmnd *))
 {
     struct mscp *my_mscp;
@@ -825,6 +825,8 @@ retry:
     return 0;
 }
 
+static DEF_SCSI_QCMD(ultrastor_queuecommand)
+
 /* This code must deal with 2 cases:
 
    1. The command has not been written to the OGM.  In this case, set
diff --git a/drivers/scsi/ultrastor.h b/drivers/scsi/ultrastor.h
index a692905f95f7..165c18b5cf5f 100644
--- a/drivers/scsi/ultrastor.h
+++ b/drivers/scsi/ultrastor.h
@@ -15,8 +15,7 @@
 
 static int ultrastor_detect(struct scsi_host_template *);
 static const char *ultrastor_info(struct Scsi_Host *shpnt);
-static int ultrastor_queuecommand(struct scsi_cmnd *,
-                                void (*done)(struct scsi_cmnd *));
+static int ultrastor_queuecommand(struct Scsi_Host *, struct scsi_cmnd *);
 static int ultrastor_abort(struct scsi_cmnd *);
 static int ultrastor_host_reset(struct scsi_cmnd *);
 static int ultrastor_biosparam(struct scsi_device *, struct block_device *,
diff --git a/drivers/scsi/vmw_pvscsi.c b/drivers/scsi/vmw_pvscsi.c
index 26894459c37f..a18996d24466 100644
--- a/drivers/scsi/vmw_pvscsi.c
+++ b/drivers/scsi/vmw_pvscsi.c
@@ -690,7 +690,7 @@ static int pvscsi_queue_ring(struct pvscsi_adapter *adapter,
         return 0;
 }
 
-static int pvscsi_queue(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
+static int pvscsi_queue_lck(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
 {
         struct Scsi_Host *host = cmd->device->host;
         struct pvscsi_adapter *adapter = shost_priv(host);
@@ -719,6 +719,8 @@ static int pvscsi_queue(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(pvscsi_queue)
+
 static int pvscsi_abort(struct scsi_cmnd *cmd)
 {
         struct pvscsi_adapter *adapter = shost_priv(cmd->device->host);
diff --git a/drivers/scsi/wd33c93.c b/drivers/scsi/wd33c93.c
index b701bf2cc187..5f697e0bd009 100644
--- a/drivers/scsi/wd33c93.c
+++ b/drivers/scsi/wd33c93.c
@@ -371,8 +371,8 @@ calc_sync_msg(unsigned int period, unsigned int offset, unsigned int fast,
         msg[1] = offset;
 }
 
-int
-wd33c93_queuecommand(struct scsi_cmnd *cmd,
+static int
+wd33c93_queuecommand_lck(struct scsi_cmnd *cmd,
                 void (*done)(struct scsi_cmnd *))
 {
         struct WD33C93_hostdata *hostdata;
@@ -468,6 +468,8 @@ wd33c93_queuecommand(struct scsi_cmnd *cmd,
         return 0;
 }
 
+DEF_SCSI_QCMD(wd33c93_queuecommand)
+
 /*
  * This routine attempts to start a scsi command. If the host_card is
  * already connected, we give up immediately. Otherwise, look through
diff --git a/drivers/scsi/wd33c93.h b/drivers/scsi/wd33c93.h
index 1ed5f3bf388e..3b463d7304dc 100644
--- a/drivers/scsi/wd33c93.h
+++ b/drivers/scsi/wd33c93.h
@@ -343,8 +343,7 @@ struct WD33C93_hostdata {
 void wd33c93_init (struct Scsi_Host *instance, const wd33c93_regs regs,
          dma_setup_t setup, dma_stop_t stop, int clock_freq);
 int wd33c93_abort (struct scsi_cmnd *cmd);
-int wd33c93_queuecommand (struct scsi_cmnd *cmd,
-                void (*done)(struct scsi_cmnd *));
+int wd33c93_queuecommand (struct Scsi_Host *h, struct scsi_cmnd *cmd);
 void wd33c93_intr (struct Scsi_Host *instance);
 int wd33c93_proc_info(struct Scsi_Host *, char *, char **, off_t, int, int);
 int wd33c93_host_reset (struct scsi_cmnd *);
diff --git a/drivers/scsi/wd7000.c b/drivers/scsi/wd7000.c
index 333580bf37c5..db451ae0a368 100644
--- a/drivers/scsi/wd7000.c
+++ b/drivers/scsi/wd7000.c
@@ -1082,7 +1082,7 @@ static irqreturn_t wd7000_intr(int irq, void *dev_id)
         return IRQ_HANDLED;
 }
 
-static int wd7000_queuecommand(struct scsi_cmnd *SCpnt,
+static int wd7000_queuecommand_lck(struct scsi_cmnd *SCpnt,
                 void (*done)(struct scsi_cmnd *))
 {
         Scb *scb;
@@ -1139,6 +1139,8 @@ static int wd7000_queuecommand(struct scsi_cmnd *SCpnt,
         return 0;
 }
 
+static DEF_SCSI_QCMD(wd7000_queuecommand)
+
 static int wd7000_diagnostics(Adapter * host, int code)
 {
         static IcbDiag icb = { ICB_OP_DIAGNOSTICS };
diff --git a/drivers/serial/crisv10.c b/drivers/serial/crisv10.c
index fa62578fcd20..bcc31f2140ac 100644
--- a/drivers/serial/crisv10.c
+++ b/drivers/serial/crisv10.c
@@ -18,7 +18,6 @@ static char *serial_version = "$Revision: 1.25 $";
 #include <linux/tty.h>
 #include <linux/tty_flip.h>
 #include <linux/major.h>
-#include <linux/smp_lock.h>
 #include <linux/string.h>
 #include <linux/fcntl.h>
 #include <linux/mm.h>
diff --git a/drivers/serial/serial_core.c b/drivers/serial/serial_core.c
index c4ea14670d44..9ffa5bee44ab 100644
--- a/drivers/serial/serial_core.c
+++ b/drivers/serial/serial_core.c
@@ -29,7 +29,6 @@
 #include <linux/console.h>
 #include <linux/proc_fs.h>
 #include <linux/seq_file.h>
-#include <linux/smp_lock.h>
 #include <linux/device.h>
 #include <linux/serial.h> /* for serial_state and serial_icounter_struct */
 #include <linux/serial_core.h>
diff --git a/drivers/sh/clk/core.c b/drivers/sh/clk/core.c
index 09615b51d591..3f5e387ed564 100644
--- a/drivers/sh/clk/core.c
+++ b/drivers/sh/clk/core.c
@@ -418,8 +418,11 @@ int clk_register(struct clk *clk)
                 list_add(&clk->sibling, &root_clks);
 
         list_add(&clk->node, &clock_list);
+
+#ifdef CONFIG_SH_CLK_CPG_LEGACY
         if (clk->ops && clk->ops->init)
                 clk->ops->init(clk);
+#endif
 
 out_unlock:
         mutex_unlock(&clock_list_sem);
@@ -455,19 +458,13 @@ EXPORT_SYMBOL_GPL(clk_get_rate);
 
 int clk_set_rate(struct clk *clk, unsigned long rate)
 {
-        return clk_set_rate_ex(clk, rate, 0);
-}
-EXPORT_SYMBOL_GPL(clk_set_rate);
-
-int clk_set_rate_ex(struct clk *clk, unsigned long rate, int algo_id)
-{
         int ret = -EOPNOTSUPP;
         unsigned long flags;
 
         spin_lock_irqsave(&clock_lock, flags);
 
         if (likely(clk->ops && clk->ops->set_rate)) {
-                ret = clk->ops->set_rate(clk, rate, algo_id);
+                ret = clk->ops->set_rate(clk, rate);
                 if (ret != 0)
                         goto out_unlock;
         } else {
@@ -485,7 +482,7 @@ out_unlock:
 
         return ret;
 }
-EXPORT_SYMBOL_GPL(clk_set_rate_ex);
+EXPORT_SYMBOL_GPL(clk_set_rate);
 
 int clk_set_parent(struct clk *clk, struct clk *parent)
 {
@@ -571,7 +568,7 @@ long clk_round_parent(struct clk *clk, unsigned long target,
                                         *best_freq = freq_max;
                         }
 
-                        pr_debug("too low freq %lu, error %lu\n", freq->frequency,
+                        pr_debug("too low freq %u, error %lu\n", freq->frequency,
                                  target - freq_max);
 
                         if (!error)
@@ -591,7 +588,7 @@ long clk_round_parent(struct clk *clk, unsigned long target,
                                         *best_freq = freq_min;
                         }
 
-                        pr_debug("too high freq %lu, error %lu\n", freq->frequency,
+                        pr_debug("too high freq %u, error %lu\n", freq->frequency,
                                  freq_min - target);
 
                         if (!error)
@@ -653,8 +650,7 @@ static int clks_sysdev_suspend(struct sys_device *dev, pm_message_t state)
                                         clkp->ops->set_parent(clkp,
                                                 clkp->parent);
                                 if (likely(clkp->ops->set_rate))
-                                        clkp->ops->set_rate(clkp,
-                                                rate, NO_CHANGE);
+                                        clkp->ops->set_rate(clkp, rate);
                                 else if (likely(clkp->ops->recalc))
                                         clkp->rate = clkp->ops->recalc(clkp);
                         }
diff --git a/drivers/sh/clk/cpg.c b/drivers/sh/clk/cpg.c
index 3aea5f0ceb09..6172335ae323 100644
--- a/drivers/sh/clk/cpg.c
+++ b/drivers/sh/clk/cpg.c
@@ -110,8 +110,7 @@ static int sh_clk_div6_set_parent(struct clk *clk, struct clk *parent)
         return 0;
 }
 
-static int sh_clk_div6_set_rate(struct clk *clk,
-                                unsigned long rate, int algo_id)
+static int sh_clk_div6_set_rate(struct clk *clk, unsigned long rate)
 {
         unsigned long value;
         int idx;
@@ -132,7 +131,7 @@ static int sh_clk_div6_enable(struct clk *clk)
         unsigned long value;
         int ret;
 
-        ret = sh_clk_div6_set_rate(clk, clk->rate, 0);
+        ret = sh_clk_div6_set_rate(clk, clk->rate);
         if (ret == 0) {
                 value = __raw_readl(clk->enable_reg);
                 value &= ~0x100; /* clear stop bit to enable clock */
@@ -253,7 +252,7 @@ static int sh_clk_div4_set_parent(struct clk *clk, struct clk *parent)
         return 0;
 }
 
-static int sh_clk_div4_set_rate(struct clk *clk, unsigned long rate, int algo_id)
+static int sh_clk_div4_set_rate(struct clk *clk, unsigned long rate)
 {
         struct clk_div4_table *d4t = clk->priv;
         unsigned long value;
diff --git a/drivers/sh/intc/virq.c b/drivers/sh/intc/virq.c
index e5bf5d3c698e..4e0ff7181164 100644
--- a/drivers/sh/intc/virq.c
+++ b/drivers/sh/intc/virq.c
@@ -215,7 +215,7 @@ restart:
                 entry = radix_tree_deref_slot((void **)entries[i]);
                 if (unlikely(!entry))
                         continue;
-                if (unlikely(entry == RADIX_TREE_RETRY))
+                if (radix_tree_deref_retry(entry))
                         goto restart;
 
                 irq = create_irq();
diff --git a/drivers/staging/easycap/easycap.h b/drivers/staging/easycap/easycap.h
index f3c827eb0abe..25961c23dc0f 100644
--- a/drivers/staging/easycap/easycap.h
+++ b/drivers/staging/easycap/easycap.h
@@ -77,7 +77,6 @@
 #include <linux/slab.h>
 #include <linux/module.h>
 #include <linux/kref.h>
-#include <linux/smp_lock.h>
 #include <linux/usb.h>
 #include <linux/uaccess.h>
 
diff --git a/drivers/staging/hv/storvsc_drv.c b/drivers/staging/hv/storvsc_drv.c
index 41d9acf4cd61..6f8d67d0d64f 100644
--- a/drivers/staging/hv/storvsc_drv.c
+++ b/drivers/staging/hv/storvsc_drv.c
@@ -72,8 +72,7 @@ struct storvsc_driver_context {
 
 /* Static decl */
 static int storvsc_probe(struct device *dev);
-static int storvsc_queuecommand(struct scsi_cmnd *scmnd,
-                                void (*done)(struct scsi_cmnd *));
+static int storvsc_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *scmnd);
 static int storvsc_device_alloc(struct scsi_device *);
 static int storvsc_device_configure(struct scsi_device *);
 static int storvsc_host_reset_handler(struct scsi_cmnd *scmnd);
@@ -595,7 +594,7 @@ static unsigned int copy_from_bounce_buffer(struct scatterlist *orig_sgl,
 /*
  * storvsc_queuecommand - Initiate command processing
  */
-static int storvsc_queuecommand(struct scsi_cmnd *scmnd,
+static int storvsc_queuecommand_lck(struct scsi_cmnd *scmnd,
                                 void (*done)(struct scsi_cmnd *))
 {
         int ret;
@@ -783,6 +782,8 @@ retry_request:
         return ret;
 }
 
+static DEF_SCSI_QCMD(storvsc_queuecommand)
+
 static int storvsc_merge_bvec(struct request_queue *q,
                               struct bvec_merge_data *bmd, struct bio_vec *bvec)
 {
diff --git a/drivers/staging/intel_sst/intel_sst_app_interface.c b/drivers/staging/intel_sst/intel_sst_app_interface.c
index 9618c7997461..991440015e92 100644
--- a/drivers/staging/intel_sst/intel_sst_app_interface.c
+++ b/drivers/staging/intel_sst/intel_sst_app_interface.c
@@ -34,7 +34,6 @@
 #include <linux/uaccess.h>
 #include <linux/firmware.h>
 #include <linux/ioctl.h>
-#include <linux/smp_lock.h>
 #ifdef CONFIG_MRST_RAR_HANDLER
 #include <linux/rar_register.h>
 #include "../../../drivers/staging/memrar/memrar.h"
diff --git a/drivers/staging/keucr/scsiglue.c b/drivers/staging/keucr/scsiglue.c
index a2671404f7ac..da4f42af3838 100644
--- a/drivers/staging/keucr/scsiglue.c
+++ b/drivers/staging/keucr/scsiglue.c
@@ -87,7 +87,7 @@ static int slave_configure(struct scsi_device *sdev)
 
 /* This is always called with scsi_lock(host) held */
 //----- queuecommand() ---------------------
-static int queuecommand(struct scsi_cmnd *srb, void (*done)(struct scsi_cmnd *))
+static int queuecommand_lck(struct scsi_cmnd *srb, void (*done)(struct scsi_cmnd *))
 {
         struct us_data *us = host_to_us(srb->device->host);
 
@@ -117,6 +117,8 @@ static int queuecommand(struct scsi_cmnd *srb, void (*done)(struct scsi_cmnd *))
         return 0;
 }
 
+static DEF_SCSI_QCMD(queuecommand)
+
 /***********************************************************************
  * Error handling functions
  ***********************************************************************/
diff --git a/drivers/staging/olpc_dcon/olpc_dcon.c b/drivers/staging/olpc_dcon/olpc_dcon.c
index 75aa7a36307d..4ca45ec7fd84 100644
--- a/drivers/staging/olpc_dcon/olpc_dcon.c
+++ b/drivers/staging/olpc_dcon/olpc_dcon.c
@@ -17,7 +17,6 @@
 #include <linux/console.h>
 #include <linux/i2c.h>
 #include <linux/platform_device.h>
-#include <linux/i2c-id.h>
 #include <linux/pci.h>
 #include <linux/pci_ids.h>
 #include <linux/interrupt.h>
@@ -733,7 +732,6 @@ static int dcon_probe(struct i2c_client *client, const struct i2c_device_id *id)
  edev:
         platform_device_unregister(dcon_device);
         dcon_device = NULL;
-        i2c_set_clientdata(client, NULL);
  eirq:
         free_irq(DCON_IRQ, &dcon_driver);
  einit:
@@ -757,8 +755,6 @@ static int dcon_remove(struct i2c_client *client)
                 platform_device_unregister(dcon_device);
         cancel_work_sync(&dcon_work);
 
-        i2c_set_clientdata(client, NULL);
-
         return 0;
 }
 
diff --git a/drivers/staging/rtl8712/osdep_service.h b/drivers/staging/rtl8712/osdep_service.h
index d1674cd282dc..831d81e0e429 100644
--- a/drivers/staging/rtl8712/osdep_service.h
+++ b/drivers/staging/rtl8712/osdep_service.h
@@ -22,7 +22,6 @@
 #include <linux/module.h>
 #include <linux/sched.h>
 #include <linux/kref.h>
-#include <linux/smp_lock.h>
 #include <linux/netdevice.h>
 #include <linux/skbuff.h>
 #include <linux/usb.h>
diff --git a/drivers/staging/speakup/buffers.c b/drivers/staging/speakup/buffers.c
index b7b60d5e8660..a2db956edd54 100644
--- a/drivers/staging/speakup/buffers.c
+++ b/drivers/staging/speakup/buffers.c
@@ -1,5 +1,4 @@
 #include <linux/console.h>
-#include <linux/smp_lock.h>
 #include <linux/types.h>
 #include <linux/wait.h>
 
diff --git a/drivers/staging/stradis/Kconfig b/drivers/staging/stradis/Kconfig
index 92e891141896..02f0fc504cf5 100644
--- a/drivers/staging/stradis/Kconfig
+++ b/drivers/staging/stradis/Kconfig
@@ -1,6 +1,6 @@
 config VIDEO_STRADIS
         tristate "Stradis 4:2:2 MPEG-2 video driver (DEPRECATED)"
-        depends on EXPERIMENTAL && PCI && VIDEO_V4L1 && VIRT_TO_BUS
+        depends on EXPERIMENTAL && PCI && VIDEO_V4L1 && VIRT_TO_BUS && BKL
         help
           Say Y here to enable support for the Stradis 4:2:2 MPEG-2 video
           driver for PCI.  There is a product page at
diff --git a/drivers/tty/sysrq.c b/drivers/tty/sysrq.c
index eaa5d3efa79d..c556ed9db13d 100644
--- a/drivers/tty/sysrq.c
+++ b/drivers/tty/sysrq.c
@@ -554,7 +554,7 @@ EXPORT_SYMBOL(handle_sysrq);
 #ifdef CONFIG_INPUT
 
 /* Simple translation table for the SysRq keys */
-static const unsigned char sysrq_xlate[KEY_MAX + 1] =
+static const unsigned char sysrq_xlate[KEY_CNT] =
         "\000\0331234567890-=\177\t"                    /* 0x00 - 0x0f */
         "qwertyuiop[]\r\000as"                          /* 0x10 - 0x1f */
         "dfghjkl;'`\000\\zxcv"                          /* 0x20 - 0x2f */
@@ -563,53 +563,129 @@ static const unsigned char sysrq_xlate[KEY_MAX + 1] =
         "230\177\000\000\213\214\000\000\000\000\000\000\000\000\000\000" /* 0x50 - 0x5f */
         "\r\000/";                                      /* 0x60 - 0x6f */
 
-static bool sysrq_down;
-static int sysrq_alt_use;
-static int sysrq_alt;
-static DEFINE_SPINLOCK(sysrq_event_lock);
+struct sysrq_state {
+        struct input_handle handle;
+        struct work_struct reinject_work;
+        unsigned long key_down[BITS_TO_LONGS(KEY_CNT)];
+        unsigned int alt;
+        unsigned int alt_use;
+        bool active;
+        bool need_reinject;
+};
+
+static void sysrq_reinject_alt_sysrq(struct work_struct *work)
+{
+        struct sysrq_state *sysrq =
+                        container_of(work, struct sysrq_state, reinject_work);
+        struct input_handle *handle = &sysrq->handle;
+        unsigned int alt_code = sysrq->alt_use;
+
+        if (sysrq->need_reinject) {
+                /* Simulate press and release of Alt + SysRq */
+                input_inject_event(handle, EV_KEY, alt_code, 1);
+                input_inject_event(handle, EV_KEY, KEY_SYSRQ, 1);
+                input_inject_event(handle, EV_SYN, SYN_REPORT, 1);
+
+                input_inject_event(handle, EV_KEY, KEY_SYSRQ, 0);
+                input_inject_event(handle, EV_KEY, alt_code, 0);
+                input_inject_event(handle, EV_SYN, SYN_REPORT, 1);
+        }
+}
 
-static bool sysrq_filter(struct input_handle *handle, unsigned int type,
-                         unsigned int code, int value)
+static bool sysrq_filter(struct input_handle *handle,
+                         unsigned int type, unsigned int code, int value)
 {
+        struct sysrq_state *sysrq = handle->private;
+        bool was_active = sysrq->active;
         bool suppress;
 
-        /* We are called with interrupts disabled, just take the lock */
-        spin_lock(&sysrq_event_lock);
+        switch (type) {
 
-        if (type != EV_KEY)
-                goto out;
+        case EV_SYN:
+                suppress = false;
+                break;
 
-        switch (code) {
+        case EV_KEY:
+                switch (code) {
 
-        case KEY_LEFTALT:
-        case KEY_RIGHTALT:
-                if (value)
-                        sysrq_alt = code;
-                else {
-                        if (sysrq_down && code == sysrq_alt_use)
-                                sysrq_down = false;
+                case KEY_LEFTALT:
+                case KEY_RIGHTALT:
+                        if (!value) {
+                                /* One of ALTs is being released */
+                                if (sysrq->active && code == sysrq->alt_use)
+                                        sysrq->active = false;
 
-                        sysrq_alt = 0;
+                                sysrq->alt = KEY_RESERVED;
+
+                        } else if (value != 2) {
+                                sysrq->alt = code;
+                                sysrq->need_reinject = false;
+                        }
+                        break;
+
+                case KEY_SYSRQ:
+                        if (value == 1 && sysrq->alt != KEY_RESERVED) {
+                                sysrq->active = true;
+                                sysrq->alt_use = sysrq->alt;
+                                /*
+                                 * If nothing else will be pressed we'll need
+                                 * to * re-inject Alt-SysRq keysroke.
+                                 */
+                                sysrq->need_reinject = true;
+                        }
+
+                        /*
+                         * Pretend that sysrq was never pressed at all. This
+                         * is needed to properly handle KGDB which will try
+                         * to release all keys after exiting debugger. If we
+                         * do not clear key bit it KGDB will end up sending
+                         * release events for Alt and SysRq, potentially
+                         * triggering print screen function.
+                         */
+                        if (sysrq->active)
+                                clear_bit(KEY_SYSRQ, handle->dev->key);
+
+                        break;
+
+                default:
+                        if (sysrq->active && value && value != 2) {
+                                sysrq->need_reinject = false;
+                                __handle_sysrq(sysrq_xlate[code], true);
+                        }
+                        break;
                 }
-                break;
 
-        case KEY_SYSRQ:
-                if (value == 1 && sysrq_alt) {
-                        sysrq_down = true;
-                        sysrq_alt_use = sysrq_alt;
+                suppress = sysrq->active;
+
+                if (!sysrq->active) {
+                        /*
+                         * If we are not suppressing key presses keep track of
+                         * keyboard state so we can release keys that have been
+                         * pressed before entering SysRq mode.
+                         */
+                        if (value)
+                                set_bit(code, sysrq->key_down);
+                        else
+                                clear_bit(code, sysrq->key_down);
+
+                        if (was_active)
+                                schedule_work(&sysrq->reinject_work);
+
+                } else if (value == 0 &&
+                           test_and_clear_bit(code, sysrq->key_down)) {
+                        /*
+                         * Pass on release events for keys that was pressed before
+                         * entering SysRq mode.
+                         */
+                        suppress = false;
                 }
                 break;
 
         default:
-                if (sysrq_down && value && value != 2)
-                        __handle_sysrq(sysrq_xlate[code], true);
+                suppress = sysrq->active;
                 break;
         }
 
-out:
-        suppress = sysrq_down;
-        spin_unlock(&sysrq_event_lock);
-
         return suppress;
 }
 
@@ -617,28 +693,28 @@ static int sysrq_connect(struct input_handler *handler,
                          struct input_dev *dev,
                          const struct input_device_id *id)
 {
-        struct input_handle *handle;
+        struct sysrq_state *sysrq;
         int error;
 
-        sysrq_down = false;
-        sysrq_alt = 0;
-
-        handle = kzalloc(sizeof(struct input_handle), GFP_KERNEL);
-        if (!handle)
+        sysrq = kzalloc(sizeof(struct sysrq_state), GFP_KERNEL);
+        if (!sysrq)
                 return -ENOMEM;
 
-        handle->dev = dev;
-        handle->handler = handler;
-        handle->name = "sysrq";
+        INIT_WORK(&sysrq->reinject_work, sysrq_reinject_alt_sysrq);
+
+        sysrq->handle.dev = dev;
+        sysrq->handle.handler = handler;
+        sysrq->handle.name = "sysrq";
+        sysrq->handle.private = sysrq;
 
-        error = input_register_handle(handle);
+        error = input_register_handle(&sysrq->handle);
         if (error) {
                 pr_err("Failed to register input sysrq handler, error %d\n",
                         error);
                 goto err_free;
         }
 
-        error = input_open_device(handle);
+        error = input_open_device(&sysrq->handle);
         if (error) {
                 pr_err("Failed to open input device, error %d\n", error);
                 goto err_unregister;
@@ -647,17 +723,20 @@ static int sysrq_connect(struct input_handler *handler,
         return 0;
 
  err_unregister:
-        input_unregister_handle(handle);
+        input_unregister_handle(&sysrq->handle);
  err_free:
-        kfree(handle);
+        kfree(sysrq);
         return error;
 }
 
 static void sysrq_disconnect(struct input_handle *handle)
 {
+        struct sysrq_state *sysrq = handle->private;
+
         input_close_device(handle);
+        cancel_work_sync(&sysrq->reinject_work);
         input_unregister_handle(handle);
-        kfree(handle);
+        kfree(sysrq);
 }
 
 /*
diff --git a/drivers/usb/atm/ueagle-atm.c b/drivers/usb/atm/ueagle-atm.c
index ea071a5b6eee..44447f54942f 100644
--- a/drivers/usb/atm/ueagle-atm.c
+++ b/drivers/usb/atm/ueagle-atm.c
@@ -2301,7 +2301,7 @@ out:
         return ret;
 }
 
-static DEVICE_ATTR(stat_status, S_IWUGO | S_IRUGO, read_status, reboot);
+static DEVICE_ATTR(stat_status, S_IWUSR | S_IRUGO, read_status, reboot);
 
 static ssize_t read_human_status(struct device *dev,
                         struct device_attribute *attr, char *buf)
@@ -2364,8 +2364,7 @@ out:
         return ret;
 }
 
-static DEVICE_ATTR(stat_human_status, S_IWUGO | S_IRUGO,
-                                read_human_status, NULL);
+static DEVICE_ATTR(stat_human_status, S_IRUGO, read_human_status, NULL);
 
 static ssize_t read_delin(struct device *dev, struct device_attribute *attr,
                 char *buf)
@@ -2397,7 +2396,7 @@ out:
         return ret;
 }
 
-static DEVICE_ATTR(stat_delin, S_IWUGO | S_IRUGO, read_delin, NULL);
+static DEVICE_ATTR(stat_delin, S_IRUGO, read_delin, NULL);
 
 #define UEA_ATTR(name, reset)                                   \
                                                                 \
diff --git a/drivers/usb/core/devices.c b/drivers/usb/core/devices.c
index ddb4dc980923..a3d2e2399655 100644
--- a/drivers/usb/core/devices.c
+++ b/drivers/usb/core/devices.c
@@ -54,7 +54,6 @@
 #include <linux/gfp.h>
 #include <linux/poll.h>
 #include <linux/usb.h>
-#include <linux/smp_lock.h>
 #include <linux/usbdevice_fs.h>
 #include <linux/usb/hcd.h>
 #include <linux/mutex.h>
diff --git a/drivers/usb/core/devio.c b/drivers/usb/core/devio.c
index 045bb4b823e1..a7131ad630f9 100644
--- a/drivers/usb/core/devio.c
+++ b/drivers/usb/core/devio.c
@@ -37,7 +37,6 @@
 #include <linux/fs.h>
 #include <linux/mm.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/signal.h>
 #include <linux/poll.h>
 #include <linux/module.h>
diff --git a/drivers/usb/core/file.c b/drivers/usb/core/file.c
index 9fe34fb78ef1..cf6a5423de09 100644
--- a/drivers/usb/core/file.c
+++ b/drivers/usb/core/file.c
@@ -19,7 +19,6 @@
 #include <linux/errno.h>
 #include <linux/rwsem.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/usb.h>
 
 #include "usb.h"
diff --git a/drivers/usb/core/inode.c b/drivers/usb/core/inode.c
index 9819a4cc3b26..b690aa35df9a 100644
--- a/drivers/usb/core/inode.c
+++ b/drivers/usb/core/inode.c
@@ -39,7 +39,6 @@
 #include <linux/parser.h>
 #include <linux/notifier.h>
 #include <linux/seq_file.h>
-#include <linux/smp_lock.h>
 #include <linux/usb/hcd.h>
 #include <asm/byteorder.h>
 #include "usb.h"
diff --git a/drivers/usb/gadget/atmel_usba_udc.c b/drivers/usb/gadget/atmel_usba_udc.c
index b5e20e873cba..717ff653fa23 100644
--- a/drivers/usb/gadget/atmel_usba_udc.c
+++ b/drivers/usb/gadget/atmel_usba_udc.c
@@ -2017,7 +2017,7 @@ static int __init usba_udc_probe(struct platform_device *pdev)
                         }
                 } else {
                         /* gpio_request fail so use -EINVAL for gpio_is_valid */
-                        ubc->vbus_pin = -EINVAL;
+                        udc->vbus_pin = -EINVAL;
                 }
         }
 
diff --git a/drivers/usb/gadget/f_fs.c b/drivers/usb/gadget/f_fs.c
index 4a830df4fc31..484c5ba5450e 100644
--- a/drivers/usb/gadget/f_fs.c
+++ b/drivers/usb/gadget/f_fs.c
@@ -30,7 +30,6 @@
 #include <linux/blkdev.h>
 #include <linux/pagemap.h>
 #include <asm/unaligned.h>
-#include <linux/smp_lock.h>
 
 #include <linux/usb/composite.h>
 #include <linux/usb/functionfs.h>
diff --git a/drivers/usb/gadget/f_hid.c b/drivers/usb/gadget/f_hid.c
index 4f891eddd060..598e7e2ab80c 100644
--- a/drivers/usb/gadget/f_hid.c
+++ b/drivers/usb/gadget/f_hid.c
@@ -25,7 +25,6 @@
 #include <linux/cdev.h>
 #include <linux/mutex.h>
 #include <linux/poll.h>
-#include <linux/smp_lock.h>
 #include <linux/uaccess.h>
 #include <linux/wait.h>
 #include <linux/usb/g_hid.h>
diff --git a/drivers/usb/host/ehci-dbg.c b/drivers/usb/host/ehci-dbg.c
index 86afdc73322f..6e2599661b5b 100644
--- a/drivers/usb/host/ehci-dbg.c
+++ b/drivers/usb/host/ehci-dbg.c
@@ -1067,7 +1067,7 @@ static inline void create_debug_files (struct ehci_hcd *ehci)
                                                     &debug_registers_fops))
                 goto file_error;
 
-        if (!debugfs_create_file("lpm", S_IRUGO|S_IWUGO, ehci->debug_dir, bus,
+        if (!debugfs_create_file("lpm", S_IRUGO|S_IWUSR, ehci->debug_dir, bus,
                                                     &debug_lpm_fops))
                 goto file_error;
 
diff --git a/drivers/usb/host/ehci-hcd.c b/drivers/usb/host/ehci-hcd.c
index 502a7e6fef42..e9062806d4a2 100644
--- a/drivers/usb/host/ehci-hcd.c
+++ b/drivers/usb/host/ehci-hcd.c
@@ -1063,10 +1063,11 @@ rescan:
                                 tmp && tmp != qh;
                                 tmp = tmp->qh_next.qh)
                         continue;
-                /* periodic qh self-unlinks on empty */
-                if (!tmp)
-                        goto nogood;
-                unlink_async (ehci, qh);
+                /* periodic qh self-unlinks on empty, and a COMPLETING qh
+                 * may already be unlinked.
+                 */
+                if (tmp)
+                        unlink_async(ehci, qh);
                 /* FALL THROUGH */
         case QH_STATE_UNLINK:           /* wait for hw to finish? */
         case QH_STATE_UNLINK_WAIT:
@@ -1083,7 +1084,6 @@ idle_timeout:
                 }
                 /* else FALL THROUGH */
         default:
-nogood:
                 /* caller was supposed to have unlinked any requests;
                  * that's not our job.  just leak this memory.
                  */
diff --git a/drivers/usb/host/ehci-mem.c b/drivers/usb/host/ehci-mem.c
index d36e4e75e08d..12f70c302b0b 100644
--- a/drivers/usb/host/ehci-mem.c
+++ b/drivers/usb/host/ehci-mem.c
@@ -141,6 +141,10 @@ static void ehci_mem_cleanup (struct ehci_hcd *ehci)
                 qh_put (ehci->async);
         ehci->async = NULL;
 
+        if (ehci->dummy)
+                qh_put(ehci->dummy);
+        ehci->dummy = NULL;
+
         /* DMA consistent memory and pools */
         if (ehci->qtd_pool)
                 dma_pool_destroy (ehci->qtd_pool);
@@ -227,8 +231,26 @@ static int ehci_mem_init (struct ehci_hcd *ehci, gfp_t flags)
         if (ehci->periodic == NULL) {
                 goto fail;
         }
-        for (i = 0; i < ehci->periodic_size; i++)
-                ehci->periodic [i] = EHCI_LIST_END(ehci);
+
+        if (ehci->use_dummy_qh) {
+                struct ehci_qh_hw       *hw;
+                ehci->dummy = ehci_qh_alloc(ehci, flags);
+                if (!ehci->dummy)
+                        goto fail;
+
+                hw = ehci->dummy->hw;
+                hw->hw_next = EHCI_LIST_END(ehci);
+                hw->hw_qtd_next = EHCI_LIST_END(ehci);
+                hw->hw_alt_next = EHCI_LIST_END(ehci);
+                hw->hw_token &= ~QTD_STS_ACTIVE;
+                ehci->dummy->hw = hw;
+
+                for (i = 0; i < ehci->periodic_size; i++)
+                        ehci->periodic[i] = ehci->dummy->qh_dma;
+        } else {
+                for (i = 0; i < ehci->periodic_size; i++)
+                        ehci->periodic[i] = EHCI_LIST_END(ehci);
+        }
 
         /* software shadow of hardware table */
         ehci->pshadow = kcalloc(ehci->periodic_size, sizeof(void *), flags);
diff --git a/drivers/usb/host/ehci-pci.c b/drivers/usb/host/ehci-pci.c
index a1e8d273103f..01bb72b71832 100644
--- a/drivers/usb/host/ehci-pci.c
+++ b/drivers/usb/host/ehci-pci.c
@@ -103,6 +103,19 @@ static int ehci_pci_setup(struct usb_hcd *hcd)
         if (retval)
                 return retval;
 
+        if ((pdev->vendor == PCI_VENDOR_ID_AMD && pdev->device == 0x7808) ||
+            (pdev->vendor == PCI_VENDOR_ID_ATI && pdev->device == 0x4396)) {
+                /* EHCI controller on AMD SB700/SB800/Hudson-2/3 platforms may
+                 * read/write memory space which does not belong to it when
+                 * there is NULL pointer with T-bit set to 1 in the frame list
+                 * table. To avoid the issue, the frame list link pointer
+                 * should always contain a valid pointer to a inactive qh.
+                 */
+                ehci->use_dummy_qh = 1;
+                ehci_info(ehci, "applying AMD SB700/SB800/Hudson-2/3 EHCI "
+                                "dummy qh workaround\n");
+        }
+
         /* data structure init */
         retval = ehci_init(hcd);
         if (retval)
diff --git a/drivers/usb/host/ehci-sched.c b/drivers/usb/host/ehci-sched.c
index a92526d6e5ae..d9f78eb26572 100644
--- a/drivers/usb/host/ehci-sched.c
+++ b/drivers/usb/host/ehci-sched.c
@@ -98,7 +98,14 @@ static void periodic_unlink (struct ehci_hcd *ehci, unsigned frame, void *ptr)
          */
         *prev_p = *periodic_next_shadow(ehci, &here,
                         Q_NEXT_TYPE(ehci, *hw_p));
-        *hw_p = *shadow_next_periodic(ehci, &here, Q_NEXT_TYPE(ehci, *hw_p));
+
+        if (!ehci->use_dummy_qh ||
+            *shadow_next_periodic(ehci, &here, Q_NEXT_TYPE(ehci, *hw_p))
+                        != EHCI_LIST_END(ehci))
+                *hw_p = *shadow_next_periodic(ehci, &here,
+                                Q_NEXT_TYPE(ehci, *hw_p));
+        else
+                *hw_p = ehci->dummy->qh_dma;
 }
 
 /* how many of the uframe's 125 usecs are allocated? */
@@ -2335,7 +2342,11 @@ restart:
                                  * pointer for much longer, if at all.
                                  */
                                 *q_p = q.itd->itd_next;
-                                *hw_p = q.itd->hw_next;
+                                if (!ehci->use_dummy_qh ||
+                                    q.itd->hw_next != EHCI_LIST_END(ehci))
+                                        *hw_p = q.itd->hw_next;
+                                else
+                                        *hw_p = ehci->dummy->qh_dma;
                                 type = Q_NEXT_TYPE(ehci, q.itd->hw_next);
                                 wmb();
                                 modified = itd_complete (ehci, q.itd);
@@ -2368,7 +2379,11 @@ restart:
                                  * URB completion.
                                  */
                                 *q_p = q.sitd->sitd_next;
-                                *hw_p = q.sitd->hw_next;
+                                if (!ehci->use_dummy_qh ||
+                                    q.sitd->hw_next != EHCI_LIST_END(ehci))
+                                        *hw_p = q.sitd->hw_next;
+                                else
+                                        *hw_p = ehci->dummy->qh_dma;
                                 type = Q_NEXT_TYPE(ehci, q.sitd->hw_next);
                                 wmb();
                                 modified = sitd_complete (ehci, q.sitd);
diff --git a/drivers/usb/host/ehci.h b/drivers/usb/host/ehci.h
index bde823f704e9..ba8eab366b82 100644
--- a/drivers/usb/host/ehci.h
+++ b/drivers/usb/host/ehci.h
@@ -73,6 +73,7 @@ struct ehci_hcd {			/* one per controller */
 
         /* async schedule support */
         struct ehci_qh          *async;
+        struct ehci_qh          *dummy;         /* For AMD quirk use */
         struct ehci_qh          *reclaim;
         unsigned                scanning : 1;
 
@@ -131,6 +132,7 @@ struct ehci_hcd {			/* one per controller */
         unsigned                need_io_watchdog:1;
         unsigned                broken_periodic:1;
         unsigned                fs_i_thresh:1;  /* Intel iso scheduling */
+        unsigned                use_dummy_qh:1; /* AMD Frame List table quirk*/
 
         /* required for usb32 quirk */
         #define OHCI_CTRL_HCFS          (3 << 6)
diff --git a/drivers/usb/host/isp1362-hcd.c b/drivers/usb/host/isp1362-hcd.c
index 8196fa11fec4..43a39eb56cc6 100644
--- a/drivers/usb/host/isp1362-hcd.c
+++ b/drivers/usb/host/isp1362-hcd.c
@@ -70,7 +70,6 @@
 #include <linux/ioport.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/errno.h>
 #include <linux/init.h>
 #include <linux/list.h>
@@ -2684,7 +2683,7 @@ static int __devexit isp1362_remove(struct platform_device *pdev)
         return 0;
 }
 
-static int __init isp1362_probe(struct platform_device *pdev)
+static int __devinit isp1362_probe(struct platform_device *pdev)
 {
         struct usb_hcd *hcd;
         struct isp1362_hcd *isp1362_hcd;
diff --git a/drivers/usb/host/uhci-debug.c b/drivers/usb/host/uhci-debug.c
index 6e7fb5f38db6..ee60cd3ea642 100644
--- a/drivers/usb/host/uhci-debug.c
+++ b/drivers/usb/host/uhci-debug.c
@@ -12,7 +12,6 @@
 #include <linux/slab.h>
 #include <linux/kernel.h>
 #include <linux/debugfs.h>
-#include <linux/smp_lock.h>
 #include <asm/io.h>
 
 #include "uhci-hcd.h"
diff --git a/drivers/usb/host/xhci-mem.c b/drivers/usb/host/xhci-mem.c
index 202770676da3..d178761c3981 100644
--- a/drivers/usb/host/xhci-mem.c
+++ b/drivers/usb/host/xhci-mem.c
@@ -1045,7 +1045,7 @@ static inline u32 xhci_get_max_esit_payload(struct xhci_hcd *xhci,
         if (udev->speed == USB_SPEED_SUPER)
                 return ep->ss_ep_comp.wBytesPerInterval;
 
-        max_packet = ep->desc.wMaxPacketSize & 0x3ff;
+        max_packet = GET_MAX_PACKET(ep->desc.wMaxPacketSize);
         max_burst = (ep->desc.wMaxPacketSize & 0x1800) >> 11;
         /* A 0 in max burst means 1 transfer per ESIT */
         return max_packet * (max_burst + 1);
@@ -1135,7 +1135,7 @@ int xhci_endpoint_init(struct xhci_hcd *xhci,
                 /* Fall through */
         case USB_SPEED_FULL:
         case USB_SPEED_LOW:
-                max_packet = ep->desc.wMaxPacketSize & 0x3ff;
+                max_packet = GET_MAX_PACKET(ep->desc.wMaxPacketSize);
                 ep_ctx->ep_info2 |= MAX_PACKET(max_packet);
                 break;
         default:
diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c
index 9f3115e729b1..df558f6f84e3 100644
--- a/drivers/usb/host/xhci-ring.c
+++ b/drivers/usb/host/xhci-ring.c
@@ -2104,7 +2104,6 @@ irqreturn_t xhci_irq(struct usb_hcd *hcd)
 
         if (!(status & STS_EINT)) {
                 spin_unlock(&xhci->lock);
-                xhci_warn(xhci, "Spurious interrupt.\n");
                 return IRQ_NONE;
         }
         xhci_dbg(xhci, "op reg status = %08x\n", status);
diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c
index 5d7d4e951ea4..06fca0835b52 100644
--- a/drivers/usb/host/xhci.c
+++ b/drivers/usb/host/xhci.c
@@ -577,6 +577,65 @@ static void xhci_restore_registers(struct xhci_hcd *xhci)
         xhci_write_64(xhci, xhci->s3.erst_base, &xhci->ir_set->erst_base);
 }
 
+static void xhci_set_cmd_ring_deq(struct xhci_hcd *xhci)
+{
+        u64     val_64;
+
+        /* step 2: initialize command ring buffer */
+        val_64 = xhci_read_64(xhci, &xhci->op_regs->cmd_ring);
+        val_64 = (val_64 & (u64) CMD_RING_RSVD_BITS) |
+                (xhci_trb_virt_to_dma(xhci->cmd_ring->deq_seg,
+                                      xhci->cmd_ring->dequeue) &
+                 (u64) ~CMD_RING_RSVD_BITS) |
+                xhci->cmd_ring->cycle_state;
+        xhci_dbg(xhci, "// Setting command ring address to 0x%llx\n",
+                        (long unsigned long) val_64);
+        xhci_write_64(xhci, val_64, &xhci->op_regs->cmd_ring);
+}
+
+/*
+ * The whole command ring must be cleared to zero when we suspend the host.
+ *
+ * The host doesn't save the command ring pointer in the suspend well, so we
+ * need to re-program it on resume.  Unfortunately, the pointer must be 64-byte
+ * aligned, because of the reserved bits in the command ring dequeue pointer
+ * register.  Therefore, we can't just set the dequeue pointer back in the
+ * middle of the ring (TRBs are 16-byte aligned).
+ */
+static void xhci_clear_command_ring(struct xhci_hcd *xhci)
+{
+        struct xhci_ring *ring;
+        struct xhci_segment *seg;
+
+        ring = xhci->cmd_ring;
+        seg = ring->deq_seg;
+        do {
+                memset(seg->trbs, 0, SEGMENT_SIZE);
+                seg = seg->next;
+        } while (seg != ring->deq_seg);
+
+        /* Reset the software enqueue and dequeue pointers */
+        ring->deq_seg = ring->first_seg;
+        ring->dequeue = ring->first_seg->trbs;
+        ring->enq_seg = ring->deq_seg;
+        ring->enqueue = ring->dequeue;
+
+        /*
+         * Ring is now zeroed, so the HW should look for change of ownership
+         * when the cycle bit is set to 1.
+         */
+        ring->cycle_state = 1;
+
+        /*
+         * Reset the hardware dequeue pointer.
+         * Yes, this will need to be re-written after resume, but we're paranoid
+         * and want to make sure the hardware doesn't access bogus memory
+         * because, say, the BIOS or an SMI started the host without changing
+         * the command ring pointers.
+         */
+        xhci_set_cmd_ring_deq(xhci);
+}
+
 /*
  * Stop HC (not bus-specific)
  *
@@ -604,6 +663,7 @@ int xhci_suspend(struct xhci_hcd *xhci)
                 spin_unlock_irq(&xhci->lock);
                 return -ETIMEDOUT;
         }
+        xhci_clear_command_ring(xhci);
 
         /* step 3: save registers */
         xhci_save_registers(xhci);
@@ -635,7 +695,6 @@ int xhci_resume(struct xhci_hcd *xhci, bool hibernated)
         u32                     command, temp = 0;
         struct usb_hcd          *hcd = xhci_to_hcd(xhci);
         struct pci_dev          *pdev = to_pci_dev(hcd->self.controller);
-        u64     val_64;
         int     old_state, retval;
 
         old_state = hcd->state;
@@ -648,15 +707,7 @@ int xhci_resume(struct xhci_hcd *xhci, bool hibernated)
                 /* step 1: restore register */
                 xhci_restore_registers(xhci);
                 /* step 2: initialize command ring buffer */
-                val_64 = xhci_read_64(xhci, &xhci->op_regs->cmd_ring);
-                val_64 = (val_64 & (u64) CMD_RING_RSVD_BITS) |
-                         (xhci_trb_virt_to_dma(xhci->cmd_ring->deq_seg,
-                                               xhci->cmd_ring->dequeue) &
-                         (u64) ~CMD_RING_RSVD_BITS) |
-                         xhci->cmd_ring->cycle_state;
-                xhci_dbg(xhci, "// Setting command ring address to 0x%llx\n",
-                                (long unsigned long) val_64);
-                xhci_write_64(xhci, val_64, &xhci->op_regs->cmd_ring);
+                xhci_set_cmd_ring_deq(xhci);
                 /* step 3: restore state and start state*/
                 /* step 3: set CRS flag */
                 command = xhci_readl(xhci, &xhci->op_regs->command);
@@ -714,6 +765,7 @@ int xhci_resume(struct xhci_hcd *xhci, bool hibernated)
                 return retval;
         }
 
+        spin_unlock_irq(&xhci->lock);
         /* Re-setup MSI-X */
         if (hcd->irq)
                 free_irq(hcd->irq, hcd);
@@ -736,6 +788,7 @@ int xhci_resume(struct xhci_hcd *xhci, bool hibernated)
                 hcd->irq = pdev->irq;
         }
 
+        spin_lock_irq(&xhci->lock);
         /* step 4: set Run/Stop bit */
         command = xhci_readl(xhci, &xhci->op_regs->command);
         command |= CMD_RUN;
diff --git a/drivers/usb/host/xhci.h b/drivers/usb/host/xhci.h
index 93d3bf4d213c..85e65647d445 100644
--- a/drivers/usb/host/xhci.h
+++ b/drivers/usb/host/xhci.h
@@ -621,6 +621,11 @@ struct xhci_ep_ctx {
 #define MAX_PACKET_MASK         (0xffff << 16)
 #define MAX_PACKET_DECODED(p)   (((p) >> 16) & 0xffff)
 
+/* Get max packet size from ep desc. Bit 10..0 specify the max packet size.
+ * USB2.0 spec 9.6.6.
+ */
+#define GET_MAX_PACKET(p)       ((p) & 0x7ff)
+
 /* tx_info bitmasks */
 #define AVG_TRB_LENGTH_FOR_EP(p)        ((p) & 0xffff)
 #define MAX_ESIT_PAYLOAD_FOR_EP(p)      (((p) & 0xffff) << 16)
diff --git a/drivers/usb/image/microtek.c b/drivers/usb/image/microtek.c
index 5a47805d9580..c90c89dc0003 100644
--- a/drivers/usb/image/microtek.c
+++ b/drivers/usb/image/microtek.c
@@ -364,7 +364,7 @@ static int mts_scsi_host_reset(struct scsi_cmnd *srb)
 }
 
 static int
-mts_scsi_queuecommand(struct scsi_cmnd *srb, mts_scsi_cmnd_callback callback);
+mts_scsi_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *srb);
 
 static void mts_transfer_cleanup( struct urb *transfer );
 static void mts_do_sg(struct urb * transfer);
@@ -573,7 +573,7 @@ mts_build_transfer_context(struct scsi_cmnd *srb, struct mts_desc* desc)
 
 
 static int
-mts_scsi_queuecommand(struct scsi_cmnd *srb, mts_scsi_cmnd_callback callback)
+mts_scsi_queuecommand_lck(struct scsi_cmnd *srb, mts_scsi_cmnd_callback callback)
 {
         struct mts_desc* desc = (struct mts_desc*)(srb->device->host->hostdata[0]);
         int err = 0;
@@ -626,6 +626,8 @@ out:
         return err;
 }
 
+static DEF_SCSI_QCMD(mts_scsi_queuecommand)
+
 static struct scsi_host_template mts_scsi_host_template = {
         .module                 = THIS_MODULE,
         .name                   = "microtekX6",
diff --git a/drivers/usb/misc/cypress_cy7c63.c b/drivers/usb/misc/cypress_cy7c63.c
index 2f43c57743c9..9251773ecef4 100644
--- a/drivers/usb/misc/cypress_cy7c63.c
+++ b/drivers/usb/misc/cypress_cy7c63.c
@@ -196,11 +196,9 @@ static ssize_t get_port1_handler(struct device *dev,
         return read_port(dev, attr, buf, 1, CYPRESS_READ_PORT_ID1);
 }
 
-static DEVICE_ATTR(port0, S_IWUGO | S_IRUGO,
-                   get_port0_handler, set_port0_handler);
+static DEVICE_ATTR(port0, S_IRUGO | S_IWUSR, get_port0_handler, set_port0_handler);
 
-static DEVICE_ATTR(port1, S_IWUGO | S_IRUGO,
-                   get_port1_handler, set_port1_handler);
+static DEVICE_ATTR(port1, S_IRUGO | S_IWUSR, get_port1_handler, set_port1_handler);
 
 
 static int cypress_probe(struct usb_interface *interface,
diff --git a/drivers/usb/misc/trancevibrator.c b/drivers/usb/misc/trancevibrator.c
index d77aba46ae85..f63776a48e2a 100644
--- a/drivers/usb/misc/trancevibrator.c
+++ b/drivers/usb/misc/trancevibrator.c
@@ -86,7 +86,7 @@ static ssize_t set_speed(struct device *dev, struct device_attribute *attr,
         return count;
 }
 
-static DEVICE_ATTR(speed, S_IWUGO | S_IRUGO, show_speed, set_speed);
+static DEVICE_ATTR(speed, S_IRUGO | S_IWUSR, show_speed, set_speed);
 
 static int tv_probe(struct usb_interface *interface,
                     const struct usb_device_id *id)
diff --git a/drivers/usb/misc/usbled.c b/drivers/usb/misc/usbled.c
index 63da2c3c838f..c96f51de1696 100644
--- a/drivers/usb/misc/usbled.c
+++ b/drivers/usb/misc/usbled.c
@@ -94,7 +94,7 @@ static ssize_t set_##value(struct device *dev, struct device_attribute *attr, co
         change_color(led);                                              \
         return count;                                                   \
 }                                                                       \
-static DEVICE_ATTR(value, S_IWUGO | S_IRUGO, show_##value, set_##value);
+static DEVICE_ATTR(value, S_IRUGO | S_IWUSR, show_##value, set_##value);
 show_set(blue);
 show_set(red);
 show_set(green);
diff --git a/drivers/usb/misc/usbsevseg.c b/drivers/usb/misc/usbsevseg.c
index de8ef945b536..417b8f207e8b 100644
--- a/drivers/usb/misc/usbsevseg.c
+++ b/drivers/usb/misc/usbsevseg.c
@@ -192,7 +192,7 @@ static ssize_t set_attr_##name(struct device *dev, 		\
                                                                 \
         return count;                                           \
 }                                                               \
-static DEVICE_ATTR(name, S_IWUGO | S_IRUGO, show_attr_##name, set_attr_##name);
+static DEVICE_ATTR(name, S_IRUGO | S_IWUSR, show_attr_##name, set_attr_##name);
 
 static ssize_t show_attr_text(struct device *dev,
         struct device_attribute *attr, char *buf)
@@ -223,7 +223,7 @@ static ssize_t set_attr_text(struct device *dev,
         return count;
 }
 
-static DEVICE_ATTR(text, S_IWUGO | S_IRUGO, show_attr_text, set_attr_text);
+static DEVICE_ATTR(text, S_IRUGO | S_IWUSR, show_attr_text, set_attr_text);
 
 static ssize_t show_attr_decimals(struct device *dev,
         struct device_attribute *attr, char *buf)
@@ -272,8 +272,7 @@ static ssize_t set_attr_decimals(struct device *dev,
         return count;
 }
 
-static DEVICE_ATTR(decimals, S_IWUGO | S_IRUGO,
-        show_attr_decimals, set_attr_decimals);
+static DEVICE_ATTR(decimals, S_IRUGO | S_IWUSR, show_attr_decimals, set_attr_decimals);
 
 static ssize_t show_attr_textmode(struct device *dev,
         struct device_attribute *attr, char *buf)
@@ -319,8 +318,7 @@ static ssize_t set_attr_textmode(struct device *dev,
         return -EINVAL;
 }
 
-static DEVICE_ATTR(textmode, S_IWUGO | S_IRUGO,
-        show_attr_textmode, set_attr_textmode);
+static DEVICE_ATTR(textmode, S_IRUGO | S_IWUSR, show_attr_textmode, set_attr_textmode);
 
 
 MYDEV_ATTR_SIMPLE_UNSIGNED(powered, update_display_powered);
diff --git a/drivers/usb/mon/mon_bin.c b/drivers/usb/mon/mon_bin.c
index 44cb37b5a4dc..c436e1e2c3b6 100644
--- a/drivers/usb/mon/mon_bin.c
+++ b/drivers/usb/mon/mon_bin.c
@@ -15,7 +15,6 @@
 #include <linux/poll.h>
 #include <linux/compat.h>
 #include <linux/mm.h>
-#include <linux/smp_lock.h>
 #include <linux/scatterlist.h>
 #include <linux/slab.h>
 
diff --git a/drivers/usb/mon/mon_stat.c b/drivers/usb/mon/mon_stat.c
index 8ec94f15a738..e5ce42bd316e 100644
--- a/drivers/usb/mon/mon_stat.c
+++ b/drivers/usb/mon/mon_stat.c
@@ -11,7 +11,6 @@
 #include <linux/slab.h>
 #include <linux/usb.h>
 #include <linux/fs.h>
-#include <linux/smp_lock.h>
 #include <asm/uaccess.h>
 
 #include "usb_mon.h"
diff --git a/drivers/usb/otg/langwell_otg.c b/drivers/usb/otg/langwell_otg.c
index bdc3ea66be69..9fea48264fa2 100644
--- a/drivers/usb/otg/langwell_otg.c
+++ b/drivers/usb/otg/langwell_otg.c
@@ -1896,7 +1896,7 @@ set_a_bus_req(struct device *dev, struct device_attribute *attr,
         }
         return count;
 }
-static DEVICE_ATTR(a_bus_req, S_IRUGO | S_IWUGO, get_a_bus_req, set_a_bus_req);
+static DEVICE_ATTR(a_bus_req, S_IRUGO | S_IWUSR, get_a_bus_req, set_a_bus_req);
 
 static ssize_t
 get_a_bus_drop(struct device *dev, struct device_attribute *attr, char *buf)
@@ -1942,8 +1942,7 @@ set_a_bus_drop(struct device *dev, struct device_attribute *attr,
         }
         return count;
 }
-static DEVICE_ATTR(a_bus_drop, S_IRUGO | S_IWUGO,
-        get_a_bus_drop, set_a_bus_drop);
+static DEVICE_ATTR(a_bus_drop, S_IRUGO | S_IWUSR, get_a_bus_drop, set_a_bus_drop);
 
 static ssize_t
 get_b_bus_req(struct device *dev, struct device_attribute *attr, char *buf)
@@ -1988,7 +1987,7 @@ set_b_bus_req(struct device *dev, struct device_attribute *attr,
         }
         return count;
 }
-static DEVICE_ATTR(b_bus_req, S_IRUGO | S_IWUGO, get_b_bus_req, set_b_bus_req);
+static DEVICE_ATTR(b_bus_req, S_IRUGO | S_IWUSR, get_b_bus_req, set_b_bus_req);
 
 static ssize_t
 set_a_clr_err(struct device *dev, struct device_attribute *attr,
@@ -2012,7 +2011,7 @@ set_a_clr_err(struct device *dev, struct device_attribute *attr,
         }
         return count;
 }
-static DEVICE_ATTR(a_clr_err, S_IWUGO, NULL, set_a_clr_err);
+static DEVICE_ATTR(a_clr_err, S_IWUSR, NULL, set_a_clr_err);
 
 static struct attribute *inputs_attrs[] = {
         &dev_attr_a_bus_req.attr,
diff --git a/drivers/usb/serial/usb-serial.c b/drivers/usb/serial/usb-serial.c
index e64da74bdcc5..861223f2af6e 100644
--- a/drivers/usb/serial/usb-serial.c
+++ b/drivers/usb/serial/usb-serial.c
@@ -21,7 +21,6 @@
 #include <linux/errno.h>
 #include <linux/init.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/tty.h>
 #include <linux/tty_driver.h>
 #include <linux/tty_flip.h>
diff --git a/drivers/usb/storage/scsiglue.c b/drivers/usb/storage/scsiglue.c
index a688b1e686ea..689ee1fb702a 100644
--- a/drivers/usb/storage/scsiglue.c
+++ b/drivers/usb/storage/scsiglue.c
@@ -285,7 +285,7 @@ static int slave_configure(struct scsi_device *sdev)
 
 /* queue a command */
 /* This is always called with scsi_lock(host) held */
-static int queuecommand(struct scsi_cmnd *srb,
+static int queuecommand_lck(struct scsi_cmnd *srb,
                         void (*done)(struct scsi_cmnd *))
 {
         struct us_data *us = host_to_us(srb->device->host);
@@ -315,6 +315,8 @@ static int queuecommand(struct scsi_cmnd *srb,
         return 0;
 }
 
+static DEF_SCSI_QCMD(queuecommand)
+
 /***********************************************************************
  * Error handling functions
  ***********************************************************************/
diff --git a/drivers/usb/storage/sierra_ms.c b/drivers/usb/storage/sierra_ms.c
index 57fc2f532cab..ceba512f84d0 100644
--- a/drivers/usb/storage/sierra_ms.c
+++ b/drivers/usb/storage/sierra_ms.c
@@ -121,7 +121,7 @@ static ssize_t show_truinst(struct device *dev, struct device_attribute *attr,
         }
         return result;
 }
-static DEVICE_ATTR(truinst, S_IWUGO | S_IRUGO, show_truinst, NULL);
+static DEVICE_ATTR(truinst, S_IRUGO, show_truinst, NULL);
 
 int sierra_ms_init(struct us_data *us)
 {
diff --git a/drivers/usb/storage/uas.c b/drivers/usb/storage/uas.c
index d1268191acbd..339fac3949df 100644
--- a/drivers/usb/storage/uas.c
+++ b/drivers/usb/storage/uas.c
@@ -430,7 +430,7 @@ static int uas_submit_urbs(struct scsi_cmnd *cmnd,
         return 0;
 }
 
-static int uas_queuecommand(struct scsi_cmnd *cmnd,
+static int uas_queuecommand_lck(struct scsi_cmnd *cmnd,
                                         void (*done)(struct scsi_cmnd *))
 {
         struct scsi_device *sdev = cmnd->device;
@@ -488,6 +488,8 @@ static int uas_queuecommand(struct scsi_cmnd *cmnd,
         return 0;
 }
 
+static DEF_SCSI_QCMD(uas_queuecommand)
+
 static int uas_eh_abort_handler(struct scsi_cmnd *cmnd)
 {
         struct scsi_device *sdev = cmnd->device;
diff --git a/drivers/video/backlight/backlight.c b/drivers/video/backlight/backlight.c
index e207810bba3c..08703299ef61 100644
--- a/drivers/video/backlight/backlight.c
+++ b/drivers/video/backlight/backlight.c
@@ -197,12 +197,12 @@ static int backlight_suspend(struct device *dev, pm_message_t state)
 {
         struct backlight_device *bd = to_backlight_device(dev);
 
-        if (bd->ops->options & BL_CORE_SUSPENDRESUME) {
-                mutex_lock(&bd->ops_lock);
+        mutex_lock(&bd->ops_lock);
+        if (bd->ops && bd->ops->options & BL_CORE_SUSPENDRESUME) {
                 bd->props.state |= BL_CORE_SUSPENDED;
                 backlight_update_status(bd);
-                mutex_unlock(&bd->ops_lock);
         }
+        mutex_unlock(&bd->ops_lock);
 
         return 0;
 }
@@ -211,12 +211,12 @@ static int backlight_resume(struct device *dev)
 {
         struct backlight_device *bd = to_backlight_device(dev);
 
-        if (bd->ops->options & BL_CORE_SUSPENDRESUME) {
-                mutex_lock(&bd->ops_lock);
+        mutex_lock(&bd->ops_lock);
+        if (bd->ops && bd->ops->options & BL_CORE_SUSPENDRESUME) {
                 bd->props.state &= ~BL_CORE_SUSPENDED;
                 backlight_update_status(bd);
-                mutex_unlock(&bd->ops_lock);
         }
+        mutex_unlock(&bd->ops_lock);
 
         return 0;
 }
diff --git a/drivers/video/console/vgacon.c b/drivers/video/console/vgacon.c
index 54e32c513610..915448ec75bf 100644
--- a/drivers/video/console/vgacon.c
+++ b/drivers/video/console/vgacon.c
@@ -47,7 +47,6 @@
 #include <linux/ioport.h>
 #include <linux/init.h>
 #include <linux/screen_info.h>
-#include <linux/smp_lock.h>
 #include <video/vga.h>
 #include <asm/io.h>
 
diff --git a/drivers/video/fbcmap.c b/drivers/video/fbcmap.c
index f53b9f1d6aba..affdf3e32cf3 100644
--- a/drivers/video/fbcmap.c
+++ b/drivers/video/fbcmap.c
@@ -88,34 +88,48 @@ static const struct fb_cmap default_16_colors = {
  *
  */
 
