15 files changed, 549 insertions, 65 deletions

diff --git a/Documentation/DMA-mapping.txt b/Documentation/DMA-mapping.txt
index 684557474c15..ee4bb73683cd 100644
--- a/Documentation/DMA-mapping.txt
+++ b/Documentation/DMA-mapping.txt
@@ -199,6 +199,8 @@ address during PCI bus mastering you might do something like:
                        "mydev: 24-bit DMA addressing not available.\n");
                 goto ignore_this_device;
         }
+[Better use DMA_24BIT_MASK instead of 0x00ffffff.
+See linux/include/dma-mapping.h for reference.]
 
 When pci_set_dma_mask() is successful, and returns zero, the PCI layer
 saves away this mask you have provided.  The PCI layer will use this
diff --git a/Documentation/RCU/whatisRCU.txt b/Documentation/RCU/whatisRCU.txt
index b4ea51ad3610..07cb93b82ba9 100644
--- a/Documentation/RCU/whatisRCU.txt
+++ b/Documentation/RCU/whatisRCU.txt
@@ -605,7 +605,7 @@ are the same as those shown in the preceding section, so they are omitted.
         {
                 int cpu;
 
-                for_each_cpu(cpu)
+                for_each_possible_cpu(cpu)
                         run_on(cpu);
         }
 
diff --git a/Documentation/arm/SA1100/Assabet b/Documentation/arm/SA1100/Assabet
index cbbe5587c78d..78bc1c1b04e5 100644
--- a/Documentation/arm/SA1100/Assabet
+++ b/Documentation/arm/SA1100/Assabet
@@ -26,7 +26,7 @@ Installing a bootloader
 
 A couple of bootloaders able to boot Linux on Assabet are available:
 
-BLOB (http://www.lart.tudelft.nl/lartware/blob/)
+BLOB (http://www.lartmaker.nl/lartware/blob/)
 
    BLOB is a bootloader used within the LART project.  Some contributed
    patches were merged into BLOB to add support for Assabet.
diff --git a/Documentation/arm/SA1100/LART b/Documentation/arm/SA1100/LART
index 2f73f513e16a..6d412b685598 100644
--- a/Documentation/arm/SA1100/LART
+++ b/Documentation/arm/SA1100/LART
@@ -11,4 +11,4 @@ is under development, with plenty of others in different stages of
 planning.
 
 The hardware designs for this board have been released under an open license;
-see the LART page at http://www.lart.tudelft.nl/ for more information.
+see the LART page at http://www.lartmaker.nl/ for more information.
diff --git a/Documentation/block/biodoc.txt b/Documentation/block/biodoc.txt
index 8e63831971d5..f989a9e839b4 100644
--- a/Documentation/block/biodoc.txt
+++ b/Documentation/block/biodoc.txt
@@ -132,8 +132,18 @@ Some new queue property settings:
                 limit. No highmem default.
 
         blk_queue_max_sectors(q, max_sectors)
-                Maximum size request you can handle in units of 512 byte
-                sectors. 255 default.
+                Sets two variables that limit the size of the request.
+
+                - The request queue's max_sectors, which is a soft size in
+                in units of 512 byte sectors, and could be dynamically varied
+                by the core kernel.
+
+                - The request queue's max_hw_sectors, which is a hard limit
+                and reflects the maximum size request a driver can handle
+                in units of 512 byte sectors.
+
+                The default for both max_sectors and max_hw_sectors is
+                255. The upper limit of max_sectors is 1024.
 
         blk_queue_max_phys_segments(q, max_segments)
                 Maximum physical segments you can handle in a request. 128
diff --git a/Documentation/cpu-hotplug.txt b/Documentation/cpu-hotplug.txt
index 57a09f99ecb0..1bcf69996c9d 100644
--- a/Documentation/cpu-hotplug.txt
+++ b/Documentation/cpu-hotplug.txt
@@ -97,13 +97,13 @@ at which time hotplug is disabled.
 
 You really dont need to manipulate any of the system cpu maps. They should
 be read-only for most use. When setting up per-cpu resources almost always use
-cpu_possible_map/for_each_cpu() to iterate.
+cpu_possible_map/for_each_possible_cpu() to iterate.
 