-int fb_alloc_cmap(struct fb_cmap *cmap, int len, int transp)
+int fb_alloc_cmap_gfp(struct fb_cmap *cmap, int len, int transp, gfp_t flags)
 {
-    int size = len*sizeof(u16);
-
-    if (cmap->len != len) {
-        fb_dealloc_cmap(cmap);
-        if (!len)
-            return 0;
-        if (!(cmap->red = kmalloc(size, GFP_ATOMIC)))
-            goto fail;
-        if (!(cmap->green = kmalloc(size, GFP_ATOMIC)))
-            goto fail;
-        if (!(cmap->blue = kmalloc(size, GFP_ATOMIC)))
-            goto fail;
-        if (transp) {
-            if (!(cmap->transp = kmalloc(size, GFP_ATOMIC)))
+        int size = len * sizeof(u16);
+        int ret = -ENOMEM;
+
+        if (cmap->len != len) {
+                fb_dealloc_cmap(cmap);
+                if (!len)
+                        return 0;
+
+                cmap->red = kmalloc(size, flags);
+                if (!cmap->red)
+                        goto fail;
+                cmap->green = kmalloc(size, flags);
+                if (!cmap->green)
+                        goto fail;
+                cmap->blue = kmalloc(size, flags);
+                if (!cmap->blue)
+                        goto fail;
+                if (transp) {
+                        cmap->transp = kmalloc(size, flags);
+                        if (!cmap->transp)
+                                goto fail;
+                } else {
+                        cmap->transp = NULL;
+                }
+        }
+        cmap->start = 0;
+        cmap->len = len;
+        ret = fb_copy_cmap(fb_default_cmap(len), cmap);
+        if (ret)
                 goto fail;
-        } else
-            cmap->transp = NULL;
-    }
-    cmap->start = 0;
-    cmap->len = len;
-    fb_copy_cmap(fb_default_cmap(len), cmap);
-    return 0;
+        return 0;
 
 fail:
-    fb_dealloc_cmap(cmap);
-    return -ENOMEM;
+        fb_dealloc_cmap(cmap);
+        return ret;
+}
+
+int fb_alloc_cmap(struct fb_cmap *cmap, int len, int transp)
+{
+        return fb_alloc_cmap_gfp(cmap, len, transp, GFP_ATOMIC);
 }
 
 /**
@@ -250,8 +264,12 @@ int fb_set_user_cmap(struct fb_cmap_user *cmap, struct fb_info *info)
         int rc, size = cmap->len * sizeof(u16);
         struct fb_cmap umap;
 
+        if (size < 0 || size < cmap->len)
+                return -E2BIG;
+
         memset(&umap, 0, sizeof(struct fb_cmap));
-        rc = fb_alloc_cmap(&umap, cmap->len, cmap->transp != NULL);
+        rc = fb_alloc_cmap_gfp(&umap, cmap->len, cmap->transp != NULL,
+                                GFP_KERNEL);
         if (rc)
                 return rc;
         if (copy_from_user(umap.red, cmap->red, size) ||
diff --git a/drivers/video/geode/lxfb_ops.c b/drivers/video/geode/lxfb_ops.c
index bc35a95e59d4..85ec7f64c42a 100644
--- a/drivers/video/geode/lxfb_ops.c
+++ b/drivers/video/geode/lxfb_ops.c
@@ -276,10 +276,10 @@ static void lx_graphics_enable(struct fb_info *info)
                 write_fp(par, FP_PT1, 0);
                 temp = FP_PT2_SCRC;
 
-                if (info->var.sync & FB_SYNC_HOR_HIGH_ACT)
+                if (!(info->var.sync & FB_SYNC_HOR_HIGH_ACT))
                         temp |= FP_PT2_HSP;
 
-                if (info->var.sync & FB_SYNC_VERT_HIGH_ACT)
+                if (!(info->var.sync & FB_SYNC_VERT_HIGH_ACT))
                         temp |= FP_PT2_VSP;
 
                 write_fp(par, FP_PT2, temp);
diff --git a/drivers/video/mx3fb.c b/drivers/video/mx3fb.c
index 7cfc170bce19..ca0f6be9d12e 100644
--- a/drivers/video/mx3fb.c
+++ b/drivers/video/mx3fb.c
@@ -27,6 +27,7 @@
 #include <linux/clk.h>
 #include <linux/mutex.h>
 
+#include <mach/dma.h>
 #include <mach/hardware.h>
 #include <mach/ipu.h>
 #include <mach/mx3fb.h>
@@ -1420,6 +1421,9 @@ static bool chan_filter(struct dma_chan *chan, void *arg)
         struct device *dev;
         struct mx3fb_platform_data *mx3fb_pdata;
 
+        if (!imx_dma_is_ipu(chan))
+                return false;
+
         if (!rq)
                 return false;
 
diff --git a/drivers/video/omap2/vram.c b/drivers/video/omap2/vram.c
index fed2a72bc6b6..2fd7e5271be9 100644
--- a/drivers/video/omap2/vram.c
+++ b/drivers/video/omap2/vram.c
@@ -554,9 +554,15 @@ void __init omap_vram_reserve_sdram_memblock(void)
         size = PAGE_ALIGN(size);
 
         if (paddr) {
-                if ((paddr & ~PAGE_MASK) ||
-                    !memblock_is_region_memory(paddr, size)) {
-                        pr_err("Illegal SDRAM region for VRAM\n");
+                if (paddr & ~PAGE_MASK) {
+                        pr_err("VRAM start address 0x%08x not page aligned\n",
+                                        paddr);
+                        return;
+                }
+
+                if (!memblock_is_region_memory(paddr, size)) {
+                        pr_err("Illegal SDRAM region 0x%08x..0x%08x for VRAM\n",
+                                        paddr, paddr + size - 1);
                         return;
                 }
 
@@ -570,9 +576,12 @@ void __init omap_vram_reserve_sdram_memblock(void)
                         return;
                 }
         } else {
-                paddr = memblock_alloc_base(size, PAGE_SIZE, MEMBLOCK_REAL_LIMIT);
+                paddr = memblock_alloc(size, PAGE_SIZE);
         }
 
+        memblock_free(paddr, size);
+        memblock_remove(paddr, size);
+
         omap_vram_add_region(paddr, size);
 
         pr_info("Reserving %u bytes SDRAM for VRAM\n", size);
diff --git a/drivers/video/riva/rivafb-i2c.c b/drivers/video/riva/rivafb-i2c.c
index a0e22ac483a3..167400e2a182 100644
--- a/drivers/video/riva/rivafb-i2c.c
+++ b/drivers/video/riva/rivafb-i2c.c
@@ -94,7 +94,6 @@ static int __devinit riva_setup_i2c_bus(struct riva_i2c_chan *chan,
 
         strcpy(chan->adapter.name, name);
         chan->adapter.owner             = THIS_MODULE;
-        chan->adapter.id                = I2C_HW_B_RIVA;
         chan->adapter.class             = i2c_class;
         chan->adapter.algo_data         = &chan->algo;
         chan->adapter.dev.parent        = &chan->par->pdev->dev;
diff --git a/drivers/video/sh_mobile_hdmi.c b/drivers/video/sh_mobile_hdmi.c
index 55b3077ff6ff..d7df10315d8d 100644
--- a/drivers/video/sh_mobile_hdmi.c
+++ b/drivers/video/sh_mobile_hdmi.c
@@ -1071,6 +1071,10 @@ static void sh_hdmi_edid_work_fn(struct work_struct *work)
                 if (!hdmi->info)
                         goto out;
 
+                hdmi->monspec.modedb_len = 0;
+                fb_destroy_modedb(hdmi->monspec.modedb);
+                hdmi->monspec.modedb = NULL;
+
                 acquire_console_sem();
 
                 /* HDMI disconnect */
@@ -1078,7 +1082,6 @@ static void sh_hdmi_edid_work_fn(struct work_struct *work)
 
                 release_console_sem();
                 pm_runtime_put(hdmi->dev);
-                fb_destroy_modedb(hdmi->monspec.modedb);
         }
 
 out:
diff --git a/drivers/video/sh_mobile_lcdcfb.c b/drivers/video/sh_mobile_lcdcfb.c
index 50963739a409..b02d97a879d6 100644
--- a/drivers/video/sh_mobile_lcdcfb.c
+++ b/drivers/video/sh_mobile_lcdcfb.c
@@ -115,15 +115,16 @@ static const struct fb_videomode default_720p = {
         .xres = 1280,
         .yres = 720,
 
-        .left_margin = 200,
-        .right_margin = 88,
-        .hsync_len = 48,
+        .left_margin = 220,
+        .right_margin = 110,
+        .hsync_len = 40,
 
         .upper_margin = 20,
         .lower_margin = 5,
         .vsync_len = 5,
 
         .pixclock = 13468,
+        .refresh = 60,
         .sync = FB_SYNC_VERT_HIGH_ACT | FB_SYNC_HOR_HIGH_ACT,
 };
 
@@ -859,7 +860,7 @@ static void sh_mobile_fb_reconfig(struct fb_info *info)
                 /* Couldn't reconfigure, hopefully, can continue as before */
                 return;
 
-        info->fix.line_length = mode2.xres * (ch->cfg.bpp / 8);
+        info->fix.line_length = mode1.xres * (ch->cfg.bpp / 8);
 
         /*
          * fb_set_var() calls the notifier change internally, only if
@@ -867,7 +868,7 @@ static void sh_mobile_fb_reconfig(struct fb_info *info)
          * user event, we have to call the chain ourselves.
          */
         event.info = info;
-        event.data = &mode2;
+        event.data = &mode1;
         fb_notifier_call_chain(evnt, &event);
 }
 
@@ -1197,6 +1198,7 @@ static int __devinit sh_mobile_lcdc_probe(struct platform_device *pdev)
                 const struct fb_videomode *mode = cfg->lcd_cfg;
                 unsigned long max_size = 0;
                 int k;
+                int num_cfg;
 
                 ch->info = framebuffer_alloc(0, &pdev->dev);
                 if (!ch->info) {
@@ -1232,8 +1234,14 @@ static int __devinit sh_mobile_lcdc_probe(struct platform_device *pdev)
                 info->fix = sh_mobile_lcdc_fix;
                 info->fix.smem_len = max_size * (cfg->bpp / 8) * 2;
 
-                if (!mode)
+                if (!mode) {
                         mode = &default_720p;
+                        num_cfg = 1;
+                } else {
+                        num_cfg = ch->cfg.num_cfg;
+                }
+
+                fb_videomode_to_modelist(mode, num_cfg, &info->modelist);
 
                 fb_videomode_to_var(var, mode);
                 /* Default Y virtual resolution is 2x panel size */
@@ -1281,10 +1289,6 @@ static int __devinit sh_mobile_lcdc_probe(struct platform_device *pdev)
 
         for (i = 0; i < j; i++) {
                 struct sh_mobile_lcdc_chan *ch = priv->ch + i;
-                const struct fb_videomode *mode = ch->cfg.lcd_cfg;
-
-                if (!mode)
-                        mode = &default_720p;
 
                 info = ch->info;
 
@@ -1297,7 +1301,6 @@ static int __devinit sh_mobile_lcdc_probe(struct platform_device *pdev)
                         }
                 }
 
-                fb_videomode_to_modelist(mode, ch->cfg.num_cfg, &info->modelist);
                 error = register_framebuffer(info);
                 if (error < 0)
                         goto err1;
diff --git a/drivers/video/sis/init.c b/drivers/video/sis/init.c
index c311ad3c3687..31137adc8fba 100644
--- a/drivers/video/sis/init.c
+++ b/drivers/video/sis/init.c
@@ -62,11 +62,11 @@
 
 #include "init.h"
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 #include "300vtbl.h"
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 #include "310vtbl.h"
 #endif
 
@@ -78,7 +78,7 @@
 /*         POINTER INITIALIZATION            */
 /*********************************************/
 
-#if defined(SIS300) || defined(SIS315H)
+#if defined(CONFIG_FB_SIS_300) || defined(CONFIG_FB_SIS_315)
 static void
 InitCommonPointer(struct SiS_Private *SiS_Pr)
 {
@@ -160,7 +160,7 @@ InitCommonPointer(struct SiS_Private *SiS_Pr)
 }
 #endif
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 static void
 InitTo300Pointer(struct SiS_Private *SiS_Pr)
 {
@@ -237,7 +237,7 @@ InitTo300Pointer(struct SiS_Private *SiS_Pr)
 }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static void
 InitTo310Pointer(struct SiS_Private *SiS_Pr)
 {
@@ -321,13 +321,13 @@ bool
 SiSInitPtr(struct SiS_Private *SiS_Pr)
 {
    if(SiS_Pr->ChipType < SIS_315H) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
       InitTo300Pointer(SiS_Pr);
 #else
       return false;
 #endif
    } else {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
       InitTo310Pointer(SiS_Pr);
 #else
       return false;
@@ -340,9 +340,7 @@ SiSInitPtr(struct SiS_Private *SiS_Pr)
 /*            HELPER: Get ModeID             */
 /*********************************************/
 
-#ifndef SIS_XORG_XF86
 static
-#endif
 unsigned short
 SiS_GetModeID(int VGAEngine, unsigned int VBFlags, int HDisplay, int VDisplay,
                 int Depth, bool FSTN, int LCDwidth, int LCDheight)
@@ -884,51 +882,51 @@ SiS_GetModeID_VGA2(int VGAEngine, unsigned int VBFlags, int HDisplay, int VDispl
 void
 SiS_SetReg(SISIOADDRESS port, unsigned short index, unsigned short data)
 {
-   OutPortByte(port, index);
-   OutPortByte(port + 1, data);
+        outb((u8)index, port);
+        outb((u8)data, port + 1);
 }
 
 void
 SiS_SetRegByte(SISIOADDRESS port, unsigned short data)
 {
-   OutPortByte(port, data);
+        outb((u8)data, port);
 }
 
 void
 SiS_SetRegShort(SISIOADDRESS port, unsigned short data)
 {
-   OutPortWord(port, data);
+        outw((u16)data, port);
 }
 
 void
 SiS_SetRegLong(SISIOADDRESS port, unsigned int data)
 {
-   OutPortLong(port, data);
+        outl((u32)data, port);
 }
 
 unsigned char
 SiS_GetReg(SISIOADDRESS port, unsigned short index)
 {
-   OutPortByte(port, index);
-   return(InPortByte(port + 1));
+        outb((u8)index, port);
+        return inb(port + 1);
 }
 
 unsigned char
 SiS_GetRegByte(SISIOADDRESS port)
 {
-   return(InPortByte(port));
+        return inb(port);
 }
 
 unsigned short
 SiS_GetRegShort(SISIOADDRESS port)
 {
-   return(InPortWord(port));
+        return inw(port);
 }
 
 unsigned int
 SiS_GetRegLong(SISIOADDRESS port)
 {
-   return(InPortLong(port));
+        return inl(port);
 }
 
 void
@@ -1089,7 +1087,7 @@ static void
 SiSInitPCIetc(struct SiS_Private *SiS_Pr)
 {
    switch(SiS_Pr->ChipType) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
    case SIS_300:
    case SIS_540:
    case SIS_630:
@@ -1108,7 +1106,7 @@ SiSInitPCIetc(struct SiS_Private *SiS_Pr)
       SiS_SetRegOR(SiS_Pr->SiS_P3c4,0x1E,0x5A);
       break;
 #endif
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    case SIS_315H:
    case SIS_315:
    case SIS_315PRO:
@@ -1152,9 +1150,7 @@ SiSInitPCIetc(struct SiS_Private *SiS_Pr)
 /*             HELPER: SetLVDSetc            */
 /*********************************************/
 
-#ifdef SIS_LINUX_KERNEL
 static
-#endif
 void
 SiSSetLVDSetc(struct SiS_Private *SiS_Pr)
 {
@@ -1174,7 +1170,7 @@ SiSSetLVDSetc(struct SiS_Private *SiS_Pr)
    if((temp == 1) || (temp == 2)) return;
 
    switch(SiS_Pr->ChipType) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
    case SIS_540:
    case SIS_630:
    case SIS_730:
@@ -1188,7 +1184,7 @@ SiSSetLVDSetc(struct SiS_Private *SiS_Pr)
         }
         break;
 #endif
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    case SIS_550:
    case SIS_650:
    case SIS_740:
@@ -1420,9 +1416,7 @@ SiS_ResetSegmentRegisters(struct SiS_Private *SiS_Pr)
 /*             HELPER: GetVBType             */
 /*********************************************/
 
-#ifdef SIS_LINUX_KERNEL
 static
-#endif
 void
 SiS_GetVBType(struct SiS_Private *SiS_Pr)
 {
@@ -1487,7 +1481,6 @@ SiS_GetVBType(struct SiS_Private *SiS_Pr)
 /*           HELPER: Check RAM size          */
 /*********************************************/
 
-#ifdef SIS_LINUX_KERNEL
 static bool
 SiS_CheckMemorySize(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
                 unsigned short ModeIdIndex)
@@ -1501,13 +1494,12 @@ SiS_CheckMemorySize(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
    if(AdapterMemSize < memorysize) return false;
    return true;
 }
-#endif
 
 /*********************************************/
 /*           HELPER: Get DRAM type           */
 /*********************************************/
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static unsigned char
 SiS_Get310DRAMType(struct SiS_Private *SiS_Pr)
 {
@@ -1574,7 +1566,6 @@ SiS_GetMCLK(struct SiS_Private *SiS_Pr)
 /*           HELPER: ClearBuffer             */
 /*********************************************/
 
-#ifdef SIS_LINUX_KERNEL
 static void
 SiS_ClearBuffer(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
 {
@@ -1587,7 +1578,7 @@ SiS_ClearBuffer(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
 
    if(SiS_Pr->SiS_ModeType >= ModeEGA) {
       if(ModeNo > 0x13) {
-         SiS_SetMemory(memaddr, memsize, 0);
+         memset_io(memaddr, 0, memsize);
       } else {
          pBuffer = (unsigned short SISIOMEMTYPE *)memaddr;
          for(i = 0; i < 0x4000; i++) writew(0x0000, &pBuffer[i]);
@@ -1596,10 +1587,9 @@ SiS_ClearBuffer(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
       pBuffer = (unsigned short SISIOMEMTYPE *)memaddr;
       for(i = 0; i < 0x4000; i++) writew(0x0720, &pBuffer[i]);
    } else {
-      SiS_SetMemory(memaddr, 0x8000, 0);
+      memset_io(memaddr, 0, 0x8000);
    }
 }
-#endif
 
 /*********************************************/
 /*           HELPER: SearchModeID            */
@@ -2132,7 +2122,7 @@ SiS_SetCRT1CRTC(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
       SiS_SetReg(SiS_Pr->SiS_P3d4,0x14,0x4F);
    }
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    if(SiS_Pr->ChipType == XGI_20) {
       SiS_SetReg(SiS_Pr->SiS_P3d4,0x04,crt1data[4] - 1);
       if(!(temp = crt1data[5] & 0x1f)) {
@@ -2215,7 +2205,7 @@ SiS_SetCRT1VCLK(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
    SiS_SetReg(SiS_Pr->SiS_P3c4,0x2c,clkb);
 
    if(SiS_Pr->ChipType >= SIS_315H) {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
       SiS_SetReg(SiS_Pr->SiS_P3c4,0x2D,0x01);
       if(SiS_Pr->ChipType == XGI_20) {
          unsigned short mf = SiS_GetModeFlag(SiS_Pr, ModeNo, ModeIdIndex);
@@ -2236,7 +2226,7 @@ SiS_SetCRT1VCLK(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
 /*                  FIFO                     */
 /*********************************************/
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 void
 SiS_GetFIFOThresholdIndex300(struct SiS_Private *SiS_Pr, unsigned short *idx1,
                 unsigned short *idx2)
@@ -2506,11 +2496,7 @@ SiS_SetCRT1FIFO_630(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
    SiS_SetRegANDOR(SiS_Pr->SiS_P3c4,0x09,0x80,data);
 
   /* Write foreground and background queue */
-#ifdef SIS_LINUX_KERNEL
    templ = sisfb_read_nbridge_pci_dword(SiS_Pr, 0x50);
-#else
-   templ = pciReadLong(0x00000000, 0x50);
-#endif
 
    if(SiS_Pr->ChipType == SIS_730) {
 
@@ -2530,13 +2516,8 @@ SiS_SetCRT1FIFO_630(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
 
    }
 
-#ifdef SIS_LINUX_KERNEL
    sisfb_write_nbridge_pci_dword(SiS_Pr, 0x50, templ);
    templ = sisfb_read_nbridge_pci_dword(SiS_Pr, 0xA0);
-#else
-   pciWriteLong(0x00000000, 0x50, templ);
-   templ = pciReadLong(0x00000000, 0xA0);
-#endif
 
    /* GUI grant timer (PCI config 0xA3) */
    if(SiS_Pr->ChipType == SIS_730) {
@@ -2552,15 +2533,11 @@ SiS_SetCRT1FIFO_630(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
 
    }
 
-#ifdef SIS_LINUX_KERNEL
    sisfb_write_nbridge_pci_dword(SiS_Pr, 0xA0, templ);
-#else
-   pciWriteLong(0x00000000, 0xA0, templ);
-#endif
 }
-#endif /* SIS300 */
+#endif /* CONFIG_FB_SIS_300 */
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static void
 SiS_SetCRT1FIFO_310(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short ModeIdIndex)
 {
@@ -2612,7 +2589,7 @@ SiS_SetVCLKState(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
    }
 
    if(SiS_Pr->ChipType < SIS_315H) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
       if(VCLK > 150) data |= 0x80;
       SiS_SetRegANDOR(SiS_Pr->SiS_P3c4,0x07,0x7B,data);
 
@@ -2621,7 +2598,7 @@ SiS_SetVCLKState(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
       SiS_SetRegANDOR(SiS_Pr->SiS_P3c4,0x32,0xF7,data);
 #endif
    } else if(SiS_Pr->ChipType < XGI_20) {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
       if(VCLK >= 166) data |= 0x0c;
       SiS_SetRegANDOR(SiS_Pr->SiS_P3c4,0x32,0xf3,data);
 
@@ -2630,7 +2607,7 @@ SiS_SetVCLKState(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
       }
 #endif
    } else {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
       if(VCLK >= 200) data |= 0x0c;
       if(SiS_Pr->ChipType == XGI_20) data &= ~0x04;
       SiS_SetRegANDOR(SiS_Pr->SiS_P3c4,0x32,0xf3,data);
@@ -2675,7 +2652,7 @@ SiS_SetCRT1ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
                 unsigned short ModeIdIndex, unsigned short RRTI)
 {
    unsigned short data, infoflag = 0, modeflag, resindex;
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    unsigned char  *ROMAddr  = SiS_Pr->VirtualRomBase;
    unsigned short data2, data3;
 #endif
@@ -2736,7 +2713,7 @@ SiS_SetCRT1ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
       SiS_SetRegANDOR(SiS_Pr->SiS_P3c4,0x0F,0xB7,data);
    }
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    if(SiS_Pr->ChipType >= SIS_315H) {
       SiS_SetRegAND(SiS_Pr->SiS_P3c4,0x31,0xfb);
    }
@@ -2826,7 +2803,7 @@ SiS_SetCRT1ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
 
    SiS_SetVCLKState(SiS_Pr, ModeNo, RRTI, ModeIdIndex);
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    if(((SiS_Pr->ChipType >= SIS_315H) && (SiS_Pr->ChipType < SIS_661)) ||
        (SiS_Pr->ChipType == XGI_40)) {
       if(SiS_GetReg(SiS_Pr->SiS_P3d4,0x31) & 0x40) {
@@ -2845,7 +2822,7 @@ SiS_SetCRT1ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
 #endif
 }
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static void
 SiS_SetupDualChip(struct SiS_Private *SiS_Pr)
 {
@@ -2999,11 +2976,6 @@ SiS_SetCRT1Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sho
    SiS_Pr->SiS_SelectCRT2Rate = 0;
    SiS_Pr->SiS_SetFlag &= (~ProgrammingCRT2);
 
-#ifdef SIS_XORG_XF86
-   xf86DrvMsgVerb(0, X_PROBED, 4, "(init: VBType=0x%04x, VBInfo=0x%04x)\n",
-                    SiS_Pr->SiS_VBType, SiS_Pr->SiS_VBInfo);
-#endif
-
    if(SiS_Pr->SiS_VBInfo & SetSimuScanMode) {
       if(SiS_Pr->SiS_VBInfo & SetInSlaveMode) {
          SiS_Pr->SiS_SetFlag |= ProgrammingCRT2;
@@ -3028,7 +3000,7 @@ SiS_SetCRT1Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sho
    }
 
    switch(SiS_Pr->ChipType) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
    case SIS_300:
       SiS_SetCRT1FIFO_300(SiS_Pr, ModeNo, RefreshRateTableIndex);
       break;
@@ -3039,7 +3011,7 @@ SiS_SetCRT1Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sho
       break;
 #endif
    default:
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
       if(SiS_Pr->ChipType == XGI_20) {
          unsigned char sr2b = 0, sr2c = 0;
          switch(ModeNo) {
@@ -3062,7 +3034,7 @@ SiS_SetCRT1Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sho
 
    SiS_SetCRT1ModeRegs(SiS_Pr, ModeNo, ModeIdIndex, RefreshRateTableIndex);
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    if(SiS_Pr->ChipType == XGI_40) {
       SiS_SetupDualChip(SiS_Pr);
    }
@@ -3070,11 +3042,9 @@ SiS_SetCRT1Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sho
 
    SiS_LoadDAC(SiS_Pr, ModeNo, ModeIdIndex);
 
-#ifdef SIS_LINUX_KERNEL
    if(SiS_Pr->SiS_flag_clearbuffer) {
       SiS_ClearBuffer(SiS_Pr, ModeNo);
    }
-#endif
 
    if(!(SiS_Pr->SiS_VBInfo & (SetSimuScanMode | SwitchCRT2 | SetCRT2ToLCDA))) {
       SiS_WaitRetrace1(SiS_Pr);
@@ -3104,7 +3074,7 @@ SiS_InitVB(struct SiS_Private *SiS_Pr)
 static void
 SiS_ResetVB(struct SiS_Private *SiS_Pr)
 {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    unsigned char  *ROMAddr = SiS_Pr->VirtualRomBase;
    unsigned short temp;
 
@@ -3139,7 +3109,7 @@ SiS_StrangeStuff(struct SiS_Private *SiS_Pr)
     * which locks CRT2 in some way to CRT1 timing. Disable
     * this here.
     */
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    if((IS_SIS651) || (IS_SISM650) ||
       SiS_Pr->ChipType == SIS_340 ||
       SiS_Pr->ChipType == XGI_40) {
@@ -3160,7 +3130,7 @@ SiS_StrangeStuff(struct SiS_Private *SiS_Pr)
 static void
 SiS_Handle760(struct SiS_Private *SiS_Pr)
 {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    unsigned int somebase;
    unsigned char temp1, temp2, temp3;
 
@@ -3170,11 +3140,7 @@ SiS_Handle760(struct SiS_Private *SiS_Pr)
        (!(SiS_Pr->SiS_SysFlags & SF_760UMA)) )
       return;
 
-#ifdef SIS_LINUX_KERNEL
    somebase = sisfb_read_mio_pci_word(SiS_Pr, 0x74);
-#else
-   somebase = pciReadWord(0x00001000, 0x74);
-#endif
    somebase &= 0xffff;
 
    if(somebase == 0) return;
@@ -3190,105 +3156,34 @@ SiS_Handle760(struct SiS_Private *SiS_Pr)
       temp2 = 0x0b;
    }
 
-#ifdef SIS_LINUX_KERNEL
    sisfb_write_nbridge_pci_byte(SiS_Pr, 0x7e, temp1);
    sisfb_write_nbridge_pci_byte(SiS_Pr, 0x8d, temp2);
-#else
-   pciWriteByte(0x00000000, 0x7e, temp1);
-   pciWriteByte(0x00000000, 0x8d, temp2);
-#endif
 
    SiS_SetRegByte((somebase + 0x85), temp3);
 #endif
 }
 
 /*********************************************/
-/*      X.org/XFree86: SET SCREEN PITCH      */
-/*********************************************/
-
-#ifdef SIS_XORG_XF86
-static void
-SiS_SetPitchCRT1(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn)
-{
-   SISPtr pSiS = SISPTR(pScrn);
-   unsigned short HDisplay = pSiS->scrnPitch >> 3;
-
-   SiS_SetReg(SiS_Pr->SiS_P3d4,0x13,(HDisplay & 0xFF));
-   SiS_SetRegANDOR(SiS_Pr->SiS_P3c4,0x0E,0xF0,(HDisplay >> 8));
-}
-
-static void
-SiS_SetPitchCRT2(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn)
-{
-   SISPtr pSiS = SISPTR(pScrn);
-   unsigned short HDisplay = pSiS->scrnPitch2 >> 3;
-
-    /* Unlock CRT2 */
-   if(pSiS->VGAEngine == SIS_315_VGA)
-      SiS_SetRegOR(SiS_Pr->SiS_Part1Port,0x2F, 0x01);
-   else
-      SiS_SetRegOR(SiS_Pr->SiS_Part1Port,0x24, 0x01);
-
-   SiS_SetReg(SiS_Pr->SiS_Part1Port,0x07,(HDisplay & 0xFF));
-   SiS_SetRegANDOR(SiS_Pr->SiS_Part1Port,0x09,0xF0,(HDisplay >> 8));
-}
-
-static void
-SiS_SetPitch(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn)
-{
-   SISPtr pSiS = SISPTR(pScrn);
-   bool isslavemode = false;
-
-   if( (pSiS->VBFlags2 & VB2_VIDEOBRIDGE) &&
-       ( ((pSiS->VGAEngine == SIS_300_VGA) &&
-          (SiS_GetReg(SiS_Pr->SiS_Part1Port,0x00) & 0xa0) == 0x20) ||
-         ((pSiS->VGAEngine == SIS_315_VGA) &&
-          (SiS_GetReg(SiS_Pr->SiS_Part1Port,0x00) & 0x50) == 0x10) ) ) {
-      isslavemode = true;
-   }
-
-   /* We need to set pitch for CRT1 if bridge is in slave mode, too */
-   if((pSiS->VBFlags & DISPTYPE_DISP1) || (isslavemode)) {
-      SiS_SetPitchCRT1(SiS_Pr, pScrn);
-   }
-   /* We must not set the pitch for CRT2 if bridge is in slave mode */
-   if((pSiS->VBFlags & DISPTYPE_DISP2) && (!isslavemode)) {
-      SiS_SetPitchCRT2(SiS_Pr, pScrn);
-   }
-}
-#endif
-
-/*********************************************/
 /*                 SiSSetMode()              */
 /*********************************************/
 
-#ifdef SIS_XORG_XF86
-/* We need pScrn for setting the pitch correctly */
-bool
-SiSSetMode(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn, unsigned short ModeNo, bool dosetpitch)
-#else
 bool
 SiSSetMode(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
-#endif
 {
    SISIOADDRESS BaseAddr = SiS_Pr->IOAddress;
    unsigned short RealModeNo, ModeIdIndex;
    unsigned char  backupreg = 0;
-#ifdef SIS_LINUX_KERNEL
    unsigned short KeepLockReg;
 
    SiS_Pr->UseCustomMode = false;
    SiS_Pr->CRT1UsesCustomMode = false;
-#endif
 
    SiS_Pr->SiS_flag_clearbuffer = 0;
 
    if(SiS_Pr->UseCustomMode) {
       ModeNo = 0xfe;
    } else {
-#ifdef SIS_LINUX_KERNEL
       if(!(ModeNo & 0x80)) SiS_Pr->SiS_flag_clearbuffer = 1;
-#endif
       ModeNo &= 0x7f;
    }
 
@@ -3301,13 +3196,8 @@ SiSSetMode(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
    SiS_GetSysFlags(SiS_Pr);
 
    SiS_Pr->SiS_VGAINFO = 0x11;
-#if defined(SIS_XORG_XF86) && (defined(i386) || defined(__i386) || defined(__i386__) || defined(__AMD64__) || defined(__amd64__) || defined(__x86_64__))
-   if(pScrn) SiS_Pr->SiS_VGAINFO = SiS_GetSetBIOSScratch(pScrn, 0x489, 0xff);
-#endif
 
-#ifdef SIS_LINUX_KERNEL
    KeepLockReg = SiS_GetReg(SiS_Pr->SiS_P3c4,0x05);
-#endif
    SiS_SetReg(SiS_Pr->SiS_P3c4,0x05,0x86);
 
    SiSInitPCIetc(SiS_Pr);
@@ -3344,12 +3234,10 @@ SiSSetMode(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
    SiS_GetLCDResInfo(SiS_Pr, ModeNo, ModeIdIndex);
    SiS_SetLowModeTest(SiS_Pr, ModeNo);
 
-#ifdef SIS_LINUX_KERNEL
    /* Check memory size (kernel framebuffer driver only) */
    if(!SiS_CheckMemorySize(SiS_Pr, ModeNo, ModeIdIndex)) {
       return false;
    }
-#endif
 
    SiS_OpenCRTC(SiS_Pr);
 
@@ -3384,7 +3272,7 @@ SiSSetMode(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
    SiS_DisplayOn(SiS_Pr);
    SiS_SetRegByte(SiS_Pr->SiS_P3c6,0xFF);
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    if(SiS_Pr->ChipType >= SIS_315H) {
       if(SiS_Pr->SiS_IF_DEF_LVDS == 1) {
          if(!(SiS_IsDualEdge(SiS_Pr))) {
@@ -3396,7 +3284,7 @@ SiSSetMode(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
 
    if(SiS_Pr->SiS_VBType & VB_SIS30xBLV) {
       if(SiS_Pr->ChipType >= SIS_315H) {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
          if(!SiS_Pr->SiS_ROMNew) {
             if(SiS_IsVAMode(SiS_Pr)) {
                SiS_SetRegOR(SiS_Pr->SiS_P3d4,0x35,0x01);
@@ -3424,424 +3312,16 @@ SiSSetMode(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
       }
    }
 
-#ifdef SIS_XORG_XF86
-   if(pScrn) {
-      /* SetPitch: Adapt to virtual size & position */
-      if((ModeNo > 0x13) && (dosetpitch)) {
-         SiS_SetPitch(SiS_Pr, pScrn);
-      }
-
-      /* Backup/Set ModeNo in BIOS scratch area */
-      SiS_GetSetModeID(pScrn, ModeNo);
-   }
-#endif
-
    SiS_CloseCRTC(SiS_Pr);
 
    SiS_Handle760(SiS_Pr);
 
-#ifdef SIS_LINUX_KERNEL
    /* We never lock registers in XF86 */
    if(KeepLockReg != 0xA1) SiS_SetReg(SiS_Pr->SiS_P3c4,0x05,0x00);
-#endif
 
    return true;
 }
 
-/*********************************************/
-/*       X.org/XFree86: SiSBIOSSetMode()     */
-/*           for non-Dual-Head mode          */
-/*********************************************/
-
-#ifdef SIS_XORG_XF86
-bool
-SiSBIOSSetMode(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn,
-               DisplayModePtr mode, bool IsCustom)
-{
-   SISPtr pSiS = SISPTR(pScrn);
-   unsigned short ModeNo = 0;
-
-   SiS_Pr->UseCustomMode = false;
-
-   if((IsCustom) && (SiS_CheckBuildCustomMode(pScrn, mode, pSiS->VBFlags))) {
-
-      xf86DrvMsgVerb(pScrn->scrnIndex, X_INFO, 3, "Setting custom mode %dx%d\n",
-                SiS_Pr->CHDisplay,
-                (mode->Flags & V_INTERLACE ? SiS_Pr->CVDisplay * 2 :
-                   (mode->Flags & V_DBLSCAN ? SiS_Pr->CVDisplay / 2 :
-                      SiS_Pr->CVDisplay)));
-
-   } else {
-
-      /* Don't need vbflags here; checks done earlier */
-      ModeNo = SiS_GetModeNumber(pScrn, mode, pSiS->VBFlags);
-      if(!ModeNo) return false;
-
-      xf86DrvMsgVerb(pScrn->scrnIndex, X_INFO, 3, "Setting standard mode 0x%x\n", ModeNo);
-
-   }
-
-   return(SiSSetMode(SiS_Pr, pScrn, ModeNo, true));
-}
-
-/*********************************************/
-/*    X.org/XFree86: SiSBIOSSetModeCRT2()    */
-/*           for Dual-Head modes             */
-/*********************************************/
-
-bool
-SiSBIOSSetModeCRT2(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn,
-               DisplayModePtr mode, bool IsCustom)
-{
-   SISIOADDRESS BaseAddr = SiS_Pr->IOAddress;
-   SISPtr  pSiS = SISPTR(pScrn);
-#ifdef SISDUALHEAD
-   SISEntPtr pSiSEnt = pSiS->entityPrivate;
-#endif
-   unsigned short ModeIdIndex;
-   unsigned short ModeNo = 0;
-   unsigned char  backupreg = 0;
-
-   SiS_Pr->UseCustomMode = false;
-
-   /* Remember: Custom modes for CRT2 are ONLY supported
-    *     -) on the 30x/B/C, and
-    *     -) if CRT2 is LCD or VGA, or CRT1 is LCDA
-    */
-
-   if((IsCustom) && (SiS_CheckBuildCustomMode(pScrn, mode, pSiS->VBFlags))) {
-
-         ModeNo = 0xfe;
-
-   } else {
-
-         ModeNo = SiS_GetModeNumber(pScrn, mode, pSiS->VBFlags);
-         if(!ModeNo) return false;
-
-   }
-
-   SiSRegInit(SiS_Pr, BaseAddr);
-   SiSInitPtr(SiS_Pr);
-   SiS_GetSysFlags(SiS_Pr);
-#if defined(i386) || defined(__i386) || defined(__i386__) || defined(__AMD64__) || defined(__amd64__) || defined(__x86_64__)
-   SiS_Pr->SiS_VGAINFO = SiS_GetSetBIOSScratch(pScrn, 0x489, 0xff);
-#else
-   SiS_Pr->SiS_VGAINFO = 0x11;
-#endif
-
-   SiS_SetReg(SiS_Pr->SiS_P3c4,0x05,0x86);
-
-   SiSInitPCIetc(SiS_Pr);
-   SiSSetLVDSetc(SiS_Pr);
-   SiSDetermineROMUsage(SiS_Pr);
-
-   /* Save mode info so we can set it from within SetMode for CRT1 */
-#ifdef SISDUALHEAD
-   if(pSiS->DualHeadMode) {
-      pSiSEnt->CRT2ModeNo = ModeNo;
-      pSiSEnt->CRT2DMode = mode;
-      pSiSEnt->CRT2IsCustom = IsCustom;
-      pSiSEnt->CRT2CR30 = SiS_GetReg(SiS_Pr->SiS_P3d4,0x30);
-      pSiSEnt->CRT2CR31 = SiS_GetReg(SiS_Pr->SiS_P3d4,0x31);
-      pSiSEnt->CRT2CR35 = SiS_GetReg(SiS_Pr->SiS_P3d4,0x35);
-      pSiSEnt->CRT2CR38 = SiS_GetReg(SiS_Pr->SiS_P3d4,0x38);
-#if 0
-      /* We can't set CRT2 mode before CRT1 mode is set - says who...? */
-      if(pSiSEnt->CRT1ModeNo == -1) {
-         xf86DrvMsgVerb(pScrn->scrnIndex, X_INFO, 3,
-                "Setting CRT2 mode delayed until after setting CRT1 mode\n");
-         return true;
-      }
-#endif
-      pSiSEnt->CRT2ModeSet = true;
-   }
-#endif
-
-   if(SiS_Pr->UseCustomMode) {
-
-      unsigned short temptemp = SiS_Pr->CVDisplay;
-
-      if(SiS_Pr->CModeFlag & DoubleScanMode)     temptemp >>= 1;
-      else if(SiS_Pr->CInfoFlag & InterlaceMode) temptemp <<= 1;
-
-      xf86DrvMsgVerb(pScrn->scrnIndex, X_INFO, 3,
-          "Setting custom mode %dx%d on CRT2\n",
-          SiS_Pr->CHDisplay, temptemp);
-
-   } else {
-
-      xf86DrvMsgVerb(pScrn->scrnIndex, X_INFO, 3,
-          "Setting standard mode 0x%x on CRT2\n", ModeNo);
-
-   }
-
-   SiS_UnLockCRT2(SiS_Pr);
-
-   if(!SiS_Pr->UseCustomMode) {
-      if(!(SiS_SearchModeID(SiS_Pr, &ModeNo, &ModeIdIndex))) return false;
-   } else {
-      ModeIdIndex = 0;
-   }
-
-   SiS_GetVBType(SiS_Pr);
-
-   SiS_InitVB(SiS_Pr);
-   if(SiS_Pr->SiS_VBType & VB_SIS30xBLV) {
-      if(SiS_Pr->ChipType >= SIS_315H) {
-         SiS_ResetVB(SiS_Pr);
-         SiS_SetRegOR(SiS_Pr->SiS_P3c4,0x32,0x10);
-         SiS_SetRegOR(SiS_Pr->SiS_Part2Port,0x00,0x0c);
-         backupreg = SiS_GetReg(SiS_Pr->SiS_P3d4,0x38);
-      } else {
-         backupreg = SiS_GetReg(SiS_Pr->SiS_P3d4,0x35);
-      }
-   }
-
-   /* Get VB information (connectors, connected devices) */
-   if(!SiS_Pr->UseCustomMode) {
-      SiS_GetVBInfo(SiS_Pr, ModeNo, ModeIdIndex, 1);
-   } else {
-      /* If this is a custom mode, we don't check the modeflag for CRT2Mode */
-      SiS_GetVBInfo(SiS_Pr, ModeNo, ModeIdIndex, 0);
-   }
-   SiS_SetYPbPr(SiS_Pr);
-   SiS_SetTVMode(SiS_Pr, ModeNo, ModeIdIndex);
-   SiS_GetLCDResInfo(SiS_Pr, ModeNo, ModeIdIndex);
-   SiS_SetLowModeTest(SiS_Pr, ModeNo);
-
-   SiS_ResetSegmentRegisters(SiS_Pr);
-
-   /* Set mode on CRT2 */
-   if( (SiS_Pr->SiS_VBType & VB_SISVB)    ||
-       (SiS_Pr->SiS_IF_DEF_LVDS     == 1) ||
-       (SiS_Pr->SiS_IF_DEF_CH70xx   != 0) ||
-       (SiS_Pr->SiS_IF_DEF_TRUMPION != 0) ) {
-      SiS_SetCRT2Group(SiS_Pr, ModeNo);
-   }
-
-   SiS_StrangeStuff(SiS_Pr);
-
-   SiS_DisplayOn(SiS_Pr);
-   SiS_SetRegByte(SiS_Pr->SiS_P3c6,0xFF);
-
-   if(SiS_Pr->ChipType >= SIS_315H) {
-      if(SiS_Pr->SiS_IF_DEF_LVDS == 1) {
-         if(!(SiS_IsDualEdge(SiS_Pr))) {
-            SiS_SetRegAND(SiS_Pr->SiS_Part1Port,0x13,0xfb);
-         }
-      }
-   }
-
-   if(SiS_Pr->SiS_VBType & VB_SIS30xBLV) {
-      if(SiS_Pr->ChipType >= SIS_315H) {
-         if(!SiS_Pr->SiS_ROMNew) {
-            if(SiS_IsVAMode(SiS_Pr)) {
-               SiS_SetRegOR(SiS_Pr->SiS_P3d4,0x35,0x01);
-            } else {
-               SiS_SetRegAND(SiS_Pr->SiS_P3d4,0x35,0xFE);
-            }
-         }
-
-         SiS_SetReg(SiS_Pr->SiS_P3d4,0x38,backupreg);
-
-         if(SiS_GetReg(SiS_Pr->SiS_P3d4,0x30) & SetCRT2ToLCD) {
-            SiS_SetRegAND(SiS_Pr->SiS_P3d4,0x38,0xfc);
-         }
-      } else if((SiS_Pr->ChipType == SIS_630) ||
-                (SiS_Pr->ChipType == SIS_730)) {
-         SiS_SetReg(SiS_Pr->SiS_P3d4,0x35,backupreg);
-      }
-   }
-
-   /* SetPitch: Adapt to virtual size & position */
-   SiS_SetPitchCRT2(SiS_Pr, pScrn);
-
-   SiS_Handle760(SiS_Pr);
-
-   return true;
-}
-
-/*********************************************/
-/*    X.org/XFree86: SiSBIOSSetModeCRT1()    */
-/*           for Dual-Head modes             */
-/*********************************************/
-
-bool
-SiSBIOSSetModeCRT1(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn,
-                   DisplayModePtr mode, bool IsCustom)
-{
-   SISIOADDRESS BaseAddr = SiS_Pr->IOAddress;
-   SISPtr  pSiS = SISPTR(pScrn);
-   unsigned short ModeIdIndex, ModeNo = 0;
-   unsigned char  backupreg = 0;
-#ifdef SISDUALHEAD
-   SISEntPtr pSiSEnt = pSiS->entityPrivate;
-   unsigned char  backupcr30, backupcr31, backupcr38, backupcr35, backupp40d=0;
-   bool backupcustom;
-#endif
-
-   SiS_Pr->UseCustomMode = false;
-
-   if((IsCustom) && (SiS_CheckBuildCustomMode(pScrn, mode, pSiS->VBFlags))) {
-
-         unsigned short temptemp = SiS_Pr->CVDisplay;
-
-         if(SiS_Pr->CModeFlag & DoubleScanMode)     temptemp >>= 1;
-         else if(SiS_Pr->CInfoFlag & InterlaceMode) temptemp <<= 1;
-
-         xf86DrvMsgVerb(pScrn->scrnIndex, X_INFO, 3,
-                "Setting custom mode %dx%d on CRT1\n",
-                SiS_Pr->CHDisplay, temptemp);
-         ModeNo = 0xfe;
-
-   } else {
-
-         ModeNo = SiS_GetModeNumber(pScrn, mode, 0); /* don't give VBFlags */
-         if(!ModeNo) return false;
-
-         xf86DrvMsgVerb(pScrn->scrnIndex, X_INFO, 3,
-                "Setting standard mode 0x%x on CRT1\n", ModeNo);
-   }
-
-   SiSInitPtr(SiS_Pr);
-   SiSRegInit(SiS_Pr, BaseAddr);
-   SiS_GetSysFlags(SiS_Pr);
-#if defined(i386) || defined(__i386) || defined(__i386__) || defined(__AMD64__) || defined(__amd64__) || defined(__x86_64__)
-   SiS_Pr->SiS_VGAINFO = SiS_GetSetBIOSScratch(pScrn, 0x489, 0xff);
-#else
-   SiS_Pr->SiS_VGAINFO = 0x11;
-#endif
-
-   SiS_SetReg(SiS_Pr->SiS_P3c4,0x05,0x86);
-
-   SiSInitPCIetc(SiS_Pr);
-   SiSSetLVDSetc(SiS_Pr);
-   SiSDetermineROMUsage(SiS_Pr);
-
-   SiS_UnLockCRT2(SiS_Pr);
-
-   if(!SiS_Pr->UseCustomMode) {
-      if(!(SiS_SearchModeID(SiS_Pr, &ModeNo, &ModeIdIndex))) return false;
-   } else {
-      ModeIdIndex = 0;
-   }
-
-   /* Determine VBType */
-   SiS_GetVBType(SiS_Pr);
-
-   SiS_InitVB(SiS_Pr);
-   if(SiS_Pr->SiS_VBType & VB_SIS30xBLV) {
-      if(SiS_Pr->ChipType >= SIS_315H) {
-         backupreg = SiS_GetReg(SiS_Pr->SiS_P3d4,0x38);
-      } else {
-         backupreg = SiS_GetReg(SiS_Pr->SiS_P3d4,0x35);
-      }
-   }
-
-   /* Get VB information (connectors, connected devices) */
-   /* (We don't care if the current mode is a CRT2 mode) */
-   SiS_GetVBInfo(SiS_Pr, ModeNo, ModeIdIndex, 0);
-   SiS_SetYPbPr(SiS_Pr);
-   SiS_SetTVMode(SiS_Pr, ModeNo, ModeIdIndex);
-   SiS_GetLCDResInfo(SiS_Pr, ModeNo, ModeIdIndex);
-   SiS_SetLowModeTest(SiS_Pr, ModeNo);
-
-   SiS_OpenCRTC(SiS_Pr);
-
-   /* Set mode on CRT1 */
-   SiS_SetCRT1Group(SiS_Pr, ModeNo, ModeIdIndex);
-   if(SiS_Pr->SiS_VBInfo & SetCRT2ToLCDA) {
-      SiS_SetCRT2Group(SiS_Pr, ModeNo);
-   }
-
-   /* SetPitch: Adapt to virtual size & position */
-   SiS_SetPitchCRT1(SiS_Pr, pScrn);
-
-   SiS_HandleCRT1(SiS_Pr);
-
-   SiS_StrangeStuff(SiS_Pr);
-
-   SiS_CloseCRTC(SiS_Pr);
-
-#ifdef SISDUALHEAD
-   if(pSiS->DualHeadMode) {
-      pSiSEnt->CRT1ModeNo = ModeNo;
-      pSiSEnt->CRT1DMode = mode;
-   }
-#endif
-
-   if(SiS_Pr->UseCustomMode) {
-      SiS_Pr->CRT1UsesCustomMode = true;
-      SiS_Pr->CSRClock_CRT1 = SiS_Pr->CSRClock;
-      SiS_Pr->CModeFlag_CRT1 = SiS_Pr->CModeFlag;
-   } else {
-      SiS_Pr->CRT1UsesCustomMode = false;
-   }
-
-   /* Reset CRT2 if changing mode on CRT1 */
-#ifdef SISDUALHEAD
-   if(pSiS->DualHeadMode) {
-      if(pSiSEnt->CRT2ModeNo != -1) {
-         xf86DrvMsgVerb(pScrn->scrnIndex, X_INFO, 3,
-                                "(Re-)Setting mode for CRT2\n");
-         backupcustom = SiS_Pr->UseCustomMode;
-         backupcr30 = SiS_GetReg(SiS_Pr->SiS_P3d4,0x30);
-         backupcr31 = SiS_GetReg(SiS_Pr->SiS_P3d4,0x31);
-         backupcr35 = SiS_GetReg(SiS_Pr->SiS_P3d4,0x35);
-         backupcr38 = SiS_GetReg(SiS_Pr->SiS_P3d4,0x38);
-         if(SiS_Pr->SiS_VBType & VB_SISVB) {
-            /* Backup LUT-enable */
-            if(pSiSEnt->CRT2ModeSet) {
-               backupp40d = SiS_GetReg(SiS_Pr->SiS_Part4Port,0x0d) & 0x08;
-            }
-         }
-         if(SiS_Pr->SiS_VBInfo & SetCRT2ToLCDA) {
-            SiS_SetReg(SiS_Pr->SiS_P3d4,0x30,pSiSEnt->CRT2CR30);
-            SiS_SetReg(SiS_Pr->SiS_P3d4,0x31,pSiSEnt->CRT2CR31);
-            SiS_SetReg(SiS_Pr->SiS_P3d4,0x35,pSiSEnt->CRT2CR35);
-            SiS_SetReg(SiS_Pr->SiS_P3d4,0x38,pSiSEnt->CRT2CR38);
-         }
-
-         SiSBIOSSetModeCRT2(SiS_Pr, pSiSEnt->pScrn_1,
-                            pSiSEnt->CRT2DMode, pSiSEnt->CRT2IsCustom);
-
-         SiS_SetReg(SiS_Pr->SiS_P3d4,0x30,backupcr30);
-         SiS_SetReg(SiS_Pr->SiS_P3d4,0x31,backupcr31);
-         SiS_SetReg(SiS_Pr->SiS_P3d4,0x35,backupcr35);
-         SiS_SetReg(SiS_Pr->SiS_P3d4,0x38,backupcr38);
-         if(SiS_Pr->SiS_VBType & VB_SISVB) {
-            SiS_SetRegANDOR(SiS_Pr->SiS_Part4Port,0x0d, ~0x08, backupp40d);
-         }
-         SiS_Pr->UseCustomMode = backupcustom;
-      }
-   }
-#endif
-
-   /* Warning: From here, the custom mode entries in SiS_Pr are
-    * possibly overwritten
-    */
-
-   SiS_DisplayOn(SiS_Pr);
-   SiS_SetRegByte(SiS_Pr->SiS_P3c6,0xFF);
-
-   if(SiS_Pr->SiS_VBType & VB_SIS30xBLV) {
-      if(SiS_Pr->ChipType >= SIS_315H) {
-         SiS_SetReg(SiS_Pr->SiS_P3d4,0x38,backupreg);
-      } else if((SiS_Pr->ChipType == SIS_630) ||
-                (SiS_Pr->ChipType == SIS_730)) {
-         SiS_SetReg(SiS_Pr->SiS_P3d4,0x35,backupreg);
-      }
-   }
-
-   SiS_Handle760(SiS_Pr);
-
-   /* Backup/Set ModeNo in BIOS scratch area */
-   SiS_GetSetModeID(pScrn,ModeNo);
-
-   return true;
-}
-#endif /* Linux_XF86 */
-
 #ifndef GETBITSTR
 #define BITMASK(h,l)            (((unsigned)(1U << ((h)-(l)+1))-1)<<(l))
 #define GENMASK(mask)           BITMASK(1?mask,0?mask)
@@ -3927,7 +3407,7 @@ SiS_CalcLCDACRT1Timing(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
    SiS_Pr->CVBlankStart = SiS_Pr->SiS_VGAVDE;
 
    if(SiS_Pr->ChipType < SIS_315H) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
       tempbx = SiS_Pr->SiS_VGAHT;
       if(SiS_Pr->SiS_LCDInfo & DontExpandLCD) {
          tempbx = SiS_Pr->PanelHT;
@@ -3936,7 +3416,7 @@ SiS_CalcLCDACRT1Timing(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
       remaining = tempbx % 8;
 #endif
    } else {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
       /* OK for LCDA, LVDS */
       tempbx = SiS_Pr->PanelHT - SiS_Pr->PanelXRes;
       tempax = SiS_Pr->SiS_VGAHDE;  /* not /2 ! */
@@ -3950,7 +3430,7 @@ SiS_CalcLCDACRT1Timing(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
    SiS_Pr->CHTotal = SiS_Pr->CHBlankEnd = tempbx;
 
    if(SiS_Pr->ChipType < SIS_315H) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
       if(SiS_Pr->SiS_VGAHDE == SiS_Pr->PanelXRes) {
          SiS_Pr->CHSyncStart = SiS_Pr->SiS_VGAHDE + ((SiS_Pr->PanelHRS + 1) & ~1);
          SiS_Pr->CHSyncEnd = SiS_Pr->CHSyncStart + SiS_Pr->PanelHRE;
@@ -3982,7 +3462,7 @@ SiS_CalcLCDACRT1Timing(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
       }
 #endif
    } else {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
       tempax = VGAHDE;
       if(SiS_Pr->SiS_LCDInfo & DontExpandLCD) {
          tempbx = SiS_Pr->PanelXRes;
@@ -4001,7 +3481,7 @@ SiS_CalcLCDACRT1Timing(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
    if(SiS_Pr->SiS_LCDInfo & DontExpandLCD) {
       tempax = SiS_Pr->PanelYRes;
    } else if(SiS_Pr->ChipType < SIS_315H) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
       /* Stupid hack for 640x400/320x200 */
       if(SiS_Pr->SiS_LCDResInfo == Panel_1024x768) {
          if((tempax + tempbx) == 438) tempbx += 16;
@@ -4054,36 +3534,12 @@ SiS_CalcLCDACRT1Timing(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
    if(modeflag & DoubleScanMode) tempax |= 0x80;
    SiS_SetRegANDOR(SiS_Pr->SiS_P3d4,0x09,0x5F,tempax);
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-   xf86DrvMsg(0, X_INFO, "%d %d %d %d  %d %d %d %d  (%d %d %d %d)\n",
-        SiS_Pr->CHDisplay, SiS_Pr->CHSyncStart, SiS_Pr->CHSyncEnd, SiS_Pr->CHTotal,
-        SiS_Pr->CVDisplay, SiS_Pr->CVSyncStart, SiS_Pr->CVSyncEnd, SiS_Pr->CVTotal,
-        SiS_Pr->CHBlankStart, SiS_Pr->CHBlankEnd, SiS_Pr->CVBlankStart, SiS_Pr->CVBlankEnd);
-   xf86DrvMsg(0, X_INFO, " {{0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,\n",
-        SiS_Pr->CCRT1CRTC[0], SiS_Pr->CCRT1CRTC[1],
-        SiS_Pr->CCRT1CRTC[2], SiS_Pr->CCRT1CRTC[3],
-        SiS_Pr->CCRT1CRTC[4], SiS_Pr->CCRT1CRTC[5],
-        SiS_Pr->CCRT1CRTC[6], SiS_Pr->CCRT1CRTC[7]);
-   xf86DrvMsg(0, X_INFO, "   0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,\n",
-        SiS_Pr->CCRT1CRTC[8], SiS_Pr->CCRT1CRTC[9],
-        SiS_Pr->CCRT1CRTC[10], SiS_Pr->CCRT1CRTC[11],
-        SiS_Pr->CCRT1CRTC[12], SiS_Pr->CCRT1CRTC[13],
-        SiS_Pr->CCRT1CRTC[14], SiS_Pr->CCRT1CRTC[15]);
-   xf86DrvMsg(0, X_INFO, "   0x%02x}},\n", SiS_Pr->CCRT1CRTC[16]);
-#endif
-#endif
 }
 
 void
 SiS_Generic_ConvertCRData(struct SiS_Private *SiS_Pr, unsigned char *crdata,
                         int xres, int yres,
-#ifdef SIS_XORG_XF86
-                        DisplayModePtr current
-#endif
-#ifdef SIS_LINUX_KERNEL
                         struct fb_var_screeninfo *var, bool writeres
-#endif
 )
 {
    unsigned short HRE, HBE, HRS, HBS, HDE, HT;
@@ -4127,25 +3583,10 @@ SiS_Generic_ConvertCRData(struct SiS_Private *SiS_Pr, unsigned char *crdata,
 
    D = B - F - C;
 
-#ifdef SIS_XORG_XF86
-   current->HDisplay   = (E * 8);
-   current->HSyncStart = (E * 8) + (F * 8);
-   current->HSyncEnd   = (E * 8) + (F * 8) + (C * 8);
-   current->HTotal     = (E * 8) + (F * 8) + (C * 8) + (D * 8);
-#ifdef TWDEBUG
-   xf86DrvMsg(0, X_INFO,
-                "H: A %d B %d C %d D %d E %d F %d  HT %d HDE %d HRS %d HBS %d HBE %d HRE %d\n",
-                A, B, C, D, E, F, HT, HDE, HRS, HBS, HBE, HRE);
-#else
-   (void)VBS;  (void)HBS;  (void)A;
-#endif
-#endif
-#ifdef SIS_LINUX_KERNEL
    if(writeres) var->xres = xres = E * 8;
    var->left_margin = D * 8;
    var->right_margin = F * 8;
    var->hsync_len = C * 8;
-#endif
 
    /* Vertical */
    sr_data = crdata[13];
@@ -4192,30 +3633,10 @@ SiS_Generic_ConvertCRData(struct SiS_Private *SiS_Pr, unsigned char *crdata,
 
    D = B - F - C;
 
-#ifdef SIS_XORG_XF86
-   current->VDisplay   = VDE + 1;
-   current->VSyncStart = VRS + 1;
-   current->VSyncEnd   = ((VRS & ~0x1f) | VRE) + 1;
-   if(VRE <= (VRS & 0x1f)) current->VSyncEnd += 32;
-   current->VTotal     = E + D + C + F;
-#if 0
-   current->VDisplay   = E;
-   current->VSyncStart = E + D;
-   current->VSyncEnd   = E + D + C;
-   current->VTotal     = E + D + C + F;
-#endif
-#ifdef TWDEBUG
-   xf86DrvMsg(0, X_INFO,
-        "V: A %d B %d C %d D %d E %d F %d  VT %d VDE %d VRS %d VBS %d VBE %d VRE %d\n",
-        A, B, C, D, E, F, VT, VDE, VRS, VBS, VBE, VRE);
-#endif
-#endif
-#ifdef SIS_LINUX_KERNEL
    if(writeres) var->yres = yres = E;
    var->upper_margin = D;
    var->lower_margin = F;
    var->vsync_len = C;
-#endif
 
    if((xres == 320) && ((yres == 200) || (yres == 240))) {
         /* Terrible hack, but correct CRTC data for
@@ -4224,17 +3645,9 @@ SiS_Generic_ConvertCRData(struct SiS_Private *SiS_Pr, unsigned char *crdata,
          * a negative D. The CRT controller does not
          * seem to like correcting HRE to 50)
          */
-#ifdef SIS_XORG_XF86
-      current->HDisplay   = 320;
-      current->HSyncStart = 328;
-      current->HSyncEnd   = 376;
-      current->HTotal     = 400;
-#endif
-#ifdef SIS_LINUX_KERNEL
       var->left_margin = (400 - 376);
       var->right_margin = (328 - 320);
       var->hsync_len = (376 - 328);
-#endif
 
    }
 
diff --git a/drivers/video/sis/init.h b/drivers/video/sis/init.h
index b96005c39c67..ee8ed3c203da 100644
--- a/drivers/video/sis/init.h
+++ b/drivers/video/sis/init.h
@@ -53,21 +53,8 @@
 #ifndef _INIT_H_
 #define _INIT_H_
 
-#include "osdef.h"
 #include "initdef.h"
 
-#ifdef SIS_XORG_XF86
-#include "sis.h"
-#define SIS_NEED_inSISREG
-#define SIS_NEED_inSISREGW
-#define SIS_NEED_inSISREGL
-#define SIS_NEED_outSISREG
-#define SIS_NEED_outSISREGW
-#define SIS_NEED_outSISREGL
-#include "sis_regs.h"
-#endif
-
-#ifdef SIS_LINUX_KERNEL
 #include "vgatypes.h"
 #include "vstruct.h"
 #ifdef SIS_CP
@@ -78,7 +65,6 @@
 #include <linux/fb.h>
 #include "sis.h"
 #include <video/sisfb.h>
-#endif
 
 /* Mode numbers */
 static const unsigned short ModeIndex_320x200[]      = {0x59, 0x41, 0x00, 0x4f};
@@ -286,7 +272,7 @@ static const struct SiS_ModeResInfo_S SiS_ModeResInfo[] =
         { 1280, 854, 8,16}    /* 0x22 */
 };
 
-#if defined(SIS300) || defined(SIS315H)
+#if defined(CONFIG_FB_SIS_300) || defined(CONFIG_FB_SIS_315)
 static const struct SiS_StandTable_S SiS_StandTable[]=
 {
 /* 0x00: MD_0_200 */
@@ -1521,10 +1507,6 @@ static const struct SiS_LVDSCRT1Data SiS_LVDSCRT1640x480_1_H[] =
 };
 
 bool            SiSInitPtr(struct SiS_Private *SiS_Pr);
-#ifdef SIS_XORG_XF86
-unsigned short  SiS_GetModeID(int VGAEngine, unsigned int VBFlags, int HDisplay, int VDisplay,
-                                int Depth, bool FSTN, int LCDwith, int LCDheight);
-#endif
 unsigned short  SiS_GetModeID_LCD(int VGAEngine, unsigned int VBFlags, int HDisplay,
                                 int VDisplay, int Depth, bool FSTN,
                                 unsigned short CustomT, int LCDwith, int LCDheight,
@@ -1550,17 +1532,11 @@ void		SiS_SetRegOR(SISIOADDRESS Port,unsigned short Index, unsigned short DataOR
 void            SiS_DisplayOn(struct SiS_Private *SiS_Pr);
 void            SiS_DisplayOff(struct SiS_Private *SiS_Pr);
 void            SiSRegInit(struct SiS_Private *SiS_Pr, SISIOADDRESS BaseAddr);
-#ifndef SIS_LINUX_KERNEL
-void            SiSSetLVDSetc(struct SiS_Private *SiS_Pr);
-#endif
 void            SiS_SetEnableDstn(struct SiS_Private *SiS_Pr, int enable);
 void            SiS_SetEnableFstn(struct SiS_Private *SiS_Pr, int enable);
 unsigned short  SiS_GetModeFlag(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
                                 unsigned short ModeIdIndex);
 bool            SiSDetermineROMLayout661(struct SiS_Private *SiS_Pr);
-#ifndef SIS_LINUX_KERNEL
-void            SiS_GetVBType(struct SiS_Private *SiS_Pr);
-#endif
 
 bool            SiS_SearchModeID(struct SiS_Private *SiS_Pr, unsigned short *ModeNo,
                                 unsigned short *ModeIdIndex);
@@ -1572,37 +1548,19 @@ unsigned short	SiS_GetColorDepth(struct SiS_Private *SiS_Pr, unsigned short Mode
                                 unsigned short ModeIdIndex);
 unsigned short  SiS_GetOffset(struct SiS_Private *SiS_Pr,unsigned short ModeNo,
                                 unsigned short ModeIdIndex, unsigned short RRTI);
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 void            SiS_GetFIFOThresholdIndex300(struct SiS_Private *SiS_Pr, unsigned short *idx1,
                                 unsigned short *idx2);
 unsigned short  SiS_GetFIFOThresholdB300(unsigned short idx1, unsigned short idx2);
 unsigned short  SiS_GetLatencyFactor630(struct SiS_Private *SiS_Pr, unsigned short index);
 #endif
 void            SiS_LoadDAC(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short ModeIdIndex);
-#ifdef SIS_XORG_XF86
-bool            SiSSetMode(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn, unsigned short ModeNo,
-                                bool dosetpitch);
-bool            SiSBIOSSetMode(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn,
-                                DisplayModePtr mode, bool IsCustom);
-bool            SiSBIOSSetModeCRT2(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn,
-                                DisplayModePtr mode, bool IsCustom);
-bool            SiSBIOSSetModeCRT1(struct SiS_Private *SiS_Pr, ScrnInfoPtr pScrn,
-                                DisplayModePtr mode, bool IsCustom);
-#endif
-#ifdef SIS_LINUX_KERNEL
 bool            SiSSetMode(struct SiS_Private *SiS_Pr, unsigned short ModeNo);
-#endif
 void            SiS_CalcCRRegisters(struct SiS_Private *SiS_Pr, int depth);
 void            SiS_CalcLCDACRT1Timing(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
                                 unsigned short ModeIdIndex);
-#ifdef SIS_XORG_XF86
-void            SiS_Generic_ConvertCRData(struct SiS_Private *SiS_Pr, unsigned char *crdata, int xres,
-                                int yres, DisplayModePtr current);
-#endif
-#ifdef SIS_LINUX_KERNEL
 void            SiS_Generic_ConvertCRData(struct SiS_Private *SiS_Pr, unsigned char *crdata, int xres,
                                 int yres, struct fb_var_screeninfo *var, bool writeres);
-#endif
 
 /* From init301.c: */
 extern void             SiS_GetVBInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
@@ -1626,29 +1584,16 @@ extern unsigned short	SiS_GetVCLK2Ptr(struct SiS_Private *SiS_Pr, unsigned short
 extern bool             SiS_IsVAMode(struct SiS_Private *);
 extern bool             SiS_IsDualEdge(struct SiS_Private *);
 
-#ifdef SIS_XORG_XF86
-/* From other modules: */
-extern unsigned short   SiS_CheckBuildCustomMode(ScrnInfoPtr pScrn, DisplayModePtr mode,
-                                unsigned int VBFlags);
-extern unsigned char    SiS_GetSetBIOSScratch(ScrnInfoPtr pScrn, unsigned short offset,
-                                unsigned char value);
-extern unsigned char    SiS_GetSetModeID(ScrnInfoPtr pScrn, unsigned char id);
-extern unsigned short   SiS_GetModeNumber(ScrnInfoPtr pScrn, DisplayModePtr mode,
-                                unsigned int VBFlags);
-#endif
-
-#ifdef SIS_LINUX_KERNEL
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 extern unsigned int     sisfb_read_nbridge_pci_dword(struct SiS_Private *SiS_Pr, int reg);
 extern void             sisfb_write_nbridge_pci_dword(struct SiS_Private *SiS_Pr, int reg,
                                 unsigned int val);
 #endif
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 extern void             sisfb_write_nbridge_pci_byte(struct SiS_Private *SiS_Pr, int reg,
                                 unsigned char val);
 extern unsigned int     sisfb_read_mio_pci_word(struct SiS_Private *SiS_Pr, int reg);
 #endif
-#endif
 
 #endif
 
diff --git a/drivers/video/sis/init301.c b/drivers/video/sis/init301.c
index da33d801c22e..9fa66fd4052a 100644
--- a/drivers/video/sis/init301.c
+++ b/drivers/video/sis/init301.c
@@ -75,11 +75,11 @@
 
 #include "init301.h"
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 #include "oem300.h"
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 #include "oem310.h"
 #endif
 
@@ -87,9 +87,7 @@
 #define SiS_I2CDELAYSHORT  150
 
 static unsigned short   SiS_GetBIOSLCDResInfo(struct SiS_Private *SiS_Pr);
-#ifdef SIS_LINUX_KERNEL
 static void             SiS_SetCH70xx(struct SiS_Private *SiS_Pr, unsigned short reg, unsigned char val);
-#endif
 
 /*********************************************/
 /*         HELPER: Lock/Unlock CRT2          */
@@ -106,9 +104,7 @@ SiS_UnLockCRT2(struct SiS_Private *SiS_Pr)
       SiS_SetRegOR(SiS_Pr->SiS_Part1Port,0x24,0x01);
 }
 
-#ifdef SIS_LINUX_KERNEL
 static
-#endif
 void
 SiS_LockCRT2(struct SiS_Private *SiS_Pr)
 {
@@ -138,7 +134,7 @@ SiS_SetRegSR11ANDOR(struct SiS_Private *SiS_Pr, unsigned short DataAND, unsigned
 /*    HELPER: Get Pointer to LCD structure   */
 /*********************************************/
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static unsigned char *
 GetLCDStructPtr661(struct SiS_Private *SiS_Pr)
 {
@@ -404,7 +400,7 @@ SiS_SaveCRT2Info(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
 /*    HELPER: GET SOME DATA FROM BIOS ROM    */
 /*********************************************/
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 static bool
 SiS_CR36BIOSWord23b(struct SiS_Private *SiS_Pr)
 {
@@ -449,7 +445,7 @@ SiS_DDC2Delay(struct SiS_Private *SiS_Pr, unsigned int delaytime)
       SiS_GetReg(SiS_Pr->SiS_P3c4, 0x05);
 }
 
-#if defined(SIS300) || defined(SIS315H)
+#if defined(CONFIG_FB_SIS_300) || defined(CONFIG_FB_SIS_315)
 static void
 SiS_GenericDelay(struct SiS_Private *SiS_Pr, unsigned short delay)
 {
@@ -457,7 +453,7 @@ SiS_GenericDelay(struct SiS_Private *SiS_Pr, unsigned short delay)
 }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static void
 SiS_LongDelay(struct SiS_Private *SiS_Pr, unsigned short delay)
 {
@@ -467,7 +463,7 @@ SiS_LongDelay(struct SiS_Private *SiS_Pr, unsigned short delay)
 }
 #endif
 
-#if defined(SIS300) || defined(SIS315H)
+#if defined(CONFIG_FB_SIS_300) || defined(CONFIG_FB_SIS_315)
 static void
 SiS_ShortDelay(struct SiS_Private *SiS_Pr, unsigned short delay)
 {
@@ -480,14 +476,14 @@ SiS_ShortDelay(struct SiS_Private *SiS_Pr, unsigned short delay)
 static void
 SiS_PanelDelay(struct SiS_Private *SiS_Pr, unsigned short DelayTime)
 {
-#if defined(SIS300) || defined(SIS315H)
+#if defined(CONFIG_FB_SIS_300) || defined(CONFIG_FB_SIS_315)
    unsigned char  *ROMAddr = SiS_Pr->VirtualRomBase;
    unsigned short PanelID, DelayIndex, Delay=0;
 #endif
 
    if(SiS_Pr->ChipType < SIS_315H) {
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 
       PanelID = SiS_GetReg(SiS_Pr->SiS_P3d4,0x36);
       if(SiS_Pr->SiS_VBType & VB_SISVB) {
@@ -513,11 +509,11 @@ SiS_PanelDelay(struct SiS_Private *SiS_Pr, unsigned short DelayTime)
       }
       SiS_ShortDelay(SiS_Pr, Delay);
 
-#endif  /* SIS300 */
+#endif  /* CONFIG_FB_SIS_300 */
 
    } else {
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 
       if((SiS_Pr->ChipType >= SIS_661)    ||
          (SiS_Pr->ChipType <= SIS_315PRO) ||
@@ -579,12 +575,12 @@ SiS_PanelDelay(struct SiS_Private *SiS_Pr, unsigned short DelayTime)
 
       }
 
-#endif /* SIS315H */
+#endif /* CONFIG_FB_SIS_315 */
 
    }
 }
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static void
 SiS_PanelDelayLoop(struct SiS_Private *SiS_Pr, unsigned short DelayTime, unsigned short DelayLoop)
 {
@@ -613,7 +609,7 @@ SiS_WaitRetrace1(struct SiS_Private *SiS_Pr)
    while((!(SiS_GetRegByte(SiS_Pr->SiS_P3da) & 0x08)) && --watchdog);
 }
 
-#if defined(SIS300) || defined(SIS315H)
+#if defined(CONFIG_FB_SIS_300) || defined(CONFIG_FB_SIS_315)
 static void
 SiS_WaitRetrace2(struct SiS_Private *SiS_Pr, unsigned short reg)
 {
@@ -630,7 +626,7 @@ static void
 SiS_WaitVBRetrace(struct SiS_Private *SiS_Pr)
 {
    if(SiS_Pr->ChipType < SIS_315H) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
       if(SiS_Pr->SiS_VBType & VB_SIS30xBLV) {
          if(!(SiS_GetReg(SiS_Pr->SiS_Part1Port,0x00) & 0x20)) return;
       }
@@ -641,7 +637,7 @@ SiS_WaitVBRetrace(struct SiS_Private *SiS_Pr)
       }
 #endif
    } else {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
       if(!(SiS_GetReg(SiS_Pr->SiS_Part1Port,0x00) & 0x40)) {
          SiS_WaitRetrace1(SiS_Pr);
       } else {
@@ -686,7 +682,7 @@ SiS_VBLongWait(struct SiS_Private *SiS_Pr)
 /*               HELPER: MISC                */
 /*********************************************/
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 static bool
 SiS_Is301B(struct SiS_Private *SiS_Pr)
 {
@@ -708,7 +704,7 @@ SiS_CRT2IsLCD(struct SiS_Private *SiS_Pr)
 bool
 SiS_IsDualEdge(struct SiS_Private *SiS_Pr)
 {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    if(SiS_Pr->ChipType >= SIS_315H) {
       if((SiS_Pr->ChipType != SIS_650) || (SiS_GetReg(SiS_Pr->SiS_P3d4,0x5f) & 0xf0)) {
          if(SiS_GetReg(SiS_Pr->SiS_P3d4,0x38) & EnableDualEdge) return true;
@@ -721,7 +717,7 @@ SiS_IsDualEdge(struct SiS_Private *SiS_Pr)
 bool
 SiS_IsVAMode(struct SiS_Private *SiS_Pr)
 {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    unsigned short flag;
 
    if(SiS_Pr->ChipType >= SIS_315H) {
@@ -732,7 +728,7 @@ SiS_IsVAMode(struct SiS_Private *SiS_Pr)
    return false;
 }
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static bool
 SiS_IsVAorLCD(struct SiS_Private *SiS_Pr)
 {
@@ -745,7 +741,7 @@ SiS_IsVAorLCD(struct SiS_Private *SiS_Pr)
 static bool
 SiS_IsDualLink(struct SiS_Private *SiS_Pr)
 {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    if(SiS_Pr->ChipType >= SIS_315H) {
       if((SiS_CRT2IsLCD(SiS_Pr)) ||
          (SiS_IsVAMode(SiS_Pr))) {
@@ -756,7 +752,7 @@ SiS_IsDualLink(struct SiS_Private *SiS_Pr)
    return false;
 }
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static bool
 SiS_TVEnabled(struct SiS_Private *SiS_Pr)
 {
@@ -768,7 +764,7 @@ SiS_TVEnabled(struct SiS_Private *SiS_Pr)
 }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static bool
 SiS_LCDAEnabled(struct SiS_Private *SiS_Pr)
 {
@@ -777,7 +773,7 @@ SiS_LCDAEnabled(struct SiS_Private *SiS_Pr)
 }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static bool
 SiS_WeHaveBacklightCtrl(struct SiS_Private *SiS_Pr)
 {
@@ -788,7 +784,7 @@ SiS_WeHaveBacklightCtrl(struct SiS_Private *SiS_Pr)
 }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static bool
 SiS_IsNotM650orLater(struct SiS_Private *SiS_Pr)
 {
@@ -804,7 +800,7 @@ SiS_IsNotM650orLater(struct SiS_Private *SiS_Pr)
 }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static bool
 SiS_IsYPbPr(struct SiS_Private *SiS_Pr)
 {
@@ -816,7 +812,7 @@ SiS_IsYPbPr(struct SiS_Private *SiS_Pr)
 }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static bool
 SiS_IsChScart(struct SiS_Private *SiS_Pr)
 {
@@ -828,7 +824,7 @@ SiS_IsChScart(struct SiS_Private *SiS_Pr)
 }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static bool
 SiS_IsTVOrYPbPrOrScart(struct SiS_Private *SiS_Pr)
 {
@@ -848,7 +844,7 @@ SiS_IsTVOrYPbPrOrScart(struct SiS_Private *SiS_Pr)
 }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static bool
 SiS_IsLCDOrLCDA(struct SiS_Private *SiS_Pr)
 {
@@ -914,7 +910,7 @@ SiS_BridgeInSlavemode(struct SiS_Private *SiS_Pr)
 /*********************************************/
 
 /* Setup general purpose IO for Chrontel communication */
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 void
 SiS_SetChrontelGPIO(struct SiS_Private *SiS_Pr, unsigned short myvbinfo)
 {
@@ -923,11 +919,7 @@ SiS_SetChrontelGPIO(struct SiS_Private *SiS_Pr, unsigned short myvbinfo)
 
    if(!(SiS_Pr->SiS_ChSW)) return;
 
-#ifdef SIS_LINUX_KERNEL
    acpibase = sisfb_read_lpc_pci_dword(SiS_Pr, 0x74);
-#else
-   acpibase = pciReadLong(0x00000800, 0x74);
-#endif
    acpibase &= 0xFFFF;
    if(!acpibase) return;
    temp = SiS_GetRegShort((acpibase + 0x3c));   /* ACPI register 0x3c: GP Event 1 I/O mode select */
@@ -969,7 +961,7 @@ SiS_GetVBInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
         tempax &= (DriverMode | LoadDACFlag | SetNotSimuMode | SetPALTV);
         tempbx |= tempax;
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
         if(SiS_Pr->ChipType >= SIS_315H) {
            if(SiS_Pr->SiS_VBType & VB_SISLCDA) {
               if(ModeNo == 0x03) {
@@ -1019,7 +1011,7 @@ SiS_GetVBInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
            }
         }
 
-#endif  /* SIS315H */
+#endif  /* CONFIG_FB_SIS_315 */
 
         if(!(SiS_Pr->SiS_VBType & VB_SISVGA2)) {
            tempbx &= ~(SetCRT2ToRAMDAC);
@@ -1154,24 +1146,16 @@ SiS_GetVBInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo,
 
    SiS_Pr->SiS_VBInfo = tempbx;
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
    if(SiS_Pr->ChipType == SIS_630) {
       SiS_SetChrontelGPIO(SiS_Pr, SiS_Pr->SiS_VBInfo);
    }
 #endif
 
-#ifdef SIS_LINUX_KERNEL
 #if 0
    printk(KERN_DEBUG "sisfb: (init301: VBInfo= 0x%04x, SetFlag=0x%04x)\n",
       SiS_Pr->SiS_VBInfo, SiS_Pr->SiS_SetFlag);
 #endif
-#endif
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-   xf86DrvMsg(0, X_PROBED, "(init301: VBInfo=0x%04x, SetFlag=0x%04x)\n",
-      SiS_Pr->SiS_VBInfo, SiS_Pr->SiS_SetFlag);
-#endif
-#endif
 }
 
 /*********************************************/
@@ -1415,12 +1399,6 @@ SiS_SetTVMode(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
    }
 
    SiS_Pr->SiS_VBInfo &= ~SetPALTV;
-
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-   xf86DrvMsg(0, X_INFO, "(init301: TVMode %x, VBInfo %x)\n", SiS_Pr->SiS_TVMode, SiS_Pr->SiS_VBInfo);
-#endif
-#endif
 }
 
 /*********************************************/
@@ -1443,22 +1421,10 @@ SiS_GetBIOSLCDResInfo(struct SiS_Private *SiS_Pr)
 static void
 SiS_GetLCDInfoBIOS(struct SiS_Private *SiS_Pr)
 {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    unsigned char  *ROMAddr;
    unsigned short temp;
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-   xf86DrvMsg(0, X_INFO, "Paneldata driver: [%d %d] [H %d %d] [V %d %d] [C %d 0x%02x 0x%02x]\n",
-        SiS_Pr->PanelHT, SiS_Pr->PanelVT,
-        SiS_Pr->PanelHRS, SiS_Pr->PanelHRE,
-        SiS_Pr->PanelVRS, SiS_Pr->PanelVRE,
-        SiS_Pr->SiS_VBVCLKData[SiS_Pr->PanelVCLKIdx315].CLOCK,
-        SiS_Pr->SiS_VBVCLKData[SiS_Pr->PanelVCLKIdx315].Part4_A,
-        SiS_Pr->SiS_VBVCLKData[SiS_Pr->PanelVCLKIdx315].Part4_B);
-#endif
-#endif
-
    if((ROMAddr = GetLCDStructPtr661(SiS_Pr))) {
       if((temp = SISGETROMW(6)) != SiS_Pr->PanelHT) {
          SiS_Pr->SiS_NeedRomModeData = true;
@@ -1480,18 +1446,6 @@ SiS_GetLCDInfoBIOS(struct SiS_Private *SiS_Pr)
       SiS_Pr->SiS_VCLKData[VCLK_CUSTOM_315].SR2C =
          SiS_Pr->SiS_VBVCLKData[VCLK_CUSTOM_315].Part4_B = ROMAddr[20];
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-      xf86DrvMsg(0, X_INFO, "Paneldata BIOS:  [%d %d] [H %d %d] [V %d %d] [C %d 0x%02x 0x%02x]\n",
-        SiS_Pr->PanelHT, SiS_Pr->PanelVT,
-        SiS_Pr->PanelHRS, SiS_Pr->PanelHRE,
-        SiS_Pr->PanelVRS, SiS_Pr->PanelVRE,
-        SiS_Pr->SiS_VBVCLKData[SiS_Pr->PanelVCLKIdx315].CLOCK,
-        SiS_Pr->SiS_VBVCLKData[SiS_Pr->PanelVCLKIdx315].Part4_A,
-        SiS_Pr->SiS_VBVCLKData[SiS_Pr->PanelVCLKIdx315].Part4_B);
-#endif
-#endif
-
    }
 #endif
 }
@@ -1517,13 +1471,13 @@ SiS_GetLCDResInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sh
 {
   unsigned short temp,modeflag,resinfo=0,modexres=0,modeyres=0;
   bool panelcanscale = false;
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
   unsigned char *ROMAddr = SiS_Pr->VirtualRomBase;
   static const unsigned char SiS300SeriesLCDRes[] =
           { 0,  1,  2,  3,  7,  4,  5,  8,
             0,  0, 10,  0,  0,  0,  0, 15 };
 #endif
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   unsigned char   *myptr = NULL;
 #endif
 
@@ -1562,7 +1516,7 @@ SiS_GetLCDResInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sh
      SiS_Pr->SiS_LCDTypeInfo = (temp & 0x0F) - 1;
   }
   temp &= 0x0f;
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
   if(SiS_Pr->ChipType < SIS_315H) {
      /* Very old BIOSes only know 7 sizes (NetVista 2179, 1.01g) */
      if(SiS_Pr->SiS_VBType & VB_SIS301) {
@@ -1574,7 +1528,7 @@ SiS_GetLCDResInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sh
 #endif
 
   /* Translate to our internal types */
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   if(SiS_Pr->ChipType == SIS_550) {
      if     (temp == Panel310_1152x768)  temp = Panel_320x240_2; /* Verified working */
      else if(temp == Panel310_320x240_2) temp = Panel_320x240_2;
@@ -1597,7 +1551,7 @@ SiS_GetLCDResInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sh
 
   SiS_Pr->SiS_LCDResInfo = temp;
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
   if(SiS_Pr->SiS_IF_DEF_LVDS == 1) {
      if(SiS_Pr->SiS_CustomT == CUT_BARCO1366) {
         SiS_Pr->SiS_LCDResInfo = Panel_Barco1366;
@@ -1639,7 +1593,7 @@ SiS_GetLCDResInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sh
   else if(SiS_Pr->UsePanelScaler == 1) SiS_Pr->SiS_LCDInfo |= DontExpandLCD;
 
   /* Dual link, Pass 1:1 BIOS default, etc. */
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   if(SiS_Pr->ChipType >= SIS_661) {
      if(SiS_Pr->SiS_LCDInfo & DontExpandLCD) {
         if(temp & 0x08) SiS_Pr->SiS_LCDInfo |= LCDPass11;
@@ -2076,7 +2030,7 @@ SiS_GetLCDResInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sh
      }
   }
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
   if(SiS_Pr->SiS_IF_DEF_LVDS == 1) {
      if(SiS_Pr->SiS_CustomT == CUT_PANEL848 || SiS_Pr->SiS_CustomT == CUT_PANEL856) {
         SiS_Pr->SiS_LCDInfo = 0x80 | 0x40 | 0x20;   /* neg h/v sync, RGB24(D0 = 0) */
@@ -2186,17 +2140,10 @@ SiS_GetLCDResInfo(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned sh
      SiS_Pr->SiS_SetFlag |= LCDVESATiming;
   }
 
-#ifdef SIS_LINUX_KERNEL
 #if 0
   printk(KERN_DEBUG "sisfb: (LCDInfo=0x%04x LCDResInfo=0x%02x LCDTypeInfo=0x%02x)\n",
         SiS_Pr->SiS_LCDInfo, SiS_Pr->SiS_LCDResInfo, SiS_Pr->SiS_LCDTypeInfo);
 #endif
-#endif
-#ifdef SIS_XORG_XF86
-  xf86DrvMsgVerb(0, X_PROBED, 4,
-        "(init301: LCDInfo=0x%04x LCDResInfo=0x%02x LCDTypeInfo=0x%02x SetFlag=0x%04x)\n",
-        SiS_Pr->SiS_LCDInfo, SiS_Pr->SiS_LCDResInfo, SiS_Pr->SiS_LCDTypeInfo, SiS_Pr->SiS_SetFlag);
-#endif
 }
 
 /*********************************************/
@@ -2359,7 +2306,7 @@ SiS_GetVCLK2Ptr(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned shor
               VCLKIndex = SiS_Pr->PanelVCLKIdx315;
            }
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
            /* Special Timing: Barco iQ Pro R series */
            if(SiS_Pr->SiS_CustomT == CUT_BARCO1366) VCLKIndex = 0x44;
 
@@ -2410,12 +2357,6 @@ SiS_GetVCLK2Ptr(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned shor
 
   }
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-  xf86DrvMsg(0, X_INFO, "VCLKIndex %d (0x%x)\n", VCLKIndex, VCLKIndex);
-#endif
-#endif
-
   return VCLKIndex;
 }
 
@@ -2428,10 +2369,10 @@ SiS_SetCRT2ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
 {
   unsigned short i, j, modeflag, tempah=0;
   short tempcl;
-#if defined(SIS300) || defined(SIS315H)
+#if defined(CONFIG_FB_SIS_300) || defined(CONFIG_FB_SIS_315)
   unsigned short tempbl;
 #endif
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   unsigned char  *ROMAddr = SiS_Pr->VirtualRomBase;
   unsigned short tempah2, tempbl2;
 #endif
@@ -2454,7 +2395,7 @@ SiS_SetCRT2ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
 
      if(SiS_Pr->ChipType < SIS_315H) {
 
-#ifdef SIS300    /* ---- 300 series ---- */
+#ifdef CONFIG_FB_SIS_300    /* ---- 300 series ---- */
 
         /* For 301BDH: (with LCD via LVDS) */
         if(SiS_Pr->SiS_VBType & VB_NoLCD) {
@@ -2477,11 +2418,11 @@ SiS_SetCRT2ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
 
         if(SiS_Pr->SiS_VBInfo & SetInSlaveMode)  tempah ^= 0xA0;
 
-#endif  /* SIS300 */
+#endif  /* CONFIG_FB_SIS_300 */
 
      } else {
 
-#ifdef SIS315H    /* ------- 315/330 series ------ */
+#ifdef CONFIG_FB_SIS_315    /* ------- 315/330 series ------ */
 
         if(ModeNo > 0x13) {
            tempcl -= ModeVGA;
@@ -2494,7 +2435,7 @@ SiS_SetCRT2ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
 
         if(SiS_Pr->SiS_VBInfo & SetInSlaveMode) tempah ^= 0x50;
 
-#endif  /* SIS315H */
+#endif  /* CONFIG_FB_SIS_315 */
 
      }
 
@@ -2503,7 +2444,7 @@ SiS_SetCRT2ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
      if(SiS_Pr->ChipType < SIS_315H) {
         SiS_SetReg(SiS_Pr->SiS_Part1Port,0x00,tempah);
      } else {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
         if(SiS_Pr->SiS_IF_DEF_LVDS == 1) {
            SiS_SetRegANDOR(SiS_Pr->SiS_Part1Port,0x00,0xa0,tempah);
         } else if(SiS_Pr->SiS_VBType & VB_SISVB) {
@@ -2584,7 +2525,7 @@ SiS_SetCRT2ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
 
         if(SiS_Pr->ChipType >= SIS_315H) {
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
            /* LVDS can only be slave in 8bpp modes */
            tempah = 0x80;
            if((modeflag & CRT2Mode) && (SiS_Pr->SiS_ModeType > ModeVGA)) {
@@ -2604,7 +2545,7 @@ SiS_SetCRT2ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
 
         } else {
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
            tempah = 0;
            if( (!(SiS_Pr->SiS_VBInfo & SetInSlaveMode)) && (SiS_Pr->SiS_ModeType > ModeVGA) ) {
               tempah |= 0x02;
@@ -2626,7 +2567,7 @@ SiS_SetCRT2ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
 
      if(SiS_Pr->ChipType >= SIS_315H) {
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
         /* unsigned char bridgerev = SiS_GetReg(SiS_Pr->SiS_Part4Port,0x01); */
 
         /* The following is nearly unpreditable and varies from machine
@@ -2718,11 +2659,11 @@ SiS_SetCRT2ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
            SiS_SetRegANDOR(SiS_Pr->SiS_Part4Port,0x23,tempbl,tempah);
         }
 
-#endif /* SIS315H */
+#endif /* CONFIG_FB_SIS_315 */
 
      } else if(SiS_Pr->SiS_VBType & VB_SIS30xBLV) {
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
         SiS_SetRegAND(SiS_Pr->SiS_Part4Port,0x21,0x3f);
 
         if((SiS_Pr->SiS_VBInfo & DisableCRT2Display) ||
@@ -2745,7 +2686,7 @@ SiS_SetCRT2ModeRegs(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
 
   } else {  /* LVDS */
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
      if(SiS_Pr->ChipType >= SIS_315H) {
 
         if(SiS_Pr->SiS_IF_DEF_CH70xx != 0) {
@@ -2931,7 +2872,7 @@ SiS_GetCRT2Ptr(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
            }
         }
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
         if(SiS_Pr->SiS_CustomT == CUT_COMPAQ1280) {
            if(SiS_Pr->SiS_LCDResInfo == Panel_1280x1024) {
               if(!(SiS_Pr->SiS_LCDInfo & DontExpandLCD)) {
@@ -3036,7 +2977,7 @@ SiS_GetCRT2Ptr(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
         case Panel_1280x1024: tempbx = 24; break;
         case Panel_1400x1050: tempbx = 26; break;
         case Panel_1600x1200: tempbx = 28; break;
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
         case Panel_Barco1366: tempbx = 80; break;
 #endif
         }
@@ -3053,7 +2994,7 @@ SiS_GetCRT2Ptr(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 
         if(SiS_Pr->SiS_LCDInfo & LCDPass11) tempbx = 30;
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
         if(SiS_Pr->SiS_CustomT == CUT_BARCO1024) {
            tempbx = 82;
            if(SiS_Pr->SiS_LCDInfo & DontExpandLCD) tempbx++;
@@ -3189,7 +3130,7 @@ SiS_GetCRT2DataLVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
 
    if((SiS_Pr->SiS_VBType & VB_SISVB) && (SiS_Pr->SiS_VBInfo & SetCRT2ToLCDA)) {
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
       SiS_CalcPanelLinkTiming(SiS_Pr, ModeNo, ModeIdIndex, RefreshRateTableIndex);
       SiS_CalcLCDACRT1Timing(SiS_Pr, ModeNo, ModeIdIndex);
 #endif
@@ -3214,7 +3155,7 @@ SiS_GetCRT2DataLVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
          case 16: LVDSData = SiS_Pr->SiS_LVDS800x600Data_1;    break;
          case 18: LVDSData = SiS_Pr->SiS_LVDS1024x600Data_1;   break;
          case 20: LVDSData = SiS_Pr->SiS_LVDS1024x768Data_1;   break;
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
          case 80: LVDSData = SiS_Pr->SiS_LVDSBARCO1366Data_1;  break;
          case 81: LVDSData = SiS_Pr->SiS_LVDSBARCO1366Data_2;  break;
          case 82: LVDSData = SiS_Pr->SiS_LVDSBARCO1024Data_1;  break;
@@ -3248,7 +3189,7 @@ SiS_GetCRT2DataLVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned
              (SiS_Pr->SiS_SetFlag & SetDOSMode) ) {
             SiS_Pr->SiS_HDE = SiS_Pr->PanelXRes;
             SiS_Pr->SiS_VDE = SiS_Pr->PanelYRes;
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
             if(SiS_Pr->SiS_CustomT == CUT_BARCO1366) {
                if(ResIndex < 0x08) {
                   SiS_Pr->SiS_HDE = 1280;
@@ -3270,7 +3211,7 @@ SiS_GetCRT2Data301(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
   unsigned short resinfo, CRT2Index, ResIndex;
   const struct SiS_LCDData *LCDPtr = NULL;
   const struct SiS_TVData  *TVPtr  = NULL;
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   short resinfo661;
 #endif
 
@@ -3283,7 +3224,7 @@ SiS_GetCRT2Data301(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
   } else {
      modeflag = SiS_Pr->SiS_EModeIDTable[ModeIdIndex].Ext_ModeFlag;
      resinfo = SiS_Pr->SiS_EModeIDTable[ModeIdIndex].Ext_RESINFO;
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
      resinfo661 = SiS_Pr->SiS_EModeIDTable[ModeIdIndex].ROMMODEIDX661;
      if( (SiS_Pr->SiS_VBInfo & SetCRT2ToLCD)   &&
          (SiS_Pr->SiS_SetFlag & LCDVESATiming) &&
@@ -3460,7 +3401,7 @@ SiS_GetCRT2Data301(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
 
         } else if( (!(SiS_Pr->SiS_LCDInfo & DontExpandLCD)) && (romptr) && (ROMAddr) ) {
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
            SiS_Pr->SiS_RVBHCMAX  = ROMAddr[romptr];
            SiS_Pr->SiS_RVBHCFACT = ROMAddr[romptr+1];
            SiS_Pr->SiS_VGAHT     = ROMAddr[romptr+2] | ((ROMAddr[romptr+3] & 0x0f) << 8);
@@ -3520,19 +3461,13 @@ SiS_GetCRT2Data301(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
               case Panel_1680x1050     :
               case Panel_1680x1050 + 32: LCDPtr = SiS_Pr->SiS_LCD1680x1050Data;     break;
               case 100                 : LCDPtr = SiS_Pr->SiS_NoScaleData;          break;
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
               case 200                 : LCDPtr = SiS310_ExtCompaq1280x1024Data;    break;
               case 201                 : LCDPtr = SiS_Pr->SiS_St2LCD1280x1024Data;  break;
 #endif
               default                  : LCDPtr = SiS_Pr->SiS_ExtLCD1024x768Data;   break;
            }
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-           xf86DrvMsg(0, X_INFO, "GetCRT2Data: Index %d ResIndex %d\n", CRT2Index, ResIndex);
-#endif
-#endif
-
            SiS_Pr->SiS_RVBHCMAX  = (LCDPtr+ResIndex)->RVBHCMAX;
            SiS_Pr->SiS_RVBHCFACT = (LCDPtr+ResIndex)->RVBHCFACT;
            SiS_Pr->SiS_VGAHT     = (LCDPtr+ResIndex)->VGAHT;
@@ -3624,7 +3559,7 @@ SiS_GetLVDSDesPtr(struct SiS_Private *SiS_Pr)
 {
    const struct SiS_LVDSDes *PanelDesPtr = NULL;
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
    if(SiS_Pr->SiS_VBInfo & SetCRT2ToLCD) {
 
       if(SiS_Pr->ChipType < SIS_315H) {
@@ -3696,7 +3631,7 @@ SiS_GetLVDSDesData(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
 
   if((SiS_Pr->SiS_VBType & VB_SIS30xBLV) && (SiS_Pr->SiS_VBInfo & SetCRT2ToLCDA)) {
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
      if(SiS_Pr->SiS_LCDInfo & DontExpandLCD) {
         /* non-pass 1:1 only, see above */
         if(SiS_Pr->SiS_VGAHDE != SiS_Pr->PanelXRes) {
@@ -3771,7 +3706,7 @@ SiS_GetLVDSDesData(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
      } else {
 
         if(SiS_Pr->ChipType < SIS_315H) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
            switch(SiS_Pr->SiS_LCDResInfo) {
            case Panel_800x600:
               if(SiS_Pr->SiS_VGAVDE == SiS_Pr->PanelYRes) {
@@ -3816,7 +3751,7 @@ SiS_GetLVDSDesData(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
            }
 #endif
         } else {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
            switch(SiS_Pr->SiS_LCDResInfo) {
            case Panel_1024x768:
            case Panel_1280x1024:
@@ -3844,7 +3779,7 @@ SiS_GetLVDSDesData(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
                  if(SiS_Pr->ChipType < SIS_315H) {
                     if(!(modeflag & HalfDCLK)) SiS_Pr->SiS_LCDHDES = 320;
                  } else {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
                     if(SiS_Pr->SiS_LCDResInfo == Panel_1024x768)  SiS_Pr->SiS_LCDHDES = 480;
                     if(SiS_Pr->SiS_LCDResInfo == Panel_1400x1050) SiS_Pr->SiS_LCDHDES = 804;
                     if(SiS_Pr->SiS_LCDResInfo == Panel_1600x1200) SiS_Pr->SiS_LCDHDES = 704;
@@ -3866,7 +3801,7 @@ SiS_GetLVDSDesData(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
 /*           DISABLE VIDEO BRIDGE            */
 /*********************************************/
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static int
 SiS_HandlePWD(struct SiS_Private *SiS_Pr)
 {
@@ -3891,11 +3826,6 @@ SiS_HandlePWD(struct SiS_Private *SiS_Pr)
          ret = 1;
       }
       SiS_SetRegANDOR(SiS_Pr->SiS_Part4Port,0x27,0x7f,temp);
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-      xf86DrvMsg(0, 0, "Setting PWD %x\n", temp);
-#endif
-#endif
    }
 #endif
    return ret;
@@ -3909,7 +3839,7 @@ SiS_HandlePWD(struct SiS_Private *SiS_Pr)
 void
 SiS_DisableBridge(struct SiS_Private *SiS_Pr)
 {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   unsigned short tempah, pushax=0, modenum;
 #endif
   unsigned short temp=0;
@@ -3920,7 +3850,7 @@ SiS_DisableBridge(struct SiS_Private *SiS_Pr)
 
         if(SiS_Pr->ChipType < SIS_315H) {
 
-#ifdef SIS300      /* 300 series */
+#ifdef CONFIG_FB_SIS_300           /* 300 series */
 
            if(!(SiS_CR36BIOSWord23b(SiS_Pr))) {
               if(SiS_Pr->SiS_VBType & VB_SISLVDS) {
@@ -3953,11 +3883,11 @@ SiS_DisableBridge(struct SiS_Private *SiS_Pr)
               }
            }
 
-#endif  /* SIS300 */
+#endif  /* CONFIG_FB_SIS_300 */
 
         } else {
 
-#ifdef SIS315H     /* 315 series */
+#ifdef CONFIG_FB_SIS_315           /* 315 series */
 
            int didpwd = 0;
            bool custom1 = (SiS_Pr->SiS_CustomT == CUT_COMPAQ1280) ||
@@ -4081,14 +4011,14 @@ SiS_DisableBridge(struct SiS_Private *SiS_Pr)
 
            }
 
-#endif /* SIS315H */
+#endif /* CONFIG_FB_SIS_315 */
 
         }
 
      } else {     /* ============ For 301 ================ */
 
         if(SiS_Pr->ChipType < SIS_315H) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
            if(!(SiS_CR36BIOSWord23b(SiS_Pr))) {
               SiS_SetRegSR11ANDOR(SiS_Pr,0xF7,0x08);
               SiS_PanelDelay(SiS_Pr, 3);
@@ -4111,7 +4041,7 @@ SiS_DisableBridge(struct SiS_Private *SiS_Pr)
             SiS_SetRegOR(SiS_Pr->SiS_P3c4,0x1E,0x20);
             SiS_SetReg(SiS_Pr->SiS_Part1Port,0x00,temp);
         } else {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
             SiS_SetRegAND(SiS_Pr->SiS_P3c4,0x1E,0xDF);            /* disable CRT2 */
             if( (!(SiS_CRT2IsLCD(SiS_Pr))) ||
                 (!(SiS_CR36BIOSWord23d(SiS_Pr))) ) {
@@ -4127,7 +4057,7 @@ SiS_DisableBridge(struct SiS_Private *SiS_Pr)
 
     if(SiS_Pr->ChipType < SIS_315H) {
 
-#ifdef SIS300   /* 300 series */
+#ifdef CONFIG_FB_SIS_300        /* 300 series */
 
         if(SiS_Pr->SiS_IF_DEF_CH70xx == 1) {
            SiS_SetCH700x(SiS_Pr,0x0E,0x09);
@@ -4171,11 +4101,11 @@ SiS_DisableBridge(struct SiS_Private *SiS_Pr)
            SiS_SetRegSR11ANDOR(SiS_Pr,0xFB,0x04);
         }
 
-#endif  /* SIS300 */
+#endif  /* CONFIG_FB_SIS_300 */
 
     } else {
 
-#ifdef SIS315H  /* 315 series */
+#ifdef CONFIG_FB_SIS_315        /* 315 series */
 
         if(!(SiS_IsNotM650orLater(SiS_Pr))) {
            /*if(SiS_Pr->ChipType < SIS_340) { */ /* XGI needs this */
@@ -4288,7 +4218,7 @@ SiS_DisableBridge(struct SiS_Private *SiS_Pr)
            }
         }
 
-#endif  /* SIS315H */
+#endif  /* CONFIG_FB_SIS_315 */
 
     }  /* 315 series */
 
@@ -4304,14 +4234,12 @@ SiS_DisableBridge(struct SiS_Private *SiS_Pr)
  * from outside the context of a mode switch!
  * MUST call getVBType before calling this
  */
-#ifdef SIS_LINUX_KERNEL
 static
-#endif
 void
 SiS_EnableBridge(struct SiS_Private *SiS_Pr)
 {
   unsigned short temp=0, tempah;
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   unsigned short temp1, pushax=0;
   bool delaylong = false;
 #endif
@@ -4322,7 +4250,7 @@ SiS_EnableBridge(struct SiS_Private *SiS_Pr)
 
       if(SiS_Pr->ChipType < SIS_315H) {
 
-#ifdef SIS300     /* 300 series */
+#ifdef CONFIG_FB_SIS_300     /* 300 series */
 
          if(SiS_CRT2IsLCD(SiS_Pr)) {
             if(SiS_Pr->SiS_VBType & VB_SISLVDS) {
@@ -4385,11 +4313,11 @@ SiS_EnableBridge(struct SiS_Private *SiS_Pr)
          }
 
 
-#endif /* SIS300 */
+#endif /* CONFIG_FB_SIS_300 */
 
       } else {
 
-#ifdef SIS315H    /* 315 series */
+#ifdef CONFIG_FB_SIS_315    /* 315 series */
 
 #ifdef SET_EMI
          unsigned char   r30=0, r31=0, r32=0, r33=0, cr36=0;
@@ -4688,7 +4616,7 @@ SiS_EnableBridge(struct SiS_Private *SiS_Pr)
             SiS_SetRegAND(SiS_Pr->SiS_Part1Port,0x00,0x7f);
          }
 
-#endif /* SIS315H */
+#endif /* CONFIG_FB_SIS_315 */
 
       }
 
@@ -4739,7 +4667,7 @@ SiS_EnableBridge(struct SiS_Private *SiS_Pr)
 
     if(SiS_Pr->ChipType < SIS_315H) {
 
-#ifdef SIS300    /* 300 series */
+#ifdef CONFIG_FB_SIS_300    /* 300 series */
 
        if(SiS_CRT2IsLCD(SiS_Pr)) {
           if(SiS_Pr->ChipType == SIS_730) {
@@ -4783,11 +4711,11 @@ SiS_EnableBridge(struct SiS_Private *SiS_Pr)
           }
        }
 
-#endif  /* SIS300 */
+#endif  /* CONFIG_FB_SIS_300 */
 
     } else {
 
-#ifdef SIS315H    /* 315 series */
+#ifdef CONFIG_FB_SIS_315    /* 315 series */
 
        if(!(SiS_IsNotM650orLater(SiS_Pr))) {
           /*if(SiS_Pr->ChipType < SIS_340) {*/  /* XGI needs this */
@@ -4881,7 +4809,7 @@ SiS_EnableBridge(struct SiS_Private *SiS_Pr)
           }
        }
 
-#endif  /* SIS315H */
+#endif  /* CONFIG_FB_SIS_315 */
 
     } /* 310 series */
 
@@ -4971,7 +4899,7 @@ SiS_SetCRT2Sync(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned shor
 
       if(SiS_Pr->ChipType < SIS_315H) {
 
-#ifdef SIS300  /* ---- 300 series --- */
+#ifdef CONFIG_FB_SIS_300  /* ---- 300 series --- */
 
          if(SiS_Pr->SiS_VBType & VB_SIS30xBLV) {                        /* 630 - 301B(-DH) */
 
@@ -5000,11 +4928,11 @@ SiS_SetCRT2Sync(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned shor
 
          }
 
-#endif /* SIS300 */
+#endif /* CONFIG_FB_SIS_300 */
 
       } else {
 
-#ifdef SIS315H  /* ------- 315 series ------ */
+#ifdef CONFIG_FB_SIS_315  /* ------- 315 series ------ */
 
          if(SiS_Pr->SiS_VBType & VB_SISLVDS) {                  /* 315 - LVDS */
 
@@ -5076,13 +5004,13 @@ SiS_SetCRT2Sync(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned shor
             }
 
          }
-#endif  /* SIS315H */
+#endif  /* CONFIG_FB_SIS_315 */
       }
    }
 }
 
 /* Set CRT2 FIFO on 300/540/630/730 */
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 static void
 SiS_SetCRT2FIFO_300(struct SiS_Private *SiS_Pr,unsigned short ModeNo)
 {
@@ -5154,13 +5082,8 @@ SiS_SetCRT2FIFO_300(struct SiS_Private *SiS_Pr,unsigned short ModeNo)
 
      } else {
 
-#ifdef SIS_LINUX_KERNEL
         pci50 = sisfb_read_nbridge_pci_dword(SiS_Pr, 0x50);
         pciA0 = sisfb_read_nbridge_pci_dword(SiS_Pr, 0xa0);
-#else
-        pci50 = pciReadLong(0x00000000, 0x50);
-        pciA0 = pciReadLong(0x00000000, 0xA0);
-#endif
 
         if(SiS_Pr->ChipType == SIS_730) {
 
@@ -5262,7 +5185,7 @@ SiS_SetCRT2FIFO_300(struct SiS_Private *SiS_Pr,unsigned short ModeNo)
 #endif
 
 /* Set CRT2 FIFO on 315/330 series */
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static void
 SiS_SetCRT2FIFO_310(struct SiS_Private *SiS_Pr)
 {
@@ -5420,27 +5343,6 @@ SiS_SetGroup1_301(struct SiS_Private *SiS_Pr, unsigned short ModeNo,unsigned sho
 
   temp = SiS_GetRegByte((SiS_Pr->SiS_P3ca+0x02));
   SiS_SetReg(SiS_Pr->SiS_Part1Port,0x1b,temp);                  /* ? */
-
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-   xf86DrvMsg(0, X_INFO, "%d %d %d %d  %d %d %d %d  (%d %d %d %d)\n",
-        SiS_Pr->CHDisplay, SiS_Pr->CHSyncStart, SiS_Pr->CHSyncEnd, SiS_Pr->CHTotal,
-        SiS_Pr->CVDisplay, SiS_Pr->CVSyncStart, SiS_Pr->CVSyncEnd, SiS_Pr->CVTotal,
-        SiS_Pr->CHBlankStart, SiS_Pr->CHBlankEnd, SiS_Pr->CVBlankStart, SiS_Pr->CVBlankEnd);
-
-   xf86DrvMsg(0, X_INFO, " {{0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,\n",
-        SiS_Pr->CCRT1CRTC[0], SiS_Pr->CCRT1CRTC[1],
-        SiS_Pr->CCRT1CRTC[2], SiS_Pr->CCRT1CRTC[3],
-        SiS_Pr->CCRT1CRTC[4], SiS_Pr->CCRT1CRTC[5],
-        SiS_Pr->CCRT1CRTC[6], SiS_Pr->CCRT1CRTC[7]);
-   xf86DrvMsg(0, X_INFO, "   0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,0x%02x,\n",
-        SiS_Pr->CCRT1CRTC[8], SiS_Pr->CCRT1CRTC[9],
-        SiS_Pr->CCRT1CRTC[10], SiS_Pr->CCRT1CRTC[11],
-        SiS_Pr->CCRT1CRTC[12], SiS_Pr->CCRT1CRTC[13],
-        SiS_Pr->CCRT1CRTC[14], SiS_Pr->CCRT1CRTC[15]);
-   xf86DrvMsg(0, X_INFO, "   0x%02x}},\n", SiS_Pr->CCRT1CRTC[16]);
-#endif
-#endif
 }
 
 /* Setup panel link
@@ -5455,17 +5357,17 @@ SiS_SetGroup1_LVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
   unsigned short push2, tempax, tempbx, tempcx, temp;
   unsigned int   tempeax = 0, tempebx, tempecx, tempvcfact = 0;
   bool islvds = false, issis  = false, chkdclkfirst = false;
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
   unsigned short crt2crtc = 0;
 #endif
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   unsigned short pushcx;
 #endif
 
   if(ModeNo <= 0x13) {
      modeflag = SiS_Pr->SiS_SModeIDTable[ModeIdIndex].St_ModeFlag;
      resinfo = SiS_Pr->SiS_SModeIDTable[ModeIdIndex].St_ResInfo;
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
      crt2crtc = SiS_Pr->SiS_SModeIDTable[ModeIdIndex].St_CRT2CRTC;
 #endif
   } else if(SiS_Pr->UseCustomMode) {
@@ -5473,7 +5375,7 @@ SiS_SetGroup1_LVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
   } else {
      modeflag = SiS_Pr->SiS_EModeIDTable[ModeIdIndex].Ext_ModeFlag;
      resinfo = SiS_Pr->SiS_EModeIDTable[ModeIdIndex].Ext_RESINFO;
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
      crt2crtc = SiS_Pr->SiS_RefIndex[RefreshRateTableIndex].Ext_CRT2CRTC;
 #endif
   }
@@ -5494,7 +5396,7 @@ SiS_SetGroup1_LVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
      }
   }
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   if((SiS_Pr->ChipType >= SIS_315H) && (SiS_Pr->SiS_VBInfo & SetCRT2ToLCDA)) {
      if(IS_SIS330) {
         SiS_SetRegOR(SiS_Pr->SiS_Part1Port,0x2D,0x10);
@@ -5744,7 +5646,7 @@ SiS_SetGroup1_LVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
 
   if(SiS_Pr->ChipType < SIS_315H) {
 
-#ifdef SIS300      /* 300 series */
+#ifdef CONFIG_FB_SIS_300      /* 300 series */
      tempeax = SiS_Pr->SiS_VGAVDE << 6;
      temp = (tempeax % (unsigned int)SiS_Pr->SiS_VDE);
      tempeax = tempeax / (unsigned int)SiS_Pr->SiS_VDE;
@@ -5755,11 +5657,11 @@ SiS_SetGroup1_LVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
      temp = (unsigned short)(tempeax & 0x00FF);
      SiS_SetReg(SiS_Pr->SiS_Part1Port,0x1E,temp);       /* BPLVCFACT */
      tempvcfact = temp;
-#endif /* SIS300 */
+#endif /* CONFIG_FB_SIS_300 */
 
   } else {
 
-#ifdef SIS315H  /* 315 series */
+#ifdef CONFIG_FB_SIS_315  /* 315 series */
      tempeax = SiS_Pr->SiS_VGAVDE << 18;
      tempebx = SiS_Pr->SiS_VDE;
      temp = (tempeax % tempebx);
@@ -5845,7 +5747,7 @@ SiS_SetGroup1_LVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
   temp = (unsigned short)(tempecx & 0x00FF);
   SiS_SetReg(SiS_Pr->SiS_Part1Port,0x23,temp);
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   if(SiS_Pr->ChipType >= SIS_315H) {
      if(SiS_Pr->SiS_VBInfo & SetCRT2ToLCDA) {
         if((islvds) || (SiS_Pr->SiS_VBInfo & VB_SISLVDS)) {
@@ -5863,7 +5765,7 @@ SiS_SetGroup1_LVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
   }
 #endif
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
   if(SiS_Pr->SiS_IF_DEF_TRUMPION) {
      unsigned char *ROMAddr = SiS_Pr->VirtualRomBase;
      unsigned char *trumpdata;
@@ -5899,7 +5801,7 @@ SiS_SetGroup1_LVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
   }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   if(SiS_Pr->SiS_IF_DEF_FSTN || SiS_Pr->SiS_IF_DEF_DSTN) {
      SiS_SetReg(SiS_Pr->SiS_Part1Port,0x25,0x00);
      SiS_SetReg(SiS_Pr->SiS_Part1Port,0x26,0x00);
@@ -5999,7 +5901,7 @@ SiS_SetGroup1_LVDS(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned s
         SiS_SetReg(SiS_Pr->SiS_Part1Port,0x45,0x0a);
      }
   }
-#endif  /* SIS315H */
+#endif  /* CONFIG_FB_SIS_315 */
 }
 
 /* Set Part 1 */
@@ -6007,12 +5909,12 @@ static void
 SiS_SetGroup1(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short ModeIdIndex,
                 unsigned short RefreshRateTableIndex)
 {
-#if defined(SIS300) || defined(SIS315H)
+#if defined(CONFIG_FB_SIS_300) || defined(CONFIG_FB_SIS_315)
   unsigned char   *ROMAddr = SiS_Pr->VirtualRomBase;
 #endif
   unsigned short  temp=0, tempax=0, tempbx=0, tempcx=0, bridgeadd=0;
   unsigned short  pushbx=0, CRT1Index=0, modeflag, resinfo=0;
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   unsigned short  tempbl=0;
 #endif
 
@@ -6038,11 +5940,11 @@ SiS_SetGroup1(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
          (SiS_Pr->SiS_VBInfo & SetInSlaveMode)) ) {
 
      if(SiS_Pr->ChipType < SIS_315H ) {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
         SiS_SetCRT2FIFO_300(SiS_Pr, ModeNo);
 #endif
      } else {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
         SiS_SetCRT2FIFO_310(SiS_Pr);
 #endif
      }
@@ -6051,7 +5953,7 @@ SiS_SetGroup1(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 
      if(SiS_Pr->ChipType < SIS_315H ) {
 
-#ifdef SIS300   /* ------------- 300 series --------------*/
+#ifdef CONFIG_FB_SIS_300   /* ------------- 300 series --------------*/
 
         temp = (SiS_Pr->SiS_VGAHT - 1) & 0x0FF;                   /* BTVGA2HT 0x08,0x09 */
         SiS_SetReg(SiS_Pr->SiS_Part1Port,0x08,temp);              /* CRT2 Horizontal Total */
@@ -6070,11 +5972,11 @@ SiS_SetGroup1(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 
         bridgeadd = 12;
 
-#endif /* SIS300 */
+#endif /* CONFIG_FB_SIS_300 */
 
      } else {
 
-#ifdef SIS315H  /* ------------------- 315/330 series --------------- */
+#ifdef CONFIG_FB_SIS_315  /* ------------------- 315/330 series --------------- */
 
         tempcx = SiS_Pr->SiS_VGAHT;                               /* BTVGA2HT 0x08,0x09 */
         if(modeflag & HalfDCLK) {
@@ -6125,7 +6027,7 @@ SiS_SetGroup1(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
            }
         }
 
-#endif  /* SIS315H */
+#endif  /* CONFIG_FB_SIS_315 */
 
      }  /* 315/330 series */
 
@@ -6256,7 +6158,7 @@ SiS_SetGroup1(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 
      if(SiS_Pr->ChipType < SIS_315H) {
 
-#ifdef SIS300  /* ---------- 300 series -------------- */
+#ifdef CONFIG_FB_SIS_300  /* ---------- 300 series -------------- */
 
         if(SiS_Pr->SiS_VBType & VB_SISVB) {
            temp = 0x20;
@@ -6310,11 +6212,11 @@ SiS_SetGroup1(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 
         SiS_SetRegANDOR(SiS_Pr->SiS_Part1Port,0x13,~0x3C,temp);   /* Panel Link Delay Compensation; (Software Command Reset; Power Saving) */
 
-#endif  /* SIS300 */
+#endif  /* CONFIG_FB_SIS_300 */
 
      } else {
 
-#ifdef SIS315H   /* --------------- 315/330 series ---------------*/
+#ifdef CONFIG_FB_SIS_315   /* --------------- 315/330 series ---------------*/
 
         if(SiS_Pr->ChipType < SIS_661) {
 
@@ -6349,7 +6251,7 @@ SiS_SetGroup1(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
         if(modeflag & HalfDCLK)       tempax |= 0x40;
         SiS_SetRegANDOR(SiS_Pr->SiS_Part1Port,0x2C,0x3f,tempax);
 
-#endif  /* SIS315H */
+#endif  /* CONFIG_FB_SIS_315 */
 
      }
 
@@ -6381,7 +6283,7 @@ SiS_SetGroup1(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 /*         SET PART 2 REGISTER GROUP         */
 /*********************************************/
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static unsigned char *
 SiS_GetGroup2CLVXPtr(struct SiS_Private *SiS_Pr, int tabletype)
 {
@@ -6478,7 +6380,7 @@ SiS_GetCRT2Part2Ptr(struct SiS_Private *SiS_Pr,unsigned short ModeNo,unsigned sh
 }
 #endif
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 static void
 SiS_Group2LCDSpecial(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short crt2crtc)
 {
@@ -6690,7 +6592,7 @@ SiS_SetGroup2(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
   unsigned int   longtemp, PhaseIndex;
   bool           newtvphase;
   const unsigned char *TimingPoint;
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   unsigned short resindex, CRT2Index;
   const struct SiS_Part2PortTbl *CRT2Part2Ptr = NULL;
 
@@ -7069,7 +6971,7 @@ SiS_SetGroup2(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
   SiS_SetRegAND(SiS_Pr->SiS_Part2Port,0x17,0xFB);
   SiS_SetRegAND(SiS_Pr->SiS_Part2Port,0x18,0xDF);
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   if(SiS_GetCRT2Part2Ptr(SiS_Pr, ModeNo, ModeIdIndex, RefreshRateTableIndex,
                                                 &CRT2Index, &resindex)) {
       switch(CRT2Index) {
@@ -7130,12 +7032,6 @@ SiS_SetGroup2(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 
     /* Non-expanding: lcdvdes = tempcx = VT-1; lcdvdee = tempbx = VDE-1 */
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-    xf86DrvMsg(0, X_INFO, "lcdvdes 0x%x lcdvdee 0x%x\n", tempcx, tempbx);
-#endif
-#endif
-
     SiS_SetReg(SiS_Pr->SiS_Part2Port,0x05,tempcx);      /* lcdvdes  */
     SiS_SetReg(SiS_Pr->SiS_Part2Port,0x06,tempbx);      /* lcdvdee  */
 
@@ -7184,12 +7080,6 @@ SiS_SetGroup2(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
        tempbx = SiS_Pr->CVSyncStart;
     }
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-    xf86DrvMsg(0, X_INFO, "lcdvrs 0x%x\n", tempbx);
-#endif
-#endif
-
     SiS_SetReg(SiS_Pr->SiS_Part2Port,0x04,tempbx);          /* lcdvrs */
 
     temp = (tempbx >> 4) & 0xF0;
@@ -7201,15 +7091,9 @@ SiS_SetGroup2(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
        temp |= (SiS_Pr->CVSyncEnd & 0x0f);
     }
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-    xf86DrvMsg(0, X_INFO, "lcdvre[3:0] 0x%x\n", (temp & 0x0f));
-#endif
-#endif
-
     SiS_SetReg(SiS_Pr->SiS_Part2Port,0x01,temp);
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
     SiS_Group2LCDSpecial(SiS_Pr, ModeNo, crt2crtc);
 #endif
 
@@ -7245,12 +7129,6 @@ SiS_SetGroup2(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
        tempax >>= 1;
     }
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-    xf86DrvMsg(0, X_INFO, "lcdhdee 0x%x\n", tempbx);
-#endif
-#endif
-
     tempbx += bridgeoffset;
 
     SiS_SetReg(SiS_Pr->SiS_Part2Port,0x23,tempbx);          /* lcdhdee */
@@ -7276,12 +7154,6 @@ SiS_SetGroup2(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
        tempbx += bridgeoffset;
     }
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-    xf86DrvMsg(0, X_INFO, "lcdhrs 0x%x\n", tempbx);
-#endif
-#endif
-
     SiS_SetReg(SiS_Pr->SiS_Part2Port,0x1C,tempbx);          /* lcdhrs */
     SiS_SetRegANDOR(SiS_Pr->SiS_Part2Port,0x1D,0x0F,((tempbx >> 4) & 0xf0));
 
@@ -7300,20 +7172,14 @@ SiS_SetGroup2(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
        tempbx += bridgeoffset;
     }
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-    xf86DrvMsg(0, X_INFO, "lcdhre 0x%x\n", tempbx);
-#endif
-#endif
-
     SiS_SetReg(SiS_Pr->SiS_Part2Port,0x21,tempbx);          /* lcdhre */
 
     SiS_SetGroup2_Tail(SiS_Pr, ModeNo);
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
     SiS_Set300Part2Regs(SiS_Pr, ModeIdIndex, RefreshRateTableIndex, ModeNo);
 #endif
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
   } /* CRT2-LCD from table */
 #endif
 }
@@ -7382,7 +7248,7 @@ SiS_SetGroup3(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 /*         SET PART 4 REGISTER GROUP         */
 /*********************************************/
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 #if 0
 static void
 SiS_ShiftXPos(struct SiS_Private *SiS_Pr, int shift)
@@ -8011,7 +7877,7 @@ SiS_SetCHTVReg(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 
    if(SiS_Pr->SiS_IF_DEF_CH70xx == 1) {
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 
       /* Chrontel 7005 - I assume that it does not come with a 315 series chip */
 
@@ -8124,7 +7990,7 @@ SiS_SetCHTVReg(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 
       /* Chrontel 7019 - assumed that it does not come with a 300 series chip */
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 
       unsigned short temp;
 
@@ -8175,7 +8041,7 @@ SiS_SetCHTVReg(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short
 
 }
 
-#ifdef SIS315H  /* ----------- 315 series only ---------- */
+#ifdef CONFIG_FB_SIS_315  /* ----------- 315 series only ---------- */
 
 void
 SiS_Chrontel701xBLOn(struct SiS_Private *SiS_Pr)
@@ -8657,7 +8523,7 @@ SiS_ChrontelDoSomething1(struct SiS_Private *SiS_Pr)
 bool
 SiS_SetCRT2Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
 {
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
    unsigned char  *ROMAddr  = SiS_Pr->VirtualRomBase;
 #endif
    unsigned short ModeIdIndex, RefreshRateTableIndex;
@@ -8703,16 +8569,6 @@ SiS_SetCRT2Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
       SiS_GetLVDSDesData(SiS_Pr, ModeNo, ModeIdIndex, RefreshRateTableIndex);
    }
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-  xf86DrvMsg(0, X_INFO, "(init301: LCDHDES 0x%03x LCDVDES 0x%03x)\n", SiS_Pr->SiS_LCDHDES, SiS_Pr->SiS_LCDVDES);
-  xf86DrvMsg(0, X_INFO, "(init301: HDE     0x%03x VDE     0x%03x)\n", SiS_Pr->SiS_HDE, SiS_Pr->SiS_VDE);
-  xf86DrvMsg(0, X_INFO, "(init301: VGAHDE  0x%03x VGAVDE  0x%03x)\n", SiS_Pr->SiS_VGAHDE, SiS_Pr->SiS_VGAVDE);
-  xf86DrvMsg(0, X_INFO, "(init301: HT      0x%03x VT      0x%03x)\n", SiS_Pr->SiS_HT, SiS_Pr->SiS_VT);
-  xf86DrvMsg(0, X_INFO, "(init301: VGAHT   0x%03x VGAVT   0x%03x)\n", SiS_Pr->SiS_VGAHT, SiS_Pr->SiS_VGAVT);
-#endif
-#endif
-
    if(SiS_Pr->SiS_SetFlag & LowModeTests) {
       SiS_SetGroup1(SiS_Pr, ModeNo, ModeIdIndex, RefreshRateTableIndex);
    }
@@ -8722,12 +8578,12 @@ SiS_SetCRT2Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
       if(SiS_Pr->SiS_SetFlag & LowModeTests) {
 
          SiS_SetGroup2(SiS_Pr, ModeNo, ModeIdIndex, RefreshRateTableIndex);
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
          SiS_SetGroup2_C_ELV(SiS_Pr, ModeNo, ModeIdIndex, RefreshRateTableIndex);
 #endif
          SiS_SetGroup3(SiS_Pr, ModeNo, ModeIdIndex);
          SiS_SetGroup4(SiS_Pr, ModeNo, ModeIdIndex, RefreshRateTableIndex);
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
          SiS_SetGroup4_C_ELV(SiS_Pr, ModeNo, ModeIdIndex);
 #endif
          SiS_SetGroup5(SiS_Pr, ModeNo, ModeIdIndex);
@@ -8758,7 +8614,7 @@ SiS_SetCRT2Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
          if(SiS_Pr->SiS_IF_DEF_CH70xx != 0) {
             if(SiS_Pr->SiS_VBInfo & (SetCRT2ToLCD | SetCRT2ToLCDA)) {
                if(SiS_Pr->SiS_IF_DEF_CH70xx == 2) {
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
                   SiS_SetCH701xForLCD(SiS_Pr);
 #endif
                }
@@ -8771,7 +8627,7 @@ SiS_SetCRT2Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
 
    }
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
    if(SiS_Pr->ChipType < SIS_315H) {
       if(SiS_Pr->SiS_SetFlag & LowModeTests) {
          if(SiS_Pr->SiS_UseOEM) {
@@ -8794,7 +8650,7 @@ SiS_SetCRT2Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo)
    }
 #endif
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
    if(SiS_Pr->ChipType >= SIS_315H) {
       if(SiS_Pr->SiS_SetFlag & LowModeTests) {
          if(SiS_Pr->ChipType < SIS_661) {
@@ -8873,7 +8729,7 @@ SiS_SetupDDCN(struct SiS_Private *SiS_Pr)
   }
 }
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 static unsigned char *
 SiS_SetTrumpBlockLoop(struct SiS_Private *SiS_Pr, unsigned char *dataptr)
 {
@@ -8923,11 +8779,6 @@ SiS_SetTrumpionBlock(struct SiS_Private *SiS_Pr, unsigned char *dataptr)
      dataptr = SiS_SetTrumpBlockLoop(SiS_Pr, dataptr);
      if(!dataptr) return false;
   }
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-  xf86DrvMsg(0, X_INFO, "Trumpion block success\n");
-#endif
-#endif
   return true;
 }
 #endif
@@ -9002,9 +8853,7 @@ SiS_SetCH701x(struct SiS_Private *SiS_Pr, unsigned short reg, unsigned char val)
   SiS_SetChReg(SiS_Pr, reg, val, 0);
 }
 
-#ifdef SIS_LINUX_KERNEL
 static
-#endif
 void
 SiS_SetCH70xx(struct SiS_Private *SiS_Pr, unsigned short reg, unsigned char val)
 {
@@ -9091,9 +8940,7 @@ SiS_GetCH701x(struct SiS_Private *SiS_Pr, unsigned short tempbx)
 
 /* Read from Chrontel 70xx */
 /* Parameter is [Register no (S7-S0)] */
-#ifdef SIS_LINUX_KERNEL
 static
-#endif
 unsigned short
 SiS_GetCH70xx(struct SiS_Private *SiS_Pr, unsigned short tempbx)
 {
@@ -9114,9 +8961,7 @@ SiS_SetCH70xxANDOR(struct SiS_Private *SiS_Pr, unsigned short reg,
 }
 
 /* Our own DDC functions */
-#ifndef SIS_XORG_XF86
 static
-#endif
 unsigned short
 SiS_InitDDCRegs(struct SiS_Private *SiS_Pr, unsigned int VBFlags, int VGAEngine,
                 unsigned short adaptnum, unsigned short DDCdatatype, bool checkcr32,
@@ -9224,12 +9069,6 @@ SiS_InitDDCRegs(struct SiS_Private *SiS_Pr, unsigned int VBFlags, int VGAEngine,
 
     SiS_SetupDDCN(SiS_Pr);
 
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-    xf86DrvMsg(0, X_INFO, "DDC Port %x Index %x Shift %d\n",
-                SiS_Pr->SiS_DDC_Port, SiS_Pr->SiS_DDC_Index, temp);
-#endif
-#endif
     return 0;
 }
 
@@ -9292,11 +9131,6 @@ SiS_DoProbeDDC(struct SiS_Private *SiS_Pr)
     SiS_SetSwitchDDC2(SiS_Pr);
     if(SiS_PrepareDDC(SiS_Pr)) {
          SiS_SetStop(SiS_Pr);
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-         xf86DrvMsg(0, X_INFO, "Probe: Prepare failed\n");
-#endif
-#endif
          return 0xFFFF;
     }
     mask = 0xf0;
@@ -9310,11 +9144,6 @@ SiS_DoProbeDDC(struct SiS_Private *SiS_Pr)
        } else {
            failed = true;
            ret = 0xFFFF;
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-           xf86DrvMsg(0, X_INFO, "Probe: Read 1 failed\n");
-#endif
-#endif
        }
     }
     if(!failed) {
@@ -9324,11 +9153,6 @@ SiS_DoProbeDDC(struct SiS_Private *SiS_Pr)
        if(temp == value) ret = 0;
        else {
           ret = 0xFFFF;
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-          xf86DrvMsg(0, X_INFO, "Probe: Read 2 failed\n");
-#endif
-#endif
           if(SiS_Pr->SiS_DDC_DeviceAddr == 0xa0) {
              if(temp == 0x30) ret = 0;
           }
@@ -9338,9 +9162,7 @@ SiS_DoProbeDDC(struct SiS_Private *SiS_Pr)
     return ret;
 }
 
-#ifndef SIS_XORG_XF86
 static
-#endif
 unsigned short
 SiS_ProbeDDC(struct SiS_Private *SiS_Pr)
 {
@@ -9357,9 +9179,7 @@ SiS_ProbeDDC(struct SiS_Private *SiS_Pr)
    return flag;
 }
 
-#ifndef SIS_XORG_XF86
 static
-#endif
 unsigned short
 SiS_ReadDDC(struct SiS_Private *SiS_Pr, unsigned short DDCdatatype, unsigned char *buffer)
 {
@@ -9606,11 +9426,6 @@ SiS_SetSCLKHigh(struct SiS_Private *SiS_Pr)
     temp = SiS_GetReg(SiS_Pr->SiS_DDC_Port,SiS_Pr->SiS_DDC_Index);
   } while((!(temp & SiS_Pr->SiS_DDC_Clk)) && --watchdog);
   if (!watchdog) {
-#ifdef SIS_XORG_XF86
-#ifdef TWDEBUG
-        xf86DrvMsg(0, X_INFO, "SetClkHigh failed\n");
-#endif
-#endif
         return 0xFFFF;
   }
   SiS_DDC2Delay(SiS_Pr,SiS_I2CDELAYSHORT);
@@ -9641,7 +9456,7 @@ SiS_CheckACK(struct SiS_Private *SiS_Pr)
 
 /* =============== SiS 315/330 O.E.M. ================= */
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 
 static unsigned short
 GetRAMDACromptr(struct SiS_Private *SiS_Pr)
@@ -10829,7 +10644,7 @@ SiS_FinalizeLCD(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned shor
 
 /*  =================  SiS 300 O.E.M. ================== */
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 
 static void
 SetOEMLCDData2(struct SiS_Private *SiS_Pr, unsigned short ModeNo,unsigned short ModeIdIndex,
diff --git a/drivers/video/sis/init301.h b/drivers/video/sis/init301.h
index 51d99222375d..e1fd31d0fddf 100644
--- a/drivers/video/sis/init301.h
+++ b/drivers/video/sis/init301.h
@@ -53,15 +53,8 @@
 #ifndef  _INIT301_H_
 #define  _INIT301_H_
 
-#include "osdef.h"
 #include "initdef.h"
 
-#ifdef SIS_XORG_XF86
-#include "sis.h"
-#include "sis_regs.h"
-#endif
-
-#ifdef SIS_LINUX_KERNEL
 #include "vgatypes.h"
 #include "vstruct.h"
 #ifdef SIS_CP
@@ -72,7 +65,6 @@
 #include <linux/fb.h>
 #include "sis.h"
 #include <video/sisfb.h>
-#endif
 
 static const unsigned char SiS_YPbPrTable[3][64] = {
   {
@@ -237,7 +229,7 @@ static const unsigned char SiS_Part2CLVX_6[] = {   /* 1080i */
     0xFF,0xFF,
 };
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 /* 661 et al LCD data structure (2.03.00) */
 static const unsigned char SiS_LCDStruct661[] = {
     /* 1024x768 */
@@ -279,7 +271,7 @@ static const unsigned char SiS_LCDStruct661[] = {
 };
 #endif
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 static unsigned char SiS300_TrumpionData[14][80] = {
   { 0x02,0x0A,0x0A,0x01,0x04,0x01,0x00,0x03,0x0D,0x00,0x0D,0x10,0x7F,0x00,0x80,0x02,
     0x20,0x03,0x0B,0x00,0x90,0x01,0xC1,0x01,0x60,0x0C,0x30,0x10,0x00,0x00,0x04,0x23,
@@ -356,9 +348,6 @@ static unsigned char SiS300_TrumpionData[14][80] = {
 #endif
 
 void            SiS_UnLockCRT2(struct SiS_Private *SiS_Pr);
-#ifndef SIS_LINUX_KERNEL
-void            SiS_LockCRT2(struct SiS_Private *SiS_Pr);
-#endif
 void            SiS_EnableCRT2(struct SiS_Private *SiS_Pr);
 unsigned short  SiS_GetRatePtr(struct SiS_Private *SiS_Pr, unsigned short ModeNo, unsigned short ModeIdIndex);
 void            SiS_WaitRetrace1(struct SiS_Private *SiS_Pr);
@@ -375,9 +364,6 @@ unsigned short	SiS_GetVCLK2Ptr(struct SiS_Private *SiS_Pr, unsigned short ModeNo
                         unsigned short RefreshRateTableIndex);
 unsigned short  SiS_GetResInfo(struct SiS_Private *SiS_Pr,unsigned short ModeNo,unsigned short ModeIdIndex);
 void            SiS_DisableBridge(struct SiS_Private *SiS_Pr);
-#ifndef SIS_LINUX_KERNEL
-void            SiS_EnableBridge(struct SiS_Private *SiS_Pr);
-#endif
 bool            SiS_SetCRT2Group(struct SiS_Private *SiS_Pr, unsigned short ModeNo);
 void            SiS_SiS30xBLOn(struct SiS_Private *SiS_Pr);
 void            SiS_SiS30xBLOff(struct SiS_Private *SiS_Pr);
@@ -386,13 +372,9 @@ void		SiS_SetCH700x(struct SiS_Private *SiS_Pr, unsigned short reg, unsigned cha
 unsigned short  SiS_GetCH700x(struct SiS_Private *SiS_Pr, unsigned short tempax);
 void            SiS_SetCH701x(struct SiS_Private *SiS_Pr, unsigned short reg, unsigned char val);
 unsigned short  SiS_GetCH701x(struct SiS_Private *SiS_Pr, unsigned short tempax);
-#ifndef SIS_LINUX_KERNEL
-void            SiS_SetCH70xx(struct SiS_Private *SiS_Pr, unsigned short reg, unsigned char val);
-unsigned short  SiS_GetCH70xx(struct SiS_Private *SiS_Pr, unsigned short tempax);
-#endif
 void            SiS_SetCH70xxANDOR(struct SiS_Private *SiS_Pr, unsigned short reg,
                         unsigned char orval,unsigned short andval);
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static void     SiS_Chrontel701xOn(struct SiS_Private *SiS_Pr);
 static void     SiS_Chrontel701xOff(struct SiS_Private *SiS_Pr);
 static void     SiS_ChrontelInitTVVSync(struct SiS_Private *SiS_Pr);
@@ -401,7 +383,7 @@ void		SiS_Chrontel701xBLOn(struct SiS_Private *SiS_Pr);
 void            SiS_Chrontel701xBLOff(struct SiS_Private *SiS_Pr);
 #endif /* 315 */
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 static  bool    SiS_SetTrumpionBlock(struct SiS_Private *SiS_Pr, unsigned char *dataptr);
 void            SiS_SetChrontelGPIO(struct SiS_Private *SiS_Pr, unsigned short myvbinfo);
 #endif
@@ -412,21 +394,12 @@ unsigned short	SiS_HandleDDC(struct SiS_Private *SiS_Pr, unsigned int VBFlags, i
                         unsigned short adaptnum, unsigned short DDCdatatype,
                         unsigned char *buffer, unsigned int VBFlags2);
 
-#ifdef SIS_XORG_XF86
-unsigned short          SiS_InitDDCRegs(struct SiS_Private *SiS_Pr, unsigned int VBFlags,
-                                int VGAEngine, unsigned short adaptnum, unsigned short DDCdatatype,
-                                bool checkcr32, unsigned int VBFlags2);
-unsigned short          SiS_ProbeDDC(struct SiS_Private *SiS_Pr);
-unsigned short          SiS_ReadDDC(struct SiS_Private *SiS_Pr, unsigned short DDCdatatype,
-                                unsigned char *buffer);
-#else
 static unsigned short   SiS_InitDDCRegs(struct SiS_Private *SiS_Pr, unsigned int VBFlags,
                                 int VGAEngine, unsigned short adaptnum, unsigned short DDCdatatype,
                                 bool checkcr32, unsigned int VBFlags2);
 static unsigned short   SiS_ProbeDDC(struct SiS_Private *SiS_Pr);
 static unsigned short   SiS_ReadDDC(struct SiS_Private *SiS_Pr, unsigned short DDCdatatype,
                                 unsigned char *buffer);
-#endif
 static void             SiS_SetSwitchDDC2(struct SiS_Private *SiS_Pr);
 static unsigned short   SiS_SetStart(struct SiS_Private *SiS_Pr);
 static unsigned short   SiS_SetStop(struct SiS_Private *SiS_Pr);
@@ -441,13 +414,13 @@ static unsigned short	SiS_PrepareDDC(struct SiS_Private *SiS_Pr);
 static void             SiS_SendACK(struct SiS_Private *SiS_Pr, unsigned short yesno);
 static unsigned short   SiS_DoProbeDDC(struct SiS_Private *SiS_Pr);
 
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 static void             SiS_OEM300Setting(struct SiS_Private *SiS_Pr,
                                 unsigned short ModeNo, unsigned short ModeIdIndex, unsigned short RefTabindex);
 static void             SetOEMLCDData2(struct SiS_Private *SiS_Pr,
                                 unsigned short ModeNo, unsigned short ModeIdIndex,unsigned short RefTableIndex);
 #endif
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
 static void             SiS_OEM310Setting(struct SiS_Private *SiS_Pr,
                                 unsigned short ModeNo,unsigned short ModeIdIndex, unsigned short RRTI);
 static void             SiS_OEM661Setting(struct SiS_Private *SiS_Pr,
@@ -482,15 +455,13 @@ extern void		SiS_CalcLCDACRT1Timing(struct SiS_Private *SiS_Pr, unsigned short M
 extern void             SiS_CalcCRRegisters(struct SiS_Private *SiS_Pr, int depth);
 extern unsigned short   SiS_GetRefCRTVCLK(struct SiS_Private *SiS_Pr, unsigned short Index, int UseWide);
 extern unsigned short   SiS_GetRefCRT1CRTC(struct SiS_Private *SiS_Pr, unsigned short Index, int UseWide);
-#ifdef SIS300
+#ifdef CONFIG_FB_SIS_300
 extern void             SiS_GetFIFOThresholdIndex300(struct SiS_Private *SiS_Pr, unsigned short *tempbx,
                                 unsigned short *tempcl);
 extern unsigned short   SiS_GetFIFOThresholdB300(unsigned short tempbx, unsigned short tempcl);
 extern unsigned short   SiS_GetLatencyFactor630(struct SiS_Private *SiS_Pr, unsigned short index);
-#ifdef SIS_LINUX_KERNEL
 extern unsigned int     sisfb_read_nbridge_pci_dword(struct SiS_Private *SiS_Pr, int reg);
 extern unsigned int     sisfb_read_lpc_pci_dword(struct SiS_Private *SiS_Pr, int reg);
 #endif
-#endif
 
 #endif
diff --git a/drivers/video/sis/initextlfb.c b/drivers/video/sis/initextlfb.c
index 99c04a4855d1..9dec64da4015 100644
--- a/drivers/video/sis/initextlfb.c
+++ b/drivers/video/sis/initextlfb.c
@@ -25,7 +25,6 @@
  * Author:      Thomas Winischhofer <thomas@winischhofer.net>
  */
 
-#include "osdef.h"
 #include "initdef.h"
 #include "vgatypes.h"
 #include "vstruct.h"
@@ -59,7 +58,7 @@ sisfb_mode_rate_to_dclock(struct SiS_Private *SiS_Pr, unsigned char modeno,
 
     if(rateindex > 0) rateindex--;
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
     switch(ModeNo) {
     case 0x5a: ModeNo = 0x50; break;
     case 0x5b: ModeNo = 0x56;
@@ -103,7 +102,7 @@ sisfb_mode_rate_to_ddata(struct SiS_Private *SiS_Pr, unsigned char modeno,
 
     if(rateindex > 0) rateindex--;
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
     switch(ModeNo) {
        case 0x5a: ModeNo = 0x50; break;
        case 0x5b: ModeNo = 0x56;
@@ -187,7 +186,7 @@ sisfb_gettotalfrommode(struct SiS_Private *SiS_Pr, unsigned char modeno, int *ht
 
     if(rateindex > 0) rateindex--;
 
-#ifdef SIS315H
+#ifdef CONFIG_FB_SIS_315
     switch(ModeNo) {
        case 0x5a: ModeNo = 0x50; break;
        case 0x5b: ModeNo = 0x56;
diff --git a/drivers/video/sis/osdef.h b/drivers/video/sis/osdef.h
deleted file mode 100644
index 6ff8f988a1a7..000000000000
--- a/drivers/video/sis/osdef.h
+++ /dev/null
@@ -1,133 +0,0 @@
-/* $XFree86$ */
-/* $XdotOrg$ */
-/*
- * OS depending defines
- *
- * Copyright (C) 2001-2005 by Thomas Winischhofer, Vienna, Austria
- *
- * If distributed as part of the Linux kernel, the following license terms
- * apply:
- *
- * * This program is free software; you can redistribute it and/or modify
- * * it under the terms of the GNU General Public License as published by
- * * the Free Software Foundation; either version 2 of the named License,
- * * or any later version.
- * *
- * * This program is distributed in the hope that it will be useful,
- * * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * * GNU General Public License for more details.
- * *
- * * You should have received a copy of the GNU General Public License
- * * along with this program; if not, write to the Free Software
- * * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307, USA
- *
- * Otherwise, the following license terms apply:
- *
- * * Redistribution and use in source and binary forms, with or without
- * * modification, are permitted provided that the following conditions
- * * are met:
- * * 1) Redistributions of source code must retain the above copyright
- * *    notice, this list of conditions and the following disclaimer.
- * * 2) Redistributions in binary form must reproduce the above copyright
- * *    notice, this list of conditions and the following disclaimer in the
- * *    documentation and/or other materials provided with the distribution.
- * * 3) The name of the author may not be used to endorse or promote products
- * *    derived from this software without specific prior written permission.
- * *
- * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR
- * * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
- * * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
- * * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT,
- * * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
- * * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
- * * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
- * * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
- * * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
- * * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
- *
- * Author:      Thomas Winischhofer <thomas@winischhofer.net>
- *              Silicon Integrated Systems, Inc. (used by permission)
- *
- */
-
-#ifndef _SIS_OSDEF_H_
-#define _SIS_OSDEF_H_
-
-/* The choices are: */
-#define SIS_LINUX_KERNEL                /* Linux kernel framebuffer */
-#undef  SIS_XORG_XF86                   /* XFree86/X.org */
-
-#ifdef OutPortByte
-#undef OutPortByte
-#endif
-
-#ifdef OutPortWord
-#undef OutPortWord
-#endif
-
-#ifdef OutPortLong
-#undef OutPortLong
-#endif
-
-#ifdef InPortByte
-#undef InPortByte
-#endif
-
-#ifdef InPortWord
-#undef InPortWord
-#endif
-
-#ifdef InPortLong
-#undef InPortLong
-#endif
-
-/**********************************************************************/
-/*  LINUX KERNEL                                                      */
-/**********************************************************************/
-
-#ifdef SIS_LINUX_KERNEL
-
-#ifdef CONFIG_FB_SIS_300
-#define SIS300
-#endif
-
-#ifdef CONFIG_FB_SIS_315
-#define SIS315H
-#endif
-
-#if !defined(SIS300) && !defined(SIS315H)
-#warning Neither CONFIG_FB_SIS_300 nor CONFIG_FB_SIS_315 is set
-#warning sisfb will not work!
-#endif
-
-#define OutPortByte(p,v) outb((u8)(v),(SISIOADDRESS)(p))
-#define OutPortWord(p,v) outw((u16)(v),(SISIOADDRESS)(p))
-#define OutPortLong(p,v) outl((u32)(v),(SISIOADDRESS)(p))
-#define InPortByte(p)    inb((SISIOADDRESS)(p))
-#define InPortWord(p)    inw((SISIOADDRESS)(p))
-#define InPortLong(p)    inl((SISIOADDRESS)(p))
-#define SiS_SetMemory(MemoryAddress,MemorySize,value) memset_io(MemoryAddress, value, MemorySize)
-
-#endif /* LINUX_KERNEL */
-
-/**********************************************************************/
-/*  XFree86/X.org                                                    */
-/**********************************************************************/
-
-#ifdef SIS_XORG_XF86
-
-#define SIS300
-#define SIS315H
-
-#define OutPortByte(p,v) outSISREG((IOADDRESS)(p),(CARD8)(v))
-#define OutPortWord(p,v) outSISREGW((IOADDRESS)(p),(CARD16)(v))
-#define OutPortLong(p,v) outSISREGL((IOADDRESS)(p),(CARD32)(v))
-#define InPortByte(p)    inSISREG((IOADDRESS)(p))
-#define InPortWord(p)    inSISREGW((IOADDRESS)(p))
-#define InPortLong(p)    inSISREGL((IOADDRESS)(p))
-#define SiS_SetMemory(MemoryAddress,MemorySize,value) memset(MemoryAddress, value, MemorySize)
-
-#endif /* XF86 */
-
-#endif  /* _OSDEF_H_ */
diff --git a/drivers/video/sis/sis.h b/drivers/video/sis/sis.h
index 7c5710e3fb56..80d89d37c414 100644
--- a/drivers/video/sis/sis.h
+++ b/drivers/video/sis/sis.h
@@ -24,7 +24,6 @@
 #ifndef _SIS_H_
 #define _SIS_H_
 
-#include "osdef.h"
 #include <video/sisfb.h>
 
 #include "vgatypes.h"
diff --git a/drivers/video/sis/sis_main.c b/drivers/video/sis/sis_main.c
index b52f8e4ef1fd..7e3370f115b6 100644
--- a/drivers/video/sis/sis_main.c
+++ b/drivers/video/sis/sis_main.c
@@ -60,6 +60,11 @@
 #include "sis.h"
 #include "sis_main.h"
 
+#if !defined(CONFIG_FB_SIS_300) && !defined(CONFIG_FB_SIS_315)
+#warning Neither CONFIG_FB_SIS_300 nor CONFIG_FB_SIS_315 is set
+#warning sisfb will not work!
+#endif
+
 static void sisfb_handle_command(struct sis_video_info *ivideo,
                                  struct sisfb_cmd *sisfb_command);
 
@@ -4114,14 +4119,6 @@ sisfb_find_rom(struct pci_dev *pdev)
                         if(sisfb_check_rom(rom_base, ivideo)) {
 
                                 if((myrombase = vmalloc(65536))) {
-
-                                        /* Work around bug in pci/rom.c: Folks forgot to check
-                                         * whether the size retrieved from the BIOS image eventually
-                                         * is larger than the mapped size
-                                         */
-                                        if(pci_resource_len(pdev, PCI_ROM_RESOURCE) < romsize)
-                                                romsize = pci_resource_len(pdev, PCI_ROM_RESOURCE);
-
                                         memcpy_fromio(myrombase, rom_base,
                                                         (romsize > 65536) ? 65536 : romsize);
                                 }
@@ -4155,23 +4152,6 @@ sisfb_find_rom(struct pci_dev *pdev)
 
         }
 
-#else
-
-        pci_read_config_dword(pdev, PCI_ROM_ADDRESS, &temp);
-        pci_write_config_dword(pdev, PCI_ROM_ADDRESS,
-                        (ivideo->video_base & PCI_ROM_ADDRESS_MASK) | PCI_ROM_ADDRESS_ENABLE);
-
-        rom_base = ioremap(ivideo->video_base, 65536);
-        if(rom_base) {
-                if(sisfb_check_rom(rom_base, ivideo)) {
-                        if((myrombase = vmalloc(65536)))
-                                memcpy_fromio(myrombase, rom_base, 65536);
-                }
-                iounmap(rom_base);
-        }
-
-        pci_write_config_dword(pdev, PCI_ROM_ADDRESS, temp);
-
 #endif
 
         return myrombase;
@@ -4181,6 +4161,9 @@ static void __devinit
 sisfb_post_map_vram(struct sis_video_info *ivideo, unsigned int *mapsize,
                         unsigned int min)
 {
+        if (*mapsize < (min << 20))
+                return;
+
         ivideo->video_vbase = ioremap(ivideo->video_base, (*mapsize));
 
         if(!ivideo->video_vbase) {
@@ -4514,7 +4497,7 @@ sisfb_post_sis300(struct pci_dev *pdev)
         } else {
 #endif
                 /* Need to map max FB size for finding out about RAM size */
-                mapsize = 64 << 20;
+                mapsize = ivideo->video_size;
                 sisfb_post_map_vram(ivideo, &mapsize, 4);
 
                 if(ivideo->video_vbase) {
@@ -4680,7 +4663,7 @@ sisfb_post_xgi_ramsize(struct sis_video_info *ivideo)
         orSISIDXREG(SISSR, 0x20, (0x80 | 0x04));
 
         /* Need to map max FB size for finding out about RAM size */
-        mapsize = 256 << 20;
+        mapsize = ivideo->video_size;
         sisfb_post_map_vram(ivideo, &mapsize, 32);
 
         if(!ivideo->video_vbase) {
@@ -5936,6 +5919,7 @@ sisfb_probe(struct pci_dev *pdev, const struct pci_device_id *ent)
         }
 
         ivideo->video_base = pci_resource_start(pdev, 0);
+        ivideo->video_size = pci_resource_len(pdev, 0);
         ivideo->mmio_base  = pci_resource_start(pdev, 1);
         ivideo->mmio_size  = pci_resource_len(pdev, 1);
         ivideo->SiS_Pr.RelIO = pci_resource_start(pdev, 2) + 0x30;
diff --git a/drivers/video/sis/vgatypes.h b/drivers/video/sis/vgatypes.h
index 81a22eaabfde..12c0dfaf2518 100644
--- a/drivers/video/sis/vgatypes.h
+++ b/drivers/video/sis/vgatypes.h
@@ -55,21 +55,10 @@
 
 #define SISIOMEMTYPE
 
-#ifdef SIS_LINUX_KERNEL
 typedef unsigned long SISIOADDRESS;
 #include <linux/types.h>  /* Need __iomem */
 #undef SISIOMEMTYPE
 #define SISIOMEMTYPE __iomem
-#endif
-
-#ifdef SIS_XORG_XF86
-#if XF86_VERSION_CURRENT < XF86_VERSION_NUMERIC(4,2,0,0,0)
-typedef unsigned long IOADDRESS;
-typedef unsigned long SISIOADDRESS;
-#else
-typedef IOADDRESS SISIOADDRESS;
-#endif
-#endif
 
 typedef enum _SIS_CHIP_TYPE {
     SIS_VGALegacy = 0,
diff --git a/drivers/video/sis/vstruct.h b/drivers/video/sis/vstruct.h
index bef4aae388d0..ea94d214dcff 100644
--- a/drivers/video/sis/vstruct.h
+++ b/drivers/video/sis/vstruct.h
@@ -233,24 +233,15 @@ struct SiS_Private
 {
         unsigned char                   ChipType;
         unsigned char                   ChipRevision;
-#ifdef SIS_XORG_XF86
-        PCITAG                          PciTag;
-#endif
-#ifdef SIS_LINUX_KERNEL
         void                            *ivideo;
-#endif
         unsigned char                   *VirtualRomBase;
         bool                            UseROM;
-#ifdef SIS_LINUX_KERNEL
         unsigned char SISIOMEMTYPE      *VideoMemoryAddress;
         unsigned int                    VideoMemorySize;
-#endif
         SISIOADDRESS                    IOAddress;
         SISIOADDRESS                    IOAddress2;  /* For dual chip XGI volari */
 
-#ifdef SIS_LINUX_KERNEL
         SISIOADDRESS                    RelIO;
-#endif
         SISIOADDRESS                    SiS_P3c4;
         SISIOADDRESS                    SiS_P3d4;
         SISIOADDRESS                    SiS_P3c0;
@@ -280,9 +271,6 @@ struct SiS_Private
         unsigned short                  SiS_IF_DEF_FSTN;
         unsigned short                  SiS_SysFlags;
         unsigned char                   SiS_VGAINFO;
-#ifdef SIS_XORG_XF86
-        unsigned short                  SiS_CP1, SiS_CP2, SiS_CP3, SiS_CP4;
-#endif
         bool                            SiS_UseROM;
         bool                            SiS_ROMNew;
         bool                            SiS_XGIROM;
diff --git a/drivers/virtio/virtio.c b/drivers/virtio/virtio.c
index 3a43ebf83a49..efb35aa8309a 100644
--- a/drivers/virtio/virtio.c
+++ b/drivers/virtio/virtio.c
@@ -9,19 +9,19 @@ static ssize_t device_show(struct device *_d,
                            struct device_attribute *attr, char *buf)
 {
         struct virtio_device *dev = container_of(_d,struct virtio_device,dev);
-        return sprintf(buf, "%hu", dev->id.device);
+        return sprintf(buf, "0x%04x\n", dev->id.device);
 }
 static ssize_t vendor_show(struct device *_d,
                            struct device_attribute *attr, char *buf)
 {
         struct virtio_device *dev = container_of(_d,struct virtio_device,dev);
-        return sprintf(buf, "%hu", dev->id.vendor);
+        return sprintf(buf, "0x%04x\n", dev->id.vendor);
 }
 static ssize_t status_show(struct device *_d,
                            struct device_attribute *attr, char *buf)
 {
         struct virtio_device *dev = container_of(_d,struct virtio_device,dev);
-        return sprintf(buf, "0x%08x", dev->config->get_status(dev));
+        return sprintf(buf, "0x%08x\n", dev->config->get_status(dev));
 }
 static ssize_t modalias_show(struct device *_d,
                              struct device_attribute *attr, char *buf)
diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c
index 1475ed6b575f..cc2f73e03475 100644
--- a/drivers/virtio/virtio_ring.c
+++ b/drivers/virtio/virtio_ring.c
@@ -230,9 +230,6 @@ add_head:
         pr_debug("Added buffer head %i to %p\n", head, vq);
         END_USE(vq);
 
-        /* If we're indirect, we can fit many (assuming not OOM). */
-        if (vq->indirect)
-                return vq->num_free ? vq->vring.num : 0;
         return vq->num_free;
 }
 EXPORT_SYMBOL_GPL(virtqueue_add_buf_gfp);
diff --git a/drivers/xen/Makefile b/drivers/xen/Makefile
index eb8a78d77d9d..533a199e7a3f 100644
--- a/drivers/xen/Makefile
+++ b/drivers/xen/Makefile
@@ -8,9 +8,12 @@ obj-$(CONFIG_BLOCK)		+= biomerge.o
 obj-$(CONFIG_HOTPLUG_CPU)       += cpu_hotplug.o
 obj-$(CONFIG_XEN_XENCOMM)       += xencomm.o
 obj-$(CONFIG_XEN_BALLOON)       += balloon.o
-obj-$(CONFIG_XEN_DEV_EVTCHN)    += evtchn.o
+obj-$(CONFIG_XEN_DEV_EVTCHN)    += xen-evtchn.o
 obj-$(CONFIG_XENFS)             += xenfs/
 obj-$(CONFIG_XEN_SYS_HYPERVISOR)        += sys-hypervisor.o
 obj-$(CONFIG_XEN_PLATFORM_PCI)  += platform-pci.o
 obj-$(CONFIG_SWIOTLB_XEN)       += swiotlb-xen.o
 obj-$(CONFIG_XEN_DOM0)          += pci.o
+
+xen-evtchn-y                    := evtchn.o
+
diff --git a/drivers/xen/balloon.c b/drivers/xen/balloon.c
index 500290b150bb..2b17ad5b4b32 100644
--- a/drivers/xen/balloon.c
+++ b/drivers/xen/balloon.c
@@ -50,6 +50,7 @@
 #include <asm/pgtable.h>
 #include <asm/uaccess.h>
 #include <asm/tlb.h>
+#include <asm/e820.h>
 
 #include <asm/xen/hypervisor.h>
 #include <asm/xen/hypercall.h>
@@ -119,7 +120,7 @@ static void scrub_page(struct page *page)
 }
 
 /* balloon_append: add the given page to the balloon. */
-static void balloon_append(struct page *page)
+static void __balloon_append(struct page *page)
 {
         /* Lowmem is re-populated first, so highmem pages go at list tail. */
         if (PageHighMem(page)) {
@@ -130,7 +131,11 @@ static void balloon_append(struct page *page)
                 list_add(&page->lru, &ballooned_pages);
                 balloon_stats.balloon_low++;
         }
+}
 
+static void balloon_append(struct page *page)
+{
+        __balloon_append(page);
         totalram_pages--;
 }
 
@@ -191,7 +196,7 @@ static unsigned long current_target(void)
 
 static int increase_reservation(unsigned long nr_pages)
 {
-        unsigned long  pfn, i, flags;
+        unsigned long  pfn, i;
         struct page   *page;
         long           rc;
         struct xen_memory_reservation reservation = {
@@ -203,8 +208,6 @@ static int increase_reservation(unsigned long nr_pages)
         if (nr_pages > ARRAY_SIZE(frame_list))
                 nr_pages = ARRAY_SIZE(frame_list);
 
-        spin_lock_irqsave(&xen_reservation_lock, flags);
-
         page = balloon_first_page();
         for (i = 0; i < nr_pages; i++) {
                 BUG_ON(page == NULL);
@@ -247,14 +250,12 @@ static int increase_reservation(unsigned long nr_pages)
         balloon_stats.current_pages += rc;
 
  out:
-        spin_unlock_irqrestore(&xen_reservation_lock, flags);
-
         return rc < 0 ? rc : rc != nr_pages;
 }
 
 static int decrease_reservation(unsigned long nr_pages)
 {
-        unsigned long  pfn, i, flags;
+        unsigned long  pfn, i;
         struct page   *page;
         int            need_sleep = 0;
         int ret;
@@ -292,8 +293,6 @@ static int decrease_reservation(unsigned long nr_pages)
         kmap_flush_unused();
         flush_tlb_all();
 
-        spin_lock_irqsave(&xen_reservation_lock, flags);
-
         /* No more mappings: invalidate P2M and add to balloon. */
         for (i = 0; i < nr_pages; i++) {
                 pfn = mfn_to_pfn(frame_list[i]);
@@ -308,8 +307,6 @@ static int decrease_reservation(unsigned long nr_pages)
 
         balloon_stats.current_pages -= nr_pages;
 
-        spin_unlock_irqrestore(&xen_reservation_lock, flags);
-
         return need_sleep;
 }
 
@@ -395,7 +392,7 @@ static struct notifier_block xenstore_notifier;
 
 static int __init balloon_init(void)
 {
-        unsigned long pfn;
+        unsigned long pfn, extra_pfn_end;
         struct page *page;
 
         if (!xen_pv_domain())
@@ -416,10 +413,15 @@ static int __init balloon_init(void)
         register_balloon(&balloon_sysdev);
 
         /* Initialise the balloon with excess memory space. */
-        for (pfn = xen_start_info->nr_pages; pfn < max_pfn; pfn++) {
+        extra_pfn_end = min(e820_end_of_ram_pfn(),
+                            (unsigned long)PFN_DOWN(xen_extra_mem_start + xen_extra_mem_size));
+        for (pfn = PFN_UP(xen_extra_mem_start);
+             pfn < extra_pfn_end;
+             pfn++) {
                 page = pfn_to_page(pfn);
-                if (!PageReserved(page))
-                        balloon_append(page);
+                /* totalram_pages doesn't include the boot-time
+                   balloon extension, so don't subtract from it. */
+                __balloon_append(page);
         }
 
         target_watch.callback = watch_target;
diff --git a/drivers/xen/events.c b/drivers/xen/events.c
index 321a0c8346e5..2811bb988ea0 100644
--- a/drivers/xen/events.c
+++ b/drivers/xen/events.c
@@ -278,17 +278,17 @@ static void bind_evtchn_to_cpu(unsigned int chn, unsigned int cpu)
         cpumask_copy(irq_to_desc(irq)->affinity, cpumask_of(cpu));
 #endif
 
-        __clear_bit(chn, cpu_evtchn_mask(cpu_from_irq(irq)));
-        __set_bit(chn, cpu_evtchn_mask(cpu));
+        clear_bit(chn, cpu_evtchn_mask(cpu_from_irq(irq)));
+        set_bit(chn, cpu_evtchn_mask(cpu));
 
         irq_info[irq].cpu = cpu;
 }
 
 static void init_evtchn_cpu_bindings(void)
 {
+        int i;
 #ifdef CONFIG_SMP
         struct irq_desc *desc;
-        int i;
 
         /* By default all event channels notify CPU#0. */
         for_each_irq_desc(i, desc) {
@@ -296,7 +296,10 @@ static void init_evtchn_cpu_bindings(void)
         }
 #endif
 
-        memset(cpu_evtchn_mask(0), ~0, sizeof(struct cpu_evtchn_s));
+        for_each_possible_cpu(i)
+                memset(cpu_evtchn_mask(i),
+                       (i == 0) ? ~0 : 0, sizeof(struct cpu_evtchn_s));
+
 }
 
 static inline void clear_evtchn(int port)
@@ -752,7 +755,7 @@ int xen_destroy_irq(int irq)
                 goto out;
 
         if (xen_initial_domain()) {
-                unmap_irq.pirq = info->u.pirq.gsi;
+                unmap_irq.pirq = info->u.pirq.pirq;
                 unmap_irq.domid = DOMID_SELF;
                 rc = HYPERVISOR_physdev_op(PHYSDEVOP_unmap_pirq, &unmap_irq);
                 if (rc) {
diff --git a/drivers/xen/evtchn.c b/drivers/xen/evtchn.c
index fec6ba3c08a8..ef11daf0cafe 100644
--- a/drivers/xen/evtchn.c
+++ b/drivers/xen/evtchn.c
@@ -69,20 +69,51 @@ struct per_user_data {
         const char *name;
 };
 
-/* Who's bound to each port? */
-static struct per_user_data *port_user[NR_EVENT_CHANNELS];
+/*
+ * Who's bound to each port?  This is logically an array of struct
+ * per_user_data *, but we encode the current enabled-state in bit 0.
+ */
+static unsigned long *port_user;
 static DEFINE_SPINLOCK(port_user_lock); /* protects port_user[] and ring_prod */
 
-irqreturn_t evtchn_interrupt(int irq, void *data)
+static inline struct per_user_data *get_port_user(unsigned port)
+{
+        return (struct per_user_data *)(port_user[port] & ~1);
+}
+
+static inline void set_port_user(unsigned port, struct per_user_data *u)
+{
+        port_user[port] = (unsigned long)u;
+}
+
+static inline bool get_port_enabled(unsigned port)
+{
+        return port_user[port] & 1;
+}
+
+static inline void set_port_enabled(unsigned port, bool enabled)
+{
+        if (enabled)
+                port_user[port] |= 1;
+        else
+                port_user[port] &= ~1;
+}
+
+static irqreturn_t evtchn_interrupt(int irq, void *data)
 {
         unsigned int port = (unsigned long)data;
         struct per_user_data *u;
 
         spin_lock(&port_user_lock);
 
-        u = port_user[port];
+        u = get_port_user(port);
+
+        WARN(!get_port_enabled(port),
+             "Interrupt for port %d, but apparently not enabled; per-user %p\n",
+             port, u);
 
         disable_irq_nosync(irq);
+        set_port_enabled(port, false);
 
         if ((u->ring_prod - u->ring_cons) < EVTCHN_RING_SIZE) {
                 u->ring[EVTCHN_RING_MASK(u->ring_prod)] = port;
@@ -92,9 +123,8 @@ irqreturn_t evtchn_interrupt(int irq, void *data)
                         kill_fasync(&u->evtchn_async_queue,
                                     SIGIO, POLL_IN);
                 }
-        } else {
+        } else
                 u->ring_overflow = 1;
-        }
 
         spin_unlock(&port_user_lock);
 
@@ -198,9 +228,18 @@ static ssize_t evtchn_write(struct file *file, const char __user *buf,
                 goto out;
 
         spin_lock_irq(&port_user_lock);
-        for (i = 0; i < (count/sizeof(evtchn_port_t)); i++)
-                if ((kbuf[i] < NR_EVENT_CHANNELS) && (port_user[kbuf[i]] == u))
-                        enable_irq(irq_from_evtchn(kbuf[i]));
+
+        for (i = 0; i < (count/sizeof(evtchn_port_t)); i++) {
+                unsigned port = kbuf[i];
+
+                if (port < NR_EVENT_CHANNELS &&
+                    get_port_user(port) == u &&
+                    !get_port_enabled(port)) {
+                        set_port_enabled(port, true);
+                        enable_irq(irq_from_evtchn(port));
+                }
+        }
+
         spin_unlock_irq(&port_user_lock);
 
         rc = count;
@@ -222,8 +261,9 @@ static int evtchn_bind_to_user(struct per_user_data *u, int port)
          * interrupt handler yet, and our caller has already
          * serialized bind operations.)
          */
-        BUG_ON(port_user[port] != NULL);
-        port_user[port] = u;
+        BUG_ON(get_port_user(port) != NULL);
+        set_port_user(port, u);
+        set_port_enabled(port, true); /* start enabled */
 
         rc = bind_evtchn_to_irqhandler(port, evtchn_interrupt, IRQF_DISABLED,
                                        u->name, (void *)(unsigned long)port);
@@ -239,10 +279,7 @@ static void evtchn_unbind_from_user(struct per_user_data *u, int port)
 
         unbind_from_irqhandler(irq, (void *)(unsigned long)port);
 
-        /* make sure we unbind the irq handler before clearing the port */
-        barrier();
-
-        port_user[port] = NULL;
+        set_port_user(port, NULL);
 }
 
 static long evtchn_ioctl(struct file *file,
@@ -333,15 +370,17 @@ static long evtchn_ioctl(struct file *file,
                 spin_lock_irq(&port_user_lock);
 
                 rc = -ENOTCONN;
-                if (port_user[unbind.port] != u) {
+                if (get_port_user(unbind.port) != u) {
                         spin_unlock_irq(&port_user_lock);
                         break;
                 }
 
-                evtchn_unbind_from_user(u, unbind.port);
+                disable_irq(irq_from_evtchn(unbind.port));
 
                 spin_unlock_irq(&port_user_lock);
 
+                evtchn_unbind_from_user(u, unbind.port);
+
                 rc = 0;
                 break;
         }
@@ -355,7 +394,7 @@ static long evtchn_ioctl(struct file *file,
 
                 if (notify.port >= NR_EVENT_CHANNELS) {
                         rc = -EINVAL;
-                } else if (port_user[notify.port] != u) {
+                } else if (get_port_user(notify.port) != u) {
                         rc = -ENOTCONN;
                 } else {
                         notify_remote_via_evtchn(notify.port);
@@ -431,7 +470,7 @@ static int evtchn_open(struct inode *inode, struct file *filp)
 
         filp->private_data = u;
 
-        return 0;
+        return nonseekable_open(inode, filp);;
 }
 
 static int evtchn_release(struct inode *inode, struct file *filp)
@@ -444,14 +483,21 @@ static int evtchn_release(struct inode *inode, struct file *filp)
         free_page((unsigned long)u->ring);
 
         for (i = 0; i < NR_EVENT_CHANNELS; i++) {
-                if (port_user[i] != u)
+                if (get_port_user(i) != u)
                         continue;
 
-                evtchn_unbind_from_user(port_user[i], i);
+                disable_irq(irq_from_evtchn(i));
         }
 
         spin_unlock_irq(&port_user_lock);
 
+        for (i = 0; i < NR_EVENT_CHANNELS; i++) {
+                if (get_port_user(i) != u)
+                        continue;
+
+                evtchn_unbind_from_user(get_port_user(i), i);
+        }
+
         kfree(u->name);
         kfree(u);
 
@@ -467,12 +513,12 @@ static const struct file_operations evtchn_fops = {
         .fasync  = evtchn_fasync,
         .open    = evtchn_open,
         .release = evtchn_release,
-        .llseek = noop_llseek,
+        .llseek  = no_llseek,
 };
 
 static struct miscdevice evtchn_miscdev = {
         .minor        = MISC_DYNAMIC_MINOR,
-        .name         = "evtchn",
+        .name         = "xen/evtchn",
         .fops         = &evtchn_fops,
 };
 static int __init evtchn_init(void)
@@ -482,8 +528,11 @@ static int __init evtchn_init(void)
         if (!xen_domain())
                 return -ENODEV;
 
+        port_user = kcalloc(NR_EVENT_CHANNELS, sizeof(*port_user), GFP_KERNEL);
+        if (port_user == NULL)
+                return -ENOMEM;
+
         spin_lock_init(&port_user_lock);
-        memset(port_user, 0, sizeof(port_user));
 
         /* Create '/dev/misc/evtchn'. */
         err = misc_register(&evtchn_miscdev);
@@ -499,6 +548,9 @@ static int __init evtchn_init(void)
 
 static void __exit evtchn_cleanup(void)
 {
+        kfree(port_user);
+        port_user = NULL;
+
         misc_deregister(&evtchn_miscdev);
 }
 
diff --git a/drivers/xen/xenfs/privcmd.c b/drivers/xen/xenfs/privcmd.c
index f80be7f6eb95..dbd3b16fd131 100644
--- a/drivers/xen/xenfs/privcmd.c
+++ b/drivers/xen/xenfs/privcmd.c
@@ -15,7 +15,6 @@
 #include <linux/mman.h>
 #include <linux/uaccess.h>
 #include <linux/swap.h>
-#include <linux/smp_lock.h>
 #include <linux/highmem.h>
 #include <linux/pagemap.h>
 #include <linux/seq_file.h>
@@ -266,9 +265,7 @@ static int mmap_return_errors(void *data, void *state)
         xen_pfn_t *mfnp = data;
         struct mmap_batch_state *st = state;
 
-        put_user(*mfnp, st->user++);
-
-        return 0;
+        return put_user(*mfnp, st->user++);
 }
 
 static struct vm_operations_struct privcmd_vm_ops;
@@ -323,10 +320,8 @@ static long privcmd_ioctl_mmap_batch(void __user *udata)
         up_write(&mm->mmap_sem);
 
         if (state.err > 0) {
-                ret = 0;
-
                 state.user = m.arr;
-                traverse_pages(m.num, sizeof(xen_pfn_t),
+                ret = traverse_pages(m.num, sizeof(xen_pfn_t),
                                &pagelist,
                                mmap_return_errors, &state);
         }
@@ -384,8 +379,9 @@ static int privcmd_mmap(struct file *file, struct vm_area_struct *vma)
         if (xen_feature(XENFEAT_auto_translated_physmap))
                 return -ENOSYS;
 
-        /* DONTCOPY is essential for Xen as copy_page_range is broken. */
-        vma->vm_flags |= VM_RESERVED | VM_IO | VM_DONTCOPY;
+        /* DONTCOPY is essential for Xen because copy_page_range doesn't know
+         * how to recreate these mappings */
+        vma->vm_flags |= VM_RESERVED | VM_IO | VM_DONTCOPY | VM_PFNMAP;
         vma->vm_ops = &privcmd_vm_ops;
         vma->vm_private_data = NULL;
 
diff --git a/drivers/xen/xenfs/super.c b/drivers/xen/xenfs/super.c
index f6339d11d59c..1aa389719846 100644
--- a/drivers/xen/xenfs/super.c
+++ b/drivers/xen/xenfs/super.c
@@ -12,8 +12,6 @@
 #include <linux/module.h>
 #include <linux/fs.h>
 #include <linux/magic.h>
-#include <linux/mm.h>
-#include <linux/backing-dev.h>
 
 #include <xen/xen.h>
 
@@ -24,28 +22,12 @@
 MODULE_DESCRIPTION("Xen filesystem");
 MODULE_LICENSE("GPL");
 
-static int xenfs_set_page_dirty(struct page *page)
-{
-        return !TestSetPageDirty(page);
-}
-
-static const struct address_space_operations xenfs_aops = {
-        .set_page_dirty = xenfs_set_page_dirty,
-};
-
-static struct backing_dev_info xenfs_backing_dev_info = {
-        .ra_pages       = 0,    /* No readahead */
-        .capabilities   = BDI_CAP_NO_ACCT_AND_WRITEBACK,
-};
-
 static struct inode *xenfs_make_inode(struct super_block *sb, int mode)
 {
         struct inode *ret = new_inode(sb);
 
         if (ret) {
                 ret->i_mode = mode;
-                ret->i_mapping->a_ops = &xenfs_aops;
-                ret->i_mapping->backing_dev_info = &xenfs_backing_dev_info;
                 ret->i_uid = ret->i_gid = 0;
                 ret->i_blocks = 0;
                 ret->i_atime = ret->i_mtime = ret->i_ctime = CURRENT_TIME;
@@ -121,9 +103,9 @@ static int xenfs_fill_super(struct super_block *sb, void *data, int silent)
         return rc;
 }
 
-static int xenfs_mount(struct file_system_type *fs_type,
-                        int flags, const char *dev_name,
-                        void *data)
+static struct dentry *xenfs_mount(struct file_system_type *fs_type,
+                                  int flags, const char *dev_name,
+                                  void *data)
 {
         return mount_single(fs_type, flags, data, xenfs_fill_super);
 }
@@ -137,25 +119,11 @@ static struct file_system_type xenfs_type = {
 
 static int __init xenfs_init(void)
 {
-        int err;
-        if (!xen_domain()) {
-                printk(KERN_INFO "xenfs: not registering filesystem on non-xen platform\n");
-                return 0;
-        }
-
-        err = register_filesystem(&xenfs_type);
-        if (err) {
-                printk(KERN_ERR "xenfs: Unable to register filesystem!\n");
-                goto out;
-        }
-
-        err = bdi_init(&xenfs_backing_dev_info);
-        if (err)
-                unregister_filesystem(&xenfs_type);
-
- out:
+        if (xen_domain())
+                return register_filesystem(&xenfs_type);
 
-        return err;
+        printk(KERN_INFO "XENFS: not registering filesystem on non-xen platform\n");
+        return 0;
 }
 
 static void __exit xenfs_exit(void)
diff --git a/drivers/zorro/proc.c b/drivers/zorro/proc.c
index cafc50454292..e0c84725d3e9 100644
--- a/drivers/zorro/proc.c
+++ b/drivers/zorro/proc.c
@@ -13,7 +13,6 @@
 #include <linux/proc_fs.h>
 #include <linux/seq_file.h>
 #include <linux/init.h>
-#include <linux/smp_lock.h>
 #include <asm/uaccess.h>
 #include <asm/amigahw.h>
 #include <asm/setup.h>
diff --git a/fs/block_dev.c b/fs/block_dev.c
index 06e8ff12b97c..4230252fd689 100644
--- a/fs/block_dev.c
+++ b/fs/block_dev.c
@@ -11,7 +11,6 @@
 #include <linux/slab.h>
 #include <linux/kmod.h>
 #include <linux/major.h>
-#include <linux/smp_lock.h>
 #include <linux/device_cgroup.h>
 #include <linux/highmem.h>
 #include <linux/blkdev.h>
diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c
index e9c874abc9e1..561438b6a50c 100644
--- a/fs/ceph/addr.c
+++ b/fs/ceph/addr.c
@@ -204,7 +204,7 @@ static int readpage_nounlock(struct file *filp, struct page *page)
         err = ceph_osdc_readpages(osdc, ceph_vino(inode), &ci->i_layout,
                                   page->index << PAGE_CACHE_SHIFT, &len,
                                   ci->i_truncate_seq, ci->i_truncate_size,
-                                  &page, 1);
+                                  &page, 1, 0);
         if (err == -ENOENT)
                 err = 0;
         if (err < 0) {
@@ -287,7 +287,7 @@ static int ceph_readpages(struct file *file, struct address_space *mapping,
         rc = ceph_osdc_readpages(osdc, ceph_vino(inode), &ci->i_layout,
                                  offset, &len,
                                  ci->i_truncate_seq, ci->i_truncate_size,
-                                 pages, nr_pages);
+                                 pages, nr_pages, 0);
         if (rc == -ENOENT)
                 rc = 0;
         if (rc < 0)
@@ -774,7 +774,7 @@ get_more_pages:
                                             snapc, do_sync,
                                             ci->i_truncate_seq,
                                             ci->i_truncate_size,
-                                            &inode->i_mtime, true, 1);
+                                            &inode->i_mtime, true, 1, 0);
                                 max_pages = req->r_num_pages;
 
                                 alloc_page_vec(fsc, req);
diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c
index 98ab13e2b71d..60d27bc9eb83 100644
--- a/fs/ceph/caps.c
+++ b/fs/ceph/caps.c
@@ -1430,8 +1430,8 @@ static int try_nonblocking_invalidate(struct inode *inode)
             invalidating_gen == ci->i_rdcache_gen) {
                 /* success. */
                 dout("try_nonblocking_invalidate %p success\n", inode);
-                ci->i_rdcache_gen = 0;
-                ci->i_rdcache_revoking = 0;
+                /* save any racing async invalidate some trouble */
+                ci->i_rdcache_revoking = ci->i_rdcache_gen - 1;
                 return 0;
         }
         dout("try_nonblocking_invalidate %p failed\n", inode);
@@ -2273,8 +2273,7 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
 {
         struct ceph_inode_info *ci = ceph_inode(inode);
         int mds = session->s_mds;
-        unsigned seq = le32_to_cpu(grant->seq);
-        unsigned issue_seq = le32_to_cpu(grant->issue_seq);
+        int seq = le32_to_cpu(grant->seq);
         int newcaps = le32_to_cpu(grant->caps);
         int issued, implemented, used, wanted, dirty;
         u64 size = le64_to_cpu(grant->size);
@@ -2286,8 +2285,8 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
         int revoked_rdcache = 0;
         int queue_invalidate = 0;
 
-        dout("handle_cap_grant inode %p cap %p mds%d seq %u/%u %s\n",
-             inode, cap, mds, seq, issue_seq, ceph_cap_string(newcaps));
+        dout("handle_cap_grant inode %p cap %p mds%d seq %d %s\n",
+             inode, cap, mds, seq, ceph_cap_string(newcaps));
         dout(" size %llu max_size %llu, i_size %llu\n", size, max_size,
                 inode->i_size);
 
@@ -2383,7 +2382,6 @@ static void handle_cap_grant(struct inode *inode, struct ceph_mds_caps *grant,
         }
 
         cap->seq = seq;
-        cap->issue_seq = issue_seq;
 
         /* file layout may have changed */
         ci->i_layout = grant->layout;
@@ -2691,6 +2689,11 @@ static void handle_cap_import(struct ceph_mds_client *mdsc,
                      NULL /* no caps context */);
         try_flush_caps(inode, session, NULL);
         up_read(&mdsc->snap_rwsem);
+
+        /* make sure we re-request max_size, if necessary */
+        spin_lock(&inode->i_lock);
+        ci->i_requested_max_size = 0;
+        spin_unlock(&inode->i_lock);
 }
 
 /*
diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c
index e0a2dc6fcafc..7d447af84ec4 100644
--- a/fs/ceph/dir.c
+++ b/fs/ceph/dir.c
@@ -336,7 +336,10 @@ more:
                 if (req->r_reply_info.dir_end) {
                         kfree(fi->last_name);
                         fi->last_name = NULL;
-                        fi->next_offset = 2;
+                        if (ceph_frag_is_rightmost(frag))
+                                fi->next_offset = 2;
+                        else
+                                fi->next_offset = 0;
                 } else {
                         rinfo = &req->r_reply_info;
                         err = note_last_dentry(fi,
@@ -355,18 +358,22 @@ more:
                 u64 pos = ceph_make_fpos(frag, off);
                 struct ceph_mds_reply_inode *in =
                         rinfo->dir_in[off - fi->offset].in;
+                struct ceph_vino vino;
+                ino_t ino;
+
                 dout("readdir off %d (%d/%d) -> %lld '%.*s' %p\n",
                      off, off - fi->offset, rinfo->dir_nr, pos,
                      rinfo->dir_dname_len[off - fi->offset],
                      rinfo->dir_dname[off - fi->offset], in);
                 BUG_ON(!in);
                 ftype = le32_to_cpu(in->mode) >> 12;
+                vino.ino = le64_to_cpu(in->ino);
+                vino.snap = le64_to_cpu(in->snapid);
+                ino = ceph_vino_to_ino(vino);
                 if (filldir(dirent,
                             rinfo->dir_dname[off - fi->offset],
                             rinfo->dir_dname_len[off - fi->offset],
-                            pos,
-                            le64_to_cpu(in->ino),
-                            ftype) < 0) {
+                            pos, ino, ftype) < 0) {
                         dout("filldir stopping us...\n");
                         return 0;
                 }
@@ -414,6 +421,7 @@ static void reset_readdir(struct ceph_file_info *fi)
                 fi->last_readdir = NULL;
         }
         kfree(fi->last_name);
+        fi->last_name = NULL;
         fi->next_offset = 2;  /* compensate for . and .. */
         if (fi->dentry) {
                 dput(fi->dentry);
diff --git a/fs/ceph/file.c b/fs/ceph/file.c
index e77c28cf3690..8d79b8912e31 100644
--- a/fs/ceph/file.c
+++ b/fs/ceph/file.c
@@ -154,11 +154,13 @@ int ceph_open(struct inode *inode, struct file *file)
         }
 
         /*
-         * No need to block if we have any caps.  Update wanted set
+         * No need to block if we have caps on the auth MDS (for
+         * write) or any MDS (for read).  Update wanted set
          * asynchronously.
          */
         spin_lock(&inode->i_lock);
-        if (__ceph_is_any_real_caps(ci)) {
+        if (__ceph_is_any_real_caps(ci) &&
+            (((fmode & CEPH_FILE_MODE_WR) == 0) || ci->i_auth_cap)) {
                 int mds_wanted = __ceph_caps_mds_wanted(ci);
                 int issued = __ceph_caps_issued(ci, NULL);
 
@@ -280,11 +282,12 @@ int ceph_release(struct inode *inode, struct file *file)
 static int striped_read(struct inode *inode,
                         u64 off, u64 len,
                         struct page **pages, int num_pages,
-                        int *checkeof)
+                        int *checkeof, bool align_to_pages)
 {
         struct ceph_fs_client *fsc = ceph_inode_to_client(inode);
         struct ceph_inode_info *ci = ceph_inode(inode);
         u64 pos, this_len;
+        int io_align, page_align;
         int page_off = off & ~PAGE_CACHE_MASK; /* first byte's offset in page */
         int left, pages_left;
         int read;
@@ -300,14 +303,19 @@ static int striped_read(struct inode *inode,
         page_pos = pages;
         pages_left = num_pages;
         read = 0;
+        io_align = off & ~PAGE_MASK;
 
 more:
+        if (align_to_pages)
+                page_align = (pos - io_align) & ~PAGE_MASK;
+        else
+                page_align = pos & ~PAGE_MASK;
         this_len = left;
         ret = ceph_osdc_readpages(&fsc->client->osdc, ceph_vino(inode),
                                   &ci->i_layout, pos, &this_len,
                                   ci->i_truncate_seq,
                                   ci->i_truncate_size,
-                                  page_pos, pages_left);
+                                  page_pos, pages_left, page_align);
         hit_stripe = this_len < left;
         was_short = ret >= 0 && ret < this_len;
         if (ret == -ENOENT)
@@ -374,26 +382,25 @@ static ssize_t ceph_sync_read(struct file *file, char __user *data,
         dout("sync_read on file %p %llu~%u %s\n", file, off, len,
              (file->f_flags & O_DIRECT) ? "O_DIRECT" : "");
 
-        if (file->f_flags & O_DIRECT) {
-                pages = ceph_get_direct_page_vector(data, num_pages, off, len);
-
-                /*
-                 * flush any page cache pages in this range.  this
-                 * will make concurrent normal and O_DIRECT io slow,
-                 * but it will at least behave sensibly when they are
-                 * in sequence.
-                 */
-        } else {
+        if (file->f_flags & O_DIRECT)
+                pages = ceph_get_direct_page_vector(data, num_pages);
+        else
                 pages = ceph_alloc_page_vector(num_pages, GFP_NOFS);
-        }
         if (IS_ERR(pages))
                 return PTR_ERR(pages);
 
+        /*
+         * flush any page cache pages in this range.  this
+         * will make concurrent normal and sync io slow,
+         * but it will at least behave sensibly when they are
+         * in sequence.
+         */
         ret = filemap_write_and_wait(inode->i_mapping);
         if (ret < 0)
                 goto done;
 
-        ret = striped_read(inode, off, len, pages, num_pages, checkeof);
+        ret = striped_read(inode, off, len, pages, num_pages, checkeof,
+                           file->f_flags & O_DIRECT);
 
         if (ret >= 0 && (file->f_flags & O_DIRECT) == 0)
                 ret = ceph_copy_page_vector_to_user(pages, data, off, ret);
@@ -448,6 +455,7 @@ static ssize_t ceph_sync_write(struct file *file, const char __user *data,
         int flags;
         int do_sync = 0;
         int check_caps = 0;
+        int page_align, io_align;
         int ret;
         struct timespec mtime = CURRENT_TIME;
 
@@ -462,6 +470,8 @@ static ssize_t ceph_sync_write(struct file *file, const char __user *data,
         else
                 pos = *offset;
 
+        io_align = pos & ~PAGE_MASK;
+
         ret = filemap_write_and_wait_range(inode->i_mapping, pos, pos + left);
         if (ret < 0)
                 return ret;
@@ -486,20 +496,26 @@ static ssize_t ceph_sync_write(struct file *file, const char __user *data,
          */
 more:
         len = left;
+        if (file->f_flags & O_DIRECT)
+                /* write from beginning of first page, regardless of
+                   io alignment */
+                page_align = (pos - io_align) & ~PAGE_MASK;
+        else
+                page_align = pos & ~PAGE_MASK;
         req = ceph_osdc_new_request(&fsc->client->osdc, &ci->i_layout,
                                     ceph_vino(inode), pos, &len,
                                     CEPH_OSD_OP_WRITE, flags,
                                     ci->i_snap_realm->cached_context,
                                     do_sync,
                                     ci->i_truncate_seq, ci->i_truncate_size,
-                                    &mtime, false, 2);
+                                    &mtime, false, 2, page_align);
         if (!req)
                 return -ENOMEM;
 
         num_pages = calc_pages_for(pos, len);
 
         if (file->f_flags & O_DIRECT) {
-                pages = ceph_get_direct_page_vector(data, num_pages, pos, len);
+                pages = ceph_get_direct_page_vector(data, num_pages);
                 if (IS_ERR(pages)) {
                         ret = PTR_ERR(pages);
                         goto out;
diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c
index 1d6a45b5a04c..bf1286588f26 100644
--- a/fs/ceph/inode.c
+++ b/fs/ceph/inode.c
@@ -2,7 +2,6 @@
 
 #include <linux/module.h>
 #include <linux/fs.h>
-#include <linux/smp_lock.h>
 #include <linux/slab.h>
 #include <linux/string.h>
 #include <linux/uaccess.h>
@@ -471,7 +470,9 @@ void ceph_fill_file_time(struct inode *inode, int issued,
 
         if (issued & (CEPH_CAP_FILE_EXCL|
                       CEPH_CAP_FILE_WR|
-                      CEPH_CAP_FILE_BUFFER)) {
+                      CEPH_CAP_FILE_BUFFER|
+                      CEPH_CAP_AUTH_EXCL|
+                      CEPH_CAP_XATTR_EXCL)) {
                 if (timespec_compare(ctime, &inode->i_ctime) > 0) {
                         dout("ctime %ld.%09ld -> %ld.%09ld inc w/ cap\n",
                              inode->i_ctime.tv_sec, inode->i_ctime.tv_nsec,
@@ -511,7 +512,7 @@ void ceph_fill_file_time(struct inode *inode, int issued,
                         warn = 1;
                 }
         } else {
-                /* we have no write caps; whatever the MDS says is true */
+                /* we have no write|excl caps; whatever the MDS says is true */
                 if (ceph_seq_cmp(time_warp_seq, ci->i_time_warp_seq) >= 0) {
                         inode->i_ctime = *ctime;
                         inode->i_mtime = *mtime;
@@ -567,12 +568,17 @@ static int fill_inode(struct inode *inode,
 
         /*
          * provided version will be odd if inode value is projected,
-         * even if stable.  skip the update if we have a newer info
-         * (e.g., due to inode info racing form multiple MDSs), or if
-         * we are getting projected (unstable) inode info.
+         * even if stable.  skip the update if we have newer stable
+         * info (ours>=theirs, e.g. due to racing mds replies), unless
+         * we are getting projected (unstable) info (in which case the
+         * version is odd, and we want ours>theirs).
+         *   us   them
+         *   2    2     skip
+         *   3    2     skip
+         *   3    3     update
          */
         if (le64_to_cpu(info->version) > 0 &&
-            (ci->i_version & ~1) > le64_to_cpu(info->version))
+            (ci->i_version & ~1) >= le64_to_cpu(info->version))
                 goto no_change;
 
         issued = __ceph_caps_issued(ci, &implemented);
@@ -606,7 +612,14 @@ static int fill_inode(struct inode *inode,
                             le32_to_cpu(info->time_warp_seq),
                             &ctime, &mtime, &atime);
 
-        ci->i_max_size = le64_to_cpu(info->max_size);
+        /* only update max_size on auth cap */
+        if ((info->cap.flags & CEPH_CAP_FLAG_AUTH) &&
+            ci->i_max_size != le64_to_cpu(info->max_size)) {
+                dout("max_size %lld -> %llu\n", ci->i_max_size,
+                     le64_to_cpu(info->max_size));
+                ci->i_max_size = le64_to_cpu(info->max_size);
+        }
+
         ci->i_layout = info->layout;
         inode->i_blkbits = fls(le32_to_cpu(info->layout.fl_stripe_unit)) - 1;
 
@@ -1055,7 +1068,8 @@ int ceph_fill_trace(struct super_block *sb, struct ceph_mds_request *req,
                 ininfo = rinfo->targeti.in;
                 vino.ino = le64_to_cpu(ininfo->ino);
                 vino.snap = le64_to_cpu(ininfo->snapid);
-                if (!dn->d_inode) {
+                in = dn->d_inode;
+                if (!in) {
                         in = ceph_get_inode(sb, vino);
                         if (IS_ERR(in)) {
                                 pr_err("fill_trace bad get_inode "
@@ -1386,11 +1400,8 @@ static void ceph_invalidate_work(struct work_struct *work)
         spin_lock(&inode->i_lock);
         dout("invalidate_pages %p gen %d revoking %d\n", inode,
              ci->i_rdcache_gen, ci->i_rdcache_revoking);
-        if (ci->i_rdcache_gen == 0 ||
-            ci->i_rdcache_revoking != ci->i_rdcache_gen) {
-                BUG_ON(ci->i_rdcache_revoking > ci->i_rdcache_gen);
+        if (ci->i_rdcache_revoking != ci->i_rdcache_gen) {
                 /* nevermind! */
-                ci->i_rdcache_revoking = 0;
                 spin_unlock(&inode->i_lock);
                 goto out;
         }
@@ -1400,15 +1411,16 @@ static void ceph_invalidate_work(struct work_struct *work)
         ceph_invalidate_nondirty_pages(inode->i_mapping);
 
         spin_lock(&inode->i_lock);
-        if (orig_gen == ci->i_rdcache_gen) {
+        if (orig_gen == ci->i_rdcache_gen &&
+            orig_gen == ci->i_rdcache_revoking) {
                 dout("invalidate_pages %p gen %d successful\n", inode,
                      ci->i_rdcache_gen);
-                ci->i_rdcache_gen = 0;
-                ci->i_rdcache_revoking = 0;
+                ci->i_rdcache_revoking--;
                 check = 1;
         } else {
-                dout("invalidate_pages %p gen %d raced, gen now %d\n",
-                     inode, orig_gen, ci->i_rdcache_gen);
+                dout("invalidate_pages %p gen %d raced, now %d revoking %d\n",
+                     inode, orig_gen, ci->i_rdcache_gen,
+                     ci->i_rdcache_revoking);
         }
         spin_unlock(&inode->i_lock);
 
@@ -1739,7 +1751,7 @@ int ceph_do_getattr(struct inode *inode, int mask)
                 return 0;
         }
 
-        dout("do_getattr inode %p mask %s\n", inode, ceph_cap_string(mask));
+        dout("do_getattr inode %p mask %s mode 0%o\n", inode, ceph_cap_string(mask), inode->i_mode);
         if (ceph_caps_issued_mask(ceph_inode(inode), mask, 1))
                 return 0;
 
diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index 3142b15940c2..098b18508479 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -6,7 +6,6 @@
 #include <linux/sched.h>
 #include <linux/debugfs.h>
 #include <linux/seq_file.h>
-#include <linux/smp_lock.h>
 
 #include "super.h"
 #include "mds_client.h"
@@ -529,6 +528,9 @@ static void __register_request(struct ceph_mds_client *mdsc,
         ceph_mdsc_get_request(req);
         __insert_request(mdsc, req);
 
+        req->r_uid = current_fsuid();
+        req->r_gid = current_fsgid();
+
         if (dir) {
                 struct ceph_inode_info *ci = ceph_inode(dir);
 
@@ -1588,8 +1590,8 @@ static struct ceph_msg *create_request_message(struct ceph_mds_client *mdsc,
 
         head->mdsmap_epoch = cpu_to_le32(mdsc->mdsmap->m_epoch);
         head->op = cpu_to_le32(req->r_op);
-        head->caller_uid = cpu_to_le32(current_fsuid());
-        head->caller_gid = cpu_to_le32(current_fsgid());
+        head->caller_uid = cpu_to_le32(req->r_uid);
+        head->caller_gid = cpu_to_le32(req->r_gid);
         head->args = req->r_args;
 
         ceph_encode_filepath(&p, end, ino1, path1);
diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h
index d66d63c72355..9341fd4f1432 100644
--- a/fs/ceph/mds_client.h
+++ b/fs/ceph/mds_client.h
@@ -170,6 +170,8 @@ struct ceph_mds_request {
 
         union ceph_mds_request_args r_args;
         int r_fmode;        /* file mode, if expecting cap */
+        uid_t r_uid;
+        gid_t r_gid;
 
         /* for choosing which mds to send this request to */
         int r_direct_mode;
diff --git a/fs/ceph/super.h b/fs/ceph/super.h
index 1886294e12f7..7f01728a4657 100644
--- a/fs/ceph/super.h
+++ b/fs/ceph/super.h
@@ -293,9 +293,7 @@ struct ceph_inode_info {
         int i_rd_ref, i_rdcache_ref, i_wr_ref;
         int i_wrbuffer_ref, i_wrbuffer_ref_head;
         u32 i_shared_gen;       /* increment each time we get FILE_SHARED */
-        u32 i_rdcache_gen;      /* we increment this each time we get
-                                   FILE_CACHE.  If it's non-zero, we
-                                   _may_ have cached pages. */
+        u32 i_rdcache_gen;      /* incremented each time we get FILE_CACHE. */
         u32 i_rdcache_revoking; /* RDCACHE gen to async invalidate, if any */
 
         struct list_head i_unsafe_writes; /* uncommitted sync writes */
diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c
index 410ed188faa1..a60579b007b0 100644
--- a/fs/compat_ioctl.c
+++ b/fs/compat_ioctl.c
@@ -19,7 +19,6 @@
 #include <linux/compiler.h>
 #include <linux/sched.h>
 #include <linux/smp.h>
-#include <linux/smp_lock.h>
 #include <linux/ioctl.h>
 #include <linux/if.h>
 #include <linux/if_bridge.h>
diff --git a/fs/ecryptfs/super.c b/fs/ecryptfs/super.c
index 253732382d37..2720178b7718 100644
--- a/fs/ecryptfs/super.c
+++ b/fs/ecryptfs/super.c
@@ -28,7 +28,6 @@
 #include <linux/key.h>
 #include <linux/slab.h>
 #include <linux/seq_file.h>
-#include <linux/smp_lock.h>
 #include <linux/file.h>
 #include <linux/crypto.h>
 #include "ecryptfs_kernel.h"
diff --git a/fs/ext3/super.c b/fs/ext3/super.c
index 2fedaf8b5012..acf8695fa8f0 100644
--- a/fs/ext3/super.c
+++ b/fs/ext3/super.c
@@ -27,7 +27,6 @@
 #include <linux/init.h>
 #include <linux/blkdev.h>
 #include <linux/parser.h>
-#include <linux/smp_lock.h>
 #include <linux/buffer_head.h>
 #include <linux/exportfs.h>
 #include <linux/vfs.h>
diff --git a/fs/ext4/ioctl.c b/fs/ext4/ioctl.c
index bf5ae883b1bd..eb3bc2fe647e 100644
--- a/fs/ext4/ioctl.c
+++ b/fs/ext4/ioctl.c
@@ -331,6 +331,30 @@ mext_out:
                 return err;
         }
 
+        case FITRIM:
+        {
+                struct super_block *sb = inode->i_sb;
+                struct fstrim_range range;
+                int ret = 0;
+
+                if (!capable(CAP_SYS_ADMIN))
+                        return -EPERM;
+
+                if (copy_from_user(&range, (struct fstrim_range *)arg,
+                    sizeof(range)))
+                        return -EFAULT;
+
+                ret = ext4_trim_fs(sb, &range);
+                if (ret < 0)
+                        return ret;
+
+                if (copy_to_user((struct fstrim_range *)arg, &range,
+                    sizeof(range)))
+                        return -EFAULT;
+
+                return 0;
+        }
+
         default:
                 return -ENOTTY;
         }
diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c
index 7f5451cd1d38..beacce11ac50 100644
--- a/fs/ext4/page-io.c
+++ b/fs/ext4/page-io.c
@@ -237,8 +237,6 @@ static void ext4_end_bio(struct bio *bio, int error)
                         } while (bh != head);
                 }
 
-                put_io_page(io_end->pages[i]);
-
                 /*
                  * If this is a partial write which happened to make
                  * all buffers uptodate then we can optimize away a
@@ -248,6 +246,8 @@ static void ext4_end_bio(struct bio *bio, int error)
                  */
                 if (!partial_write)
                         SetPageUptodate(page);
+
+                put_io_page(io_end->pages[i]);
         }
         io_end->num_io_pages = 0;
         inode = io_end->inode;
diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 61182fe6254e..e32195d6aac3 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1197,7 +1197,6 @@ static const struct super_operations ext4_sops = {
         .quota_write    = ext4_quota_write,
 #endif
         .bdev_try_to_free_page = bdev_try_to_free_page,
-        .trim_fs        = ext4_trim_fs
 };
 
 static const struct super_operations ext4_nojournal_sops = {
@@ -2799,9 +2798,6 @@ static void ext4_clear_request_list(void)
         struct ext4_li_request *elr;
 
         mutex_lock(&ext4_li_info->li_list_mtx);
-        if (list_empty(&ext4_li_info->li_request_list))
-                return;
-
         list_for_each_safe(pos, n, &ext4_li_info->li_request_list) {
                 elr = list_entry(pos, struct ext4_li_request,
                                  lr_request);
@@ -3268,13 +3264,14 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
          * Test whether we have more sectors than will fit in sector_t,
          * and whether the max offset is addressable by the page cache.
          */
-        ret = generic_check_addressable(sb->s_blocksize_bits,
+        err = generic_check_addressable(sb->s_blocksize_bits,
                                         ext4_blocks_count(es));
-        if (ret) {
+        if (err) {
                 ext4_msg(sb, KERN_ERR, "filesystem"
                          " too large to mount safely on this system");
                 if (sizeof(sector_t) < 8)
                         ext4_msg(sb, KERN_WARNING, "CONFIG_LBDAF not enabled");
+                ret = err;
                 goto failed_mount;
         }
 
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index c8224587123f..9242d294fe90 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -134,6 +134,7 @@ EXPORT_SYMBOL_GPL(fuse_do_open);
 void fuse_finish_open(struct inode *inode, struct file *file)
 {
         struct fuse_file *ff = file->private_data;
+        struct fuse_conn *fc = get_fuse_conn(inode);
 
         if (ff->open_flags & FOPEN_DIRECT_IO)
                 file->f_op = &fuse_direct_io_file_operations;
@@ -141,6 +142,15 @@ void fuse_finish_open(struct inode *inode, struct file *file)
                 invalidate_inode_pages2(inode->i_mapping);
         if (ff->open_flags & FOPEN_NONSEEKABLE)
                 nonseekable_open(inode, file);
+        if (fc->atomic_o_trunc && (file->f_flags & O_TRUNC)) {
+                struct fuse_inode *fi = get_fuse_inode(inode);
+
+                spin_lock(&fc->lock);
+                fi->attr_version = ++fc->attr_version;
+                i_size_write(inode, 0);
+                spin_unlock(&fc->lock);
+                fuse_invalidate_attr(inode);
+        }
 }
 
 int fuse_open_common(struct inode *inode, struct file *file, bool isdir)
diff --git a/fs/gfs2/export.c b/fs/gfs2/export.c
index 06d582732d34..5ab3839dfcb9 100644
--- a/fs/gfs2/export.c
+++ b/fs/gfs2/export.c
@@ -138,10 +138,8 @@ static struct dentry *gfs2_get_dentry(struct super_block *sb,
                                       struct gfs2_inum_host *inum)
 {
         struct gfs2_sbd *sdp = sb->s_fs_info;
-        struct gfs2_holder i_gh;
         struct inode *inode;
         struct dentry *dentry;
-        int error;
 
         inode = gfs2_ilookup(sb, inum->no_addr);
         if (inode) {
@@ -152,52 +150,16 @@ static struct dentry *gfs2_get_dentry(struct super_block *sb,
                 goto out_inode;
         }
 
-        error = gfs2_glock_nq_num(sdp, inum->no_addr, &gfs2_inode_glops,
-                                  LM_ST_SHARED, LM_FLAG_ANY, &i_gh);
-        if (error)
-                return ERR_PTR(error);
-
-        error = gfs2_check_blk_type(sdp, inum->no_addr, GFS2_BLKST_DINODE);
-        if (error)
-                goto fail;
-
-        inode = gfs2_inode_lookup(sb, DT_UNKNOWN, inum->no_addr, 0);
-        if (IS_ERR(inode)) {
-                error = PTR_ERR(inode);
-                goto fail;
-        }
-
-        error = gfs2_inode_refresh(GFS2_I(inode));
-        if (error) {
-                iput(inode);
-                goto fail;
-        }
-
-        /* Pick up the works we bypass in gfs2_inode_lookup */
-        if (inode->i_state & I_NEW) 
-                gfs2_set_iop(inode);
-
-        if (GFS2_I(inode)->i_no_formal_ino != inum->no_formal_ino) {
-                iput(inode);
-                goto fail;
-        }
-
-        error = -EIO;
-        if (GFS2_I(inode)->i_diskflags & GFS2_DIF_SYSTEM) {
-                iput(inode);
-                goto fail;
-        }
-
-        gfs2_glock_dq_uninit(&i_gh);
+        inode = gfs2_lookup_by_inum(sdp, inum->no_addr, &inum->no_formal_ino,
+                                    GFS2_BLKST_DINODE);
+        if (IS_ERR(inode))
+                return ERR_CAST(inode);
 
 out_inode:
         dentry = d_obtain_alias(inode);
         if (!IS_ERR(dentry))
                 dentry->d_op = &gfs2_dops;
         return dentry;
-fail:
-        gfs2_glock_dq_uninit(&i_gh);
-        return ERR_PTR(error);
 }
 
 static struct dentry *gfs2_fh_to_dentry(struct super_block *sb, struct fid *fid,
diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 87778857f099..f92c17704169 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -686,21 +686,20 @@ static void delete_work_func(struct work_struct *work)
 {
         struct gfs2_glock *gl = container_of(work, struct gfs2_glock, gl_delete);
         struct gfs2_sbd *sdp = gl->gl_sbd;
-        struct gfs2_inode *ip = NULL;
+        struct gfs2_inode *ip;
         struct inode *inode;
-        u64 no_addr = 0;
+        u64 no_addr = gl->gl_name.ln_number;
+
+        ip = gl->gl_object;
+        /* Note: Unsafe to dereference ip as we don't hold right refs/locks */
 
-        spin_lock(&gl->gl_spin);
-        ip = (struct gfs2_inode *)gl->gl_object;
         if (ip)
-                no_addr = ip->i_no_addr;
-        spin_unlock(&gl->gl_spin);
-        if (ip) {
                 inode = gfs2_ilookup(sdp->sd_vfs, no_addr);
-                if (inode) {
-                        d_prune_aliases(inode);
-                        iput(inode);
-                }
+        else
+                inode = gfs2_lookup_by_inum(sdp, no_addr, NULL, GFS2_BLKST_UNLINKED);
+        if (inode && !IS_ERR(inode)) {
+                d_prune_aliases(inode);
+                iput(inode);
         }
         gfs2_glock_put(gl);
 }
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index 06370f8bd8cf..e1213f7f9217 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -73,49 +73,6 @@ static struct inode *gfs2_iget(struct super_block *sb, u64 no_addr)
         return iget5_locked(sb, hash, iget_test, iget_set, &no_addr);
 }
 
-struct gfs2_skip_data {
-        u64     no_addr;
-        int     skipped;
-};
-
-static int iget_skip_test(struct inode *inode, void *opaque)
-{
-        struct gfs2_inode *ip = GFS2_I(inode);
-        struct gfs2_skip_data *data = opaque;
-
-        if (ip->i_no_addr == data->no_addr) {
-                if (inode->i_state & (I_FREEING|I_WILL_FREE)){
-                        data->skipped = 1;
-                        return 0;
-                }
-                return 1;
-        }
-        return 0;
-}
-
-static int iget_skip_set(struct inode *inode, void *opaque)
-{
-        struct gfs2_inode *ip = GFS2_I(inode);
-        struct gfs2_skip_data *data = opaque;
-
-        if (data->skipped)
-                return 1;
-        inode->i_ino = (unsigned long)(data->no_addr);
-        ip->i_no_addr = data->no_addr;
-        return 0;
-}
-
-static struct inode *gfs2_iget_skip(struct super_block *sb,
-                                    u64 no_addr)
-{
-        struct gfs2_skip_data data;
-        unsigned long hash = (unsigned long)no_addr;
-
-        data.no_addr = no_addr;
-        data.skipped = 0;
-        return iget5_locked(sb, hash, iget_skip_test, iget_skip_set, &data);
-}
-
 /**
  * GFS2 lookup code fills in vfs inode contents based on info obtained
  * from directory entry inside gfs2_inode_lookup(). This has caused issues
@@ -243,93 +200,54 @@ fail:
         return ERR_PTR(error);
 }
 
-/**
- * gfs2_process_unlinked_inode - Lookup an unlinked inode for reclamation
- *                               and try to reclaim it by doing iput.
- *
- * This function assumes no rgrp locks are currently held.
- *
- * @sb: The super block
- * no_addr: The inode number
- *
- */
-
-void gfs2_process_unlinked_inode(struct super_block *sb, u64 no_addr)
+struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr,
+                                  u64 *no_formal_ino, unsigned int blktype)
 {
-        struct gfs2_sbd *sdp;
-        struct gfs2_inode *ip;
-        struct gfs2_glock *io_gl = NULL;
-        int error;
-        struct gfs2_holder gh;
+        struct super_block *sb = sdp->sd_vfs;
+        struct gfs2_holder i_gh;
         struct inode *inode;
+        int error;
 
-        inode = gfs2_iget_skip(sb, no_addr);
-
-        if (!inode)
-                return;
-
-        /* If it's not a new inode, someone's using it, so leave it alone. */
-        if (!(inode->i_state & I_NEW)) {
-                iput(inode);
-                return;
-        }
-
-        ip = GFS2_I(inode);
-        sdp = GFS2_SB(inode);
-        ip->i_no_formal_ino = -1;
+        error = gfs2_glock_nq_num(sdp, no_addr, &gfs2_inode_glops,
+                                  LM_ST_SHARED, LM_FLAG_ANY, &i_gh);
+        if (error)
+                return ERR_PTR(error);
 
-        error = gfs2_glock_get(sdp, no_addr, &gfs2_inode_glops, CREATE, &ip->i_gl);
-        if (unlikely(error))
+        error = gfs2_check_blk_type(sdp, no_addr, blktype);
+        if (error)
                 goto fail;
-        ip->i_gl->gl_object = ip;
 
-        error = gfs2_glock_get(sdp, no_addr, &gfs2_iopen_glops, CREATE, &io_gl);
-        if (unlikely(error))
-                goto fail_put;
-
-        set_bit(GIF_INVALID, &ip->i_flags);
-        error = gfs2_glock_nq_init(io_gl, LM_ST_SHARED, LM_FLAG_TRY | GL_EXACT,
-                                   &ip->i_iopen_gh);
-        if (unlikely(error))
-                goto fail_iopen;
+        inode = gfs2_inode_lookup(sb, DT_UNKNOWN, no_addr, 0);
+        if (IS_ERR(inode))
+                goto fail;
 
-        ip->i_iopen_gh.gh_gl->gl_object = ip;
-        gfs2_glock_put(io_gl);
-        io_gl = NULL;
+        error = gfs2_inode_refresh(GFS2_I(inode));
+        if (error)
+                goto fail_iput;
 
-        inode->i_mode = DT2IF(DT_UNKNOWN);
+        /* Pick up the works we bypass in gfs2_inode_lookup */
+        if (inode->i_state & I_NEW) 
+                gfs2_set_iop(inode);
 
-        /*
-         * We must read the inode in order to work out its type in
-         * this case. Note that this doesn't happen often as we normally
-         * know the type beforehand. This code path only occurs during
-         * unlinked inode recovery (where it is safe to do this glock,
-         * which is not true in the general case).
-         */
-        error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, LM_FLAG_TRY,
-                                   &gh);
-        if (unlikely(error))
-                goto fail_glock;
+        /* Two extra checks for NFS only */
+        if (no_formal_ino) {
+                error = -ESTALE;
+                if (GFS2_I(inode)->i_no_formal_ino != *no_formal_ino)
+                        goto fail_iput;
 
-        /* Inode is now uptodate */
-        gfs2_glock_dq_uninit(&gh);
-        gfs2_set_iop(inode);
+                error = -EIO;
+                if (GFS2_I(inode)->i_diskflags & GFS2_DIF_SYSTEM)
+                        goto fail_iput;
 
-        /* The iput will cause it to be deleted. */
-        iput(inode);
-        return;
+                error = 0;
+        }
 
-fail_glock:
-        gfs2_glock_dq(&ip->i_iopen_gh);
-fail_iopen:
-        if (io_gl)
-                gfs2_glock_put(io_gl);
-fail_put:
-        ip->i_gl->gl_object = NULL;
-        gfs2_glock_put(ip->i_gl);
 fail:
-        iget_failed(inode);
-        return;
+        gfs2_glock_dq_uninit(&i_gh);
+        return error ? ERR_PTR(error) : inode;
+fail_iput:
+        iput(inode);
+        goto fail;
 }
 
 static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf)
diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h
index 6720d7d5fbc6..d8499fadcc53 100644
--- a/fs/gfs2/inode.h
+++ b/fs/gfs2/inode.h
@@ -99,7 +99,9 @@ err:
 extern void gfs2_set_iop(struct inode *inode);
 extern struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned type, 
                                        u64 no_addr, u64 no_formal_ino);
-extern void gfs2_process_unlinked_inode(struct super_block *sb, u64 no_addr);
+extern struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr,
+                                         u64 *no_formal_ino,
+                                         unsigned int blktype);
 extern struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr);
 
 extern int gfs2_inode_refresh(struct gfs2_inode *ip);
diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c
index bef3ab6cf5c1..33c8407b876f 100644
--- a/fs/gfs2/rgrp.c
+++ b/fs/gfs2/rgrp.c
@@ -963,17 +963,18 @@ static int try_rgrp_fit(struct gfs2_rgrpd *rgd, struct gfs2_alloc *al)
  *          The inode, if one has been found, in inode.
  */
 
-static u64 try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked,
-                           u64 skip)
+static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip)
 {
         u32 goal = 0, block;
         u64 no_addr;
         struct gfs2_sbd *sdp = rgd->rd_sbd;
         unsigned int n;
+        struct gfs2_glock *gl;
+        struct gfs2_inode *ip;
+        int error;
+        int found = 0;
 
-        for(;;) {
-                if (goal >= rgd->rd_data)
-                        break;
+        while (goal < rgd->rd_data) {
                 down_write(&sdp->sd_log_flush_lock);
                 n = 1;
                 block = rgblk_search(rgd, goal, GFS2_BLKST_UNLINKED,
@@ -990,11 +991,32 @@ static u64 try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked,
                 if (no_addr == skip)
                         continue;
                 *last_unlinked = no_addr;
-                return no_addr;
+
+                error = gfs2_glock_get(sdp, no_addr, &gfs2_inode_glops, CREATE, &gl);
+                if (error)
+                        continue;
+
+                /* If the inode is already in cache, we can ignore it here
+                 * because the existing inode disposal code will deal with
+                 * it when all refs have gone away. Accessing gl_object like
+                 * this is not safe in general. Here it is ok because we do
+                 * not dereference the pointer, and we only need an approx
+                 * answer to whether it is NULL or not.
+                 */
+                ip = gl->gl_object;
+
+                if (ip || queue_work(gfs2_delete_workqueue, &gl->gl_delete) == 0)
+                        gfs2_glock_put(gl);
+                else
+                        found++;
+
+                /* Limit reclaim to sensible number of tasks */
+                if (found > 2*NR_CPUS)
+                        return;
         }
 
         rgd->rd_flags &= ~GFS2_RDF_CHECK;
-        return 0;
+        return;
 }
 
 /**
@@ -1075,11 +1097,9 @@ static void forward_rgrp_set(struct gfs2_sbd *sdp, struct gfs2_rgrpd *rgd)
  * Try to acquire rgrp in way which avoids contending with others.
  *
  * Returns: errno
- *          unlinked: the block address of an unlinked block to be reclaimed
  */
 
-static int get_local_rgrp(struct gfs2_inode *ip, u64 *unlinked,
-                          u64 *last_unlinked)
+static int get_local_rgrp(struct gfs2_inode *ip, u64 *last_unlinked)
 {
         struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
         struct gfs2_rgrpd *rgd, *begin = NULL;
@@ -1089,7 +1109,6 @@ static int get_local_rgrp(struct gfs2_inode *ip, u64 *unlinked,
         int loops = 0;
         int error, rg_locked;
 
-        *unlinked = 0;
         rgd = gfs2_blk2rgrpd(sdp, ip->i_goal);
 
         while (rgd) {
@@ -1106,17 +1125,10 @@ static int get_local_rgrp(struct gfs2_inode *ip, u64 *unlinked,
                 case 0:
                         if (try_rgrp_fit(rgd, al))
                                 goto out;
-                        /* If the rg came in already locked, there's no
-                           way we can recover from a failed try_rgrp_unlink
-                           because that would require an iput which can only
-                           happen after the rgrp is unlocked. */
-                        if (!rg_locked && rgd->rd_flags & GFS2_RDF_CHECK)
-                                *unlinked = try_rgrp_unlink(rgd, last_unlinked,
-                                                           ip->i_no_addr);
+                        if (rgd->rd_flags & GFS2_RDF_CHECK)
+                                try_rgrp_unlink(rgd, last_unlinked, ip->i_no_addr);
                         if (!rg_locked)
                                 gfs2_glock_dq_uninit(&al->al_rgd_gh);
-                        if (*unlinked)
-                                return -EAGAIN;
                         /* fall through */
                 case GLR_TRYFAILED:
                         rgd = recent_rgrp_next(rgd);
@@ -1145,13 +1157,10 @@ static int get_local_rgrp(struct gfs2_inode *ip, u64 *unlinked,
                 case 0:
                         if (try_rgrp_fit(rgd, al))
                                 goto out;
-                        if (!rg_locked && rgd->rd_flags & GFS2_RDF_CHECK)
-                                *unlinked = try_rgrp_unlink(rgd, last_unlinked,
-                                                            ip->i_no_addr);
+                        if (rgd->rd_flags & GFS2_RDF_CHECK)
+                                try_rgrp_unlink(rgd, last_unlinked, ip->i_no_addr);
                         if (!rg_locked)
                                 gfs2_glock_dq_uninit(&al->al_rgd_gh);
-                        if (*unlinked)
-                                return -EAGAIN;
                         break;
 
                 case GLR_TRYFAILED:
@@ -1204,12 +1213,12 @@ int gfs2_inplace_reserve_i(struct gfs2_inode *ip, int hold_rindex,
         struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
         struct gfs2_alloc *al = ip->i_alloc;
         int error = 0;
-        u64 last_unlinked = NO_BLOCK, unlinked;
+        u64 last_unlinked = NO_BLOCK;
+        int tries = 0;
 
         if (gfs2_assert_warn(sdp, al->al_requested))
                 return -EINVAL;
 
-try_again:
         if (hold_rindex) {
                 /* We need to hold the rindex unless the inode we're using is
                    the rindex itself, in which case it's already held. */
@@ -1218,31 +1227,23 @@ try_again:
                 else if (!sdp->sd_rgrps) /* We may not have the rindex read
                                             in, so: */
                         error = gfs2_ri_update_special(ip);
+                if (error)
+                        return error;
         }
 
-        if (error)
-                return error;
+        do {
+                error = get_local_rgrp(ip, &last_unlinked);
+                /* If there is no space, flushing the log may release some */
+                if (error)
+                        gfs2_log_flush(sdp, NULL);
+        } while (error && tries++ < 3);
 
-        /* Find an rgrp suitable for allocation.  If it encounters any unlinked
-           dinodes along the way, error will equal -EAGAIN and unlinked will
-           contains it block address. We then need to look up that inode and
-           try to free it, and try the allocation again. */
-        error = get_local_rgrp(ip, &unlinked, &last_unlinked);
         if (error) {
                 if (hold_rindex && ip != GFS2_I(sdp->sd_rindex))
                         gfs2_glock_dq_uninit(&al->al_ri_gh);
-                if (error != -EAGAIN)
-                        return error;
-
-                gfs2_process_unlinked_inode(ip->i_inode.i_sb, unlinked);
-                /* regardless of whether or not gfs2_process_unlinked_inode
-                   was successful, we don't want to repeat it again. */
-                last_unlinked = unlinked;
-                gfs2_log_flush(sdp, NULL);
-                error = 0;
-
-                goto try_again;
+                return error;
         }
+
         /* no error, so we have the rgrp set in the inode's allocation. */
         al->al_file = file;
         al->al_line = line;
diff --git a/fs/ioctl.c b/fs/ioctl.c
index e92fdbb3bc3a..d6cc16476620 100644
--- a/fs/ioctl.c
+++ b/fs/ioctl.c
@@ -6,7 +6,6 @@
 
 #include <linux/syscalls.h>
 #include <linux/mm.h>
-#include <linux/smp_lock.h>
 #include <linux/capability.h>
 #include <linux/file.h>
 #include <linux/fs.h>
@@ -530,41 +529,6 @@ static int ioctl_fsthaw(struct file *filp)
         return thaw_super(sb);
 }
 
-static int ioctl_fstrim(struct file *filp, void __user *argp)
-{
-        struct super_block *sb = filp->f_path.dentry->d_inode->i_sb;
-        struct fstrim_range range;
-        int ret = 0;
-
-        if (!capable(CAP_SYS_ADMIN))
-                return -EPERM;
-
-        /* If filesystem doesn't support trim feature, return. */
-        if (sb->s_op->trim_fs == NULL)
-                return -EOPNOTSUPP;
-
-        /* If a blockdevice-backed filesystem isn't specified, return EINVAL. */
-        if (sb->s_bdev == NULL)
-                return -EINVAL;
-
-        if (argp == NULL) {
-                range.start = 0;
-                range.len = ULLONG_MAX;
-                range.minlen = 0;
-        } else if (copy_from_user(&range, argp, sizeof(range)))
-                return -EFAULT;
-
-        ret = sb->s_op->trim_fs(sb, &range);
-        if (ret < 0)
-                return ret;
-
-        if ((argp != NULL) &&
-            (copy_to_user(argp, &range, sizeof(range))))
-                return -EFAULT;
-
-        return 0;
-}
-
 /*
  * When you add any new common ioctls to the switches above and below
  * please update compat_sys_ioctl() too.
@@ -615,10 +579,6 @@ int do_vfs_ioctl(struct file *filp, unsigned int fd, unsigned int cmd,
                 error = ioctl_fsthaw(filp);
                 break;
 
-        case FITRIM:
-                error = ioctl_fstrim(filp, argp);
-                break;
-
         case FS_IOC_FIEMAP:
                 return ioctl_fiemap(filp, arg);
 
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index c590d155c095..f837ba953529 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -899,6 +899,14 @@ journal_t * jbd2_journal_init_dev(struct block_device *bdev,
 
         /* journal descriptor can store up to n blocks -bzzz */
         journal->j_blocksize = blocksize;
+        journal->j_dev = bdev;
+        journal->j_fs_dev = fs_dev;
+        journal->j_blk_offset = start;
+        journal->j_maxlen = len;
+        bdevname(journal->j_dev, journal->j_devname);
+        p = journal->j_devname;
+        while ((p = strchr(p, '/')))
+                *p = '!';
         jbd2_stats_proc_init(journal);
         n = journal->j_blocksize / sizeof(journal_block_tag_t);
         journal->j_wbufsize = n;
@@ -908,14 +916,6 @@ journal_t * jbd2_journal_init_dev(struct block_device *bdev,
                         __func__);
                 goto out_err;
         }
-        journal->j_dev = bdev;
-        journal->j_fs_dev = fs_dev;
-        journal->j_blk_offset = start;
-        journal->j_maxlen = len;
-        bdevname(journal->j_dev, journal->j_devname);
-        p = journal->j_devname;
-        while ((p = strchr(p, '/')))
-                *p = '!';
 
         bh = __getblk(journal->j_dev, start, journal->j_blocksize);
         if (!bh) {
diff --git a/fs/lockd/clntlock.c b/fs/lockd/clntlock.c
index d5bb86866e6c..25509eb28fd7 100644
--- a/fs/lockd/clntlock.c
+++ b/fs/lockd/clntlock.c
@@ -14,7 +14,6 @@
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/svc.h>
 #include <linux/lockd/lockd.h>
-#include <linux/smp_lock.h>
 #include <linux/kthread.h>
 
 #define NLMDBG_FACILITY         NLMDBG_CLIENT
diff --git a/fs/lockd/clntproc.c b/fs/lockd/clntproc.c
index 47ea1e1925b8..332c54cf75e0 100644
--- a/fs/lockd/clntproc.c
+++ b/fs/lockd/clntproc.c
@@ -7,7 +7,6 @@
  */
 
 #include <linux/module.h>
-#include <linux/smp_lock.h>
 #include <linux/slab.h>
 #include <linux/types.h>
 #include <linux/errno.h>
diff --git a/fs/lockd/host.c b/fs/lockd/host.c
index 25e21e4023b2..ed0c59fe23ce 100644
--- a/fs/lockd/host.c
+++ b/fs/lockd/host.c
@@ -124,7 +124,7 @@ static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni)
                         continue;
                 if (host->h_server != ni->server)
                         continue;
-                if (ni->server &&
+                if (ni->server && ni->src_len != 0 &&
                     !rpc_cmp_addr(nlm_srcaddr(host), ni->src_sap))
                         continue;
 
@@ -167,6 +167,7 @@ static struct nlm_host *nlm_lookup_host(struct nlm_lookup_host_info *ni)
         host->h_addrlen = ni->salen;
         rpc_set_port(nlm_addr(host), 0);
         memcpy(nlm_srcaddr(host), ni->src_sap, ni->src_len);
+        host->h_srcaddrlen = ni->src_len;
         host->h_version    = ni->version;
         host->h_proto      = ni->protocol;
         host->h_rpcclnt    = NULL;
@@ -238,9 +239,6 @@ struct nlm_host *nlmclnt_lookup_host(const struct sockaddr *sap,
                                      const char *hostname,
                                      int noresvport)
 {
-        const struct sockaddr source = {
-                .sa_family      = AF_UNSPEC,
-        };
         struct nlm_lookup_host_info ni = {
                 .server         = 0,
                 .sap            = sap,
@@ -249,8 +247,6 @@ struct nlm_host *nlmclnt_lookup_host(const struct sockaddr *sap,
                 .version        = version,
                 .hostname       = hostname,
                 .hostname_len   = strlen(hostname),
-                .src_sap        = &source,
-                .src_len        = sizeof(source),
                 .noresvport     = noresvport,
         };
 
@@ -357,7 +353,6 @@ nlm_bind_host(struct nlm_host *host)
                         .protocol       = host->h_proto,
                         .address        = nlm_addr(host),
                         .addrsize       = host->h_addrlen,
-                        .saddress       = nlm_srcaddr(host),
                         .timeout        = &timeparms,
                         .servername     = host->h_name,
                         .program        = &nlm_program,
@@ -376,6 +371,8 @@ nlm_bind_host(struct nlm_host *host)
                         args.flags |= RPC_CLNT_CREATE_HARDRTRY;
                 if (host->h_noresvport)
                         args.flags |= RPC_CLNT_CREATE_NONPRIVPORT;
+                if (host->h_srcaddrlen)
+                        args.saddress = nlm_srcaddr(host);
 
                 clnt = rpc_create(&args);
                 if (!IS_ERR(clnt))
diff --git a/fs/lockd/svc4proc.c b/fs/lockd/svc4proc.c
index a336e832475d..38d261192453 100644
--- a/fs/lockd/svc4proc.c
+++ b/fs/lockd/svc4proc.c
@@ -9,7 +9,6 @@
 
 #include <linux/types.h>
 #include <linux/time.h>
-#include <linux/smp_lock.h>
 #include <linux/lockd/lockd.h>
 #include <linux/lockd/share.h>
 
diff --git a/fs/lockd/svclock.c b/fs/lockd/svclock.c
index c462d346acbd..ef5659b211e9 100644
--- a/fs/lockd/svclock.c
+++ b/fs/lockd/svclock.c
@@ -25,7 +25,6 @@
 #include <linux/errno.h>
 #include <linux/kernel.h>
 #include <linux/sched.h>
-#include <linux/smp_lock.h>
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/svc.h>
 #include <linux/lockd/nlm.h>
diff --git a/fs/lockd/svcproc.c b/fs/lockd/svcproc.c
index c3069f38d602..0caea5310ac3 100644
--- a/fs/lockd/svcproc.c
+++ b/fs/lockd/svcproc.c
@@ -9,7 +9,6 @@
 
 #include <linux/types.h>
 #include <linux/time.h>
-#include <linux/smp_lock.h>
 #include <linux/lockd/lockd.h>
 #include <linux/lockd/share.h>
 
diff --git a/fs/locks.c b/fs/locks.c
index 0e62dd35d088..8729347bcd1a 100644
--- a/fs/locks.c
+++ b/fs/locks.c
@@ -122,7 +122,6 @@
 #include <linux/module.h>
 #include <linux/security.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/syscalls.h>
 #include <linux/time.h>
 #include <linux/rcupdate.h>
diff --git a/fs/namespace.c b/fs/namespace.c
index 8a415c9c5e55..3dbfc072ec70 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -13,7 +13,6 @@
 #include <linux/sched.h>
 #include <linux/spinlock.h>
 #include <linux/percpu.h>
-#include <linux/smp_lock.h>
 #include <linux/init.h>
 #include <linux/kernel.h>
 #include <linux/acct.h>
diff --git a/fs/ncpfs/dir.c b/fs/ncpfs/dir.c
index aac8832e919e..f22b12e7d337 100644
--- a/fs/ncpfs/dir.c
+++ b/fs/ncpfs/dir.c
@@ -19,7 +19,6 @@
 #include <linux/mm.h>
 #include <asm/uaccess.h>
 #include <asm/byteorder.h>
-#include <linux/smp_lock.h>
 
 #include <linux/ncp_fs.h>
 
diff --git a/fs/ncpfs/file.c b/fs/ncpfs/file.c
index 6c754f70c529..cb50aaf981df 100644
--- a/fs/ncpfs/file.c
+++ b/fs/ncpfs/file.c
@@ -17,7 +17,6 @@
 #include <linux/mm.h>
 #include <linux/vmalloc.h>
 #include <linux/sched.h>
-#include <linux/smp_lock.h>
 
 #include <linux/ncp_fs.h>
 #include "ncplib_kernel.h"
diff --git a/fs/ncpfs/inode.c b/fs/ncpfs/inode.c
index d290545aa0c4..8fb93b604e73 100644
--- a/fs/ncpfs/inode.c
+++ b/fs/ncpfs/inode.c
@@ -26,7 +26,6 @@
 #include <linux/slab.h>
 #include <linux/vmalloc.h>
 #include <linux/init.h>
-#include <linux/smp_lock.h>
 #include <linux/vfs.h>
 #include <linux/mount.h>
 #include <linux/seq_file.h>
diff --git a/fs/ncpfs/ioctl.c b/fs/ncpfs/ioctl.c
index c2a1f9a155c3..d40a547e3377 100644
--- a/fs/ncpfs/ioctl.c
+++ b/fs/ncpfs/ioctl.c
@@ -17,7 +17,6 @@
 #include <linux/mount.h>
 #include <linux/slab.h>
 #include <linux/highuid.h>
-#include <linux/smp_lock.h>
 #include <linux/vmalloc.h>
 #include <linux/sched.h>
 
diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
index aeec017fe814..93a8b3bd69e3 100644
--- a/fs/nfs/callback.c
+++ b/fs/nfs/callback.c
@@ -9,7 +9,6 @@
 #include <linux/completion.h>
 #include <linux/ip.h>
 #include <linux/module.h>
-#include <linux/smp_lock.h>
 #include <linux/sunrpc/svc.h>
 #include <linux/sunrpc/svcsock.h>
 #include <linux/nfs_fs.h>
diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index 232a7eead33a..1fd62fc49be3 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -11,7 +11,6 @@
 #include <linux/module.h>
 #include <linux/sched.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/spinlock.h>
 
 #include <linux/nfs4.h>
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
index 07ac3847e562..662df2a5fad5 100644
--- a/fs/nfs/dir.c
+++ b/fs/nfs/dir.c
@@ -34,6 +34,7 @@
 #include <linux/mount.h>
 #include <linux/sched.h>
 #include <linux/vmalloc.h>
+#include <linux/kmemleak.h>
 
 #include "delegation.h"
 #include "iostat.h"
@@ -194,9 +195,13 @@ typedef struct {
 static
 struct nfs_cache_array *nfs_readdir_get_array(struct page *page)
 {
+        void *ptr;
         if (page == NULL)
                 return ERR_PTR(-EIO);
-        return (struct nfs_cache_array *)kmap(page);
+        ptr = kmap(page);
+        if (ptr == NULL)
+                return ERR_PTR(-ENOMEM);
+        return ptr;
 }
 
 static
@@ -213,6 +218,9 @@ int nfs_readdir_clear_array(struct page *page, gfp_t mask)
 {
         struct nfs_cache_array *array = nfs_readdir_get_array(page);
         int i;
+
+        if (IS_ERR(array))
+                return PTR_ERR(array);
         for (i = 0; i < array->size; i++)
                 kfree(array->array[i].string.name);
         nfs_readdir_release_array(page);
@@ -231,6 +239,11 @@ int nfs_readdir_make_qstr(struct qstr *string, const char *name, unsigned int le
         string->name = kmemdup(name, len, GFP_KERNEL);
         if (string->name == NULL)
                 return -ENOMEM;
+        /*
+         * Avoid a kmemleak false positive. The pointer to the name is stored
+         * in a page cache page which kmemleak does not scan.
+         */
+        kmemleak_not_leak(string->name);
         string->hash = full_name_hash(name, len);
         return 0;
 }
@@ -244,7 +257,7 @@ int nfs_readdir_add_to_array(struct nfs_entry *entry, struct page *page)
 
         if (IS_ERR(array))
                 return PTR_ERR(array);
-        ret = -EIO;
+        ret = -ENOSPC;
         if (array->size >= MAX_READDIR_ARRAY)
                 goto out;
 
@@ -255,9 +268,9 @@ int nfs_readdir_add_to_array(struct nfs_entry *entry, struct page *page)
         if (ret)
                 goto out;
         array->last_cookie = entry->cookie;
+        array->size++;
         if (entry->eof == 1)
                 array->eof_index = array->size;
-        array->size++;
 out:
         nfs_readdir_release_array(page);
         return ret;
@@ -272,7 +285,7 @@ int nfs_readdir_search_for_pos(struct nfs_cache_array *array, nfs_readdir_descri
         if (diff < 0)
                 goto out_eof;
         if (diff >= array->size) {
-                if (array->eof_index > 0)
+                if (array->eof_index >= 0)
                         goto out_eof;
                 desc->current_index += array->size;
                 return -EAGAIN;
@@ -281,8 +294,6 @@ int nfs_readdir_search_for_pos(struct nfs_cache_array *array, nfs_readdir_descri
         index = (unsigned int)diff;
         *desc->dir_cookie = array->array[index].cookie;
         desc->cache_entry_index = index;
-        if (index == array->eof_index)
-                desc->eof = 1;
         return 0;
 out_eof:
         desc->eof = 1;
@@ -296,17 +307,17 @@ int nfs_readdir_search_for_cookie(struct nfs_cache_array *array, nfs_readdir_des
         int status = -EAGAIN;
 
         for (i = 0; i < array->size; i++) {
-                if (i == array->eof_index) {
-                        desc->eof = 1;
-                        status = -EBADCOOKIE;
-                }
                 if (array->array[i].cookie == *desc->dir_cookie) {
                         desc->cache_entry_index = i;
                         status = 0;
-                        break;
+                        goto out;
                 }
         }
-
+        if (i == array->eof_index) {
+                desc->eof = 1;
+                status = -EBADCOOKIE;
+        }
+out:
         return status;
 }
 
@@ -449,7 +460,7 @@ out:
 
 /* Perform conversion from xdr to cache array */
 static
-void nfs_readdir_page_filler(nfs_readdir_descriptor_t *desc, struct nfs_entry *entry,
+int nfs_readdir_page_filler(nfs_readdir_descriptor_t *desc, struct nfs_entry *entry,
                                 void *xdr_page, struct page *page, unsigned int buflen)
 {
         struct xdr_stream stream;
@@ -471,21 +482,29 @@ void nfs_readdir_page_filler(nfs_readdir_descriptor_t *desc, struct nfs_entry *e
 
         do {
                 status = xdr_decode(desc, entry, &stream);
-                if (status != 0)
+                if (status != 0) {
+                        if (status == -EAGAIN)
+                                status = 0;
                         break;
+                }
 
-                if (nfs_readdir_add_to_array(entry, page) == -1)
-                        break;
                 if (desc->plus == 1)
                         nfs_prime_dcache(desc->file->f_path.dentry, entry);
+
+                status = nfs_readdir_add_to_array(entry, page);
+                if (status != 0)
+                        break;
         } while (!entry->eof);
 
         if (status == -EBADCOOKIE && entry->eof) {
                 array = nfs_readdir_get_array(page);
-                array->eof_index = array->size - 1;
-                status = 0;
-                nfs_readdir_release_array(page);
+                if (!IS_ERR(array)) {
+                        array->eof_index = array->size;
+                        status = 0;
+                        nfs_readdir_release_array(page);
+                }
         }
+        return status;
 }
 
 static
@@ -537,7 +556,7 @@ int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page,
         struct nfs_entry entry;
         struct file     *file = desc->file;
         struct nfs_cache_array *array;
-        int status = 0;
+        int status = -ENOMEM;
         unsigned int array_size = ARRAY_SIZE(pages);
 
         entry.prev_cookie = 0;
@@ -549,6 +568,10 @@ int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page,
                 goto out;
 
         array = nfs_readdir_get_array(page);
+        if (IS_ERR(array)) {
+                status = PTR_ERR(array);
+                goto out;
+        }
         memset(array, 0, sizeof(struct nfs_cache_array));
         array->eof_index = -1;
 
@@ -556,12 +579,19 @@ int nfs_readdir_xdr_to_array(nfs_readdir_descriptor_t *desc, struct page *page,
         if (!pages_ptr)
                 goto out_release_array;
         do {
+                unsigned int pglen;
                 status = nfs_readdir_xdr_filler(pages, desc, &entry, file, inode);
 
                 if (status < 0)
                         break;
-                nfs_readdir_page_filler(desc, &entry, pages_ptr, page, array_size * PAGE_SIZE);
-        } while (array->eof_index < 0 && array->size < MAX_READDIR_ARRAY);
+                pglen = status;
+                status = nfs_readdir_page_filler(desc, &entry, pages_ptr, page, pglen);
+                if (status < 0) {
+                        if (status == -ENOSPC)
+                                status = 0;
+                        break;
+                }
+        } while (array->eof_index < 0);
 
         nfs_readdir_free_large_page(pages_ptr, pages, array_size);
 out_release_array:
@@ -582,8 +612,10 @@ static
 int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page* page)
 {
         struct inode    *inode = desc->file->f_path.dentry->d_inode;
+        int ret;
 
-        if (nfs_readdir_xdr_to_array(desc, page, inode) < 0)
+        ret = nfs_readdir_xdr_to_array(desc, page, inode);
+        if (ret < 0)
                 goto error;
         SetPageUptodate(page);
 
@@ -595,7 +627,7 @@ int nfs_readdir_filler(nfs_readdir_descriptor_t *desc, struct page* page)
         return 0;
  error:
         unlock_page(page);
-        return -EIO;
+        return ret;
 }
 
 static
@@ -608,12 +640,8 @@ void cache_page_release(nfs_readdir_descriptor_t *desc)
 static
 struct page *get_cache_page(nfs_readdir_descriptor_t *desc)
 {
-        struct page *page;
-        page = read_cache_page(desc->file->f_path.dentry->d_inode->i_mapping,
+        return read_cache_page(desc->file->f_path.dentry->d_inode->i_mapping,
                         desc->page_index, (filler_t *)nfs_readdir_filler, desc);
-        if (IS_ERR(page))
-                desc->eof = 1;
-        return page;
 }
 
 /*
@@ -639,8 +667,10 @@ int find_cache_page(nfs_readdir_descriptor_t *desc)
 static inline
 int readdir_search_pagecache(nfs_readdir_descriptor_t *desc)
 {
-        int res = -EAGAIN;
+        int res;
 
+        if (desc->page_index == 0)
+                desc->current_index = 0;
         while (1) {
                 res = find_cache_page(desc);
                 if (res != -EAGAIN)
@@ -670,6 +700,8 @@ int nfs_do_filldir(nfs_readdir_descriptor_t *desc, void *dirent,
         struct dentry *dentry = NULL;
 
         array = nfs_readdir_get_array(desc->page);
+        if (IS_ERR(array))
+                return PTR_ERR(array);
 
         for (i = desc->cache_entry_index; i < array->size; i++) {
                 d_type = DT_UNKNOWN;
@@ -685,11 +717,9 @@ int nfs_do_filldir(nfs_readdir_descriptor_t *desc, void *dirent,
                         *desc->dir_cookie = array->array[i+1].cookie;
                 else
                         *desc->dir_cookie = array->last_cookie;
-                if (i == array->eof_index) {
-                        desc->eof = 1;
-                        break;
-                }
         }
+        if (i == array->eof_index)
+                desc->eof = 1;
 
         nfs_readdir_release_array(desc->page);
         cache_page_release(desc);
@@ -1345,12 +1375,12 @@ static struct dentry *nfs_atomic_lookup(struct inode *dir, struct dentry *dentry
                                 res = NULL;
                                 goto out;
                         /* This turned out not to be a regular file */
-                        case -EISDIR:
                         case -ENOTDIR:
                                 goto no_open;
                         case -ELOOP:
                                 if (!(nd->intent.open.flags & O_NOFOLLOW))
                                         goto no_open;
+                        /* case -EISDIR: */
                         /* case -EINVAL: */
                         default:
                                 res = ERR_CAST(inode);
diff --git a/fs/nfs/nfs2xdr.c b/fs/nfs/nfs2xdr.c
index e6bf45710cc7..2563f765c9b4 100644
--- a/fs/nfs/nfs2xdr.c
+++ b/fs/nfs/nfs2xdr.c
@@ -423,7 +423,7 @@ nfs_xdr_readdirres(struct rpc_rqst *req, __be32 *p, void *dummy)
         struct page **page;
         size_t hdrlen;
         unsigned int pglen, recvd;
-        int status, nr = 0;
+        int status;
 
         if ((status = ntohl(*p++)))
                 return nfs_stat_to_errno(status);
@@ -443,7 +443,7 @@ nfs_xdr_readdirres(struct rpc_rqst *req, __be32 *p, void *dummy)
         if (pglen > recvd)
                 pglen = recvd;
         page = rcvbuf->pages;
-        return nr;
+        return pglen;
 }
 
 static void print_overflow_msg(const char *func, const struct xdr_stream *xdr)
diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c
index d9a5e832c257..748dc91a4a14 100644
--- a/fs/nfs/nfs3xdr.c
+++ b/fs/nfs/nfs3xdr.c
@@ -555,7 +555,7 @@ nfs3_xdr_readdirres(struct rpc_rqst *req, __be32 *p, struct nfs3_readdirres *res
         struct page **page;
         size_t hdrlen;
         u32 recvd, pglen;
-        int status, nr = 0;
+        int status;
 
         status = ntohl(*p++);
         /* Decode post_op_attrs */
@@ -586,7 +586,7 @@ nfs3_xdr_readdirres(struct rpc_rqst *req, __be32 *p, struct nfs3_readdirres *res
                 pglen = recvd;
         page = rcvbuf->pages;
 
-        return nr;
+        return pglen;
 }
 
 __be32 *
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 0f24cdf2cb13..6a653ffd8e4e 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -2852,8 +2852,10 @@ static int _nfs4_proc_readdir(struct dentry *dentry, struct rpc_cred *cred,
         nfs4_setup_readdir(cookie, NFS_COOKIEVERF(dir), dentry, &args);
         res.pgbase = args.pgbase;
         status = nfs4_call_sync(NFS_SERVER(dir), &msg, &args, &res, 0);
-        if (status == 0)
+        if (status >= 0) {
                 memcpy(NFS_COOKIEVERF(dir), res.verifier.data, NFS4_VERIFIER_SIZE);
+                status += args.pgbase;
+        }
 
         nfs_invalidate_atime(dir);
 
diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c
index f313c4cce7e4..b7a204ff6fe1 100644
--- a/fs/nfs/nfs4xdr.c
+++ b/fs/nfs/nfs4xdr.c
@@ -4518,7 +4518,7 @@ static int decode_readdir(struct xdr_stream *xdr, struct rpc_rqst *req, struct n
         xdr_read_pages(xdr, pglen);
 
 
-        return 0;
+        return pglen;
 }
 
 static int decode_readlink(struct xdr_stream *xdr, struct rpc_rqst *req)
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index 0a42e8f4adcb..3c045044fca2 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -39,7 +39,6 @@
 #include <linux/nfs_mount.h>
 #include <linux/nfs4_mount.h>
 #include <linux/lockd/bind.h>
-#include <linux/smp_lock.h>
 #include <linux/seq_file.h>
 #include <linux/mount.h>
 #include <linux/mnt_namespace.h>
@@ -67,6 +66,12 @@
 
 #define NFSDBG_FACILITY         NFSDBG_VFS
 
+#ifdef CONFIG_NFS_V3
+#define NFS_DEFAULT_VERSION 3
+#else
+#define NFS_DEFAULT_VERSION 2
+#endif
+
 enum {
         /* Mount options that take no arguments */
         Opt_soft, Opt_hard,
@@ -2277,7 +2282,7 @@ static int nfs_get_sb(struct file_system_type *fs_type,
         };
         int error = -ENOMEM;
 
-        data = nfs_alloc_parsed_mount_data(3);
+        data = nfs_alloc_parsed_mount_data(NFS_DEFAULT_VERSION);
         mntfh = nfs_alloc_fhandle();
         if (data == NULL || mntfh == NULL)
                 goto out_free_fh;
diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index ad2bfa68d534..116cab970e0f 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -2262,7 +2262,7 @@ nfs4_file_downgrade(struct nfs4_file *fp, unsigned int share_access)
  * Spawn a thread to perform a recall on the delegation represented
  * by the lease (file_lock)
  *
- * Called from break_lease() with lock_kernel() held.
+ * Called from break_lease() with lock_flocks() held.
  * Note: we assume break_lease will only call this *once* for any given
  * lease.
  */
@@ -2286,7 +2286,7 @@ void nfsd_break_deleg_cb(struct file_lock *fl)
         list_add_tail(&dp->dl_recall_lru, &del_recall_lru);
         spin_unlock(&recall_lock);
 
-        /* only place dl_time is set. protected by lock_kernel*/
+        /* only place dl_time is set. protected by lock_flocks*/
         dp->dl_time = get_seconds();
 
         /*
@@ -2303,7 +2303,7 @@ void nfsd_break_deleg_cb(struct file_lock *fl)
 /*
  * The file_lock is being reapd.
  *
- * Called by locks_free_lock() with lock_kernel() held.
+ * Called by locks_free_lock() with lock_flocks() held.
  */
 static
 void nfsd_release_deleg_cb(struct file_lock *fl)
@@ -2318,7 +2318,7 @@ void nfsd_release_deleg_cb(struct file_lock *fl)
 }
 
 /*
- * Called from setlease() with lock_kernel() held
+ * Called from setlease() with lock_flocks() held
  */
 static
 int nfsd_same_client_deleg_cb(struct file_lock *onlist, struct file_lock *try)
diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h
index d8408217e3bd..1efea3615589 100644
--- a/fs/ocfs2/ocfs2.h
+++ b/fs/ocfs2/ocfs2.h
@@ -159,7 +159,9 @@ struct ocfs2_lock_res {
         char                     l_name[OCFS2_LOCK_ID_MAX_LEN];
         unsigned int             l_ro_holders;
         unsigned int             l_ex_holders;
-        unsigned char            l_level;
+        char                     l_level;
+        char                     l_requested;
+        char                     l_blocking;
 
         /* Data packed - type enum ocfs2_lock_type */
         unsigned char            l_type;
@@ -169,8 +171,6 @@ struct ocfs2_lock_res {
         unsigned char            l_action;
         /* Data packed - enum type ocfs2_unlock_action */
         unsigned char            l_unlock_action;
-        unsigned char            l_requested;
-        unsigned char            l_blocking;
         unsigned int             l_pending_gen;
 
         spinlock_t               l_lock;
diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c
index f02c0ef31578..cfeab7ce3697 100644
--- a/fs/ocfs2/super.c
+++ b/fs/ocfs2/super.c
@@ -41,7 +41,6 @@
 #include <linux/mount.h>
 #include <linux/seq_file.h>
 #include <linux/quotaops.h>
-#include <linux/smp_lock.h>
 
 #define MLOG_MASK_PREFIX ML_SUPER
 #include <cluster/masklog.h>
diff --git a/fs/proc/inode.c b/fs/proc/inode.c
index 9c2b5f484879..3ddb6068177c 100644
--- a/fs/proc/inode.c
+++ b/fs/proc/inode.c
@@ -16,7 +16,6 @@
 #include <linux/limits.h>
 #include <linux/init.h>
 #include <linux/module.h>
-#include <linux/smp_lock.h>
 #include <linux/sysctl.h>
 #include <linux/slab.h>
 
diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
index da6b01d70f01..c126c83b9a45 100644
--- a/fs/proc/task_mmu.c
+++ b/fs/proc/task_mmu.c
@@ -706,6 +706,7 @@ static int pagemap_hugetlb_range(pte_t *pte, unsigned long hmask,
  * skip over unmapped regions.
  */
 #define PAGEMAP_WALK_SIZE       (PMD_SIZE)
+#define PAGEMAP_WALK_MASK       (PMD_MASK)
 static ssize_t pagemap_read(struct file *file, char __user *buf,
                             size_t count, loff_t *ppos)
 {
@@ -776,7 +777,7 @@ static ssize_t pagemap_read(struct file *file, char __user *buf,
                 unsigned long end;
 
                 pm.pos = 0;
-                end = start_vaddr + PAGEMAP_WALK_SIZE;
+                end = (start_vaddr + PAGEMAP_WALK_SIZE) & PAGEMAP_WALK_MASK;
                 /* overflow ? */
                 if (end < start_vaddr || end > end_vaddr)
                         end = end_vaddr;
diff --git a/fs/read_write.c b/fs/read_write.c
index 431a0ed610c8..5d431bacbea9 100644
--- a/fs/read_write.c
+++ b/fs/read_write.c
@@ -9,7 +9,6 @@
 #include <linux/fcntl.h>
 #include <linux/file.h>
 #include <linux/uio.h>
-#include <linux/smp_lock.h>
 #include <linux/fsnotify.h>
 #include <linux/security.h>
 #include <linux/module.h>
diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c
index 41656d40dc5c..0bae036831e2 100644
--- a/fs/reiserfs/inode.c
+++ b/fs/reiserfs/inode.c
@@ -8,7 +8,6 @@
 #include <linux/reiserfs_acl.h>
 #include <linux/reiserfs_xattr.h>
 #include <linux/exportfs.h>
-#include <linux/smp_lock.h>
 #include <linux/pagemap.h>
 #include <linux/highmem.h>
 #include <linux/slab.h>
diff --git a/fs/reiserfs/ioctl.c b/fs/reiserfs/ioctl.c
index adf22b485cea..79265fdc317a 100644
--- a/fs/reiserfs/ioctl.c
+++ b/fs/reiserfs/ioctl.c
@@ -9,7 +9,6 @@
 #include <linux/time.h>
 #include <asm/uaccess.h>
 #include <linux/pagemap.h>
-#include <linux/smp_lock.h>
 #include <linux/compat.h>
 
 /*
@@ -184,12 +183,11 @@ int reiserfs_unpack(struct inode *inode, struct file *filp)
                 return 0;
         }
 
-        /* we need to make sure nobody is changing the file size beneath
-         ** us
-         */
-        reiserfs_mutex_lock_safe(&inode->i_mutex, inode->i_sb);
         depth = reiserfs_write_lock_once(inode->i_sb);
 
+        /* we need to make sure nobody is changing the file size beneath us */
+        reiserfs_mutex_lock_safe(&inode->i_mutex, inode->i_sb);
+
         write_from = inode->i_size & (blocksize - 1);
         /* if we are on a block boundary, we are already unpacked.  */
         if (write_from == 0) {
diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c
index 076c8b194682..d31bce1a9f90 100644
--- a/fs/reiserfs/journal.c
+++ b/fs/reiserfs/journal.c
@@ -43,7 +43,6 @@
 #include <linux/fcntl.h>
 #include <linux/stat.h>
 #include <linux/string.h>
-#include <linux/smp_lock.h>
 #include <linux/buffer_head.h>
 #include <linux/workqueue.h>
 #include <linux/writeback.h>
diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c
index 3bf7a6457f4d..b243117b8752 100644
--- a/fs/reiserfs/super.c
+++ b/fs/reiserfs/super.c
@@ -28,7 +28,6 @@
 #include <linux/mount.h>
 #include <linux/namei.h>
 #include <linux/crc32.h>
-#include <linux/smp_lock.h>
 
 struct file_system_type reiserfs_fs_type;
 
diff --git a/include/drm/nouveau_drm.h b/include/drm/nouveau_drm.h
index 01a714119506..bc5590b1a1ac 100644
--- a/include/drm/nouveau_drm.h
+++ b/include/drm/nouveau_drm.h
@@ -80,6 +80,7 @@ struct drm_nouveau_gpuobj_free {
 #define NOUVEAU_GETPARAM_VM_VRAM_BASE    12
 #define NOUVEAU_GETPARAM_GRAPH_UNITS     13
 #define NOUVEAU_GETPARAM_PTIMER_TIME     14
+#define NOUVEAU_GETPARAM_HAS_BO_USAGE    15
 struct drm_nouveau_getparam {
         uint64_t param;
         uint64_t value;
@@ -95,6 +96,12 @@ struct drm_nouveau_setparam {
 #define NOUVEAU_GEM_DOMAIN_GART      (1 << 2)
 #define NOUVEAU_GEM_DOMAIN_MAPPABLE  (1 << 3)
 
+#define NOUVEAU_GEM_TILE_LAYOUT_MASK 0x0000ff00
+#define NOUVEAU_GEM_TILE_16BPP       0x00000001
+#define NOUVEAU_GEM_TILE_32BPP       0x00000002
+#define NOUVEAU_GEM_TILE_ZETA        0x00000004
+#define NOUVEAU_GEM_TILE_NONCONTIG   0x00000008
+
 struct drm_nouveau_gem_info {
         uint32_t handle;
         uint32_t domain;
diff --git a/include/linux/ceph/libceph.h b/include/linux/ceph/libceph.h
index f22b2e941686..9e76d35670d2 100644
--- a/include/linux/ceph/libceph.h
+++ b/include/linux/ceph/libceph.h
@@ -227,8 +227,7 @@ extern int ceph_open_session(struct ceph_client *client);
 extern void ceph_release_page_vector(struct page **pages, int num_pages);
 
 extern struct page **ceph_get_direct_page_vector(const char __user *data,
-                                            int num_pages,
-                                            loff_t off, size_t len);
+                                                 int num_pages);
 extern void ceph_put_page_vector(struct page **pages, int num_pages);
 extern void ceph_release_page_vector(struct page **pages, int num_pages);
 extern struct page **ceph_alloc_page_vector(int num_pages, gfp_t flags);
diff --git a/include/linux/ceph/messenger.h b/include/linux/ceph/messenger.h
index 5956d62c3057..a108b425fee2 100644
--- a/include/linux/ceph/messenger.h
+++ b/include/linux/ceph/messenger.h
@@ -82,6 +82,7 @@ struct ceph_msg {
         struct ceph_buffer *middle;
         struct page **pages;            /* data payload.  NOT OWNER. */
         unsigned nr_pages;              /* size of page array */
+        unsigned page_alignment;        /* io offset in first page */
         struct ceph_pagelist *pagelist; /* instead of pages */
         struct list_head list_head;
         struct kref kref;
diff --git a/include/linux/ceph/osd_client.h b/include/linux/ceph/osd_client.h
index 6c91fb032c39..a1af29648fb5 100644
--- a/include/linux/ceph/osd_client.h
+++ b/include/linux/ceph/osd_client.h
@@ -79,6 +79,7 @@ struct ceph_osd_request {
         struct ceph_file_layout r_file_layout;
         struct ceph_snap_context *r_snapc;    /* snap context for writes */
         unsigned          r_num_pages;        /* size of page array (follows) */
+        unsigned          r_page_alignment;   /* io offset in first page */
         struct page     **r_pages;            /* pages for data payload */
         int               r_pages_from_pool;
         int               r_own_pages;        /* if true, i own page list */
@@ -194,7 +195,8 @@ extern struct ceph_osd_request *ceph_osdc_new_request(struct ceph_osd_client *,
                                       int do_sync, u32 truncate_seq,
                                       u64 truncate_size,
                                       struct timespec *mtime,
-                                      bool use_mempool, int num_reply);
+                                      bool use_mempool, int num_reply,
+                                      int page_align);
 
 static inline void ceph_osdc_get_request(struct ceph_osd_request *req)
 {
@@ -218,7 +220,8 @@ extern int ceph_osdc_readpages(struct ceph_osd_client *osdc,
                                struct ceph_file_layout *layout,
                                u64 off, u64 *plen,
                                u32 truncate_seq, u64 truncate_size,
-                               struct page **pages, int nr_pages);
+                               struct page **pages, int nr_pages,
+                               int page_align);
 
 extern int ceph_osdc_writepages(struct ceph_osd_client *osdc,
                                 struct ceph_vino vino,
diff --git a/include/linux/fb.h b/include/linux/fb.h
index 7fca3dc4e475..d1631d37e9e0 100644
--- a/include/linux/fb.h
+++ b/include/linux/fb.h
@@ -1122,6 +1122,7 @@ extern const struct fb_videomode *fb_find_best_display(const struct fb_monspecs
 
 /* drivers/video/fbcmap.c */
 extern int fb_alloc_cmap(struct fb_cmap *cmap, int len, int transp);
+extern int fb_alloc_cmap_gfp(struct fb_cmap *cmap, int len, int transp, gfp_t flags);
 extern void fb_dealloc_cmap(struct fb_cmap *cmap);
 extern int fb_copy_cmap(const struct fb_cmap *from, struct fb_cmap *to);
 extern int fb_cmap_to_user(const struct fb_cmap *from, struct fb_cmap_user *to);
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 334d68a17108..c9e06cc70dad 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -34,9 +34,9 @@
 #define SEEK_MAX        SEEK_END
 
 struct fstrim_range {
-        uint64_t start;
-        uint64_t len;
-        uint64_t minlen;
+        __u64 start;
+        __u64 len;
+        __u64 minlen;
 };
 
 /* And dynamically-tunable limits and defaults: */
@@ -1612,7 +1612,6 @@ struct super_operations {
         ssize_t (*quota_write)(struct super_block *, int, const char *, size_t, loff_t);
 #endif
         int (*bdev_try_to_free_page)(struct super_block*, struct page*, gfp_t);
-        int (*trim_fs) (struct super_block *, struct fstrim_range *);
 };
 
 /*
diff --git a/include/linux/fsl-diu-fb.h b/include/linux/fsl-diu-fb.h
index fc295d7ea463..781d4671415f 100644
--- a/include/linux/fsl-diu-fb.h
+++ b/include/linux/fsl-diu-fb.h
@@ -54,7 +54,6 @@ struct aoi_display_offset {
 };
 
 #define MFB_SET_CHROMA_KEY      _IOW('M', 1, struct mfb_chroma_key)
-#define MFB_WAIT_FOR_VSYNC      _IOW('F', 0x20, u_int32_t)
 #define MFB_SET_BRIGHTNESS      _IOW('M', 3, __u8)
 
 #define MFB_SET_ALPHA           0x80014d00
diff --git a/include/linux/hardirq.h b/include/linux/hardirq.h
index 41cb31f14ee3..32f9fd6619b4 100644
--- a/include/linux/hardirq.h
+++ b/include/linux/hardirq.h
@@ -2,9 +2,6 @@
 #define LINUX_HARDIRQ_H
 
 #include <linux/preempt.h>
-#ifdef CONFIG_PREEMPT
-#include <linux/smp_lock.h>
-#endif
 #include <linux/lockdep.h>
 #include <linux/ftrace_irq.h>
 #include <asm/hardirq.h>
@@ -97,7 +94,8 @@
 #define in_nmi()        (preempt_count() & NMI_MASK)
 
 #if defined(CONFIG_PREEMPT) && defined(CONFIG_BKL)
-# define PREEMPT_INATOMIC_BASE kernel_locked()
+# include <linux/sched.h>
+# define PREEMPT_INATOMIC_BASE (current->lock_depth >= 0)
 #else
 # define PREEMPT_INATOMIC_BASE 0
 #endif
diff --git a/include/linux/i2c-id.h b/include/linux/i2c-id.h
index e844a0b18695..4bef5c557160 100644
--- a/include/linux/i2c-id.h
+++ b/include/linux/i2c-id.h
@@ -32,28 +32,6 @@
  */
 
 /* --- Bit algorithm adapters                                           */
-#define I2C_HW_B_BT848          0x010005 /* BT848 video boards */
-#define I2C_HW_B_RIVA           0x010010 /* Riva based graphics cards */
-#define I2C_HW_B_ZR36067        0x010019 /* Zoran-36057/36067 based boards */
 #define I2C_HW_B_CX2388x        0x01001b /* connexant 2388x based tv cards */
-#define I2C_HW_B_EM28XX         0x01001f /* em28xx video capture cards */
-#define I2C_HW_B_CX2341X        0x010020 /* Conexant CX2341X MPEG encoder cards */
-#define I2C_HW_B_CX23885        0x010022 /* conexant 23885 based tv cards (bus1) */
-#define I2C_HW_B_AU0828         0x010023 /* auvitek au0828 usb bridge */
-#define I2C_HW_B_CX231XX        0x010024 /* Conexant CX231XX USB based cards */
-#define I2C_HW_B_HDPVR          0x010025 /* Hauppauge HD PVR */
-
-/* --- SGI adapters                                                     */
-#define I2C_HW_SGI_VINO         0x160000
-
-/* --- SMBus only adapters                                              */
-#define I2C_HW_SMBUS_W9968CF    0x04000d
-#define I2C_HW_SMBUS_OV511      0x04000e /* OV511(+) USB 1.1 webcam ICs */
-#define I2C_HW_SMBUS_OV518      0x04000f /* OV518(+) USB 1.1 webcam ICs */
-#define I2C_HW_SMBUS_CAFE       0x040012 /* Marvell 88ALP01 "CAFE" cam  */
-
-/* --- Miscellaneous adapters */
-#define I2C_HW_SAA7146          0x060000 /* SAA7146 video decoder bus */
-#define I2C_HW_SAA7134          0x090000 /* SAA7134 video decoder bus */
 
 #endif /* LINUX_I2C_ID_H */
diff --git a/include/linux/i2c.h b/include/linux/i2c.h
index 889b35abaeda..56cfe23ffb39 100644
--- a/include/linux/i2c.h
+++ b/include/linux/i2c.h
@@ -353,7 +353,7 @@ struct i2c_algorithm {
  */
 struct i2c_adapter {
         struct module *owner;
-        unsigned int id;
+        unsigned int id __deprecated;
         unsigned int class;               /* classes to allow probing for */
         const struct i2c_algorithm *algo; /* the algorithm to access the bus */
         void *algo_data;
diff --git a/include/linux/kernel.h b/include/linux/kernel.h
index fc3da9e4da19..b6de9a6f7018 100644
--- a/include/linux/kernel.h
+++ b/include/linux/kernel.h
@@ -17,13 +17,11 @@
 #include <linux/bitops.h>
 #include <linux/log2.h>
 #include <linux/typecheck.h>
+#include <linux/printk.h>
 #include <linux/dynamic_debug.h>
 #include <asm/byteorder.h>
 #include <asm/bug.h>
 
-extern const char linux_banner[];
-extern const char linux_proc_banner[];
-
 #define USHRT_MAX       ((u16)(~0U))
 #define SHRT_MAX        ((s16)(USHRT_MAX>>1))
 #define SHRT_MIN        ((s16)(-SHRT_MAX - 1))
@@ -110,31 +108,6 @@ extern const char linux_proc_banner[];
  */
 #define lower_32_bits(n) ((u32)(n))
 
-#define KERN_EMERG      "<0>"   /* system is unusable                   */
-#define KERN_ALERT      "<1>"   /* action must be taken immediately     */
-#define KERN_CRIT       "<2>"   /* critical conditions                  */
-#define KERN_ERR        "<3>"   /* error conditions                     */
-#define KERN_WARNING    "<4>"   /* warning conditions                   */
-#define KERN_NOTICE     "<5>"   /* normal but significant condition     */
-#define KERN_INFO       "<6>"   /* informational                        */
-#define KERN_DEBUG      "<7>"   /* debug-level messages                 */
-
-/* Use the default kernel loglevel */
-#define KERN_DEFAULT    "<d>"
-/*
- * Annotation for a "continued" line of log printout (only done after a
- * line that had no enclosing \n). Only to be used by core/arch code
- * during early bootup (a continued line is not SMP-safe otherwise).
- */
-#define KERN_CONT       "<c>"
-
-extern int console_printk[];
-
-#define console_loglevel (console_printk[0])
-#define default_message_loglevel (console_printk[1])
-#define minimum_console_loglevel (console_printk[2])
-#define default_console_loglevel (console_printk[3])
-
 struct completion;
 struct pt_regs;
 struct user;
@@ -187,11 +160,6 @@ static inline void might_fault(void)
 }
 #endif
 
-struct va_format {
-        const char *fmt;
-        va_list *va;
-};
-
 extern struct atomic_notifier_head panic_notifier_list;
 extern long (*panic_blink)(int state);
 NORET_TYPE void panic(const char * fmt, ...)
@@ -245,115 +213,8 @@ extern int func_ptr_is_kernel_text(void *ptr);
 struct pid;
 extern struct pid *session_of_pgrp(struct pid *pgrp);
 
-/*
- * FW_BUG
- * Add this to a message where you are sure the firmware is buggy or behaves
- * really stupid or out of spec. Be aware that the responsible BIOS developer
- * should be able to fix this issue or at least get a concrete idea of the
- * problem by reading your message without the need of looking at the kernel
- * code.
- * 
- * Use it for definite and high priority BIOS bugs.
- *
- * FW_WARN
- * Use it for not that clear (e.g. could the kernel messed up things already?)
- * and medium priority BIOS bugs.
- *
- * FW_INFO
- * Use this one if you want to tell the user or vendor about something
- * suspicious, but generally harmless related to the firmware.
- *
- * Use it for information or very low priority BIOS bugs.
- */
-#define FW_BUG          "[Firmware Bug]: "
-#define FW_WARN         "[Firmware Warn]: "
-#define FW_INFO         "[Firmware Info]: "
-
-/*
- * HW_ERR
- * Add this to a message for hardware errors, so that user can report
- * it to hardware vendor instead of LKML or software vendor.
- */
-#define HW_ERR          "[Hardware Error]: "
-
-#ifdef CONFIG_PRINTK
-asmlinkage int vprintk(const char *fmt, va_list args)
-        __attribute__ ((format (printf, 1, 0)));
-asmlinkage int printk(const char * fmt, ...)
-        __attribute__ ((format (printf, 1, 2))) __cold;
-
-/*
- * Please don't use printk_ratelimit(), because it shares ratelimiting state
- * with all other unrelated printk_ratelimit() callsites.  Instead use
- * printk_ratelimited() or plain old __ratelimit().
- */
-extern int __printk_ratelimit(const char *func);
-#define printk_ratelimit() __printk_ratelimit(__func__)
-extern bool printk_timed_ratelimit(unsigned long *caller_jiffies,
-                                   unsigned int interval_msec);
-
-extern int printk_delay_msec;
-extern int dmesg_restrict;
-
-/*
- * Print a one-time message (analogous to WARN_ONCE() et al):
- */
-#define printk_once(x...) ({                    \
-        static bool __print_once;               \
-                                                \
-        if (!__print_once) {                    \
-                __print_once = true;            \
-                printk(x);                      \
-        }                                       \
-})
-
-void log_buf_kexec_setup(void);
-#else
-static inline int vprintk(const char *s, va_list args)
-        __attribute__ ((format (printf, 1, 0)));
-static inline int vprintk(const char *s, va_list args) { return 0; }
-static inline int printk(const char *s, ...)
-        __attribute__ ((format (printf, 1, 2)));
-static inline int __cold printk(const char *s, ...) { return 0; }
-static inline int printk_ratelimit(void) { return 0; }
-static inline bool printk_timed_ratelimit(unsigned long *caller_jiffies, \
-                                          unsigned int interval_msec)   \
-                { return false; }
-
-/* No effect, but we still get type checking even in the !PRINTK case: */
-#define printk_once(x...) printk(x)
-
-static inline void log_buf_kexec_setup(void)
-{
-}
-#endif
-
-/*
- * Dummy printk for disabled debugging statements to use whilst maintaining
- * gcc's format and side-effect checking.
- */
-static inline __attribute__ ((format (printf, 1, 2)))
-int no_printk(const char *s, ...) { return 0; }
-
-extern int printk_needs_cpu(int cpu);
-extern void printk_tick(void);
-
-extern void asmlinkage __attribute__((format(printf, 1, 2)))
-        early_printk(const char *fmt, ...);
-
 unsigned long int_sqrt(unsigned long);
 
-static inline void console_silent(void)
-{
-        console_loglevel = 0;
-}
-
-static inline void console_verbose(void)
-{
-        if (console_loglevel)
-                console_loglevel = 15;
-}
-
 extern void bust_spinlocks(int yes);
 extern void wake_up_klogd(void);
 extern int oops_in_progress;            /* If set, an oops, panic(), BUG() or die() is in progress */
@@ -390,22 +251,6 @@ extern enum system_states {
 #define TAINT_CRAP                      10
 #define TAINT_FIRMWARE_WORKAROUND       11
 
-extern void dump_stack(void) __cold;
-
-enum {
-        DUMP_PREFIX_NONE,
-        DUMP_PREFIX_ADDRESS,
-        DUMP_PREFIX_OFFSET
-};
-extern void hex_dump_to_buffer(const void *buf, size_t len,
-                                int rowsize, int groupsize,
-                                char *linebuf, size_t linebuflen, bool ascii);
-extern void print_hex_dump(const char *level, const char *prefix_str,
-                                int prefix_type, int rowsize, int groupsize,
-                                const void *buf, size_t len, bool ascii);
-extern void print_hex_dump_bytes(const char *prefix_str, int prefix_type,
-                        const void *buf, size_t len);
-
 extern const char hex_asc[];
 #define hex_asc_lo(x)   hex_asc[((x) & 0x0f)]
 #define hex_asc_hi(x)   hex_asc[((x) & 0xf0) >> 4]
@@ -419,94 +264,6 @@ static inline char *pack_hex_byte(char *buf, u8 byte)
 
 extern int hex_to_bin(char ch);
 
-#ifndef pr_fmt
-#define pr_fmt(fmt) fmt
-#endif
-
-#define pr_emerg(fmt, ...) \
-        printk(KERN_EMERG pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_alert(fmt, ...) \
-        printk(KERN_ALERT pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_crit(fmt, ...) \
-        printk(KERN_CRIT pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_err(fmt, ...) \
-        printk(KERN_ERR pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_warning(fmt, ...) \
-        printk(KERN_WARNING pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_warn pr_warning
-#define pr_notice(fmt, ...) \
-        printk(KERN_NOTICE pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_info(fmt, ...) \
-        printk(KERN_INFO pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_cont(fmt, ...) \
-        printk(KERN_CONT fmt, ##__VA_ARGS__)
-
-/* pr_devel() should produce zero code unless DEBUG is defined */
-#ifdef DEBUG
-#define pr_devel(fmt, ...) \
-        printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__)
-#else
-#define pr_devel(fmt, ...) \
-        ({ if (0) printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__); 0; })
-#endif
-
-/* If you are writing a driver, please use dev_dbg instead */
-#if defined(DEBUG)
-#define pr_debug(fmt, ...) \
-        printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__)
-#elif defined(CONFIG_DYNAMIC_DEBUG)
-/* dynamic_pr_debug() uses pr_fmt() internally so we don't need it here */
-#define pr_debug(fmt, ...) \
-        dynamic_pr_debug(fmt, ##__VA_ARGS__)
-#else
-#define pr_debug(fmt, ...) \
-        ({ if (0) printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__); 0; })
-#endif
-
-/*
- * ratelimited messages with local ratelimit_state,
- * no local ratelimit_state used in the !PRINTK case
- */
-#ifdef CONFIG_PRINTK
-#define printk_ratelimited(fmt, ...)  ({                                \
-        static DEFINE_RATELIMIT_STATE(_rs,                              \
-                                      DEFAULT_RATELIMIT_INTERVAL,       \
-                                      DEFAULT_RATELIMIT_BURST);         \
-                                                                        \
-        if (__ratelimit(&_rs))                                          \
-                printk(fmt, ##__VA_ARGS__);                             \
-})
-#else
-/* No effect, but we still get type checking even in the !PRINTK case: */
-#define printk_ratelimited printk
-#endif
-
-#define pr_emerg_ratelimited(fmt, ...) \
-        printk_ratelimited(KERN_EMERG pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_alert_ratelimited(fmt, ...) \
-        printk_ratelimited(KERN_ALERT pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_crit_ratelimited(fmt, ...) \
-        printk_ratelimited(KERN_CRIT pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_err_ratelimited(fmt, ...) \
-        printk_ratelimited(KERN_ERR pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_warning_ratelimited(fmt, ...) \
-        printk_ratelimited(KERN_WARNING pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_warn_ratelimited pr_warning_ratelimited
-#define pr_notice_ratelimited(fmt, ...) \
-        printk_ratelimited(KERN_NOTICE pr_fmt(fmt), ##__VA_ARGS__)
-#define pr_info_ratelimited(fmt, ...) \
-        printk_ratelimited(KERN_INFO pr_fmt(fmt), ##__VA_ARGS__)
-/* no pr_cont_ratelimited, don't do that... */
-/* If you are writing a driver, please use dev_dbg instead */
-#if defined(DEBUG)
-#define pr_debug_ratelimited(fmt, ...) \
-        printk_ratelimited(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__)
-#else
-#define pr_debug_ratelimited(fmt, ...) \
-        ({ if (0) printk_ratelimited(KERN_DEBUG pr_fmt(fmt), \
-                                     ##__VA_ARGS__); 0; })
-#endif
-
 /*
  * General tracing related utility functions - trace_printk(),
  * tracing_on/tracing_off and tracing_start()/tracing_stop
diff --git a/include/linux/libata.h b/include/linux/libata.h
index 15b77b8dc7e1..d947b1231662 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -986,7 +986,7 @@ extern void ata_host_init(struct ata_host *, struct device *,
                           unsigned long, struct ata_port_operations *);
 extern int ata_scsi_detect(struct scsi_host_template *sht);
 extern int ata_scsi_ioctl(struct scsi_device *dev, int cmd, void __user *arg);
-extern int ata_scsi_queuecmd(struct scsi_cmnd *cmd, void (*done)(struct scsi_cmnd *));
+extern int ata_scsi_queuecmd(struct Scsi_Host *h, struct scsi_cmnd *cmd);
 extern int ata_sas_scsi_ioctl(struct ata_port *ap, struct scsi_device *dev,
                             int cmd, void __user *arg);
 extern void ata_sas_port_destroy(struct ata_port *);
diff --git a/include/linux/lockd/lockd.h b/include/linux/lockd/lockd.h
index a34dea46b629..2dee05e5119a 100644
--- a/include/linux/lockd/lockd.h
+++ b/include/linux/lockd/lockd.h
@@ -43,6 +43,7 @@ struct nlm_host {
         struct sockaddr_storage h_addr;         /* peer address */
         size_t                  h_addrlen;
         struct sockaddr_storage h_srcaddr;      /* our address (optional) */
+        size_t                  h_srcaddrlen;
         struct rpc_clnt         *h_rpcclnt;     /* RPC client to talk to peer */
         char                    *h_name;                /* remote hostname */
         u32                     h_version;      /* interface version */
diff --git a/include/linux/marvell_phy.h b/include/linux/marvell_phy.h
index 1ff81b51b656..dd3c34ebca9a 100644
--- a/include/linux/marvell_phy.h
+++ b/include/linux/marvell_phy.h
@@ -11,6 +11,7 @@
 #define MARVELL_PHY_ID_88E1118          0x01410e10
 #define MARVELL_PHY_ID_88E1121R         0x01410cb0
 #define MARVELL_PHY_ID_88E1145          0x01410cd0
+#define MARVELL_PHY_ID_88E1149R         0x01410e50
 #define MARVELL_PHY_ID_88E1240          0x01410e30
 #define MARVELL_PHY_ID_88E1318S         0x01410e90
 
diff --git a/include/linux/mfd/wm8350/audio.h b/include/linux/mfd/wm8350/audio.h
index a95141eafce3..bd581c6fa085 100644
--- a/include/linux/mfd/wm8350/audio.h
+++ b/include/linux/mfd/wm8350/audio.h
@@ -522,9 +522,6 @@
 #define WM8350_MCLK_SEL_PLL_32K                 3
 #define WM8350_MCLK_SEL_MCLK                    5
 
-#define WM8350_MCLK_DIR_OUT                     0
-#define WM8350_MCLK_DIR_IN                      1
-
 /* clock divider id's */
 #define WM8350_ADC_CLKDIV                       0
 #define WM8350_DAC_CLKDIV                       1
diff --git a/include/linux/module.h b/include/linux/module.h
index b29e7458b966..7575bbbdf2a2 100644
--- a/include/linux/module.h
+++ b/include/linux/module.h
@@ -517,7 +517,7 @@ static inline void __module_get(struct module *module)
 #define symbol_put_addr(p) do { } while(0)
 
 #endif /* CONFIG_MODULE_UNLOAD */
-int use_module(struct module *a, struct module *b);
+int ref_module(struct module *a, struct module *b);
 
 /* This is a #define so the string doesn't get put in every .o file */
 #define module_name(mod)                        \
diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h
index bba26684acdc..c66fdb7d6998 100644
--- a/include/linux/nfs_fs.h
+++ b/include/linux/nfs_fs.h
@@ -593,12 +593,6 @@ nfs_fileid_to_ino_t(u64 fileid)
         return ino;
 }
 
-#define nfs_wait_event(clnt, wq, condition)                             \
-({                                                                      \
-        int __retval = wait_event_killable(wq, condition);              \
-        __retval;                                                       \
-})
-
 #define NFS_JUKEBOX_RETRY_TIME (5 * HZ)
 
 #endif /* __KERNEL__ */
diff --git a/include/linux/page_cgroup.h b/include/linux/page_cgroup.h
index 5bb13b3db84d..b02195dfc1b0 100644
--- a/include/linux/page_cgroup.h
+++ b/include/linux/page_cgroup.h
@@ -59,8 +59,6 @@ static inline void ClearPageCgroup##uname(struct page_cgroup *pc)	\
 static inline int TestClearPageCgroup##uname(struct page_cgroup *pc)    \
         { return test_and_clear_bit(PCG_##lname, &pc->flags);  }
 
-TESTPCGFLAG(Locked, LOCK)
-
 /* Cache flag is set only once (at allocation) */
 TESTPCGFLAG(Cache, CACHE)
 CLEARPCGFLAG(Cache, CACHE)
@@ -104,6 +102,11 @@ static inline void unlock_page_cgroup(struct page_cgroup *pc)
         bit_spin_unlock(PCG_LOCK, &pc->flags);
 }
 
+static inline int page_is_cgroup_locked(struct page_cgroup *pc)
+{
+        return bit_spin_is_locked(PCG_LOCK, &pc->flags);
+}
+
 #else /* CONFIG_CGROUP_MEM_RES_CTLR */
 struct page_cgroup;
 
diff --git a/include/linux/printk.h b/include/linux/printk.h
new file mode 100644
index 000000000000..b772ca5fbdf0
--- /dev/null
+++ b/include/linux/printk.h
@@ -0,0 +1,248 @@
+#ifndef __KERNEL_PRINTK__
+#define __KERNEL_PRINTK__
+
+extern const char linux_banner[];
+extern const char linux_proc_banner[];
+
+#define KERN_EMERG      "<0>"   /* system is unusable                   */
+#define KERN_ALERT      "<1>"   /* action must be taken immediately     */
+#define KERN_CRIT       "<2>"   /* critical conditions                  */
+#define KERN_ERR        "<3>"   /* error conditions                     */
+#define KERN_WARNING    "<4>"   /* warning conditions                   */
+#define KERN_NOTICE     "<5>"   /* normal but significant condition     */
+#define KERN_INFO       "<6>"   /* informational                        */
+#define KERN_DEBUG      "<7>"   /* debug-level messages                 */
+
+/* Use the default kernel loglevel */
+#define KERN_DEFAULT    "<d>"
+/*
+ * Annotation for a "continued" line of log printout (only done after a
+ * line that had no enclosing \n). Only to be used by core/arch code
+ * during early bootup (a continued line is not SMP-safe otherwise).
+ */
+#define KERN_CONT       "<c>"
+
+extern int console_printk[];
+
+#define console_loglevel (console_printk[0])
+#define default_message_loglevel (console_printk[1])
+#define minimum_console_loglevel (console_printk[2])
+#define default_console_loglevel (console_printk[3])
+
+struct va_format {
+        const char *fmt;
+        va_list *va;
+};
+
+/*
+ * FW_BUG
+ * Add this to a message where you are sure the firmware is buggy or behaves
+ * really stupid or out of spec. Be aware that the responsible BIOS developer
+ * should be able to fix this issue or at least get a concrete idea of the
+ * problem by reading your message without the need of looking at the kernel
+ * code.
+ *
+ * Use it for definite and high priority BIOS bugs.
+ *
+ * FW_WARN
+ * Use it for not that clear (e.g. could the kernel messed up things already?)
+ * and medium priority BIOS bugs.
+ *
+ * FW_INFO
+ * Use this one if you want to tell the user or vendor about something
+ * suspicious, but generally harmless related to the firmware.
+ *
+ * Use it for information or very low priority BIOS bugs.
+ */
+#define FW_BUG          "[Firmware Bug]: "
+#define FW_WARN         "[Firmware Warn]: "
+#define FW_INFO         "[Firmware Info]: "
+
+/*
+ * HW_ERR
+ * Add this to a message for hardware errors, so that user can report
+ * it to hardware vendor instead of LKML or software vendor.
+ */
+#define HW_ERR          "[Hardware Error]: "
+
+#ifdef CONFIG_PRINTK
+asmlinkage int vprintk(const char *fmt, va_list args)
+        __attribute__ ((format (printf, 1, 0)));
+asmlinkage int printk(const char * fmt, ...)
+        __attribute__ ((format (printf, 1, 2))) __cold;
+
+/*
+ * Please don't use printk_ratelimit(), because it shares ratelimiting state
+ * with all other unrelated printk_ratelimit() callsites.  Instead use
+ * printk_ratelimited() or plain old __ratelimit().
+ */
+extern int __printk_ratelimit(const char *func);
+#define printk_ratelimit() __printk_ratelimit(__func__)
+extern bool printk_timed_ratelimit(unsigned long *caller_jiffies,
+                                   unsigned int interval_msec);
+
+extern int printk_delay_msec;
+extern int dmesg_restrict;
+
+/*
+ * Print a one-time message (analogous to WARN_ONCE() et al):
+ */
+#define printk_once(x...) ({                    \
+        static bool __print_once;               \
+                                                \
+        if (!__print_once) {                    \
+                __print_once = true;            \
+                printk(x);                      \
+        }                                       \
+})
+
+void log_buf_kexec_setup(void);
+#else
+static inline int vprintk(const char *s, va_list args)
+        __attribute__ ((format (printf, 1, 0)));
+static inline int vprintk(const char *s, va_list args) { return 0; }
+static inline int printk(const char *s, ...)
+        __attribute__ ((format (printf, 1, 2)));
+static inline int __cold printk(const char *s, ...) { return 0; }
+static inline int printk_ratelimit(void) { return 0; }
+static inline bool printk_timed_ratelimit(unsigned long *caller_jiffies, \
+                                          unsigned int interval_msec)   \
+                { return false; }
+
+/* No effect, but we still get type checking even in the !PRINTK case: */
+#define printk_once(x...) printk(x)
+
+static inline void log_buf_kexec_setup(void)
+{
+}
+#endif
+
+/*
+ * Dummy printk for disabled debugging statements to use whilst maintaining
+ * gcc's format and side-effect checking.
+ */
+static inline __attribute__ ((format (printf, 1, 2)))
+int no_printk(const char *s, ...) { return 0; }
+
+extern int printk_needs_cpu(int cpu);
+extern void printk_tick(void);
+
+extern void asmlinkage __attribute__((format(printf, 1, 2)))
+        early_printk(const char *fmt, ...);
+
+static inline void console_silent(void)
+{
+        console_loglevel = 0;
+}
+
+static inline void console_verbose(void)
+{
+        if (console_loglevel)
+                console_loglevel = 15;
+}
+
+extern void dump_stack(void) __cold;
+
+enum {
+        DUMP_PREFIX_NONE,
+        DUMP_PREFIX_ADDRESS,
+        DUMP_PREFIX_OFFSET
+};
+extern void hex_dump_to_buffer(const void *buf, size_t len,
+                                int rowsize, int groupsize,
+                                char *linebuf, size_t linebuflen, bool ascii);
+extern void print_hex_dump(const char *level, const char *prefix_str,
+                                int prefix_type, int rowsize, int groupsize,
+                                const void *buf, size_t len, bool ascii);
+extern void print_hex_dump_bytes(const char *prefix_str, int prefix_type,
+                        const void *buf, size_t len);
+
+#ifndef pr_fmt
+#define pr_fmt(fmt) fmt
+#endif
+
+#define pr_emerg(fmt, ...) \
+        printk(KERN_EMERG pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_alert(fmt, ...) \
+        printk(KERN_ALERT pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_crit(fmt, ...) \
+        printk(KERN_CRIT pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_err(fmt, ...) \
+        printk(KERN_ERR pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_warning(fmt, ...) \
+        printk(KERN_WARNING pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_warn pr_warning
+#define pr_notice(fmt, ...) \
+        printk(KERN_NOTICE pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_info(fmt, ...) \
+        printk(KERN_INFO pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_cont(fmt, ...) \
+        printk(KERN_CONT fmt, ##__VA_ARGS__)
+
+/* pr_devel() should produce zero code unless DEBUG is defined */
+#ifdef DEBUG
+#define pr_devel(fmt, ...) \
+        printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__)
+#else
+#define pr_devel(fmt, ...) \
+        ({ if (0) printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__); 0; })
+#endif
+
+/* If you are writing a driver, please use dev_dbg instead */
+#if defined(DEBUG)
+#define pr_debug(fmt, ...) \
+        printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__)
+#elif defined(CONFIG_DYNAMIC_DEBUG)
+/* dynamic_pr_debug() uses pr_fmt() internally so we don't need it here */
+#define pr_debug(fmt, ...) \
+        dynamic_pr_debug(fmt, ##__VA_ARGS__)
+#else
+#define pr_debug(fmt, ...) \
+        ({ if (0) printk(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__); 0; })
+#endif
+
+/*
+ * ratelimited messages with local ratelimit_state,
+ * no local ratelimit_state used in the !PRINTK case
+ */
+#ifdef CONFIG_PRINTK
+#define printk_ratelimited(fmt, ...)  ({                                \
+        static DEFINE_RATELIMIT_STATE(_rs,                              \
+                                      DEFAULT_RATELIMIT_INTERVAL,       \
+                                      DEFAULT_RATELIMIT_BURST);         \
+                                                                        \
+        if (__ratelimit(&_rs))                                          \
+                printk(fmt, ##__VA_ARGS__);                             \
+})
+#else
+/* No effect, but we still get type checking even in the !PRINTK case: */
+#define printk_ratelimited printk
+#endif
+
+#define pr_emerg_ratelimited(fmt, ...) \
+        printk_ratelimited(KERN_EMERG pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_alert_ratelimited(fmt, ...) \
+        printk_ratelimited(KERN_ALERT pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_crit_ratelimited(fmt, ...) \
+        printk_ratelimited(KERN_CRIT pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_err_ratelimited(fmt, ...) \
+        printk_ratelimited(KERN_ERR pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_warning_ratelimited(fmt, ...) \
+        printk_ratelimited(KERN_WARNING pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_warn_ratelimited pr_warning_ratelimited
+#define pr_notice_ratelimited(fmt, ...) \
+        printk_ratelimited(KERN_NOTICE pr_fmt(fmt), ##__VA_ARGS__)
+#define pr_info_ratelimited(fmt, ...) \
+        printk_ratelimited(KERN_INFO pr_fmt(fmt), ##__VA_ARGS__)
+/* no pr_cont_ratelimited, don't do that... */
+/* If you are writing a driver, please use dev_dbg instead */
+#if defined(DEBUG)
+#define pr_debug_ratelimited(fmt, ...) \
+        printk_ratelimited(KERN_DEBUG pr_fmt(fmt), ##__VA_ARGS__)
+#else
+#define pr_debug_ratelimited(fmt, ...) \
+        ({ if (0) printk_ratelimited(KERN_DEBUG pr_fmt(fmt), \
+                                     ##__VA_ARGS__); 0; })
+#endif
+
+#endif
diff --git a/include/linux/reiserfs_fs.h b/include/linux/reiserfs_fs.h
index 5ca47e59b727..c21072adbfad 100644
--- a/include/linux/reiserfs_fs.h
+++ b/include/linux/reiserfs_fs.h
@@ -22,7 +22,6 @@
 #include <asm/unaligned.h>
 #include <linux/bitops.h>
 #include <linux/proc_fs.h>
-#include <linux/smp_lock.h>
 #include <linux/buffer_head.h>
 #include <linux/reiserfs_fs_i.h>
 #include <linux/reiserfs_fs_sb.h>
diff --git a/include/linux/rtnetlink.h b/include/linux/rtnetlink.h
index d42f274418b8..bbad657a3725 100644
--- a/include/linux/rtnetlink.h
+++ b/include/linux/rtnetlink.h
@@ -6,7 +6,6 @@
 #include <linux/if_link.h>
 #include <linux/if_addr.h>
 #include <linux/neighbour.h>
-#include <linux/netdevice.h>
 
 /* rtnetlink families. Values up to 127 are reserved for real address
  * families, values above 128 may be used arbitrarily.
@@ -606,6 +605,7 @@ struct tcamsg {
 #ifdef __KERNEL__
 
 #include <linux/mutex.h>
+#include <linux/netdevice.h>
 
 static __inline__ int rtattr_strcmp(const struct rtattr *rta, const char *str)
 {
diff --git a/include/linux/sched.h b/include/linux/sched.h
index d0036e52a24a..2c79e921a68b 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -862,6 +862,7 @@ struct sched_group {
          * single CPU.
          */
         unsigned int cpu_power, cpu_power_orig;
+        unsigned int group_weight;
 
         /*
          * The CPUs this group covers.
diff --git a/include/linux/security.h b/include/linux/security.h
index b8246a8df7d2..fd4d55fb8845 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -77,7 +77,6 @@ extern int cap_task_prctl(int option, unsigned long arg2, unsigned long arg3,
 extern int cap_task_setscheduler(struct task_struct *p);
 extern int cap_task_setioprio(struct task_struct *p, int ioprio);
 extern int cap_task_setnice(struct task_struct *p, int nice);
-extern int cap_syslog(int type, bool from_file);
 extern int cap_vm_enough_memory(struct mm_struct *mm, long pages);
 
 struct msghdr;
@@ -1388,7 +1387,7 @@ struct security_operations {
         int (*sysctl) (struct ctl_table *table, int op);
         int (*quotactl) (int cmds, int type, int id, struct super_block *sb);
         int (*quota_on) (struct dentry *dentry);
-        int (*syslog) (int type, bool from_file);
+        int (*syslog) (int type);
         int (*settime) (struct timespec *ts, struct timezone *tz);
         int (*vm_enough_memory) (struct mm_struct *mm, long pages);
 
@@ -1671,7 +1670,7 @@ int security_real_capable_noaudit(struct task_struct *tsk, int cap);
 int security_sysctl(struct ctl_table *table, int op);
 int security_quotactl(int cmds, int type, int id, struct super_block *sb);
 int security_quota_on(struct dentry *dentry);
-int security_syslog(int type, bool from_file);
+int security_syslog(int type);
 int security_settime(struct timespec *ts, struct timezone *tz);
 int security_vm_enough_memory(long pages);
 int security_vm_enough_memory_mm(struct mm_struct *mm, long pages);
@@ -1901,9 +1900,9 @@ static inline int security_quota_on(struct dentry *dentry)
         return 0;
 }
 
-static inline int security_syslog(int type, bool from_file)
+static inline int security_syslog(int type)
 {
-        return cap_syslog(type, from_file);
+        return 0;
 }
 
 static inline int security_settime(struct timespec *ts, struct timezone *tz)
diff --git a/include/linux/sh_clk.h b/include/linux/sh_clk.h
index cea0c38e7a63..9a52f72527dc 100644
--- a/include/linux/sh_clk.h
+++ b/include/linux/sh_clk.h
@@ -19,11 +19,13 @@ struct clk_mapping {
 };
 
 struct clk_ops {
+#ifdef CONFIG_SH_CLK_CPG_LEGACY
         void (*init)(struct clk *clk);
+#endif
         int (*enable)(struct clk *clk);
         void (*disable)(struct clk *clk);
         unsigned long (*recalc)(struct clk *clk);
-        int (*set_rate)(struct clk *clk, unsigned long rate, int algo_id);
+        int (*set_rate)(struct clk *clk, unsigned long rate);
         int (*set_parent)(struct clk *clk, struct clk *parent);
         long (*round_rate)(struct clk *clk, unsigned long rate);
 };
@@ -67,36 +69,6 @@ int clk_register(struct clk *);
 void clk_unregister(struct clk *);
 void clk_enable_init_clocks(void);
 
-/**
- * clk_set_rate_ex - set the clock rate for a clock source, with additional parameter
- * @clk: clock source
- * @rate: desired clock rate in Hz
- * @algo_id: algorithm id to be passed down to ops->set_rate
- *
- * Returns success (0) or negative errno.
- */
-int clk_set_rate_ex(struct clk *clk, unsigned long rate, int algo_id);
-
-enum clk_sh_algo_id {
-        NO_CHANGE = 0,
-
-        IUS_N1_N1,
-        IUS_322,
-        IUS_522,
-        IUS_N11,
-
-        SB_N1,
-
-        SB3_N1,
-        SB3_32,
-        SB3_43,
-        SB3_54,
-
-        BP_N1,
-
-        IP_N1,
-};
-
 struct clk_div_mult_table {
         unsigned int *divisors;
         unsigned int nr_divisors;
diff --git a/include/linux/sh_intc.h b/include/linux/sh_intc.h
index f656d1a43dc0..5812fefbcedf 100644
--- a/include/linux/sh_intc.h
+++ b/include/linux/sh_intc.h
@@ -79,7 +79,7 @@ struct intc_hw_desc {
         unsigned int nr_subgroups;
 };
 
-#define _INTC_ARRAY(a) a, a == NULL ? 0 : sizeof(a)/sizeof(*a)
+#define _INTC_ARRAY(a) a, __same_type(a, NULL) ? 0 : sizeof(a)/sizeof(*a)
 
 #define INTC_HW_DESC(vectors, groups, mask_regs,        \
                      prio_regs, sense_regs, ack_regs)   \
diff --git a/include/linux/smp_lock.h b/include/linux/smp_lock.h
index 291f721144c2..3a1988202731 100644
--- a/include/linux/smp_lock.h
+++ b/include/linux/smp_lock.h
@@ -4,8 +4,6 @@
 #ifdef CONFIG_LOCK_KERNEL
 #include <linux/sched.h>
 
-#define kernel_locked()         (current->lock_depth >= 0)
-
 extern int __lockfunc __reacquire_kernel_lock(void);
 extern void __lockfunc __release_kernel_lock(void);
 
@@ -58,7 +56,6 @@ static inline void cycle_kernel_lock(void)
 #define lock_kernel()
 #define unlock_kernel()
 #define cycle_kernel_lock()                     do { } while(0)
-#define kernel_locked()                         1
 #endif /* CONFIG_BKL */
 
 #define release_kernel_lock(task)               do { } while(0)
diff --git a/include/linux/tty.h b/include/linux/tty.h
index c7ea9bc8897c..032d79ff1d9d 100644
--- a/include/linux/tty.h
+++ b/include/linux/tty.h
@@ -13,7 +13,6 @@
 #include <linux/tty_driver.h>
 #include <linux/tty_ldisc.h>
 #include <linux/mutex.h>
-#include <linux/smp_lock.h>
 
 #include <asm/system.h>
 
diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h
index 2a7936d7851d..97b8b7c9b63c 100644
--- a/include/net/cfg80211.h
+++ b/include/net/cfg80211.h
@@ -1355,7 +1355,7 @@ enum wiphy_flags {
         WIPHY_FLAG_4ADDR_AP                     = BIT(5),
         WIPHY_FLAG_4ADDR_STATION                = BIT(6),
         WIPHY_FLAG_CONTROL_PORT_PROTOCOL        = BIT(7),
-        WIPHY_FLAG_IBSS_RSN                     = BIT(7),
+        WIPHY_FLAG_IBSS_RSN                     = BIT(8),
 };
 
 struct mac_address {
diff --git a/include/net/neighbour.h b/include/net/neighbour.h
index 55590ab16b3e..6beb1ffc2b7f 100644
--- a/include/net/neighbour.h
+++ b/include/net/neighbour.h
@@ -303,7 +303,7 @@ static inline void neigh_confirm(struct neighbour *neigh)
 
 static inline int neigh_event_send(struct neighbour *neigh, struct sk_buff *skb)
 {
-        unsigned long now = ACCESS_ONCE(jiffies);
+        unsigned long now = jiffies;
         
         if (neigh->used != now)
                 neigh->used = now;
diff --git a/include/scsi/libfc.h b/include/scsi/libfc.h
index f986ab7ffe6f..5c4c1678f7be 100644
--- a/include/scsi/libfc.h
+++ b/include/scsi/libfc.h
@@ -1006,8 +1006,7 @@ void fc_fcp_destroy(struct fc_lport *);
 /*
  * SCSI INTERACTION LAYER
  *****************************/
-int fc_queuecommand(struct scsi_cmnd *,
-                    void (*done)(struct scsi_cmnd *));
+int fc_queuecommand(struct Scsi_Host *, struct scsi_cmnd *);
 int fc_eh_abort(struct scsi_cmnd *);
 int fc_eh_device_reset(struct scsi_cmnd *);
 int fc_eh_host_reset(struct scsi_cmnd *);
diff --git a/include/scsi/libiscsi.h b/include/scsi/libiscsi.h
index ae5196aae1a5..b81d969ddc67 100644
--- a/include/scsi/libiscsi.h
+++ b/include/scsi/libiscsi.h
@@ -341,8 +341,7 @@ extern int iscsi_eh_abort(struct scsi_cmnd *sc);
 extern int iscsi_eh_recover_target(struct scsi_cmnd *sc);
 extern int iscsi_eh_session_reset(struct scsi_cmnd *sc);
 extern int iscsi_eh_device_reset(struct scsi_cmnd *sc);
-extern int iscsi_queuecommand(struct scsi_cmnd *sc,
-                              void (*done)(struct scsi_cmnd *));
+extern int iscsi_queuecommand(struct Scsi_Host *h, struct scsi_cmnd *sc);
 
 /*
  * iSCSI host helpers.
diff --git a/include/scsi/libsas.h b/include/scsi/libsas.h
index 3dec1949f69c..90ce527ecf3d 100644
--- a/include/scsi/libsas.h
+++ b/include/scsi/libsas.h
@@ -621,8 +621,7 @@ int sas_set_phy_speed(struct sas_phy *phy,
 int sas_phy_enable(struct sas_phy *phy, int enabled);
 int sas_phy_reset(struct sas_phy *phy, int hard_reset);
 int sas_queue_up(struct sas_task *task);
-extern int sas_queuecommand(struct scsi_cmnd *,
-                     void (*scsi_done)(struct scsi_cmnd *));
+extern int sas_queuecommand(struct Scsi_Host * ,struct scsi_cmnd *);
 extern int sas_target_alloc(struct scsi_target *);
 extern int sas_slave_alloc(struct scsi_device *);
 extern int sas_slave_configure(struct scsi_device *);
diff --git a/include/scsi/scsi_host.h b/include/scsi/scsi_host.h
index d0a6a845f204..e7e385842a38 100644
--- a/include/scsi/scsi_host.h
+++ b/include/scsi/scsi_host.h
@@ -127,8 +127,7 @@ struct scsi_host_template {
          *
          * STATUS: REQUIRED
          */
-        int (* queuecommand)(struct scsi_cmnd *,
-                             void (*done)(struct scsi_cmnd *));
+        int (* queuecommand)(struct Scsi_Host *, struct scsi_cmnd *);
 
         /*
          * The transfer functions are used to queue a scsi command to
@@ -505,6 +504,25 @@ struct scsi_host_template {
 };
 
 /*
+ * Temporary #define for host lock push down. Can be removed when all
+ * drivers have been updated to take advantage of unlocked
+ * queuecommand.
+ *
+ */
+#define DEF_SCSI_QCMD(func_name) \
+        int func_name(struct Scsi_Host *shost, struct scsi_cmnd *cmd)   \
+        {                                                               \
+                unsigned long irq_flags;                                \
+                int rc;                                                 \
+                spin_lock_irqsave(shost->host_lock, irq_flags);         \
+                scsi_cmd_get_serial(shost, cmd);                        \
+                rc = func_name##_lck (cmd, cmd->scsi_done);                     \
+                spin_unlock_irqrestore(shost->host_lock, irq_flags);    \
+                return rc;                                              \
+        }
+
+
+/*
  * shost state: If you alter this, you also need to alter scsi_sysfs.c
  * (for the ascii descriptions) and the state model enforcer:
  * scsi_host_set_state()
@@ -752,6 +770,7 @@ extern struct Scsi_Host *scsi_host_get(struct Scsi_Host *);
 extern void scsi_host_put(struct Scsi_Host *t);
 extern struct Scsi_Host *scsi_host_lookup(unsigned short);
 extern const char *scsi_host_state_name(enum scsi_host_state);
+extern void scsi_cmd_get_serial(struct Scsi_Host *, struct scsi_cmnd *);
 
 extern u64 scsi_calculate_bounce_limit(struct Scsi_Host *);
 
diff --git a/include/video/da8xx-fb.h b/include/video/da8xx-fb.h
index 6316cdabf73f..89d43b3d4cb9 100644
--- a/include/video/da8xx-fb.h
+++ b/include/video/da8xx-fb.h
@@ -99,7 +99,6 @@ struct lcd_sync_arg {
 #define FBIPUT_COLOR            _IOW('F', 6, int)
 #define FBIPUT_HSYNC            _IOW('F', 9, int)
 #define FBIPUT_VSYNC            _IOW('F', 10, int)
-#define FBIO_WAITFORVSYNC       _IOW('F', 0x20, u_int32_t)
 
 #endif  /* ifndef DA8XX_FB_H */
 
diff --git a/include/xen/interface/memory.h b/include/xen/interface/memory.h
index d7a6c13bde69..eac3ce153719 100644
--- a/include/xen/interface/memory.h
+++ b/include/xen/interface/memory.h
@@ -141,6 +141,19 @@ struct xen_machphys_mfn_list {
 DEFINE_GUEST_HANDLE_STRUCT(xen_machphys_mfn_list);
 
 /*
+ * Returns the location in virtual address space of the machine_to_phys
+ * mapping table. Architectures which do not have a m2p table, or which do not
+ * map it by default into guest address space, do not implement this command.
+ * arg == addr of xen_machphys_mapping_t.
+ */
+#define XENMEM_machphys_mapping     12
+struct xen_machphys_mapping {
+    unsigned long v_start, v_end; /* Start and end virtual addresses.   */
+    unsigned long max_mfn;        /* Maximum MFN that can be looked up. */
+};
+DEFINE_GUEST_HANDLE_STRUCT(xen_machphys_mapping_t);
+
+/*
  * Sets the GPFN at which a particular page appears in the specified guest's
  * pseudophysical address space.
  * arg == addr of xen_add_to_physmap_t.
diff --git a/include/xen/page.h b/include/xen/page.h
index eaf85fab1263..0be36b976f4b 100644
--- a/include/xen/page.h
+++ b/include/xen/page.h
@@ -1 +1,8 @@
+#ifndef _XEN_PAGE_H
+#define _XEN_PAGE_H
+
 #include <asm/xen/page.h>
+
+extern phys_addr_t xen_extra_mem_start, xen_extra_mem_size;
+
+#endif  /* _XEN_PAGE_H */
diff --git a/include/xen/privcmd.h b/include/xen/privcmd.h
index b42cdfd92fee..17857fb4d550 100644
--- a/include/xen/privcmd.h
+++ b/include/xen/privcmd.h
@@ -34,13 +34,10 @@
 #define __LINUX_PUBLIC_PRIVCMD_H__
 
 #include <linux/types.h>
+#include <linux/compiler.h>
 
 typedef unsigned long xen_pfn_t;
 
-#ifndef __user
-#define __user
-#endif
-
 struct privcmd_hypercall {
         __u64 op;
         __u64 arg[5];
diff --git a/init/Kconfig b/init/Kconfig
index 88c10468db46..c9728992a776 100644
--- a/init/Kconfig
+++ b/init/Kconfig
@@ -613,6 +613,19 @@ config CGROUP_MEM_RES_CTLR_SWAP
           if boot option "noswapaccount" is set, swap will not be accounted.
           Now, memory usage of swap_cgroup is 2 bytes per entry. If swap page
           size is 4096bytes, 512k per 1Gbytes of swap.
+config CGROUP_MEM_RES_CTLR_SWAP_ENABLED
+        bool "Memory Resource Controller Swap Extension enabled by default"
+        depends on CGROUP_MEM_RES_CTLR_SWAP
+        default y
+        help
+          Memory Resource Controller Swap Extension comes with its price in
+          a bigger memory consumption. General purpose distribution kernels
+          which want to enable the feautre but keep it disabled by default
+          and let the user enable it by swapaccount boot command line
+          parameter should have this option unselected.
+          For those who want to have the feature enabled by default should
+          select this option (if, for some reason, they need to disable it
+          then noswapaccount does the trick).
 
 menuconfig CGROUP_SCHED
         bool "Group CPU scheduler"
diff --git a/init/main.c b/init/main.c
index e59af24a0b7c..8646401f7a0e 100644
--- a/init/main.c
+++ b/init/main.c
@@ -20,7 +20,6 @@
 #include <linux/delay.h>
 #include <linux/ioport.h>
 #include <linux/init.h>
-#include <linux/smp_lock.h>
 #include <linux/initrd.h>
 #include <linux/bootmem.h>
 #include <linux/acpi.h>
diff --git a/kernel/debug/kdb/kdb_main.c b/kernel/debug/kdb/kdb_main.c
index 37755d621924..a6e729766821 100644
--- a/kernel/debug/kdb/kdb_main.c
+++ b/kernel/debug/kdb/kdb_main.c
@@ -82,7 +82,7 @@ static kdbtab_t kdb_base_commands[50];
 #define for_each_kdbcmd(cmd, num)                                       \
         for ((cmd) = kdb_base_commands, (num) = 0;                      \
              num < kdb_max_commands;                                    \
-             num == KDB_BASE_CMD_MAX ? cmd = kdb_commands : cmd++, num++)
+             num++, num == KDB_BASE_CMD_MAX ? cmd = kdb_commands : cmd++)
 
 typedef struct _kdbmsg {
         int     km_diag;        /* kdb diagnostic */
@@ -646,7 +646,7 @@ static int kdb_defcmd2(const char *cmdstr, const char *argv0)
         }
         if (!s->usable)
                 return KDB_NOTIMP;
-        s->command = kmalloc((s->count + 1) * sizeof(*(s->command)), GFP_KDB);
+        s->command = kzalloc((s->count + 1) * sizeof(*(s->command)), GFP_KDB);
         if (!s->command) {
                 kdb_printf("Could not allocate new kdb_defcmd table for %s\n",
                            cmdstr);
@@ -2361,7 +2361,7 @@ static int kdb_pid(int argc, const char **argv)
  */
 static int kdb_ll(int argc, const char **argv)
 {
-        int diag;
+        int diag = 0;
         unsigned long addr;
         long offset = 0;
         unsigned long va;
@@ -2400,20 +2400,21 @@ static int kdb_ll(int argc, const char **argv)
                 char buf[80];
 
                 if (KDB_FLAG(CMD_INTERRUPT))
-                        return 0;
+                        goto out;
 
                 sprintf(buf, "%s " kdb_machreg_fmt "\n", command, va);
                 diag = kdb_parse(buf);
                 if (diag)
-                        return diag;
+                        goto out;
 
                 addr = va + linkoffset;
                 if (kdb_getword(&va, addr, sizeof(va)))
-                        return 0;
+                        goto out;
         }
-        kfree(command);
 
-        return 0;
+out:
+        kfree(command);
+        return diag;
 }
 
 static int kdb_kgdb(int argc, const char **argv)
@@ -2739,13 +2740,13 @@ int kdb_register_repeat(char *cmd,
                 }
                 if (kdb_commands) {
                         memcpy(new, kdb_commands,
-                               kdb_max_commands * sizeof(*new));
+                          (kdb_max_commands - KDB_BASE_CMD_MAX) * sizeof(*new));
                         kfree(kdb_commands);
                 }
                 memset(new + kdb_max_commands, 0,
                        kdb_command_extend * sizeof(*new));
                 kdb_commands = new;
-                kp = kdb_commands + kdb_max_commands;
+                kp = kdb_commands + kdb_max_commands - KDB_BASE_CMD_MAX;
                 kdb_max_commands += kdb_command_extend;
         }
 
diff --git a/kernel/futex.c b/kernel/futex.c
index 6c683b37f2ce..40a8777a27d0 100644
--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -2489,7 +2489,8 @@ void exit_robust_list(struct task_struct *curr)
 {
         struct robust_list_head __user *head = curr->robust_list;
         struct robust_list __user *entry, *next_entry, *pending;
-        unsigned int limit = ROBUST_LIST_LIMIT, pi, next_pi, pip;
+        unsigned int limit = ROBUST_LIST_LIMIT, pi, pip;
+        unsigned int uninitialized_var(next_pi);
         unsigned long futex_offset;
         int rc;
 
diff --git a/kernel/futex_compat.c b/kernel/futex_compat.c
index 06da4dfc339b..a7934ac75e5b 100644
--- a/kernel/futex_compat.c
+++ b/kernel/futex_compat.c
@@ -49,7 +49,8 @@ void compat_exit_robust_list(struct task_struct *curr)
 {
         struct compat_robust_list_head __user *head = curr->compat_robust_list;
         struct robust_list __user *entry, *next_entry, *pending;
-        unsigned int limit = ROBUST_LIST_LIMIT, pi, next_pi, pip;
+        unsigned int limit = ROBUST_LIST_LIMIT, pi, pip;
+        unsigned int uninitialized_var(next_pi);
         compat_uptr_t uentry, next_uentry, upending;
         compat_long_t futex_offset;
         int rc;
diff --git a/kernel/pm_qos_params.c b/kernel/pm_qos_params.c
index c7a8f453919e..aeaa7f846821 100644
--- a/kernel/pm_qos_params.c
+++ b/kernel/pm_qos_params.c
@@ -121,10 +121,10 @@ static inline int pm_qos_get_value(struct pm_qos_object *o)
 
         switch (o->type) {
         case PM_QOS_MIN:
-                return plist_last(&o->requests)->prio;
+                return plist_first(&o->requests)->prio;
 
         case PM_QOS_MAX:
-                return plist_first(&o->requests)->prio;
+                return plist_last(&o->requests)->prio;
 
         default:
                 /* runtime check for not using enum */
diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig
index 29bff6117abc..a5aff3ebad38 100644
--- a/kernel/power/Kconfig
+++ b/kernel/power/Kconfig
@@ -246,9 +246,13 @@ config PM_OPS
         depends on PM_SLEEP || PM_RUNTIME
         default y
 
+config ARCH_HAS_OPP
+        bool
+
 config PM_OPP
         bool "Operating Performance Point (OPP) Layer library"
         depends on PM
+        depends on ARCH_HAS_OPP
         ---help---
           SOCs have a standard set of tuples consisting of frequency and
           voltage pairs that the device will support per voltage domain. This
diff --git a/kernel/printk.c b/kernel/printk.c
index 38e7d5868d60..9a2264fc42ca 100644
--- a/kernel/printk.c
+++ b/kernel/printk.c
@@ -274,7 +274,20 @@ int do_syslog(int type, char __user *buf, int len, bool from_file)
         char c;
         int error = 0;
 
-        error = security_syslog(type, from_file);
+        /*
+         * If this is from /proc/kmsg we only do the capabilities checks
+         * at open time.
+         */
+        if (type == SYSLOG_ACTION_OPEN || !from_file) {
+                if (dmesg_restrict && !capable(CAP_SYS_ADMIN))
+                        return -EPERM;
+                if ((type != SYSLOG_ACTION_READ_ALL &&
+                     type != SYSLOG_ACTION_SIZE_BUFFER) &&
+                    !capable(CAP_SYS_ADMIN))
+                        return -EPERM;
+        }
+
+        error = security_syslog(type);
         if (error)
                 return error;
 
diff --git a/kernel/sched.c b/kernel/sched.c
index aa14a56f9d03..dc91a4d09ac3 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -560,18 +560,8 @@ struct rq {
 
 static DEFINE_PER_CPU_SHARED_ALIGNED(struct rq, runqueues);
 
-static inline
-void check_preempt_curr(struct rq *rq, struct task_struct *p, int flags)
-{
-        rq->curr->sched_class->check_preempt_curr(rq, p, flags);
 
-        /*
-         * A queue event has occurred, and we're going to schedule.  In
-         * this case, we can save a useless back to back clock update.
-         */
-        if (test_tsk_need_resched(p))
-                rq->skip_clock_update = 1;
-}
+static void check_preempt_curr(struct rq *rq, struct task_struct *p, int flags);
 
 static inline int cpu_of(struct rq *rq)
 {
@@ -2118,6 +2108,31 @@ static inline void check_class_changed(struct rq *rq, struct task_struct *p,
                 p->sched_class->prio_changed(rq, p, oldprio, running);
 }
 
+static void check_preempt_curr(struct rq *rq, struct task_struct *p, int flags)
+{
+        const struct sched_class *class;
+
+        if (p->sched_class == rq->curr->sched_class) {
+                rq->curr->sched_class->check_preempt_curr(rq, p, flags);
+        } else {
+                for_each_class(class) {
+                        if (class == rq->curr->sched_class)
+                                break;
+                        if (class == p->sched_class) {
+                                resched_task(rq->curr);
+                                break;
+                        }
+                }
+        }
+
+        /*
+         * A queue event has occurred, and we're going to schedule.  In
+         * this case, we can save a useless back to back clock update.
+         */
+        if (test_tsk_need_resched(rq->curr))
+                rq->skip_clock_update = 1;
+}
+
 #ifdef CONFIG_SMP
 /*
  * Is this task likely cache-hot:
@@ -6960,6 +6975,8 @@ static void init_sched_groups_power(int cpu, struct sched_domain *sd)
         if (cpu != group_first_cpu(sd->groups))
                 return;
 
+        sd->groups->group_weight = cpumask_weight(sched_group_cpus(sd->groups));
+
         child = sd->child;
 
         sd->groups->cpu_power = 0;
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index f4f6a8326dd0..52ab113d8bb9 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -1654,12 +1654,6 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_
         struct cfs_rq *cfs_rq = task_cfs_rq(curr);
         int scale = cfs_rq->nr_running >= sched_nr_latency;
 
-        if (unlikely(rt_prio(p->prio)))
-                goto preempt;
-
-        if (unlikely(p->sched_class != &fair_sched_class))
-                return;
-
         if (unlikely(se == pse))
                 return;
 
@@ -2035,13 +2029,16 @@ struct sd_lb_stats {
         unsigned long this_load_per_task;
         unsigned long this_nr_running;
         unsigned long this_has_capacity;
+        unsigned int  this_idle_cpus;
 
         /* Statistics of the busiest group */
+        unsigned int  busiest_idle_cpus;
         unsigned long max_load;
         unsigned long busiest_load_per_task;
         unsigned long busiest_nr_running;
         unsigned long busiest_group_capacity;
         unsigned long busiest_has_capacity;
+        unsigned int  busiest_group_weight;
 
         int group_imb; /* Is there imbalance in this sd */
 #if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT)
@@ -2063,6 +2060,8 @@ struct sg_lb_stats {
         unsigned long sum_nr_running; /* Nr tasks running in the group */
         unsigned long sum_weighted_load; /* Weighted load of group's tasks */
         unsigned long group_capacity;
+        unsigned long idle_cpus;
+        unsigned long group_weight;
         int group_imb; /* Is there an imbalance in the group ? */
         int group_has_capacity; /* Is there extra capacity in the group? */
 };
@@ -2431,7 +2430,8 @@ static inline void update_sg_lb_stats(struct sched_domain *sd,
                 sgs->group_load += load;
                 sgs->sum_nr_running += rq->nr_running;
                 sgs->sum_weighted_load += weighted_cpuload(i);
-
+                if (idle_cpu(i))
+                        sgs->idle_cpus++;
         }
 
         /*
@@ -2469,6 +2469,7 @@ static inline void update_sg_lb_stats(struct sched_domain *sd,
         sgs->group_capacity = DIV_ROUND_CLOSEST(group->cpu_power, SCHED_LOAD_SCALE);
         if (!sgs->group_capacity)
                 sgs->group_capacity = fix_small_capacity(sd, group);
+        sgs->group_weight = group->group_weight;
 
         if (sgs->group_capacity > sgs->sum_nr_running)
                 sgs->group_has_capacity = 1;
@@ -2576,13 +2577,16 @@ static inline void update_sd_lb_stats(struct sched_domain *sd, int this_cpu,
                         sds->this_nr_running = sgs.sum_nr_running;
                         sds->this_load_per_task = sgs.sum_weighted_load;
                         sds->this_has_capacity = sgs.group_has_capacity;
+                        sds->this_idle_cpus = sgs.idle_cpus;
                 } else if (update_sd_pick_busiest(sd, sds, sg, &sgs, this_cpu)) {
                         sds->max_load = sgs.avg_load;
                         sds->busiest = sg;
                         sds->busiest_nr_running = sgs.sum_nr_running;
+                        sds->busiest_idle_cpus = sgs.idle_cpus;
                         sds->busiest_group_capacity = sgs.group_capacity;
                         sds->busiest_load_per_task = sgs.sum_weighted_load;
                         sds->busiest_has_capacity = sgs.group_has_capacity;
+                        sds->busiest_group_weight = sgs.group_weight;
                         sds->group_imb = sgs.group_imb;
                 }
 
@@ -2860,8 +2864,26 @@ find_busiest_group(struct sched_domain *sd, int this_cpu,
         if (sds.this_load >= sds.avg_load)
                 goto out_balanced;
 
-        if (100 * sds.max_load <= sd->imbalance_pct * sds.this_load)
-                goto out_balanced;
+        /*
+         * In the CPU_NEWLY_IDLE, use imbalance_pct to be conservative.
+         * And to check for busy balance use !idle_cpu instead of
+         * CPU_NOT_IDLE. This is because HT siblings will use CPU_NOT_IDLE
+         * even when they are idle.
+         */
+        if (idle == CPU_NEWLY_IDLE || !idle_cpu(this_cpu)) {
+                if (100 * sds.max_load <= sd->imbalance_pct * sds.this_load)
+                        goto out_balanced;
+        } else {
+                /*
+                 * This cpu is idle. If the busiest group load doesn't
+                 * have more tasks than the number of available cpu's and
+                 * there is no imbalance between this and busiest group
+                 * wrt to idle cpu's, it is balanced.
+                 */
+                if ((sds.this_idle_cpus  <= sds.busiest_idle_cpus + 1) &&
+                    sds.busiest_nr_running <= sds.busiest_group_weight)
+                        goto out_balanced;
+        }
 
 force_balance:
         /* Looks like there is an imbalance. Compute it */
diff --git a/kernel/sched_stoptask.c b/kernel/sched_stoptask.c
index 45bddc0c1048..2bf6b47058c1 100644
--- a/kernel/sched_stoptask.c
+++ b/kernel/sched_stoptask.c
@@ -19,14 +19,14 @@ select_task_rq_stop(struct rq *rq, struct task_struct *p,
 static void
 check_preempt_curr_stop(struct rq *rq, struct task_struct *p, int flags)
 {
-        resched_task(rq->curr); /* we preempt everything */
+        /* we're never preempted */
 }
 
 static struct task_struct *pick_next_task_stop(struct rq *rq)
 {
         struct task_struct *stop = rq->stop;
 
-        if (stop && stop->state == TASK_RUNNING)
+        if (stop && stop->se.on_rq)
                 return stop;
 
         return NULL;
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index b65bf634035e..5abfa1518554 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -702,7 +702,6 @@ static struct ctl_table kern_table[] = {
                 .extra1         = &zero,
                 .extra2         = &ten_thousand,
         },
-#endif
         {
                 .procname       = "dmesg_restrict",
                 .data           = &dmesg_restrict,
@@ -712,6 +711,7 @@ static struct ctl_table kern_table[] = {
                 .extra1         = &zero,
                 .extra2         = &one,
         },
+#endif
         {
                 .procname       = "ngroups_max",
                 .data           = &ngroups_max,
diff --git a/kernel/trace/Kconfig b/kernel/trace/Kconfig
index e04b8bcdef88..ea37e2ff4164 100644
--- a/kernel/trace/Kconfig
+++ b/kernel/trace/Kconfig
@@ -126,7 +126,7 @@ if FTRACE
 config FUNCTION_TRACER
         bool "Kernel Function Tracer"
         depends on HAVE_FUNCTION_TRACER
-        select FRAME_POINTER if (!ARM_UNWIND)
+        select FRAME_POINTER if !ARM_UNWIND && !S390
         select KALLSYMS
         select GENERIC_TRACER
         select CONTEXT_SWITCH_TRACER
diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
index 82d9b8106cd0..042084157980 100644
--- a/kernel/trace/trace.c
+++ b/kernel/trace/trace.c
@@ -17,7 +17,6 @@
 #include <linux/writeback.h>
 #include <linux/kallsyms.h>
 #include <linux/seq_file.h>
-#include <linux/smp_lock.h>
 #include <linux/notifier.h>
 #include <linux/irqflags.h>
 #include <linux/debugfs.h>
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 2efa8ea07ff7..7a22b4129211 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -61,7 +61,14 @@ struct mem_cgroup *root_mem_cgroup __read_mostly;
 #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
 /* Turned on only when memory cgroup is enabled && really_do_swap_account = 1 */
 int do_swap_account __read_mostly;
-static int really_do_swap_account __initdata = 1; /* for remember boot option*/
+
+/* for remember boot option*/
+#ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP_ENABLED
+static int really_do_swap_account __initdata = 1;
+#else
+static int really_do_swap_account __initdata = 0;
+#endif
+
 #else
 #define do_swap_account         (0)
 #endif
@@ -278,13 +285,14 @@ enum move_type {
 
 /* "mc" and its members are protected by cgroup_mutex */
 static struct move_charge_struct {
-        spinlock_t        lock; /* for from, to, moving_task */
+        spinlock_t        lock; /* for from, to */
         struct mem_cgroup *from;
         struct mem_cgroup *to;
         unsigned long precharge;
         unsigned long moved_charge;
         unsigned long moved_swap;
         struct task_struct *moving_task;        /* a task moving charges */
+        struct mm_struct *mm;
         wait_queue_head_t waitq;                /* a waitq for other context */
 } mc = {
         .lock = __SPIN_LOCK_UNLOCKED(mc.lock),
@@ -2152,7 +2160,7 @@ static void __mem_cgroup_move_account(struct page_cgroup *pc,
 {
         VM_BUG_ON(from == to);
         VM_BUG_ON(PageLRU(pc->page));
-        VM_BUG_ON(!PageCgroupLocked(pc));
+        VM_BUG_ON(!page_is_cgroup_locked(pc));
         VM_BUG_ON(!PageCgroupUsed(pc));
         VM_BUG_ON(pc->mem_cgroup != from);
 
@@ -4631,7 +4639,7 @@ static unsigned long mem_cgroup_count_precharge(struct mm_struct *mm)
         unsigned long precharge;
         struct vm_area_struct *vma;
 
-        down_read(&mm->mmap_sem);
+        /* We've already held the mmap_sem */
         for (vma = mm->mmap; vma; vma = vma->vm_next) {
                 struct mm_walk mem_cgroup_count_precharge_walk = {
                         .pmd_entry = mem_cgroup_count_precharge_pte_range,
@@ -4643,7 +4651,6 @@ static unsigned long mem_cgroup_count_precharge(struct mm_struct *mm)
                 walk_page_range(vma->vm_start, vma->vm_end,
                                         &mem_cgroup_count_precharge_walk);
         }
-        up_read(&mm->mmap_sem);
 
         precharge = mc.precharge;
         mc.precharge = 0;
@@ -4694,11 +4701,16 @@ static void mem_cgroup_clear_mc(void)
 
                 mc.moved_swap = 0;
         }
+        if (mc.mm) {
+                up_read(&mc.mm->mmap_sem);
+                mmput(mc.mm);
+        }
         spin_lock(&mc.lock);
         mc.from = NULL;
         mc.to = NULL;
-        mc.moving_task = NULL;
         spin_unlock(&mc.lock);
+        mc.moving_task = NULL;
+        mc.mm = NULL;
         mem_cgroup_end_move(from);
         memcg_oom_recover(from);
         memcg_oom_recover(to);
@@ -4724,12 +4736,21 @@ static int mem_cgroup_can_attach(struct cgroup_subsys *ss,
                         return 0;
                 /* We move charges only when we move a owner of the mm */
                 if (mm->owner == p) {
+                        /*
+                         * We do all the move charge works under one mmap_sem to
+                         * avoid deadlock with down_write(&mmap_sem)
+                         * -> try_charge() -> if (mc.moving_task) -> sleep.
+                         */
+                        down_read(&mm->mmap_sem);
+
                         VM_BUG_ON(mc.from);
                         VM_BUG_ON(mc.to);
                         VM_BUG_ON(mc.precharge);
                         VM_BUG_ON(mc.moved_charge);
                         VM_BUG_ON(mc.moved_swap);
                         VM_BUG_ON(mc.moving_task);
+                        VM_BUG_ON(mc.mm);
+
                         mem_cgroup_start_move(from);
                         spin_lock(&mc.lock);
                         mc.from = from;
@@ -4737,14 +4758,16 @@ static int mem_cgroup_can_attach(struct cgroup_subsys *ss,
                         mc.precharge = 0;
                         mc.moved_charge = 0;
                         mc.moved_swap = 0;
-                        mc.moving_task = current;
                         spin_unlock(&mc.lock);
+                        mc.moving_task = current;
+                        mc.mm = mm;
 
                         ret = mem_cgroup_precharge_mc(mm);
                         if (ret)
                                 mem_cgroup_clear_mc();
-                }
-                mmput(mm);
+                        /* We call up_read() and mmput() in clear_mc(). */
+                } else
+                        mmput(mm);
         }
         return ret;
 }
@@ -4832,7 +4855,7 @@ static void mem_cgroup_move_charge(struct mm_struct *mm)
         struct vm_area_struct *vma;
 
         lru_add_drain_all();
-        down_read(&mm->mmap_sem);
+        /* We've already held the mmap_sem */
         for (vma = mm->mmap; vma; vma = vma->vm_next) {
                 int ret;
                 struct mm_walk mem_cgroup_move_charge_walk = {
@@ -4851,7 +4874,6 @@ static void mem_cgroup_move_charge(struct mm_struct *mm)
                          */
                         break;
         }
-        up_read(&mm->mmap_sem);
 }
 
 static void mem_cgroup_move_task(struct cgroup_subsys *ss,
@@ -4860,17 +4882,11 @@ static void mem_cgroup_move_task(struct cgroup_subsys *ss,
                                 struct task_struct *p,
                                 bool threadgroup)
 {
-        struct mm_struct *mm;
-
-        if (!mc.to)
+        if (!mc.mm)
                 /* no need to move charge */
                 return;
 
-        mm = get_task_mm(p);
-        if (mm) {
-                mem_cgroup_move_charge(mm);
-                mmput(mm);
-        }
+        mem_cgroup_move_charge(mc.mm);
         mem_cgroup_clear_mc();
 }
 #else   /* !CONFIG_MMU */
@@ -4911,10 +4927,20 @@ struct cgroup_subsys mem_cgroup_subsys = {
 };
 
 #ifdef CONFIG_CGROUP_MEM_RES_CTLR_SWAP
+static int __init enable_swap_account(char *s)
+{
+        /* consider enabled if no parameter or 1 is given */
+        if (!s || !strcmp(s, "1"))
+                really_do_swap_account = 1;
+        else if (!strcmp(s, "0"))
+                really_do_swap_account = 0;
+        return 1;
+}
+__setup("swapaccount", enable_swap_account);
 
 static int __init disable_swap_account(char *s)
 {
-        really_do_swap_account = 0;
+        enable_swap_account("0");
         return 1;
 }
 __setup("noswapaccount", disable_swap_account);
diff --git a/mm/nommu.c b/mm/nommu.c
index 3613517c7592..27a9ac588516 100644
--- a/mm/nommu.c
+++ b/mm/nommu.c
@@ -1717,6 +1717,7 @@ void exit_mmap(struct mm_struct *mm)
                 mm->mmap = vma->vm_next;
                 delete_vma_from_mm(vma);
                 delete_vma(mm, vma);
+                cond_resched();
         }
 
         kleave("");
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 07a654486f75..e4092704c1a9 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -3008,14 +3008,6 @@ static __init_refok int __build_all_zonelists(void *data)
                 build_zonelist_cache(pgdat);
         }
 
-#ifdef CONFIG_MEMORY_HOTPLUG
-        /* Setup real pagesets for the new zone */
-        if (data) {
-                struct zone *zone = data;
-                setup_zone_pageset(zone);
-        }
-#endif
-
         /*
          * Initialize the boot_pagesets that are going to be used
          * for bootstrapping processors. The real pagesets for
@@ -3064,7 +3056,11 @@ void build_all_zonelists(void *data)
         } else {
                 /* we have to stop all cpus to guarantee there is no user
                    of zonelist */
-                stop_machine(__build_all_zonelists, data, NULL);
+#ifdef CONFIG_MEMORY_HOTPLUG
+                if (data)
+                        setup_zone_pageset((struct zone *)data);
+#endif
+                stop_machine(__build_all_zonelists, NULL, NULL);
                 /* cpuset refresh routine should be here */
         }
         vm_total_pages = nr_free_pagecache_pages();
diff --git a/mm/pagewalk.c b/mm/pagewalk.c
index 8b1a2ce21ee5..38cc58b8b2b0 100644
--- a/mm/pagewalk.c
+++ b/mm/pagewalk.c
@@ -139,7 +139,6 @@ int walk_page_range(unsigned long addr, unsigned long end,
         pgd_t *pgd;
         unsigned long next;
         int err = 0;
-        struct vm_area_struct *vma;
 
         if (addr >= end)
                 return err;
@@ -149,15 +148,17 @@ int walk_page_range(unsigned long addr, unsigned long end,
 
         pgd = pgd_offset(walk->mm, addr);
         do {
+                struct vm_area_struct *uninitialized_var(vma);
+
                 next = pgd_addr_end(addr, end);
 
+#ifdef CONFIG_HUGETLB_PAGE
                 /*
                  * handle hugetlb vma individually because pagetable walk for
                  * the hugetlb page is dependent on the architecture and
                  * we can't handled it in the same manner as non-huge pages.
                  */
                 vma = find_vma(walk->mm, addr);
-#ifdef CONFIG_HUGETLB_PAGE
                 if (vma && is_vm_hugetlb_page(vma)) {
                         if (vma->vm_end < next)
                                 next = vma->vm_end;
diff --git a/mm/slub.c b/mm/slub.c
index 8fd5401bb071..981fb730aa04 100644
--- a/mm/slub.c
+++ b/mm/slub.c
@@ -3273,9 +3273,9 @@ struct kmem_cache *kmem_cache_create(const char *name, size_t size,
                 kfree(n);
                 kfree(s);
         }
+err:
         up_write(&slub_lock);
 
-err:
         if (flags & SLAB_PANIC)
                 panic("Cannot create slabcache %s\n", name);
         else
@@ -3862,6 +3862,7 @@ static ssize_t show_slab_objects(struct kmem_cache *s,
                         x += sprintf(buf + x, " N%d=%lu",
                                         node, nodes[node]);
 #endif
+        up_read(&slub_lock);
         kfree(nodes);
         return x + sprintf(buf + x, "\n");
 }
diff --git a/net/ceph/buffer.c b/net/ceph/buffer.c
index 53d8abfa25d5..bf3e6a13c215 100644
--- a/net/ceph/buffer.c
+++ b/net/ceph/buffer.c
@@ -19,7 +19,7 @@ struct ceph_buffer *ceph_buffer_new(size_t len, gfp_t gfp)
         if (b->vec.iov_base) {
                 b->is_vmalloc = false;
         } else {
-                b->vec.iov_base = __vmalloc(len, gfp, PAGE_KERNEL);
+                b->vec.iov_base = __vmalloc(len, gfp | __GFP_HIGHMEM, PAGE_KERNEL);
                 if (!b->vec.iov_base) {
                         kfree(b);
                         return NULL;
diff --git a/net/ceph/messenger.c b/net/ceph/messenger.c
index 0e8157ee5d43..1c7a2ec4f3cc 100644
--- a/net/ceph/messenger.c
+++ b/net/ceph/messenger.c
@@ -540,8 +540,7 @@ static void prepare_write_message(struct ceph_connection *con)
                 /* initialize page iterator */
                 con->out_msg_pos.page = 0;
                 if (m->pages)
-                        con->out_msg_pos.page_pos =
-                                le16_to_cpu(m->hdr.data_off) & ~PAGE_MASK;
+                        con->out_msg_pos.page_pos = m->page_alignment;
                 else
                         con->out_msg_pos.page_pos = 0;
                 con->out_msg_pos.data_pos = 0;
@@ -1491,7 +1490,7 @@ static int read_partial_message(struct ceph_connection *con)
         struct ceph_msg *m = con->in_msg;
         int ret;
         int to, left;
-        unsigned front_len, middle_len, data_len, data_off;
+        unsigned front_len, middle_len, data_len;
         int datacrc = con->msgr->nocrc;
         int skip;
         u64 seq;
@@ -1527,19 +1526,17 @@ static int read_partial_message(struct ceph_connection *con)
         data_len = le32_to_cpu(con->in_hdr.data_len);
         if (data_len > CEPH_MSG_MAX_DATA_LEN)
                 return -EIO;
-        data_off = le16_to_cpu(con->in_hdr.data_off);
 
         /* verify seq# */
         seq = le64_to_cpu(con->in_hdr.seq);
         if ((s64)seq - (s64)con->in_seq < 1) {
-                pr_info("skipping %s%lld %s seq %lld, expected %lld\n",
+                pr_info("skipping %s%lld %s seq %lld expected %lld\n",
                         ENTITY_NAME(con->peer_name),
                         ceph_pr_addr(&con->peer_addr.in_addr),
                         seq, con->in_seq + 1);
                 con->in_base_pos = -front_len - middle_len - data_len -
                         sizeof(m->footer);
                 con->in_tag = CEPH_MSGR_TAG_READY;
-                con->in_seq++;
                 return 0;
         } else if ((s64)seq - (s64)con->in_seq > 1) {
                 pr_err("read_partial_message bad seq %lld expected %lld\n",
@@ -1576,7 +1573,7 @@ static int read_partial_message(struct ceph_connection *con)
 
                 con->in_msg_pos.page = 0;
                 if (m->pages)
-                        con->in_msg_pos.page_pos = data_off & ~PAGE_MASK;
+                        con->in_msg_pos.page_pos = m->page_alignment;
                 else
                         con->in_msg_pos.page_pos = 0;
                 con->in_msg_pos.data_pos = 0;
@@ -2301,6 +2298,7 @@ struct ceph_msg *ceph_msg_new(int type, int front_len, gfp_t flags)
 
         /* data */
         m->nr_pages = 0;
+        m->page_alignment = 0;
         m->pages = NULL;
         m->pagelist = NULL;
         m->bio = NULL;
@@ -2370,6 +2368,7 @@ static struct ceph_msg *ceph_alloc_msg(struct ceph_connection *con,
                                type, front_len);
                         return NULL;
                 }
+                msg->page_alignment = le16_to_cpu(hdr->data_off);
         }
         memcpy(&msg->hdr, &con->in_hdr, sizeof(con->in_hdr));
 
diff --git a/net/ceph/osd_client.c b/net/ceph/osd_client.c
index 79391994b3ed..3e20a122ffa2 100644
--- a/net/ceph/osd_client.c
+++ b/net/ceph/osd_client.c
@@ -71,6 +71,7 @@ void ceph_calc_raw_layout(struct ceph_osd_client *osdc,
                 op->extent.length = objlen;
         }
         req->r_num_pages = calc_pages_for(off, *plen);
+        req->r_page_alignment = off & ~PAGE_MASK;
         if (op->op == CEPH_OSD_OP_WRITE)
                 op->payload_len = *plen;
 
@@ -390,6 +391,8 @@ void ceph_osdc_build_request(struct ceph_osd_request *req,
                 req->r_request->hdr.data_len = cpu_to_le32(data_len);
         }
 
+        req->r_request->page_alignment = req->r_page_alignment;
+
         BUG_ON(p > msg->front.iov_base + msg->front.iov_len);
         msg_size = p - msg->front.iov_base;
         msg->front.iov_len = msg_size;
@@ -419,7 +422,8 @@ struct ceph_osd_request *ceph_osdc_new_request(struct ceph_osd_client *osdc,
                                                u32 truncate_seq,
                                                u64 truncate_size,
                                                struct timespec *mtime,
-                                               bool use_mempool, int num_reply)
+                                               bool use_mempool, int num_reply,
+                                               int page_align)
 {
         struct ceph_osd_req_op ops[3];
         struct ceph_osd_request *req;
@@ -447,6 +451,10 @@ struct ceph_osd_request *ceph_osdc_new_request(struct ceph_osd_client *osdc,
         calc_layout(osdc, vino, layout, off, plen, req, ops);
         req->r_file_layout = *layout;  /* keep a copy */
 
+        /* in case it differs from natural alignment that calc_layout
+           filled in for us */
+        req->r_page_alignment = page_align;
+
         ceph_osdc_build_request(req, off, plen, ops,
                                 snapc,
                                 mtime,
@@ -1489,7 +1497,7 @@ int ceph_osdc_readpages(struct ceph_osd_client *osdc,
                         struct ceph_vino vino, struct ceph_file_layout *layout,
                         u64 off, u64 *plen,
                         u32 truncate_seq, u64 truncate_size,
-                        struct page **pages, int num_pages)
+                        struct page **pages, int num_pages, int page_align)
 {
         struct ceph_osd_request *req;
         int rc = 0;
@@ -1499,15 +1507,15 @@ int ceph_osdc_readpages(struct ceph_osd_client *osdc,
         req = ceph_osdc_new_request(osdc, layout, vino, off, plen,
                                     CEPH_OSD_OP_READ, CEPH_OSD_FLAG_READ,
                                     NULL, 0, truncate_seq, truncate_size, NULL,
-                                    false, 1);
+                                    false, 1, page_align);
         if (!req)
                 return -ENOMEM;
 
         /* it may be a short read due to an object boundary */
         req->r_pages = pages;
 
-        dout("readpages  final extent is %llu~%llu (%d pages)\n",
-             off, *plen, req->r_num_pages);
+        dout("readpages  final extent is %llu~%llu (%d pages align %d)\n",
+             off, *plen, req->r_num_pages, page_align);
 
         rc = ceph_osdc_start_request(osdc, req, false);
         if (!rc)
@@ -1533,6 +1541,7 @@ int ceph_osdc_writepages(struct ceph_osd_client *osdc, struct ceph_vino vino,
 {
         struct ceph_osd_request *req;
         int rc = 0;
+        int page_align = off & ~PAGE_MASK;
 
         BUG_ON(vino.snap != CEPH_NOSNAP);
         req = ceph_osdc_new_request(osdc, layout, vino, off, &len,
@@ -1541,7 +1550,7 @@ int ceph_osdc_writepages(struct ceph_osd_client *osdc, struct ceph_vino vino,
                                             CEPH_OSD_FLAG_WRITE,
                                     snapc, do_sync,
                                     truncate_seq, truncate_size, mtime,
-                                    nofail, 1);
+                                    nofail, 1, page_align);
         if (!req)
                 return -ENOMEM;
 
@@ -1638,8 +1647,7 @@ static struct ceph_msg *get_reply(struct ceph_connection *con,
         m = ceph_msg_get(req->r_reply);
 
         if (data_len > 0) {
-                unsigned data_off = le16_to_cpu(hdr->data_off);
-                int want = calc_pages_for(data_off & ~PAGE_MASK, data_len);
+                int want = calc_pages_for(req->r_page_alignment, data_len);
 
                 if (unlikely(req->r_num_pages < want)) {
                         pr_warning("tid %lld reply %d > expected %d pages\n",
@@ -1651,6 +1659,7 @@ static struct ceph_msg *get_reply(struct ceph_connection *con,
                 }
                 m->pages = req->r_pages;
                 m->nr_pages = req->r_num_pages;
+                m->page_alignment = req->r_page_alignment;
 #ifdef CONFIG_BLOCK
                 m->bio = req->r_bio;
 #endif
diff --git a/net/ceph/pagevec.c b/net/ceph/pagevec.c
index 54caf0687155..ac34feeb2b3a 100644
--- a/net/ceph/pagevec.c
+++ b/net/ceph/pagevec.c
@@ -13,8 +13,7 @@
  * build a vector of user pages
  */
 struct page **ceph_get_direct_page_vector(const char __user *data,
-                                                 int num_pages,
-                                                 loff_t off, size_t len)
+                                          int num_pages)
 {
         struct page **pages;
         int rc;
diff --git a/net/core/filter.c b/net/core/filter.c
index 23e9b2a6b4c8..c1ee800bc080 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -589,7 +589,7 @@ int sk_chk_filter(struct sock_filter *filter, int flen)
 EXPORT_SYMBOL(sk_chk_filter);
 
 /**
- *      sk_filter_rcu_release: Release a socket filter by rcu_head
+ *      sk_filter_rcu_release - Release a socket filter by rcu_head
  *      @rcu: rcu_head that contains the sk_filter to free
  */
 static void sk_filter_rcu_release(struct rcu_head *rcu)
diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
index a5ff5a89f376..7f902cad10f8 100644
--- a/net/core/net-sysfs.c
+++ b/net/core/net-sysfs.c
@@ -712,15 +712,21 @@ static void rx_queue_release(struct kobject *kobj)
 
 
         map = rcu_dereference_raw(queue->rps_map);
-        if (map)
+        if (map) {
+                RCU_INIT_POINTER(queue->rps_map, NULL);
                 call_rcu(&map->rcu, rps_map_release);
+        }
 
         flow_table = rcu_dereference_raw(queue->rps_flow_table);
-        if (flow_table)
+        if (flow_table) {
+                RCU_INIT_POINTER(queue->rps_flow_table, NULL);
                 call_rcu(&flow_table->rcu, rps_dev_flow_table_release);
+        }
 
         if (atomic_dec_and_test(&first->count))
                 kfree(first);
+        else
+                memset(kobj, 0, sizeof(*kobj));
 }
 
 static struct kobj_type rx_queue_ktype = {
diff --git a/net/core/request_sock.c b/net/core/request_sock.c
index 7552495aff7a..fceeb37d7161 100644
--- a/net/core/request_sock.c
+++ b/net/core/request_sock.c
@@ -45,9 +45,7 @@ int reqsk_queue_alloc(struct request_sock_queue *queue,
         nr_table_entries = roundup_pow_of_two(nr_table_entries + 1);
         lopt_size += nr_table_entries * sizeof(struct request_sock *);
         if (lopt_size > PAGE_SIZE)
-                lopt = __vmalloc(lopt_size,
-                        GFP_KERNEL | __GFP_HIGHMEM | __GFP_ZERO,
-                        PAGE_KERNEL);
+                lopt = vzalloc(lopt_size);
         else
                 lopt = kzalloc(lopt_size, GFP_KERNEL);
         if (lopt == NULL)
diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c
index 200eb538fbb3..0f280348e0fd 100644
--- a/net/ipv4/fib_trie.c
+++ b/net/ipv4/fib_trie.c
@@ -365,7 +365,7 @@ static struct tnode *tnode_alloc(size_t size)
         if (size <= PAGE_SIZE)
                 return kzalloc(size, GFP_KERNEL);
         else
-                return __vmalloc(size, GFP_KERNEL | __GFP_ZERO, PAGE_KERNEL);
+                return vzalloc(size);
 }
 
 static void __tnode_vfree(struct work_struct *arg)
diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c
index 96bc7f9475a3..e5d1a44bcbdf 100644
--- a/net/ipv4/icmp.c
+++ b/net/ipv4/icmp.c
@@ -569,6 +569,9 @@ void icmp_send(struct sk_buff *skb_in, int type, int code, __be32 info)
                 /* No need to clone since we're just using its address. */
                 rt2 = rt;
 
+                if (!fl.nl_u.ip4_u.saddr)
+                        fl.nl_u.ip4_u.saddr = rt->rt_src;
+
                 err = xfrm_lookup(net, (struct dst_entry **)&rt, &fl, NULL, 0);
                 switch (err) {
                 case 0:
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index b41ce0f0d514..23cc8e1ce8d4 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -98,7 +98,11 @@
 #endif
 
 #define INFINITY_LIFE_TIME      0xFFFFFFFF
-#define TIME_DELTA(a, b) ((unsigned long)((long)(a) - (long)(b)))
+
+static inline u32 cstamp_delta(unsigned long cstamp)
+{
+        return (cstamp - INITIAL_JIFFIES) * 100UL / HZ;
+}
 
 #define ADDRCONF_TIMER_FUZZ_MINUS       (HZ > 50 ? HZ/50 : 1)
 #define ADDRCONF_TIMER_FUZZ             (HZ / 4)
@@ -2754,13 +2758,13 @@ static int addrconf_ifdown(struct net_device *dev, int how)
                         ifa->state = INET6_IFADDR_STATE_DEAD;
                         spin_unlock_bh(&ifa->state_lock);
 
-                        if (state == INET6_IFADDR_STATE_DEAD) {
-                                in6_ifa_put(ifa);
-                        } else {
+                        if (state != INET6_IFADDR_STATE_DEAD) {
                                 __ipv6_ifa_notify(RTM_DELADDR, ifa);
                                 atomic_notifier_call_chain(&inet6addr_chain,
                                                            NETDEV_DOWN, ifa);
                         }
+
+                        in6_ifa_put(ifa);
                         write_lock_bh(&idev->lock);
                 }
         }
@@ -3444,10 +3448,8 @@ static int put_cacheinfo(struct sk_buff *skb, unsigned long cstamp,
 {
         struct ifa_cacheinfo ci;
 
-        ci.cstamp = (u32)(TIME_DELTA(cstamp, INITIAL_JIFFIES) / HZ * 100
-                        + TIME_DELTA(cstamp, INITIAL_JIFFIES) % HZ * 100 / HZ);
-        ci.tstamp = (u32)(TIME_DELTA(tstamp, INITIAL_JIFFIES) / HZ * 100
-                        + TIME_DELTA(tstamp, INITIAL_JIFFIES) % HZ * 100 / HZ);
+        ci.cstamp = cstamp_delta(cstamp);
+        ci.tstamp = cstamp_delta(tstamp);
         ci.ifa_prefered = preferred;
         ci.ifa_valid = valid;
 
@@ -3798,8 +3800,10 @@ static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
         array[DEVCONF_AUTOCONF] = cnf->autoconf;
         array[DEVCONF_DAD_TRANSMITS] = cnf->dad_transmits;
         array[DEVCONF_RTR_SOLICITS] = cnf->rtr_solicits;
-        array[DEVCONF_RTR_SOLICIT_INTERVAL] = cnf->rtr_solicit_interval;
-        array[DEVCONF_RTR_SOLICIT_DELAY] = cnf->rtr_solicit_delay;
+        array[DEVCONF_RTR_SOLICIT_INTERVAL] =
+                jiffies_to_msecs(cnf->rtr_solicit_interval);
+        array[DEVCONF_RTR_SOLICIT_DELAY] =
+                jiffies_to_msecs(cnf->rtr_solicit_delay);
         array[DEVCONF_FORCE_MLD_VERSION] = cnf->force_mld_version;
 #ifdef CONFIG_IPV6_PRIVACY
         array[DEVCONF_USE_TEMPADDR] = cnf->use_tempaddr;
@@ -3813,7 +3817,8 @@ static inline void ipv6_store_devconf(struct ipv6_devconf *cnf,
         array[DEVCONF_ACCEPT_RA_PINFO] = cnf->accept_ra_pinfo;
 #ifdef CONFIG_IPV6_ROUTER_PREF
         array[DEVCONF_ACCEPT_RA_RTR_PREF] = cnf->accept_ra_rtr_pref;
-        array[DEVCONF_RTR_PROBE_INTERVAL] = cnf->rtr_probe_interval;
+        array[DEVCONF_RTR_PROBE_INTERVAL] =
+                jiffies_to_msecs(cnf->rtr_probe_interval);
 #ifdef CONFIG_IPV6_ROUTE_INFO
         array[DEVCONF_ACCEPT_RA_RT_INFO_MAX_PLEN] = cnf->accept_ra_rt_info_max_plen;
 #endif
@@ -3929,10 +3934,9 @@ static int inet6_fill_ifinfo(struct sk_buff *skb, struct inet6_dev *idev,
         NLA_PUT_U32(skb, IFLA_INET6_FLAGS, idev->if_flags);
 
         ci.max_reasm_len = IPV6_MAXPLEN;
-        ci.tstamp = (__u32)(TIME_DELTA(idev->tstamp, INITIAL_JIFFIES) / HZ * 100
-                    + TIME_DELTA(idev->tstamp, INITIAL_JIFFIES) % HZ * 100 / HZ);
-        ci.reachable_time = idev->nd_parms->reachable_time;
-        ci.retrans_time = idev->nd_parms->retrans_time;
+        ci.tstamp = cstamp_delta(idev->tstamp);
+        ci.reachable_time = jiffies_to_msecs(idev->nd_parms->reachable_time);
+        ci.retrans_time = jiffies_to_msecs(idev->nd_parms->retrans_time);
         NLA_PUT(skb, IFLA_INET6_CACHEINFO, sizeof(ci), &ci);
 
         nla = nla_reserve(skb, IFLA_INET6_CONF, DEVCONF_MAX * sizeof(s32));
diff --git a/net/irda/af_irda.c b/net/irda/af_irda.c
index 7f097989cde2..a6de3059746d 100644
--- a/net/irda/af_irda.c
+++ b/net/irda/af_irda.c
@@ -45,7 +45,6 @@
 #include <linux/capability.h>
 #include <linux/module.h>
 #include <linux/types.h>
-#include <linux/smp_lock.h>
 #include <linux/socket.h>
 #include <linux/sockios.h>
 #include <linux/slab.h>
diff --git a/net/irda/irnet/irnet_ppp.c b/net/irda/irnet/irnet_ppp.c
index 7fa86373de41..7c567b8aa89a 100644
--- a/net/irda/irnet/irnet_ppp.c
+++ b/net/irda/irnet/irnet_ppp.c
@@ -15,7 +15,6 @@
 
 #include <linux/sched.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include "irnet_ppp.h"          /* Private header */
 /* Please put other headers in irnet.h - Thanks */
 
diff --git a/net/irda/irttp.c b/net/irda/irttp.c
index 285761e77d90..f6054f9ccbe3 100644
--- a/net/irda/irttp.c
+++ b/net/irda/irttp.c
@@ -550,22 +550,30 @@ EXPORT_SYMBOL(irttp_close_tsap);
  */
 int irttp_udata_request(struct tsap_cb *self, struct sk_buff *skb)
 {
+        int ret;
+
         IRDA_ASSERT(self != NULL, return -1;);
         IRDA_ASSERT(self->magic == TTP_TSAP_MAGIC, return -1;);
         IRDA_ASSERT(skb != NULL, return -1;);
 
         IRDA_DEBUG(4, "%s()\n", __func__);
 
+        /* Take shortcut on zero byte packets */
+        if (skb->len == 0) {
+                ret = 0;
+                goto err;
+        }
+
         /* Check that nothing bad happens */
-        if ((skb->len == 0) || (!self->connected)) {
-                IRDA_DEBUG(1, "%s(), No data, or not connected\n",
-                           __func__);
+        if (!self->connected) {
+                IRDA_WARNING("%s(), Not connected\n", __func__);
+                ret = -ENOTCONN;
                 goto err;
         }
 
         if (skb->len > self->max_seg_size) {
-                IRDA_DEBUG(1, "%s(), UData is too large for IrLAP!\n",
-                           __func__);
+                IRDA_ERROR("%s(), UData is too large for IrLAP!\n", __func__);
+                ret = -EMSGSIZE;
                 goto err;
         }
 
@@ -576,7 +584,7 @@ int irttp_udata_request(struct tsap_cb *self, struct sk_buff *skb)
 
 err:
         dev_kfree_skb(skb);
-        return -1;
+        return ret;
 }
 EXPORT_SYMBOL(irttp_udata_request);
 
@@ -599,9 +607,15 @@ int irttp_data_request(struct tsap_cb *self, struct sk_buff *skb)
         IRDA_DEBUG(2, "%s() : queue len = %d\n", __func__,
                    skb_queue_len(&self->tx_queue));
 
+        /* Take shortcut on zero byte packets */
+        if (skb->len == 0) {
+                ret = 0;
+                goto err;
+        }
+
         /* Check that nothing bad happens */
-        if ((skb->len == 0) || (!self->connected)) {
-                IRDA_WARNING("%s: No data, or not connected\n", __func__);
+        if (!self->connected) {
+                IRDA_WARNING("%s: Not connected\n", __func__);
                 ret = -ENOTCONN;
                 goto err;
         }
diff --git a/net/netfilter/ipvs/Kconfig b/net/netfilter/ipvs/Kconfig
index a22dac227055..70bd1d0774c6 100644
--- a/net/netfilter/ipvs/Kconfig
+++ b/net/netfilter/ipvs/Kconfig
@@ -4,6 +4,7 @@
 menuconfig IP_VS
         tristate "IP virtual server support"
         depends on NET && INET && NETFILTER
+        depends on (NF_CONNTRACK || NF_CONNTRACK=n)
         ---help---
           IP Virtual Server support will let you build a high-performance
           virtual server based on cluster of two or more real servers. This
diff --git a/net/rds/rdma.c b/net/rds/rdma.c
index 8920f2a83327..4e37c1cbe8b2 100644
--- a/net/rds/rdma.c
+++ b/net/rds/rdma.c
@@ -567,7 +567,7 @@ int rds_cmsg_rdma_args(struct rds_sock *rs, struct rds_message *rm,
                 goto out;
         }
 
-        if (args->nr_local > (u64)UINT_MAX) {
+        if (args->nr_local > UIO_MAXIOV) {
                 ret = -EMSGSIZE;
                 goto out;
         }
diff --git a/net/sunrpc/stats.c b/net/sunrpc/stats.c
index f71a73107ae9..80df89d957ba 100644
--- a/net/sunrpc/stats.c
+++ b/net/sunrpc/stats.c
@@ -115,9 +115,7 @@ EXPORT_SYMBOL_GPL(svc_seq_show);
  */
 struct rpc_iostats *rpc_alloc_iostats(struct rpc_clnt *clnt)
 {
-        struct rpc_iostats *new;
-        new = kcalloc(clnt->cl_maxproc, sizeof(struct rpc_iostats), GFP_KERNEL);
-        return new;
+        return kcalloc(clnt->cl_maxproc, sizeof(struct rpc_iostats), GFP_KERNEL);
 }
 EXPORT_SYMBOL_GPL(rpc_alloc_iostats);
 
diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c
index c82fe739fbdc..ea2ff78dcf7b 100644
--- a/net/sunrpc/svc_xprt.c
+++ b/net/sunrpc/svc_xprt.c
@@ -5,7 +5,6 @@
  */
 
 #include <linux/sched.h>
-#include <linux/smp_lock.h>
 #include <linux/errno.h>
 #include <linux/freezer.h>
 #include <linux/kthread.h>
diff --git a/net/wireless/chan.c b/net/wireless/chan.c
index d0c92dddb26b..17cd0c04d139 100644
--- a/net/wireless/chan.c
+++ b/net/wireless/chan.c
@@ -44,6 +44,38 @@ rdev_freq_to_chan(struct cfg80211_registered_device *rdev,
         return chan;
 }
 
+static bool can_beacon_sec_chan(struct wiphy *wiphy,
+                                struct ieee80211_channel *chan,
+                                enum nl80211_channel_type channel_type)
+{
+        struct ieee80211_channel *sec_chan;
+        int diff;
+
+        switch (channel_type) {
+        case NL80211_CHAN_HT40PLUS:
+                diff = 20;
+                break;
+        case NL80211_CHAN_HT40MINUS:
+                diff = -20;
+                break;
+        default:
+                return false;
+        }
+
+        sec_chan = ieee80211_get_channel(wiphy, chan->center_freq + diff);
+        if (!sec_chan)
+                return false;
+
+        /* we'll need a DFS capability later */
+        if (sec_chan->flags & (IEEE80211_CHAN_DISABLED |
+                               IEEE80211_CHAN_PASSIVE_SCAN |
+                               IEEE80211_CHAN_NO_IBSS |
+                               IEEE80211_CHAN_RADAR))
+                return false;
+
+        return true;
+}
+
 int cfg80211_set_freq(struct cfg80211_registered_device *rdev,
                       struct wireless_dev *wdev, int freq,
                       enum nl80211_channel_type channel_type)
@@ -68,6 +100,28 @@ int cfg80211_set_freq(struct cfg80211_registered_device *rdev,
         if (!chan)
                 return -EINVAL;
 
+        /* Both channels should be able to initiate communication */
+        if (wdev && (wdev->iftype == NL80211_IFTYPE_ADHOC ||
+                     wdev->iftype == NL80211_IFTYPE_AP ||
+                     wdev->iftype == NL80211_IFTYPE_AP_VLAN ||
+                     wdev->iftype == NL80211_IFTYPE_MESH_POINT ||
+                     wdev->iftype == NL80211_IFTYPE_P2P_GO)) {
+                switch (channel_type) {
+                case NL80211_CHAN_HT40PLUS:
+                case NL80211_CHAN_HT40MINUS:
+                        if (!can_beacon_sec_chan(&rdev->wiphy, chan,
+                                                 channel_type)) {
+                                printk(KERN_DEBUG
+                                       "cfg80211: Secondary channel not "
+                                       "allowed to initiate communication\n");
+                                return -EINVAL;
+                        }
+                        break;
+                default:
+                        break;
+                }
+        }
+
         result = rdev->ops->set_channel(&rdev->wiphy,
                                         wdev ? wdev->netdev : NULL,
                                         chan, channel_type);
diff --git a/net/xfrm/xfrm_hash.c b/net/xfrm/xfrm_hash.c
index a2023ec52329..1e98bc0fe0a5 100644
--- a/net/xfrm/xfrm_hash.c
+++ b/net/xfrm/xfrm_hash.c
@@ -19,7 +19,7 @@ struct hlist_head *xfrm_hash_alloc(unsigned int sz)
         if (sz <= PAGE_SIZE)
                 n = kzalloc(sz, GFP_KERNEL);
         else if (hashdist)
-                n = __vmalloc(sz, GFP_KERNEL | __GFP_ZERO, PAGE_KERNEL);
+                n = vzalloc(sz);
         else
                 n = (struct hlist_head *)
                         __get_free_pages(GFP_KERNEL | __GFP_NOWARN | __GFP_ZERO,
diff --git a/scripts/gfp-translate b/scripts/gfp-translate
index d81b968d864e..c9230e158a8f 100644
--- a/scripts/gfp-translate
+++ b/scripts/gfp-translate
@@ -63,7 +63,12 @@ fi
 
 # Extract GFP flags from the kernel source
 TMPFILE=`mktemp -t gfptranslate-XXXXXX` || exit 1
-grep "^#define __GFP" $SOURCE/include/linux/gfp.h | sed -e 's/(__force gfp_t)//' | sed -e 's/u)/)/' | grep -v GFP_BITS | sed -e 's/)\//) \//' > $TMPFILE
+grep -q ___GFP $SOURCE/include/linux/gfp.h
+if [ $? -eq 0 ]; then
+        grep "^#define ___GFP" $SOURCE/include/linux/gfp.h | sed -e 's/u$//' | grep -v GFP_BITS > $TMPFILE
+else
+        grep "^#define __GFP" $SOURCE/include/linux/gfp.h | sed -e 's/(__force gfp_t)//' | sed -e 's/u)/)/' | grep -v GFP_BITS | sed -e 's/)\//) \//' > $TMPFILE
+fi
 
 # Parse the flags
 IFS="
diff --git a/scripts/kernel-doc b/scripts/kernel-doc
index cdb6dc1f6458..39580a5dc5df 100755
--- a/scripts/kernel-doc
+++ b/scripts/kernel-doc
@@ -5,7 +5,7 @@ use strict;
 ## Copyright (c) 1998 Michael Zucchi, All Rights Reserved        ##
 ## Copyright (C) 2000, 1  Tim Waugh <twaugh@redhat.com>          ##
 ## Copyright (C) 2001  Simon Huggins                             ##
-## Copyright (C) 2005-2009  Randy Dunlap                         ##
+## Copyright (C) 2005-2010  Randy Dunlap                         ##
 ##                                                               ##
 ## #define enhancements by Armin Kuster <akuster@mvista.com>     ##
 ## Copyright (c) 2000 MontaVista Software, Inc.                  ##
@@ -453,7 +453,7 @@ sub output_highlight {
     if ($output_mode eq "html" || $output_mode eq "xml") {
         $contents = local_unescape($contents);
         # convert data read & converted thru xml_escape() into &xyz; format:
-        $contents =~ s/\\\\\\/&/g;
+        $contents =~ s/\\\\\\/\&/g;
     }
 #   print STDERR "contents b4:$contents\n";
     eval $dohighlight;
@@ -770,7 +770,11 @@ sub output_struct_xml(%) {
     print $args{'type'} . " " . $args{'struct'} . " {\n";
     foreach $parameter (@{$args{'parameterlist'}}) {
         if ($parameter =~ /^#/) {
-            print "$parameter\n";
+            my $prm = $parameter;
+            # convert data read & converted thru xml_escape() into &xyz; format:
+            # This allows us to have #define macros interspersed in a struct.
+            $prm =~ s/\\\\\\/\&/g;
+            print "$prm\n";
             next;
         }
 
@@ -1701,6 +1705,8 @@ sub push_parameter($$$) {
         }
         }
 
+        $param = xml_escape($param);
+
         # strip spaces from $param so that it is one continous string
         # on @parameterlist;
         # this fixes a problem where check_sections() cannot find
diff --git a/security/capability.c b/security/capability.c
index 30ae00fbecd5..c773635ca3a0 100644
--- a/security/capability.c
+++ b/security/capability.c
@@ -17,6 +17,11 @@ static int cap_sysctl(ctl_table *table, int op)
         return 0;
 }
 
+static int cap_syslog(int type)
+{
+        return 0;
+}
+
 static int cap_quotactl(int cmds, int type, int id, struct super_block *sb)
 {
         return 0;
diff --git a/security/commoncap.c b/security/commoncap.c
index 04b80f9912bf..64c2ed9c9015 100644
--- a/security/commoncap.c
+++ b/security/commoncap.c
@@ -27,7 +27,6 @@
 #include <linux/sched.h>
 #include <linux/prctl.h>
 #include <linux/securebits.h>
-#include <linux/syslog.h>
 
 /*
  * If a non-root user executes a setuid-root binary in
@@ -884,26 +883,6 @@ error:
 }
 
 /**
- * cap_syslog - Determine whether syslog function is permitted
- * @type: Function requested
- * @from_file: Whether this request came from an open file (i.e. /proc)
- *
- * Determine whether the current process is permitted to use a particular
- * syslog function, returning 0 if permission is granted, -ve if not.
- */
-int cap_syslog(int type, bool from_file)
-{
-        if (type != SYSLOG_ACTION_OPEN && from_file)
-                return 0;
-        if (dmesg_restrict && !capable(CAP_SYS_ADMIN))
-                return -EPERM;
-        if ((type != SYSLOG_ACTION_READ_ALL &&
-             type != SYSLOG_ACTION_SIZE_BUFFER) && !capable(CAP_SYS_ADMIN))
-                return -EPERM;
-        return 0;
-}
-
-/**
  * cap_vm_enough_memory - Determine whether a new virtual mapping is permitted
  * @mm: The VM space in which the new mapping is to be made
  * @pages: The size of the mapping
diff --git a/security/security.c b/security/security.c
index 3ef5e2a7a741..1b798d3df710 100644
--- a/security/security.c
+++ b/security/security.c
@@ -197,9 +197,9 @@ int security_quota_on(struct dentry *dentry)
         return security_ops->quota_on(dentry);
 }
 
-int security_syslog(int type, bool from_file)
+int security_syslog(int type)
 {
-        return security_ops->syslog(type, from_file);
+        return security_ops->syslog(type);
 }
 
 int security_settime(struct timespec *ts, struct timezone *tz)
diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
index d9154cf90ae1..65fa8bf596f5 100644
--- a/security/selinux/hooks.c
+++ b/security/selinux/hooks.c
@@ -1973,14 +1973,10 @@ static int selinux_quota_on(struct dentry *dentry)
         return dentry_has_perm(cred, NULL, dentry, FILE__QUOTAON);
 }
 
-static int selinux_syslog(int type, bool from_file)
+static int selinux_syslog(int type)
 {
         int rc;
 
-        rc = cap_syslog(type, from_file);
-        if (rc)
-                return rc;
-
         switch (type) {
         case SYSLOG_ACTION_READ_ALL:    /* Read last kernel messages */
         case SYSLOG_ACTION_SIZE_BUFFER: /* Return size of the log buffer */
diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c
index bc39f4067af6..489a85afa477 100644
--- a/security/smack/smack_lsm.c
+++ b/security/smack/smack_lsm.c
@@ -157,15 +157,11 @@ static int smack_ptrace_traceme(struct task_struct *ptp)
  *
  * Returns 0 on success, error code otherwise.
  */
-static int smack_syslog(int type, bool from_file)
+static int smack_syslog(int typefrom_file)
 {
-        int rc;
+        int rc = 0;
         char *sp = current_security();
 
-        rc = cap_syslog(type, from_file);
-        if (rc != 0)
-                return rc;
-
         if (capable(CAP_MAC_OVERRIDE))
                 return 0;
 
diff --git a/sound/atmel/abdac.c b/sound/atmel/abdac.c
index f2f41c854221..6e2409181895 100644
--- a/sound/atmel/abdac.c
+++ b/sound/atmel/abdac.c
@@ -420,9 +420,9 @@ static int __devinit atmel_abdac_probe(struct platform_device *pdev)
                 return PTR_ERR(pclk);
         }
         sample_clk = clk_get(&pdev->dev, "sample_clk");
-        if (IS_ERR(pclk)) {
+        if (IS_ERR(sample_clk)) {
                 dev_dbg(&pdev->dev, "no sample clock\n");
-                retval = PTR_ERR(pclk);
+                retval = PTR_ERR(sample_clk);
                 goto out_put_pclk;
         }
         clk_enable(pclk);
diff --git a/sound/core/info.c b/sound/core/info.c
index b70564ed8b37..7077f601da5a 100644
--- a/sound/core/info.c
+++ b/sound/core/info.c
@@ -23,7 +23,6 @@
 #include <linux/time.h>
 #include <linux/mm.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/string.h>
 #include <sound/core.h>
 #include <sound/minors.h>
diff --git a/sound/core/pcm_lib.c b/sound/core/pcm_lib.c
index a1707cca9c66..b75db8e9cc0f 100644
--- a/sound/core/pcm_lib.c
+++ b/sound/core/pcm_lib.c
@@ -223,7 +223,7 @@ static void xrun_log(struct snd_pcm_substream *substream,
         entry->jiffies = jiffies;
         entry->pos = pos;
         entry->period_size = runtime->period_size;
-        entry->buffer_size = runtime->buffer_size;;
+        entry->buffer_size = runtime->buffer_size;
         entry->old_hw_ptr = runtime->status->hw_ptr;
         entry->hw_ptr_base = runtime->hw_ptr_base;
         log->idx = (log->idx + 1) % XRUN_LOG_CNT;
diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c
index 8bc7cb3db330..e82c1f97d99e 100644
--- a/sound/core/pcm_native.c
+++ b/sound/core/pcm_native.c
@@ -22,7 +22,6 @@
 #include <linux/mm.h>
 #include <linux/file.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/time.h>
 #include <linux/pm_qos_params.h>
 #include <linux/uio.h>
diff --git a/sound/core/sound.c b/sound/core/sound.c
index 62a093efb453..66691fe437e6 100644
--- a/sound/core/sound.c
+++ b/sound/core/sound.c
@@ -21,7 +21,6 @@
 
 #include <linux/init.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/time.h>
 #include <linux/device.h>
 #include <linux/moduleparam.h>
diff --git a/sound/oss/dev_table.c b/sound/oss/dev_table.c
index 727bdb9ba2dc..d8cf3e58dc76 100644
--- a/sound/oss/dev_table.c
+++ b/sound/oss/dev_table.c
@@ -71,7 +71,7 @@ int sound_install_audiodrv(int vers, char *name, struct audio_driver *driver,
         if (sound_nblocks >= MAX_MEM_BLOCKS)
                 sound_nblocks = MAX_MEM_BLOCKS - 1;
 
-        op = (struct audio_operations *) (sound_mem_blocks[sound_nblocks] = vmalloc(sizeof(struct audio_operations)));
+        op = (struct audio_operations *) (sound_mem_blocks[sound_nblocks] = vzalloc(sizeof(struct audio_operations)));
         sound_nblocks++;
         if (sound_nblocks >= MAX_MEM_BLOCKS)
                 sound_nblocks = MAX_MEM_BLOCKS - 1;
@@ -81,7 +81,6 @@ int sound_install_audiodrv(int vers, char *name, struct audio_driver *driver,
                 sound_unload_audiodev(num);
                 return -(ENOMEM);
         }
-        memset((char *) op, 0, sizeof(struct audio_operations));
         init_waitqueue_head(&op->in_sleeper);
         init_waitqueue_head(&op->out_sleeper);  
         init_waitqueue_head(&op->poll_sleeper);
@@ -128,7 +127,7 @@ int sound_install_mixer(int vers, char *name, struct mixer_operations *driver,
         /* FIXME: This leaks a mixer_operations struct every time its called
            until you unload sound! */
            
-        op = (struct mixer_operations *) (sound_mem_blocks[sound_nblocks] = vmalloc(sizeof(struct mixer_operations)));
+        op = (struct mixer_operations *) (sound_mem_blocks[sound_nblocks] = vzalloc(sizeof(struct mixer_operations)));
         sound_nblocks++;
         if (sound_nblocks >= MAX_MEM_BLOCKS)
                 sound_nblocks = MAX_MEM_BLOCKS - 1;
@@ -137,7 +136,6 @@ int sound_install_mixer(int vers, char *name, struct mixer_operations *driver,
                 printk(KERN_ERR "Sound: Can't allocate mixer driver for (%s)\n", name);
                 return -ENOMEM;
         }
-        memset((char *) op, 0, sizeof(struct mixer_operations));
         memcpy((char *) op, (char *) driver, driver_size);
 
         strlcpy(op->name, name, sizeof(op->name));
diff --git a/sound/oss/midibuf.c b/sound/oss/midibuf.c
index 782b3b84dac6..ceedb1eff203 100644
--- a/sound/oss/midibuf.c
+++ b/sound/oss/midibuf.c
@@ -178,7 +178,7 @@ int MIDIbuf_open(int dev, struct file *file)
                 return err;
 
         parms[dev].prech_timeout = MAX_SCHEDULE_TIMEOUT;
-        midi_in_buf[dev] = (struct midi_buf *) vmalloc(sizeof(struct midi_buf));
+        midi_in_buf[dev] = vmalloc(sizeof(struct midi_buf));
 
         if (midi_in_buf[dev] == NULL)
         {
@@ -188,7 +188,7 @@ int MIDIbuf_open(int dev, struct file *file)
         }
         midi_in_buf[dev]->len = midi_in_buf[dev]->head = midi_in_buf[dev]->tail = 0;
 
-        midi_out_buf[dev] = (struct midi_buf *) vmalloc(sizeof(struct midi_buf));
+        midi_out_buf[dev] = vmalloc(sizeof(struct midi_buf));
 
         if (midi_out_buf[dev] == NULL)
         {
diff --git a/sound/oss/pss.c b/sound/oss/pss.c
index e19dd5dcc2de..9b800ce5100e 100644
--- a/sound/oss/pss.c
+++ b/sound/oss/pss.c
@@ -859,7 +859,7 @@ static int pss_coproc_ioctl(void *dev_info, unsigned int cmd, void __user *arg,
                         return 0;
 
                 case SNDCTL_COPR_LOAD:
-                        buf = (copr_buffer *) vmalloc(sizeof(copr_buffer));
+                        buf = vmalloc(sizeof(copr_buffer));
                         if (buf == NULL)
                                 return -ENOSPC;
                         if (copy_from_user(buf, arg, sizeof(copr_buffer))) {
@@ -871,7 +871,7 @@ static int pss_coproc_ioctl(void *dev_info, unsigned int cmd, void __user *arg,
                         return err;
                 
                 case SNDCTL_COPR_SENDMSG:
-                        mbuf = (copr_msg *)vmalloc(sizeof(copr_msg));
+                        mbuf = vmalloc(sizeof(copr_msg));
                         if (mbuf == NULL)
                                 return -ENOSPC;
                         if (copy_from_user(mbuf, arg, sizeof(copr_msg))) {
@@ -895,7 +895,7 @@ static int pss_coproc_ioctl(void *dev_info, unsigned int cmd, void __user *arg,
 
                 case SNDCTL_COPR_RCVMSG:
                         err = 0;
-                        mbuf = (copr_msg *)vmalloc(sizeof(copr_msg));
+                        mbuf = vmalloc(sizeof(copr_msg));
                         if (mbuf == NULL)
                                 return -ENOSPC;
                         data = (unsigned short *)mbuf->data;
diff --git a/sound/oss/sequencer.c b/sound/oss/sequencer.c
index e85789e53816..5ea1098ac427 100644
--- a/sound/oss/sequencer.c
+++ b/sound/oss/sequencer.c
@@ -1646,13 +1646,13 @@ void sequencer_init(void)
 {
         if (sequencer_ok)
                 return;
-        queue = (unsigned char *)vmalloc(SEQ_MAX_QUEUE * EV_SZ);
+        queue = vmalloc(SEQ_MAX_QUEUE * EV_SZ);
         if (queue == NULL)
         {
                 printk(KERN_ERR "sequencer: Can't allocate memory for sequencer output queue\n");
                 return;
         }
-        iqueue = (unsigned char *)vmalloc(SEQ_MAX_QUEUE * IEV_SZ);
+        iqueue = vmalloc(SEQ_MAX_QUEUE * IEV_SZ);
         if (iqueue == NULL)
         {
                 printk(KERN_ERR "sequencer: Can't allocate memory for sequencer input queue\n");
diff --git a/sound/pci/asihpi/hpioctl.c b/sound/pci/asihpi/hpioctl.c
index 62895a719fcb..22dbd91811a4 100644
--- a/sound/pci/asihpi/hpioctl.c
+++ b/sound/pci/asihpi/hpioctl.c
@@ -435,7 +435,7 @@ void __devexit asihpi_adapter_remove(struct pci_dev *pci_dev)
         struct hpi_message hm;
         struct hpi_response hr;
         struct hpi_adapter *pa;
-        pa = (struct hpi_adapter *)pci_get_drvdata(pci_dev);
+        pa = pci_get_drvdata(pci_dev);
 
         hpi_init_message_response(&hm, &hr, HPI_OBJ_SUBSYSTEM,
                 HPI_SUBSYS_DELETE_ADAPTER);
diff --git a/sound/pci/azt3328.c b/sound/pci/azt3328.c
index 4679ed83a43b..2f3cacbd5528 100644
--- a/sound/pci/azt3328.c
+++ b/sound/pci/azt3328.c
@@ -1129,10 +1129,11 @@ snd_azf3328_codec_setdmaa(struct snd_azf3328 *chip,
 
                 count_areas = size/2;
                 addr_area2 = addr+count_areas;
-                count_areas--; /* max. index */
                 snd_azf3328_dbgcodec("setdma: buffers %08lx[%u] / %08lx[%u]\n",
                                 addr, count_areas, addr_area2, count_areas);
 
+                count_areas--; /* max. index */
+
                 /* build combined I/O buffer length word */
                 lengths = (count_areas << 16) | (count_areas);
                 spin_lock_irqsave(&chip->reg_lock, flags);
@@ -1740,11 +1741,15 @@ static const struct snd_pcm_hardware snd_azf3328_hardware =
         .rate_max =             AZF_FREQ_66200,
         .channels_min =         1,
         .channels_max =         2,
-        .buffer_bytes_max =     65536,
-        .period_bytes_min =     64,
-        .period_bytes_max =     65536,
-        .periods_min =          1,
-        .periods_max =          1024,
+        .buffer_bytes_max =     (64*1024),
+        .period_bytes_min =     1024,
+        .period_bytes_max =     (32*1024),
+        /* We simply have two DMA areas (instead of a list of descriptors
+           such as other cards); I believe that this is a fixed hardware
+           attribute and there isn't much driver magic to be done to expand it.
+           Thus indicate that we have at least and at most 2 periods. */
+        .periods_min =          2,
+        .periods_max =          2,
         /* FIXME: maybe that card actually has a FIFO?
          * Hmm, it seems newer revisions do have one, but we still don't know
          * its size... */
@@ -1980,8 +1985,13 @@ snd_azf3328_timer_stop(struct snd_timer *timer)
         chip = snd_timer_chip(timer);
         spin_lock_irqsave(&chip->reg_lock, flags);
         /* disable timer countdown and interrupt */
-        /* FIXME: should we write TIMER_IRQ_ACK here? */
-        snd_azf3328_ctrl_outb(chip, IDX_IO_TIMER_VALUE + 3, 0);
+        /* Hmm, should we write TIMER_IRQ_ACK here?
+           YES indeed, otherwise a rogue timer operation - which prompts
+           ALSA(?) to call repeated stop() in vain, but NOT start() -
+           will never end (value 0x03 is kept shown in control byte).
+           Simply manually poking 0x04 _once_ immediately successfully stops
+           the hardware/ALSA interrupt activity. */
+        snd_azf3328_ctrl_outb(chip, IDX_IO_TIMER_VALUE + 3, 0x04);
         spin_unlock_irqrestore(&chip->reg_lock, flags);
         snd_azf3328_dbgcallleave();
         return 0;
diff --git a/sound/pci/ctxfi/ctpcm.c b/sound/pci/ctxfi/ctpcm.c
index 85ab43e89212..457d21189b0d 100644
--- a/sound/pci/ctxfi/ctpcm.c
+++ b/sound/pci/ctxfi/ctpcm.c
@@ -129,8 +129,6 @@ static int ct_pcm_playback_open(struct snd_pcm_substream *substream)
 
         apcm->substream = substream;
         apcm->interrupt = ct_atc_pcm_interrupt;
-        runtime->private_data = apcm;
-        runtime->private_free = ct_atc_pcm_free_substream;
         if (IEC958 == substream->pcm->device) {
                 runtime->hw = ct_spdif_passthru_playback_hw;
                 atc->spdif_out_passthru(atc, 1);
@@ -155,8 +153,12 @@ static int ct_pcm_playback_open(struct snd_pcm_substream *substream)
         }
 
         apcm->timer = ct_timer_instance_new(atc->timer, apcm);
-        if (!apcm->timer)
+        if (!apcm->timer) {
+                kfree(apcm);
                 return -ENOMEM;
+        }
+        runtime->private_data = apcm;
+        runtime->private_free = ct_atc_pcm_free_substream;
 
         return 0;
 }
@@ -278,8 +280,6 @@ static int ct_pcm_capture_open(struct snd_pcm_substream *substream)
         apcm->started = 0;
         apcm->substream = substream;
         apcm->interrupt = ct_atc_pcm_interrupt;
-        runtime->private_data = apcm;
-        runtime->private_free = ct_atc_pcm_free_substream;
         runtime->hw = ct_pcm_capture_hw;
         runtime->hw.rate_max = atc->rsr * atc->msr;
 
@@ -298,8 +298,12 @@ static int ct_pcm_capture_open(struct snd_pcm_substream *substream)
         }
 
         apcm->timer = ct_timer_instance_new(atc->timer, apcm);
-        if (!apcm->timer)
+        if (!apcm->timer) {
+                kfree(apcm);
                 return -ENOMEM;
+        }
+        runtime->private_data = apcm;
+        runtime->private_free = ct_atc_pcm_free_substream;
 
         return 0;
 }
diff --git a/sound/pci/hda/patch_conexant.c b/sound/pci/hda/patch_conexant.c
index 6361f752b5f3..846d1ead47fd 100644
--- a/sound/pci/hda/patch_conexant.c
+++ b/sound/pci/hda/patch_conexant.c
@@ -3100,6 +3100,7 @@ static struct snd_pci_quirk cxt5066_cfg_tbl[] = {
         SND_PCI_QUIRK(0x1028, 0x0402, "Dell Vostro", CXT5066_DELL_VOSTRO),
         SND_PCI_QUIRK(0x1028, 0x0408, "Dell Inspiron One 19T", CXT5066_IDEAPAD),
         SND_PCI_QUIRK(0x103c, 0x360b, "HP G60", CXT5066_HP_LAPTOP),
+        SND_PCI_QUIRK(0x1043, 0x13f3, "Asus A52J", CXT5066_HP_LAPTOP),
         SND_PCI_QUIRK(0x1179, 0xff1e, "Toshiba Satellite C650D", CXT5066_IDEAPAD),
         SND_PCI_QUIRK(0x1179, 0xff50, "Toshiba Satellite P500-PSPGSC-01800T", CXT5066_OLPC_XO_1_5),
         SND_PCI_QUIRK(0x1179, 0xffe0, "Toshiba Satellite Pro T130-15F", CXT5066_OLPC_XO_1_5),
@@ -3110,6 +3111,7 @@ static struct snd_pci_quirk cxt5066_cfg_tbl[] = {
         SND_PCI_QUIRK(0x17aa, 0x21b2, "Thinkpad X100e", CXT5066_IDEAPAD),
         SND_PCI_QUIRK(0x17aa, 0x21b3, "Thinkpad Edge 13 (197)", CXT5066_IDEAPAD),
         SND_PCI_QUIRK(0x17aa, 0x21b4, "Thinkpad Edge", CXT5066_IDEAPAD),
+        SND_PCI_QUIRK(0x17aa, 0x21c8, "Thinkpad Edge 11", CXT5066_IDEAPAD),
         SND_PCI_QUIRK(0x17aa, 0x215e, "Lenovo Thinkpad", CXT5066_THINKPAD),
         SND_PCI_QUIRK(0x17aa, 0x38af, "Lenovo G series", CXT5066_IDEAPAD),
         SND_PCI_QUIRK(0x17aa, 0x390a, "Lenovo S10-3t", CXT5066_IDEAPAD),
diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c
index 5f00589cb791..0ac6aed0c889 100644
--- a/sound/pci/hda/patch_realtek.c
+++ b/sound/pci/hda/patch_realtek.c
@@ -14623,7 +14623,10 @@ static int alc275_setup_dual_adc(struct hda_codec *codec)
 /* different alc269-variants */
 enum {
         ALC269_TYPE_NORMAL,
+        ALC269_TYPE_ALC258,
         ALC269_TYPE_ALC259,
+        ALC269_TYPE_ALC269VB,
+        ALC269_TYPE_ALC270,
         ALC269_TYPE_ALC271X,
 };
 
@@ -15023,7 +15026,7 @@ static int alc269_fill_coef(struct hda_codec *codec)
 static int patch_alc269(struct hda_codec *codec)
 {
         struct alc_spec *spec;
-        int board_config;
+        int board_config, coef;
         int err;
 
         spec = kzalloc(sizeof(*spec), GFP_KERNEL);
@@ -15034,14 +15037,23 @@ static int patch_alc269(struct hda_codec *codec)
 
         alc_auto_parse_customize_define(codec);
 
-        if ((alc_read_coef_idx(codec, 0) & 0x00f0) == 0x0010){
+        coef = alc_read_coef_idx(codec, 0);
+        if ((coef & 0x00f0) == 0x0010) {
                 if (codec->bus->pci->subsystem_vendor == 0x1025 &&
                     spec->cdefine.platform_type == 1) {
                         alc_codec_rename(codec, "ALC271X");
                         spec->codec_variant = ALC269_TYPE_ALC271X;
-                } else {
+                } else if ((coef & 0xf000) == 0x1000) {
+                        spec->codec_variant = ALC269_TYPE_ALC270;
+                } else if ((coef & 0xf000) == 0x2000) {
                         alc_codec_rename(codec, "ALC259");
                         spec->codec_variant = ALC269_TYPE_ALC259;
+                } else if ((coef & 0xf000) == 0x3000) {
+                        alc_codec_rename(codec, "ALC258");
+                        spec->codec_variant = ALC269_TYPE_ALC258;
+                } else {
+                        alc_codec_rename(codec, "ALC269VB");
+                        spec->codec_variant = ALC269_TYPE_ALC269VB;
                 }
         } else
                 alc_fix_pll_init(codec, 0x20, 0x04, 15);
@@ -15104,7 +15116,7 @@ static int patch_alc269(struct hda_codec *codec)
         spec->stream_digital_capture = &alc269_pcm_digital_capture;
 
         if (!spec->adc_nids) { /* wasn't filled automatically? use default */
-                if (spec->codec_variant != ALC269_TYPE_NORMAL) {
+                if (spec->codec_variant == ALC269_TYPE_NORMAL) {
                         spec->adc_nids = alc269_adc_nids;
                         spec->num_adc_nids = ARRAY_SIZE(alc269_adc_nids);
                         spec->capsrc_nids = alc269_capsrc_nids;
@@ -19298,6 +19310,7 @@ static const struct alc_fixup alc662_fixups[] = {
 
 static struct snd_pci_quirk alc662_fixup_tbl[] = {
         SND_PCI_QUIRK(0x1025, 0x038b, "Acer Aspire 8943G", ALC662_FIXUP_ASPIRE),
+        SND_PCI_QUIRK(0x144d, 0xc051, "Samsung R720", ALC662_FIXUP_IDEAPAD),
         SND_PCI_QUIRK(0x17aa, 0x38af, "Lenovo Ideapad Y550P", ALC662_FIXUP_IDEAPAD),
         SND_PCI_QUIRK(0x17aa, 0x3a0d, "Lenovo Ideapad Y550", ALC662_FIXUP_IDEAPAD),
         {}
@@ -19419,7 +19432,10 @@ static int patch_alc888(struct hda_codec *codec)
 {
         if ((alc_read_coef_idx(codec, 0) & 0x00f0)==0x0030){
                 kfree(codec->chip_name);
-                codec->chip_name = kstrdup("ALC888-VD", GFP_KERNEL);
+                if (codec->vendor_id == 0x10ec0887)
+                        codec->chip_name = kstrdup("ALC887-VD", GFP_KERNEL);
+                else
+                        codec->chip_name = kstrdup("ALC888-VD", GFP_KERNEL);
                 if (!codec->chip_name) {
                         alc_free(codec);
                         return -ENOMEM;
@@ -19909,7 +19925,7 @@ static struct hda_codec_preset snd_hda_preset_realtek[] = {
         { .id = 0x10ec0885, .rev = 0x100103, .name = "ALC889A",
           .patch = patch_alc882 },
         { .id = 0x10ec0885, .name = "ALC885", .patch = patch_alc882 },
-        { .id = 0x10ec0887, .name = "ALC887", .patch = patch_alc882 },
+        { .id = 0x10ec0887, .name = "ALC887", .patch = patch_alc888 },
         { .id = 0x10ec0888, .rev = 0x100101, .name = "ALC1200",
           .patch = patch_alc882 },
         { .id = 0x10ec0888, .name = "ALC888", .patch = patch_alc888 },
diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c
index 93fa59cc60ef..5c710807dfe5 100644
--- a/sound/pci/hda/patch_sigmatel.c
+++ b/sound/pci/hda/patch_sigmatel.c
@@ -389,6 +389,11 @@ static hda_nid_t stac92hd83xxx_dmic_nids[STAC92HD83XXX_NUM_DMICS + 1] = {
         0x11, 0x20, 0
 };
 
+#define STAC92HD87B_NUM_DMICS    1
+static hda_nid_t stac92hd87b_dmic_nids[STAC92HD87B_NUM_DMICS + 1] = {
+        0x11, 0
+};
+
 #define STAC92HD83XXX_NUM_CAPS  2
 static unsigned long stac92hd83xxx_capvols[] = {
         HDA_COMPOSE_AMP_VAL(0x17, 3, 0, HDA_OUTPUT),
@@ -3486,10 +3491,8 @@ static int stac92xx_auto_create_dmic_input_ctls(struct hda_codec *codec,
                                 return err;
                 }
 
-                if (snd_hda_get_bool_hint(codec, "separate_dmux") != 1) {
+                if (snd_hda_get_bool_hint(codec, "separate_dmux") != 1)
                         snd_hda_add_imux_item(imux, label, index, NULL);
-                        spec->num_analog_muxes++;
-                }
         }
 
         return 0;
@@ -5452,12 +5455,17 @@ again:
                                 stac92hd83xxx_brd_tbl[spec->board_config]);
 
         switch (codec->vendor_id) {
+        case 0x111d76d1:
+        case 0x111d76d9:
+                spec->dmic_nids = stac92hd87b_dmic_nids;
+                spec->num_dmics = stac92xx_connected_ports(codec,
+                                stac92hd87b_dmic_nids,
+                                STAC92HD87B_NUM_DMICS);
+                /* Fall through */
         case 0x111d7666:
         case 0x111d7667:
         case 0x111d7668:
         case 0x111d7669:
-        case 0x111d76d1:
-        case 0x111d76d9:
                 spec->num_pins = ARRAY_SIZE(stac92hd88xxx_pin_nids);
                 spec->pin_nids = stac92hd88xxx_pin_nids;
                 spec->mono_nid = 0;
diff --git a/sound/pci/intel8x0.c b/sound/pci/intel8x0.c
index 400f9ebd243e..629a5494347a 100644
--- a/sound/pci/intel8x0.c
+++ b/sound/pci/intel8x0.c
@@ -1866,6 +1866,12 @@ static struct ac97_quirk ac97_quirks[] __devinitdata = {
         },
         {
                 .subvendor = 0x1028,
+                .subdevice = 0x0182,
+                .name = "Dell Latitude D610",   /* STAC9750/51 */
+                .type = AC97_TUNE_HP_ONLY
+        },
+        {
+                .subvendor = 0x1028,
                 .subdevice = 0x0186,
                 .name = "Dell Latitude D810", /* cf. Malone #41015 */
                 .type = AC97_TUNE_HP_MUTE_LED
diff --git a/sound/pci/mixart/mixart_hwdep.h b/sound/pci/mixart/mixart_hwdep.h
index a46f5083db99..812e288ef2e7 100644
--- a/sound/pci/mixart/mixart_hwdep.h
+++ b/sound/pci/mixart/mixart_hwdep.h
@@ -25,11 +25,21 @@
 
 #include <sound/hwdep.h>
 
+#ifndef readl_be
 #define readl_be(x) be32_to_cpu(__raw_readl(x))
+#endif
+
+#ifndef writel_be
 #define writel_be(data,addr) __raw_writel(cpu_to_be32(data),addr)
+#endif
 
+#ifndef readl_le
 #define readl_le(x) le32_to_cpu(__raw_readl(x))
+#endif
+
+#ifndef writel_le
 #define writel_le(data,addr) __raw_writel(cpu_to_le32(data),addr)
+#endif
 
 #define MIXART_MEM(mgr,x)       ((mgr)->mem[0].virt + (x))
 #define MIXART_REG(mgr,x)       ((mgr)->mem[1].virt + (x))
diff --git a/sound/ppc/pmac.c b/sound/ppc/pmac.c
index 85081172403f..b47cfd45b3b9 100644
--- a/sound/ppc/pmac.c
+++ b/sound/ppc/pmac.c
@@ -1228,10 +1228,8 @@ int __devinit snd_pmac_new(struct snd_card *card, struct snd_pmac **chip_return)
                                                chip->rsrc[i].start + 1,
                                                rnames[i]) == NULL) {
                                 printk(KERN_ERR "snd: can't request rsrc "
-                                       " %d (%s: 0x%016llx:%016llx)\n",
-                                       i, rnames[i],
-                                       (unsigned long long)chip->rsrc[i].start,
-                                       (unsigned long long)chip->rsrc[i].end);
+                                       " %d (%s: %pR)\n",
+                                       i, rnames[i], &chip->rsrc[i]);
                                 err = -ENODEV;
                                 goto __error;
                         }
@@ -1256,10 +1254,8 @@ int __devinit snd_pmac_new(struct snd_card *card, struct snd_pmac **chip_return)
                                                chip->rsrc[i].start + 1,
                                                rnames[i]) == NULL) {
                                 printk(KERN_ERR "snd: can't request rsrc "
-                                       " %d (%s: 0x%016llx:%016llx)\n",
-                                       i, rnames[i],
-                                       (unsigned long long)chip->rsrc[i].start,
-                                       (unsigned long long)chip->rsrc[i].end);
+                                       " %d (%s: %pR)\n",
+                                       i, rnames[i], &chip->rsrc[i]);
                                 err = -ENODEV;
                                 goto __error;
                         }
diff --git a/sound/soc/atmel/Kconfig b/sound/soc/atmel/Kconfig
index e720d5e6f04c..bee3c94f58b0 100644
--- a/sound/soc/atmel/Kconfig
+++ b/sound/soc/atmel/Kconfig
@@ -16,7 +16,8 @@ config SND_ATMEL_SOC_SSC
 
 config SND_AT91_SOC_SAM9G20_WM8731
         tristate "SoC Audio support for WM8731-based At91sam9g20 evaluation board"
-        depends on ATMEL_SSC && ARCH_AT91SAM9G20 && SND_ATMEL_SOC
+        depends on ATMEL_SSC && ARCH_AT91SAM9G20 && SND_ATMEL_SOC && \
+                   AT91_PROGRAMMABLE_CLOCKS
         select SND_ATMEL_SOC_SSC
         select SND_SOC_WM8731
         help
@@ -25,7 +26,7 @@ config SND_AT91_SOC_SAM9G20_WM8731
 
 config SND_AT32_SOC_PLAYPAQ
         tristate "SoC Audio support for PlayPaq with WM8510"
-        depends on SND_ATMEL_SOC && BOARD_PLAYPAQ
+        depends on SND_ATMEL_SOC && BOARD_PLAYPAQ && AT91_PROGRAMMABLE_CLOCKS
         select SND_ATMEL_SOC_SSC
         select SND_SOC_WM8510
         help
diff --git a/sound/soc/codecs/max98088.c b/sound/soc/codecs/max98088.c
index bc22ee93a75d..470cb93b1d1f 100644
--- a/sound/soc/codecs/max98088.c
+++ b/sound/soc/codecs/max98088.c
@@ -28,6 +28,11 @@
 #include <sound/max98088.h>
 #include "max98088.h"
 
+enum max98088_type {
+       MAX98088,
+       MAX98089,
+};
+
 struct max98088_cdata {
        unsigned int rate;
        unsigned int fmt;
@@ -36,6 +41,7 @@ struct max98088_cdata {
 
 struct max98088_priv {
        u8 reg_cache[M98088_REG_CNT];
+       enum max98088_type devtype;
        void *control_data;
        struct max98088_pdata *pdata;
        unsigned int sysclk;
@@ -2040,6 +2046,8 @@ static int max98088_i2c_probe(struct i2c_client *i2c,
        if (max98088 == NULL)
                return -ENOMEM;
 
+       max98088->devtype = id->driver_data;
+
        i2c_set_clientdata(i2c, max98088);
        max98088->control_data = i2c;
        max98088->pdata = i2c->dev.platform_data;
@@ -2059,7 +2067,8 @@ static int __devexit max98088_i2c_remove(struct i2c_client *client)
 }
 
 static const struct i2c_device_id max98088_i2c_id[] = {
-       { "max98088", 0 },
+       { "max98088", MAX98088 },
+       { "max98089", MAX98089 },
        { }
 };
 MODULE_DEVICE_TABLE(i2c, max98088_i2c_id);
diff --git a/sound/soc/codecs/uda134x.c b/sound/soc/codecs/uda134x.c
index 7540a509a6f5..464f0cfa4c7a 100644
--- a/sound/soc/codecs/uda134x.c
+++ b/sound/soc/codecs/uda134x.c
@@ -597,6 +597,7 @@ static struct snd_soc_codec_driver soc_codec_dev_uda134x = {
         .resume =       uda134x_soc_resume,
         .reg_cache_size = sizeof(uda134x_reg),
         .reg_word_size = sizeof(u8),
+        .reg_cache_default = uda134x_reg,
         .reg_cache_step = 1,
         .read = uda134x_read_reg_cache,
         .write = uda134x_write,
diff --git a/sound/soc/codecs/wm8350.c b/sound/soc/codecs/wm8350.c
index f4f1fba38eb9..7611add7f8c3 100644
--- a/sound/soc/codecs/wm8350.c
+++ b/sound/soc/codecs/wm8350.c
@@ -831,7 +831,7 @@ static int wm8350_set_dai_sysclk(struct snd_soc_dai *codec_dai,
         }
 
         /* MCLK direction */
-        if (dir == WM8350_MCLK_DIR_OUT)
+        if (dir == SND_SOC_CLOCK_OUT)
                 wm8350_set_bits(wm8350, WM8350_CLOCK_CONTROL_2,
                                 WM8350_MCLK_DIR);
         else
@@ -1586,6 +1586,13 @@ static  int wm8350_codec_probe(struct snd_soc_codec *codec)
         wm8350_set_bits(wm8350, WM8350_ROUT2_VOLUME,
                         WM8350_OUT2_VU | WM8350_OUT2R_MUTE);
 
+        /* Make sure AIF tristating is disabled by default */
+        wm8350_clear_bits(wm8350, WM8350_AI_FORMATING, WM8350_AIF_TRI);
+
+        /* Make sure we've got a sane companding setup too */
+        wm8350_clear_bits(wm8350, WM8350_ADC_DAC_COMP,
+                          WM8350_DAC_COMP | WM8350_LOOPBACK);
+
         /* Make sure jack detect is disabled to start off with */
         wm8350_clear_bits(wm8350, WM8350_JACK_DETECT,
                           WM8350_JDL_ENA | WM8350_JDR_ENA);
diff --git a/sound/soc/codecs/wm8776.c b/sound/soc/codecs/wm8776.c
index 04182c464e35..0132a27140ae 100644
--- a/sound/soc/codecs/wm8776.c
+++ b/sound/soc/codecs/wm8776.c
@@ -34,7 +34,6 @@
 /* codec private data */
 struct wm8776_priv {
         enum snd_soc_control_type control_type;
-        u16 reg_cache[WM8776_CACHEREGNUM];
         int sysclk[2];
 };
 
diff --git a/sound/soc/codecs/wm8962.c b/sound/soc/codecs/wm8962.c
index 894d0cd3aa9b..e8092745a207 100644
--- a/sound/soc/codecs/wm8962.c
+++ b/sound/soc/codecs/wm8962.c
@@ -3500,8 +3500,11 @@ static ssize_t wm8962_beep_set(struct device *dev,
 {
         struct wm8962_priv *wm8962 = dev_get_drvdata(dev);
         long int time;
+        int ret;
 
-        strict_strtol(buf, 10, &time);
+        ret = strict_strtol(buf, 10, &time);
+        if (ret != 0)
+                return ret;
 
         input_event(wm8962->beep, EV_SND, SND_TONE, time);
 
diff --git a/sound/soc/codecs/wm8994.c b/sound/soc/codecs/wm8994.c
index 0db59c3aa5d4..830dfdd66c5f 100644
--- a/sound/soc/codecs/wm8994.c
+++ b/sound/soc/codecs/wm8994.c
@@ -3903,6 +3903,8 @@ static int wm8994_codec_probe(struct snd_soc_codec *codec)
                 return -ENOMEM;
         snd_soc_codec_set_drvdata(codec, wm8994);
 
+        codec->reg_cache = &wm8994->reg_cache;
+
         wm8994->pdata = dev_get_platdata(codec->dev->parent);
         wm8994->codec = codec;
 
diff --git a/sound/soc/davinci/davinci-evm.c b/sound/soc/davinci/davinci-evm.c
index 2b07b17a6b2d..bc9e6b0b3f6f 100644
--- a/sound/soc/davinci/davinci-evm.c
+++ b/sound/soc/davinci/davinci-evm.c
@@ -157,12 +157,23 @@ static int evm_aic3x_init(struct snd_soc_pcm_runtime *rtd)
 }
 
 /* davinci-evm digital audio interface glue - connects codec <--> CPU */
-static struct snd_soc_dai_link evm_dai = {
+static struct snd_soc_dai_link dm6446_evm_dai = {
         .name = "TLV320AIC3X",
         .stream_name = "AIC3X",
-        .cpu_dai_name = "davinci-mcasp.0",
+        .cpu_dai_name = "davinci-mcbsp",
         .codec_dai_name = "tlv320aic3x-hifi",
-        .codec_name = "tlv320aic3x-codec.0-001a",
+        .codec_name = "tlv320aic3x-codec.1-001b",
+        .platform_name = "davinci-pcm-audio",
+        .init = evm_aic3x_init,
+        .ops = &evm_ops,
+};
+
+static struct snd_soc_dai_link dm355_evm_dai = {
+        .name = "TLV320AIC3X",
+        .stream_name = "AIC3X",
+        .cpu_dai_name = "davinci-mcbsp.1",
+        .codec_dai_name = "tlv320aic3x-hifi",
+        .codec_name = "tlv320aic3x-codec.1-001b",
         .platform_name = "davinci-pcm-audio",
         .init = evm_aic3x_init,
         .ops = &evm_ops,
@@ -172,10 +183,10 @@ static struct snd_soc_dai_link dm365_evm_dai = {
 #ifdef CONFIG_SND_DM365_AIC3X_CODEC
         .name = "TLV320AIC3X",
         .stream_name = "AIC3X",
-        .cpu_dai_name = "davinci-i2s",
+        .cpu_dai_name = "davinci-mcbsp",
         .codec_dai_name = "tlv320aic3x-hifi",
         .init = evm_aic3x_init,
-        .codec_name = "tlv320aic3x-codec.0-001a",
+        .codec_name = "tlv320aic3x-codec.1-0018",
         .ops = &evm_ops,
 #elif defined(CONFIG_SND_DM365_VOICE_CODEC)
         .name = "Voice Codec - CQ93VC",
@@ -219,10 +230,17 @@ static struct snd_soc_dai_link da8xx_evm_dai = {
         .ops = &evm_ops,
 };
 
-/* davinci dm6446, dm355 evm audio machine driver */
-static struct snd_soc_card snd_soc_card_evm = {
-        .name = "DaVinci EVM",
-        .dai_link = &evm_dai,
+/* davinci dm6446 evm audio machine driver */
+static struct snd_soc_card dm6446_snd_soc_card_evm = {
+        .name = "DaVinci DM6446 EVM",
+        .dai_link = &dm6446_evm_dai,
+        .num_links = 1,
+};
+
+/* davinci dm355 evm audio machine driver */
+static struct snd_soc_card dm355_snd_soc_card_evm = {
+        .name = "DaVinci DM355 EVM",
+        .dai_link = &dm355_evm_dai,
         .num_links = 1,
 };
 
@@ -261,10 +279,10 @@ static int __init evm_init(void)
         int ret;
 
         if (machine_is_davinci_evm()) {
-                evm_snd_dev_data = &snd_soc_card_evm;
+                evm_snd_dev_data = &dm6446_snd_soc_card_evm;
                 index = 0;
         } else if (machine_is_davinci_dm355_evm()) {
-                evm_snd_dev_data = &snd_soc_card_evm;
+                evm_snd_dev_data = &dm355_snd_soc_card_evm;
                 index = 1;
         } else if (machine_is_davinci_dm365_evm()) {
                 evm_snd_dev_data = &dm365_snd_soc_card_evm;
diff --git a/sound/soc/davinci/davinci-i2s.c b/sound/soc/davinci/davinci-i2s.c
index d46b545d41f4..9e0e565e6ed9 100644
--- a/sound/soc/davinci/davinci-i2s.c
+++ b/sound/soc/davinci/davinci-i2s.c
@@ -426,9 +426,6 @@ static int davinci_i2s_hw_params(struct snd_pcm_substream *substream,
         snd_pcm_format_t fmt;
         unsigned element_cnt = 1;
 
-        dai->capture_dma_data = dev->dma_params;
-        dai->playback_dma_data = dev->dma_params;
-
         /* general line settings */
         spcr = davinci_mcbsp_read_reg(dev, DAVINCI_MCBSP_SPCR_REG);
         if (substream->stream == SNDRV_PCM_STREAM_CAPTURE) {
@@ -601,6 +598,15 @@ static int davinci_i2s_trigger(struct snd_pcm_substream *substream, int cmd,
         return ret;
 }
 
+static int davinci_i2s_startup(struct snd_pcm_substream *substream,
+                               struct snd_soc_dai *dai)
+{
+        struct davinci_mcbsp_dev *dev = snd_soc_dai_get_drvdata(dai);
+
+        snd_soc_dai_set_dma_data(dai, substream, dev->dma_params);
+        return 0;
+}
+
 static void davinci_i2s_shutdown(struct snd_pcm_substream *substream,
                 struct snd_soc_dai *dai)
 {
@@ -612,6 +618,7 @@ static void davinci_i2s_shutdown(struct snd_pcm_substream *substream,
 #define DAVINCI_I2S_RATES       SNDRV_PCM_RATE_8000_96000
 
 static struct snd_soc_dai_ops davinci_i2s_dai_ops = {
+        .startup        = davinci_i2s_startup,
         .shutdown       = davinci_i2s_shutdown,
         .prepare        = davinci_i2s_prepare,
         .trigger        = davinci_i2s_trigger,
@@ -749,7 +756,7 @@ static struct platform_driver davinci_mcbsp_driver = {
         .probe          = davinci_i2s_probe,
         .remove         = davinci_i2s_remove,
         .driver         = {
-                .name   = "davinci-i2s",
+                .name   = "davinci-mcbsp",
                 .owner  = THIS_MODULE,
         },
 };
diff --git a/sound/soc/davinci/davinci-mcasp.c b/sound/soc/davinci/davinci-mcasp.c
index 86918ee12419..fb55d2c5d704 100644
--- a/sound/soc/davinci/davinci-mcasp.c
+++ b/sound/soc/davinci/davinci-mcasp.c
@@ -715,9 +715,6 @@ static int davinci_mcasp_hw_params(struct snd_pcm_substream *substream,
         int word_length;
         u8 fifo_level;
 
-        cpu_dai->capture_dma_data = dev->dma_params;
-        cpu_dai->playback_dma_data = dev->dma_params;
-
         davinci_hw_common_param(dev, substream->stream);
         if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK)
                 fifo_level = dev->txnumevt;
@@ -799,7 +796,17 @@ static int davinci_mcasp_trigger(struct snd_pcm_substream *substream,
         return ret;
 }
 
+static int davinci_mcasp_startup(struct snd_pcm_substream *substream,
+                                 struct snd_soc_dai *dai)
+{
+        struct davinci_audio_dev *dev = snd_soc_dai_get_drvdata(dai);
+
+        snd_soc_dai_set_dma_data(dai, substream, dev->dma_params);
+        return 0;
+}
+
 static struct snd_soc_dai_ops davinci_mcasp_dai_ops = {
+        .startup        = davinci_mcasp_startup,
         .trigger        = davinci_mcasp_trigger,
         .hw_params      = davinci_mcasp_hw_params,
         .set_fmt        = davinci_mcasp_set_dai_fmt,
diff --git a/sound/soc/davinci/davinci-sffsdr.c b/sound/soc/davinci/davinci-sffsdr.c
index 009b6521a1bf..6c6666a1f942 100644
--- a/sound/soc/davinci/davinci-sffsdr.c
+++ b/sound/soc/davinci/davinci-sffsdr.c
@@ -84,7 +84,7 @@ static struct snd_soc_ops sffsdr_ops = {
 static struct snd_soc_dai_link sffsdr_dai = {
         .name = "PCM3008", /* Codec name */
         .stream_name = "PCM3008 HiFi",
-        .cpu_dai_name = "davinci-asp.0",
+        .cpu_dai_name = "davinci-mcbsp",
         .codec_dai_name = "pcm3008-hifi",
         .codec_name = "pcm3008-codec",
         .platform_name = "davinci-pcm-audio",
diff --git a/sound/soc/davinci/davinci-vcif.c b/sound/soc/davinci/davinci-vcif.c
index ea232f6a2c21..fb4cc1edf339 100644
--- a/sound/soc/davinci/davinci-vcif.c
+++ b/sound/soc/davinci/davinci-vcif.c
@@ -97,9 +97,6 @@ static int davinci_vcif_hw_params(struct snd_pcm_substream *substream,
                         &davinci_vcif_dev->dma_params[substream->stream];
         u32 w;
 
-        dai->capture_dma_data = davinci_vcif_dev->dma_params;
-        dai->playback_dma_data = davinci_vcif_dev->dma_params;
-
         /* Restart the codec before setup */
         davinci_vcif_stop(substream);
         davinci_vcif_start(substream);
@@ -174,9 +171,19 @@ static int davinci_vcif_trigger(struct snd_pcm_substream *substream, int cmd,
         return ret;
 }
 
+static int davinci_vcif_startup(struct snd_pcm_substream *substream,
+                                struct snd_soc_dai *dai)
+{
+        struct davinci_vcif_dev *dev = snd_soc_dai_get_drvdata(dai);
+
+        snd_soc_dai_set_dma_data(dai, substream, dev->dma_params);
+        return 0;
+}
+
 #define DAVINCI_VCIF_RATES      SNDRV_PCM_RATE_8000_48000
 
 static struct snd_soc_dai_ops davinci_vcif_dai_ops = {
+        .startup        = davinci_vcif_startup,
         .trigger        = davinci_vcif_trigger,
         .hw_params      = davinci_vcif_hw_params,
 };
diff --git a/sound/soc/fsl/mpc5200_psc_i2s.c b/sound/soc/fsl/mpc5200_psc_i2s.c
index 74ffed41340f..9018fa5bf0db 100644
--- a/sound/soc/fsl/mpc5200_psc_i2s.c
+++ b/sound/soc/fsl/mpc5200_psc_i2s.c
@@ -160,7 +160,7 @@ static int __devinit psc_i2s_of_probe(struct platform_device *op,
         rc = snd_soc_register_dais(&op->dev, psc_i2s_dai, ARRAY_SIZE(psc_i2s_dai));
         if (rc != 0) {
                 pr_err("Failed to register DAI\n");
-                return 0;
+                return rc;
         }
 
         psc_dma = dev_get_drvdata(&op->dev);
diff --git a/sound/soc/imx/eukrea-tlv320.c b/sound/soc/imx/eukrea-tlv320.c
index b59675257ce5..dd4fffdbd177 100644
--- a/sound/soc/imx/eukrea-tlv320.c
+++ b/sound/soc/imx/eukrea-tlv320.c
@@ -34,8 +34,8 @@ static int eukrea_tlv320_hw_params(struct snd_pcm_substream *substream,
                             struct snd_pcm_hw_params *params)
 {
         struct snd_soc_pcm_runtime *rtd = substream->private_data;
-        struct snd_soc_dai *codec_dai = rtd->dai->codec_dai;
-        struct snd_soc_dai *cpu_dai = rtd->dai->cpu_dai;
+        struct snd_soc_dai *codec_dai = rtd->codec_dai;
+        struct snd_soc_dai *cpu_dai = rtd->cpu_dai;
         int ret;
 
         ret = snd_soc_dai_set_fmt(cpu_dai, SND_SOC_DAIFMT_I2S |
@@ -79,10 +79,10 @@ static struct snd_soc_ops eukrea_tlv320_snd_ops = {
 static struct snd_soc_dai_link eukrea_tlv320_dai = {
         .name           = "tlv320aic23",
         .stream_name    = "TLV320AIC23",
-        .codec_dai      = "tlv320aic23-hifi",
+        .codec_dai_name = "tlv320aic23-hifi",
         .platform_name  = "imx-pcm-audio.0",
         .codec_name     = "tlv320aic23-codec.0-001a",
-        .cpu_dai = "imx-ssi.0",
+        .cpu_dai_name   = "imx-ssi.0",
         .ops            = &eukrea_tlv320_snd_ops,
 };
 
diff --git a/sound/soc/imx/imx-pcm-dma-mx2.c b/sound/soc/imx/imx-pcm-dma-mx2.c
index fd493ee1428e..671ef8dd524c 100644
--- a/sound/soc/imx/imx-pcm-dma-mx2.c
+++ b/sound/soc/imx/imx-pcm-dma-mx2.c
@@ -20,6 +20,7 @@
 #include <linux/module.h>
 #include <linux/platform_device.h>
 #include <linux/slab.h>
+#include <linux/dmaengine.h>
 
 #include <sound/core.h>
 #include <sound/initval.h>
@@ -27,165 +28,146 @@
 #include <sound/pcm_params.h>
 #include <sound/soc.h>
 
-#include <mach/dma-mx1-mx2.h>
+#include <mach/dma.h>
 
 #include "imx-ssi.h"
 
 struct imx_pcm_runtime_data {
-        int sg_count;
-        struct scatterlist *sg_list;
-        int period;
+        int period_bytes;
         int periods;
-        unsigned long dma_addr;
         int dma;
-        struct snd_pcm_substream *substream;
         unsigned long offset;
         unsigned long size;
-        unsigned long period_cnt;
         void *buf;
         int period_time;
+        struct dma_async_tx_descriptor *desc;
+        struct dma_chan *dma_chan;
+        struct imx_dma_data dma_data;
 };
 
-/* Called by the DMA framework when a period has elapsed */
-static void imx_ssi_dma_progression(int channel, void *data,
-                                        struct scatterlist *sg)
+static void audio_dma_irq(void *data)
 {
-        struct snd_pcm_substream *substream = data;
+        struct snd_pcm_substream *substream = (struct snd_pcm_substream *)data;
         struct snd_pcm_runtime *runtime = substream->runtime;
         struct imx_pcm_runtime_data *iprtd = runtime->private_data;
 
-        if (!sg)
-                return;
-
-        runtime = iprtd->substream->runtime;
+        iprtd->offset += iprtd->period_bytes;
+        iprtd->offset %= iprtd->period_bytes * iprtd->periods;
 
-        iprtd->offset = sg->dma_address - runtime->dma_addr;
-
-        snd_pcm_period_elapsed(iprtd->substream);
+        snd_pcm_period_elapsed(substream);
 }
 
-static void imx_ssi_dma_callback(int channel, void *data)
+static bool filter(struct dma_chan *chan, void *param)
 {
-        pr_err("%s shouldn't be called\n", __func__);
-}
+        struct imx_pcm_runtime_data *iprtd = param;
 
-static void snd_imx_dma_err_callback(int channel, void *data, int err)
-{
-        struct snd_pcm_substream *substream = data;
-        struct snd_soc_pcm_runtime *rtd = substream->private_data;
-        struct imx_pcm_dma_params *dma_params = 
-                snd_soc_dai_get_dma_data(rtd->dai->cpu_dai, substream);
-        struct snd_pcm_runtime *runtime = substream->runtime;
-        struct imx_pcm_runtime_data *iprtd = runtime->private_data;
-        int ret;
+        if (!imx_dma_is_general_purpose(chan))
+                return false;
 
-        pr_err("DMA timeout on channel %d -%s%s%s%s\n",
-                 channel,
-                 err & IMX_DMA_ERR_BURST ?    " burst" : "",
-                 err & IMX_DMA_ERR_REQUEST ?  " request" : "",
-                 err & IMX_DMA_ERR_TRANSFER ? " transfer" : "",
-                 err & IMX_DMA_ERR_BUFFER ?   " buffer" : "");
+        chan->private = &iprtd->dma_data;
 
-        imx_dma_disable(iprtd->dma);
-        ret = imx_dma_setup_sg(iprtd->dma, iprtd->sg_list, iprtd->sg_count,
-                        IMX_DMA_LENGTH_LOOP, dma_params->dma_addr,
-                        substream->stream == SNDRV_PCM_STREAM_PLAYBACK ?
-                        DMA_MODE_WRITE : DMA_MODE_READ);
-        if (!ret)
-                imx_dma_enable(iprtd->dma);
+        return true;
 }
 
-static int imx_ssi_dma_alloc(struct snd_pcm_substream *substream)
+static int imx_ssi_dma_alloc(struct snd_pcm_substream *substream,
+                                struct snd_pcm_hw_params *params)
 {
         struct snd_soc_pcm_runtime *rtd = substream->private_data;
         struct imx_pcm_dma_params *dma_params;
         struct snd_pcm_runtime *runtime = substream->runtime;
         struct imx_pcm_runtime_data *iprtd = runtime->private_data;
+        struct dma_slave_config slave_config;
+        dma_cap_mask_t mask;
+        enum dma_slave_buswidth buswidth;
         int ret;
 
         dma_params = snd_soc_dai_get_dma_data(rtd->cpu_dai, substream);
 
-        iprtd->dma = imx_dma_request_by_prio(DRV_NAME, DMA_PRIO_HIGH);
-        if (iprtd->dma < 0) {
-                pr_err("Failed to claim the audio DMA\n");
-                return -ENODEV;
-        }
+        iprtd->dma_data.peripheral_type = IMX_DMATYPE_SSI;
+        iprtd->dma_data.priority = DMA_PRIO_HIGH;
+        iprtd->dma_data.dma_request = dma_params->dma;
 
-        ret = imx_dma_setup_handlers(iprtd->dma,
-                                imx_ssi_dma_callback,
-                                snd_imx_dma_err_callback, substream);
-        if (ret)
-                goto out;
+        /* Try to grab a DMA channel */
+        dma_cap_zero(mask);
+        dma_cap_set(DMA_SLAVE, mask);
+        iprtd->dma_chan = dma_request_channel(mask, filter, iprtd);
+        if (!iprtd->dma_chan)
+                return -EINVAL;
 
-        ret = imx_dma_setup_progression_handler(iprtd->dma,
-                        imx_ssi_dma_progression);
-        if (ret) {
-                pr_err("Failed to setup the DMA handler\n");
-                goto out;
+        switch (params_format(params)) {
+        case SNDRV_PCM_FORMAT_S16_LE:
+                buswidth = DMA_SLAVE_BUSWIDTH_2_BYTES;
+                break;
+        case SNDRV_PCM_FORMAT_S20_3LE:
+        case SNDRV_PCM_FORMAT_S24_LE:
+                buswidth = DMA_SLAVE_BUSWIDTH_4_BYTES;
+                break;
+        default:
+                return 0;
         }
 
-        ret = imx_dma_config_channel(iprtd->dma,
-                        IMX_DMA_MEMSIZE_16 | IMX_DMA_TYPE_FIFO,
-                        IMX_DMA_MEMSIZE_32 | IMX_DMA_TYPE_LINEAR,
-                        dma_params->dma, 1);
-        if (ret < 0) {
-                pr_err("Cannot configure DMA channel: %d\n", ret);
-                goto out;
+        if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) {
+                slave_config.direction = DMA_TO_DEVICE;
+                slave_config.dst_addr = dma_params->dma_addr;
+                slave_config.dst_addr_width = buswidth;
+                slave_config.dst_maxburst = dma_params->burstsize;
+        } else {
+                slave_config.direction = DMA_FROM_DEVICE;
+                slave_config.src_addr = dma_params->dma_addr;
+                slave_config.src_addr_width = buswidth;
+                slave_config.src_maxburst = dma_params->burstsize;
         }
 
-        imx_dma_config_burstlen(iprtd->dma, dma_params->burstsize * 2);
+        ret = dmaengine_slave_config(iprtd->dma_chan, &slave_config);
+        if (ret)
+                return ret;
 
         return 0;
-out:
-        imx_dma_free(iprtd->dma);
-        return ret;
 }
 
 static int snd_imx_pcm_hw_params(struct snd_pcm_substream *substream,
                                 struct snd_pcm_hw_params *params)
 {
+        struct snd_soc_pcm_runtime *rtd = substream->private_data;
         struct snd_pcm_runtime *runtime = substream->runtime;
         struct imx_pcm_runtime_data *iprtd = runtime->private_data;
-        int i;
         unsigned long dma_addr;
+        struct dma_chan *chan;
+        struct imx_pcm_dma_params *dma_params;
+        int ret;
 
-        imx_ssi_dma_alloc(substream);
+        dma_params = snd_soc_dai_get_dma_data(rtd->cpu_dai, substream);
+        ret = imx_ssi_dma_alloc(substream, params);
+        if (ret)
+                return ret;
+        chan = iprtd->dma_chan;
 
         iprtd->size = params_buffer_bytes(params);
         iprtd->periods = params_periods(params);
-        iprtd->period = params_period_bytes(params);
+        iprtd->period_bytes = params_period_bytes(params);
         iprtd->offset = 0;
         iprtd->period_time = HZ / (params_rate(params) /
                         params_period_size(params));
 
         snd_pcm_set_runtime_buffer(substream, &substream->dma_buffer);
 
-        if (iprtd->sg_count != iprtd->periods) {
-                kfree(iprtd->sg_list);
-
-                iprtd->sg_list = kcalloc(iprtd->periods + 1,
-                                sizeof(struct scatterlist), GFP_KERNEL);
-                if (!iprtd->sg_list)
-                        return -ENOMEM;
-                iprtd->sg_count = iprtd->periods + 1;
-        }
-
-        sg_init_table(iprtd->sg_list, iprtd->sg_count);
         dma_addr = runtime->dma_addr;
 
-        for (i = 0; i < iprtd->periods; i++) {
-                iprtd->sg_list[i].page_link = 0;
-                iprtd->sg_list[i].offset = 0;
-                iprtd->sg_list[i].dma_address = dma_addr;
-                iprtd->sg_list[i].length = iprtd->period;
-                dma_addr += iprtd->period;
+        iprtd->buf = (unsigned int *)substream->dma_buffer.area;
+
+        iprtd->desc = chan->device->device_prep_dma_cyclic(chan, dma_addr,
+                        iprtd->period_bytes * iprtd->periods,
+                        iprtd->period_bytes,
+                        substream->stream == SNDRV_PCM_STREAM_PLAYBACK ?
+                        DMA_TO_DEVICE : DMA_FROM_DEVICE);
+        if (!iprtd->desc) {
+                dev_err(&chan->dev->device, "cannot prepare slave dma\n");
+                return -EINVAL;
         }
 
-        /* close the loop */
-        iprtd->sg_list[iprtd->sg_count - 1].offset = 0;
-        iprtd->sg_list[iprtd->sg_count - 1].length = 0;
-        iprtd->sg_list[iprtd->sg_count - 1].page_link =
-                        ((unsigned long) iprtd->sg_list | 0x01) & ~0x02;
+        iprtd->desc->callback = audio_dma_irq;
+        iprtd->desc->callback_param = substream;
+
         return 0;
 }
 
@@ -194,41 +176,21 @@ static int snd_imx_pcm_hw_free(struct snd_pcm_substream *substream)
         struct snd_pcm_runtime *runtime = substream->runtime;
         struct imx_pcm_runtime_data *iprtd = runtime->private_data;
 
-        if (iprtd->dma >= 0) {
-                imx_dma_free(iprtd->dma);
-                iprtd->dma = -EINVAL;
+        if (iprtd->dma_chan) {
+                dma_release_channel(iprtd->dma_chan);
+                iprtd->dma_chan = NULL;
         }
 
-        kfree(iprtd->sg_list);
-        iprtd->sg_list = NULL;
-
         return 0;
 }
 
 static int snd_imx_pcm_prepare(struct snd_pcm_substream *substream)
 {
-        struct snd_pcm_runtime *runtime = substream->runtime;
         struct snd_soc_pcm_runtime *rtd = substream->private_data;
         struct imx_pcm_dma_params *dma_params;
-        struct imx_pcm_runtime_data *iprtd = runtime->private_data;
-        int err;
 
         dma_params = snd_soc_dai_get_dma_data(rtd->cpu_dai, substream);
 
-        iprtd->substream = substream;
-        iprtd->buf = (unsigned int *)substream->dma_buffer.area;
-        iprtd->period_cnt = 0;
-
-        pr_debug("%s: buf: %p period: %d periods: %d\n",
-                        __func__, iprtd->buf, iprtd->period, iprtd->periods);
-
-        err = imx_dma_setup_sg(iprtd->dma, iprtd->sg_list, iprtd->sg_count,
-                        IMX_DMA_LENGTH_LOOP, dma_params->dma_addr,
-                        substream->stream == SNDRV_PCM_STREAM_PLAYBACK ?
-                        DMA_MODE_WRITE : DMA_MODE_READ);
-        if (err)
-                return err;
-
         return 0;
 }
 
@@ -241,14 +203,14 @@ static int snd_imx_pcm_trigger(struct snd_pcm_substream *substream, int cmd)
         case SNDRV_PCM_TRIGGER_START:
         case SNDRV_PCM_TRIGGER_RESUME:
         case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:
-                imx_dma_enable(iprtd->dma);
+                dmaengine_submit(iprtd->desc);
 
                 break;
 
         case SNDRV_PCM_TRIGGER_STOP:
         case SNDRV_PCM_TRIGGER_SUSPEND:
         case SNDRV_PCM_TRIGGER_PAUSE_PUSH:
-                imx_dma_disable(iprtd->dma);
+                dmaengine_terminate_all(iprtd->dma_chan);
 
                 break;
         default:
@@ -263,6 +225,9 @@ static snd_pcm_uframes_t snd_imx_pcm_pointer(struct snd_pcm_substream *substream
         struct snd_pcm_runtime *runtime = substream->runtime;
         struct imx_pcm_runtime_data *iprtd = runtime->private_data;
 
+        pr_debug("%s: %ld %ld\n", __func__, iprtd->offset,
+                        bytes_to_frames(substream->runtime, iprtd->offset));
+
         return bytes_to_frames(substream->runtime, iprtd->offset);
 }
 
@@ -279,7 +244,7 @@ static struct snd_pcm_hardware snd_imx_hardware = {
         .channels_max = 2,
         .buffer_bytes_max = IMX_SSI_DMABUF_SIZE,
         .period_bytes_min = 128,
-        .period_bytes_max = 16 * 1024,
+        .period_bytes_max = 65535, /* Limited by SDMA engine */
         .periods_min = 2,
         .periods_max = 255,
         .fifo_size = 0,
@@ -304,11 +269,23 @@ static int snd_imx_open(struct snd_pcm_substream *substream)
         }
 
         snd_soc_set_runtime_hwparams(substream, &snd_imx_hardware);
+
+        return 0;
+}
+
+static int snd_imx_close(struct snd_pcm_substream *substream)
+{
+        struct snd_pcm_runtime *runtime = substream->runtime;
+        struct imx_pcm_runtime_data *iprtd = runtime->private_data;
+
+        kfree(iprtd);
+
         return 0;
 }
 
 static struct snd_pcm_ops imx_pcm_ops = {
         .open           = snd_imx_open,
+        .close          = snd_imx_close,
         .ioctl          = snd_pcm_lib_ioctl,
         .hw_params      = snd_imx_pcm_hw_params,
         .hw_free        = snd_imx_pcm_hw_free,
@@ -340,7 +317,6 @@ static struct platform_driver imx_pcm_driver = {
                         .name = "imx-pcm-audio",
                         .owner = THIS_MODULE,
         },
-
         .probe = imx_soc_platform_probe,
         .remove = __devexit_p(imx_soc_platform_remove),
 };
@@ -356,4 +332,3 @@ static void __exit snd_imx_pcm_exit(void)
         platform_driver_unregister(&imx_pcm_driver);
 }
 module_exit(snd_imx_pcm_exit);
-
diff --git a/sound/soc/imx/imx-ssi.c b/sound/soc/imx/imx-ssi.c
index d4bd345b0a8d..d2d98c75ee8a 100644
--- a/sound/soc/imx/imx-ssi.c
+++ b/sound/soc/imx/imx-ssi.c
@@ -439,7 +439,22 @@ void imx_pcm_free(struct snd_pcm *pcm)
 }
 EXPORT_SYMBOL_GPL(imx_pcm_free);
 
+static int imx_ssi_dai_probe(struct snd_soc_dai *dai)
+{
+        struct imx_ssi *ssi = dev_get_drvdata(dai->dev);
+        uint32_t val;
+
+        snd_soc_dai_set_drvdata(dai, ssi);
+
+        val = SSI_SFCSR_TFWM0(ssi->dma_params_tx.burstsize) |
+                SSI_SFCSR_RFWM0(ssi->dma_params_rx.burstsize);
+        writel(val, ssi->base + SSI_SFCSR);
+
+        return 0;
+}
+
 static struct snd_soc_dai_driver imx_ssi_dai = {
+        .probe = imx_ssi_dai_probe,
         .playback = {
                 .channels_min = 2,
                 .channels_max = 2,
@@ -455,20 +470,6 @@ static struct snd_soc_dai_driver imx_ssi_dai = {
         .ops = &imx_ssi_pcm_dai_ops,
 };
 
-static int imx_ssi_dai_probe(struct snd_soc_dai *dai)
-{
-        struct imx_ssi *ssi = dev_get_drvdata(dai->dev);
-        uint32_t val;
-
-        snd_soc_dai_set_drvdata(dai, ssi);
-
-        val = SSI_SFCSR_TFWM0(ssi->dma_params_tx.burstsize) |
-                SSI_SFCSR_RFWM0(ssi->dma_params_rx.burstsize);
-        writel(val, ssi->base + SSI_SFCSR);
-
-        return 0;
-}
-
 static struct snd_soc_dai_driver imx_ac97_dai = {
         .probe = imx_ssi_dai_probe,
         .ac97_control = 1,
@@ -677,7 +678,17 @@ static int imx_ssi_probe(struct platform_device *pdev)
                 goto failed_register;
         }
 
-        ssi->soc_platform_pdev = platform_device_alloc("imx-fiq-pcm-audio", pdev->id);
+        ssi->soc_platform_pdev_fiq = platform_device_alloc("imx-fiq-pcm-audio", pdev->id);
+        if (!ssi->soc_platform_pdev_fiq)
+                goto failed_pdev_fiq_alloc;
+        platform_set_drvdata(ssi->soc_platform_pdev_fiq, ssi);
+        ret = platform_device_add(ssi->soc_platform_pdev_fiq);
+        if (ret) {
+                dev_err(&pdev->dev, "failed to add platform device\n");
+                goto failed_pdev_fiq_add;
+        }
+
+        ssi->soc_platform_pdev = platform_device_alloc("imx-pcm-audio", pdev->id);
         if (!ssi->soc_platform_pdev)
                 goto failed_pdev_alloc;
         platform_set_drvdata(ssi->soc_platform_pdev, ssi);
@@ -692,6 +703,9 @@ static int imx_ssi_probe(struct platform_device *pdev)
 failed_pdev_add:
         platform_device_put(ssi->soc_platform_pdev);
 failed_pdev_alloc:
+failed_pdev_fiq_add:
+        platform_device_put(ssi->soc_platform_pdev_fiq);
+failed_pdev_fiq_alloc:
         snd_soc_unregister_dai(&pdev->dev);
 failed_register:
 failed_ac97:
diff --git a/sound/soc/imx/imx-ssi.h b/sound/soc/imx/imx-ssi.h
index 53b780d9b2b0..a4406a134892 100644
--- a/sound/soc/imx/imx-ssi.h
+++ b/sound/soc/imx/imx-ssi.h
@@ -185,6 +185,9 @@
 
 #define DRV_NAME "imx-ssi"
 
+#include <linux/dmaengine.h>
+#include <mach/dma.h>
+
 struct imx_pcm_dma_params {
         int dma;
         unsigned long dma_addr;
@@ -212,6 +215,7 @@ struct imx_ssi {
         int enabled;
 
         struct platform_device *soc_platform_pdev;
+        struct platform_device *soc_platform_pdev_fiq;
 };
 
 struct snd_soc_platform *imx_ssi_fiq_init(struct platform_device *pdev,
diff --git a/sound/soc/imx/phycore-ac97.c b/sound/soc/imx/phycore-ac97.c
index 6a65dd705519..39f23734781a 100644
--- a/sound/soc/imx/phycore-ac97.c
+++ b/sound/soc/imx/phycore-ac97.c
@@ -20,9 +20,6 @@
 #include <sound/soc-dapm.h>
 #include <asm/mach-types.h>
 
-#include "../codecs/wm9712.h"
-#include "imx-ssi.h"
-
 static struct snd_soc_card imx_phycore;
 
 static struct snd_soc_ops imx_phycore_hifi_ops = {
@@ -41,7 +38,7 @@ static struct snd_soc_dai_link imx_phycore_dai_ac97[] = {
 };
 
 static struct snd_soc_card imx_phycore = {
-        .name           = "PhyCORE-audio",
+        .name           = "PhyCORE-ac97-audio",
         .dai_link       = imx_phycore_dai_ac97,
         .num_links      = ARRAY_SIZE(imx_phycore_dai_ac97),
 };
diff --git a/sound/soc/omap/omap-mcbsp.c b/sound/soc/omap/omap-mcbsp.c
index d211c9fa5a91..7e84f24b9a88 100644
--- a/sound/soc/omap/omap-mcbsp.c
+++ b/sound/soc/omap/omap-mcbsp.c
@@ -644,15 +644,23 @@ static int omap_mcbsp_dai_set_dai_sysclk(struct snd_soc_dai *cpu_dai,
 
 
         case OMAP_MCBSP_CLKR_SRC_CLKR:
+                if (cpu_class_is_omap1())
+                        break;
                 omap2_mcbsp1_mux_clkr_src(CLKR_SRC_CLKR);
                 break;
         case OMAP_MCBSP_CLKR_SRC_CLKX:
+                if (cpu_class_is_omap1())
+                        break;
                 omap2_mcbsp1_mux_clkr_src(CLKR_SRC_CLKX);
                 break;
         case OMAP_MCBSP_FSR_SRC_FSR:
+                if (cpu_class_is_omap1())
+                        break;
                 omap2_mcbsp1_mux_fsr_src(FSR_SRC_FSR);
                 break;
         case OMAP_MCBSP_FSR_SRC_FSX:
+                if (cpu_class_is_omap1())
+                        break;
                 omap2_mcbsp1_mux_fsr_src(FSR_SRC_FSX);
                 break;
         default:
diff --git a/sound/soc/pxa/corgi.c b/sound/soc/pxa/corgi.c
index 97e9423615c9..f451acd4935b 100644
--- a/sound/soc/pxa/corgi.c
+++ b/sound/soc/pxa/corgi.c
@@ -100,8 +100,13 @@ static int corgi_startup(struct snd_pcm_substream *substream)
         struct snd_soc_pcm_runtime *rtd = substream->private_data;
         struct snd_soc_codec *codec = rtd->codec;
 
+        mutex_lock(&codec->mutex);
+
         /* check the jack status at stream startup */
         corgi_ext_control(codec);
+
+        mutex_unlock(&codec->mutex);
+
         return 0;
 }
 
diff --git a/sound/soc/pxa/magician.c b/sound/soc/pxa/magician.c
index b8207ced4072..5ef0526924b9 100644
--- a/sound/soc/pxa/magician.c
+++ b/sound/soc/pxa/magician.c
@@ -72,9 +72,13 @@ static int magician_startup(struct snd_pcm_substream *substream)
         struct snd_soc_pcm_runtime *rtd = substream->private_data;
         struct snd_soc_codec *codec = rtd->codec;
 
+        mutex_lock(&codec->mutex);
+
         /* check the jack status at stream startup */
         magician_ext_control(codec);
 
+        mutex_unlock(&codec->mutex);
+
         return 0;
 }
 
diff --git a/sound/soc/pxa/poodle.c b/sound/soc/pxa/poodle.c
index af84ee9c5e11..84edd0385a21 100644
--- a/sound/soc/pxa/poodle.c
+++ b/sound/soc/pxa/poodle.c
@@ -77,8 +77,13 @@ static int poodle_startup(struct snd_pcm_substream *substream)
         struct snd_soc_pcm_runtime *rtd = substream->private_data;
         struct snd_soc_codec *codec = rtd->codec;
 
+        mutex_lock(&codec->mutex);
+
         /* check the jack status at stream startup */
         poodle_ext_control(codec);
+
+        mutex_unlock(&codec->mutex);
+
         return 0;
 }
 
diff --git a/sound/soc/pxa/spitz.c b/sound/soc/pxa/spitz.c
index f470f360f4dd..0b30d7de24ec 100644
--- a/sound/soc/pxa/spitz.c
+++ b/sound/soc/pxa/spitz.c
@@ -108,8 +108,13 @@ static int spitz_startup(struct snd_pcm_substream *substream)
         struct snd_soc_pcm_runtime *rtd = substream->private_data;
         struct snd_soc_codec *codec = rtd->codec;
 
+        mutex_lock(&codec->mutex);
+
         /* check the jack status at stream startup */
         spitz_ext_control(codec);
+
+        mutex_unlock(&codec->mutex);
+
         return 0;
 }
 
diff --git a/sound/soc/pxa/tosa.c b/sound/soc/pxa/tosa.c
index 73d0edd8ded9..7b983f935454 100644
--- a/sound/soc/pxa/tosa.c
+++ b/sound/soc/pxa/tosa.c
@@ -81,8 +81,13 @@ static int tosa_startup(struct snd_pcm_substream *substream)
         struct snd_soc_pcm_runtime *rtd = substream->private_data;
         struct snd_soc_codec *codec = rtd->codec;
 
+        mutex_lock(&codec->mutex);
+
         /* check the jack status at stream startup */
         tosa_ext_control(codec);
+
+        mutex_unlock(&codec->mutex);
+
         return 0;
 }
 
diff --git a/sound/soc/s3c24xx/Kconfig b/sound/soc/s3c24xx/Kconfig
index 8a6b53ccd203..d85bf8a0abb2 100644
--- a/sound/soc/s3c24xx/Kconfig
+++ b/sound/soc/s3c24xx/Kconfig
@@ -2,6 +2,7 @@ config SND_S3C24XX_SOC
         tristate "SoC Audio for the Samsung S3CXXXX chips"
         depends on ARCH_S3C2410 || ARCH_S3C64XX || ARCH_S5PC100 || ARCH_S5PV210
         select S3C64XX_DMA if ARCH_S3C64XX
+        select S3C2410_DMA if ARCH_S3C2410
         help
           Say Y or M if you want to add support for codecs attached to
           the S3C24XX AC97 or I2S interfaces. You will also need to
diff --git a/sound/soc/s3c24xx/rx1950_uda1380.c b/sound/soc/s3c24xx/rx1950_uda1380.c
index ffd5cf2fb0a9..468cc11fdf47 100644
--- a/sound/soc/s3c24xx/rx1950_uda1380.c
+++ b/sound/soc/s3c24xx/rx1950_uda1380.c
@@ -50,7 +50,6 @@ static unsigned int rates[] = {
         16000,
         44100,
         48000,
-        88200,
 };
 
 static struct snd_pcm_hw_constraint_list hw_rates = {
@@ -130,7 +129,6 @@ static const struct snd_soc_dapm_route audio_map[] = {
 };
 
 static struct platform_device *s3c24xx_snd_device;
-static struct clk *xtal;
 
 static int rx1950_startup(struct snd_pcm_substream *substream)
 {
@@ -179,10 +177,8 @@ static int rx1950_hw_params(struct snd_pcm_substream *substream,
         case 44100:
         case 88200:
                 clk_source = S3C24XX_CLKSRC_MPLL;
-                fs_mode = S3C2410_IISMOD_256FS;
-                div = clk_get_rate(xtal) / (256 * rate);
-                if (clk_get_rate(xtal) % (256 * rate) > (128 * rate))
-                        div++;
+                fs_mode = S3C2410_IISMOD_384FS;
+                div = 1;
                 break;
         default:
                 printk(KERN_ERR "%s: rate %d is not supported\n",
@@ -210,7 +206,7 @@ static int rx1950_hw_params(struct snd_pcm_substream *substream,
 
         /* set MCLK division for sample rate */
         ret = snd_soc_dai_set_clkdiv(cpu_dai, S3C24XX_DIV_MCLK,
-                S3C2410_IISMOD_384FS);
+                fs_mode);
         if (ret < 0)
                 return ret;
 
@@ -295,17 +291,8 @@ static int __init rx1950_init(void)
                 goto err_plat_add;
         }
 
-        xtal = clk_get(&s3c24xx_snd_device->dev, "xtal");
-
-        if (IS_ERR(xtal)) {
-                ret = PTR_ERR(xtal);
-                platform_device_unregister(s3c24xx_snd_device);
-                goto err_clk;
-        }
-
         return 0;
 
-err_clk:
 err_plat_add:
 err_plat_alloc:
 err_gpio_conf:
@@ -320,7 +307,6 @@ static void __exit rx1950_exit(void)
         platform_device_unregister(s3c24xx_snd_device);
         snd_soc_jack_free_gpios(&hp_jack, ARRAY_SIZE(hp_jack_gpios),
                 hp_jack_gpios);
-        clk_put(xtal);
         gpio_free(S3C2410_GPA(1));
 }
 
diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c
index 614a8b30d87b..441285ade024 100644
--- a/sound/soc/soc-core.c
+++ b/sound/soc/soc-core.c
@@ -3043,8 +3043,10 @@ int snd_soc_register_dais(struct device *dev,
         for (i = 0; i < count; i++) {
 
                 dai = kzalloc(sizeof(struct snd_soc_dai), GFP_KERNEL);
-                if (dai == NULL)
-                        return -ENOMEM;
+                if (dai == NULL) {
+                        ret = -ENOMEM;
+                        goto err;
+                }
 
                 /* create DAI component name */
                 dai->name = fmt_multiple_name(dev, &dai_drv[i]);
@@ -3263,9 +3265,6 @@ int snd_soc_register_codec(struct device *dev,
         return 0;
 
 error:
-        for (i--; i >= 0; i--)
-                snd_soc_unregister_dai(dev);
-
         if (codec->reg_cache)
                 kfree(codec->reg_cache);
         kfree(codec->name);
diff --git a/sound/soc/soc-dapm.c b/sound/soc/soc-dapm.c
index 7d85c6496afa..75ed6491222d 100644
--- a/sound/soc/soc-dapm.c
+++ b/sound/soc/soc-dapm.c
@@ -683,12 +683,12 @@ static int dapm_seq_compare(struct snd_soc_dapm_widget *a,
                             struct snd_soc_dapm_widget *b,
                             int sort[])
 {
-        if (a->codec != b->codec)
-                return (unsigned long)a - (unsigned long)b;
         if (sort[a->id] != sort[b->id])
                 return sort[a->id] - sort[b->id];
         if (a->reg != b->reg)
                 return a->reg - b->reg;
+        if (a->codec != b->codec)
+                return (unsigned long)a->codec - (unsigned long)b->codec;
 
         return 0;
 }
diff --git a/sound/sound_core.c b/sound/sound_core.c
index c03bbaefdbc3..5580aced8730 100644
--- a/sound/sound_core.c
+++ b/sound/sound_core.c
@@ -104,7 +104,6 @@ module_exit(cleanup_soundcore);
 
 #include <linux/init.h>
 #include <linux/slab.h>
-#include <linux/smp_lock.h>
 #include <linux/types.h>
 #include <linux/kernel.h>
 #include <linux/sound.h>
diff --git a/sound/spi/at73c213.c b/sound/spi/at73c213.c
index 1bc56b2b94e2..337a00241a1f 100644
--- a/sound/spi/at73c213.c
+++ b/sound/spi/at73c213.c
@@ -155,7 +155,7 @@ static int snd_at73c213_set_bitrate(struct snd_at73c213 *chip)
         if (max_tries < 1)
                 max_tries = 1;
 
-        /* ssc_div must be a power of 2. */
+        /* ssc_div must be even. */
         ssc_div = (ssc_div + 1) & ~1UL;
 
         if ((ssc_rate / (ssc_div * 2 * 16)) < BITRATE_MIN) {