 Never use anything other than cpumask_t to represent bitmap of CPUs.
 
 #include <linux/cpumask.h>
 
-for_each_cpu              - Iterate over cpu_possible_map
+for_each_possible_cpu     - Iterate over cpu_possible_map
 for_each_online_cpu       - Iterate over cpu_online_map
 for_each_present_cpu      - Iterate over cpu_present_map
 for_each_cpu_mask(x,mask) - Iterate over some random collection of cpu mask.
diff --git a/Documentation/cputopology.txt b/Documentation/cputopology.txt
index ff280e2e1613..2b28e9ec4e3a 100644
--- a/Documentation/cputopology.txt
+++ b/Documentation/cputopology.txt
@@ -1,5 +1,5 @@
 
-Export cpu topology info by sysfs. Items (attributes) are similar
+Export cpu topology info via sysfs. Items (attributes) are similar
 to /proc/cpuinfo.
 
 1) /sys/devices/system/cpu/cpuX/topology/physical_package_id:
@@ -12,7 +12,7 @@ represent the thread siblings to cpu X in the same core;
 represent the thread siblings to cpu X in the same physical package;
 
 To implement it in an architecture-neutral way, a new source file,
-driver/base/topology.c, is to export the 5 attributes.
+drivers/base/topology.c, is to export the 4 attributes.
 
 If one architecture wants to support this feature, it just needs to
 implement 4 defines, typically in file include/asm-XXX/topology.h.
diff --git a/Documentation/filesystems/00-INDEX b/Documentation/filesystems/00-INDEX
index 74052d22d868..66fdc0744fe0 100644
--- a/Documentation/filesystems/00-INDEX
+++ b/Documentation/filesystems/00-INDEX
@@ -1,27 +1,47 @@
 00-INDEX
         - this file (info on some of the filesystems supported by linux).
+Exporting
+        - explanation of how to make filesystems exportable.
 Locking
         - info on locking rules as they pertain to Linux VFS.
 adfs.txt
         - info and mount options for the Acorn Advanced Disc Filing System.
+afs.txt
+        - info and examples for the distributed AFS (Andrew File System) fs.
 affs.txt
         - info and mount options for the Amiga Fast File System.
+automount-support.txt
+        - information about filesystem automount support.
+befs.txt
+        - information about the BeOS filesystem for Linux.
 bfs.txt
         - info for the SCO UnixWare Boot Filesystem (BFS).
 cifs.txt
-        - description of the CIFS filesystem
+        - description of the CIFS filesystem.
 coda.txt
         - description of the CODA filesystem.
 configfs/
         - directory containing configfs documentation and example code.
 cramfs.txt
-        - info on the cram filesystem for small storage (ROMs etc)
+        - info on the cram filesystem for small storage (ROMs etc).
+dentry-locking.txt
+        - info on the RCU-based dcache locking model.
 devfs/
         - directory containing devfs documentation.
+directory-locking
+        - info about the locking scheme used for directory operations.
 dlmfs.txt
         - info on the userspace interface to the OCFS2 DLM.
 ext2.txt
         - info, mount options and specifications for the Ext2 filesystem.
+ext3.txt
+        - info, mount options and specifications for the Ext3 filesystem.
+files.txt
+        - info on file management in the Linux kernel.
+fuse.txt
+        - info on the Filesystem in User SpacE including mount options.
+hfs.txt
+        - info on the Macintosh HFS Filesystem for Linux.
 hpfs.txt
         - info and mount options for the OS/2 HPFS.
 isofs.txt
@@ -32,23 +52,43 @@ ncpfs.txt
         - info on Novell Netware(tm) filesystem using NCP protocol.
 ntfs.txt
         - info and mount options for the NTFS filesystem (Windows NT).
-proc.txt
-        - info on Linux's /proc filesystem.
 ocfs2.txt
         - info and mount options for the OCFS2 clustered filesystem.
+porting
+        - various information on filesystem porting.
+proc.txt
+        - info on Linux's /proc filesystem.
+ramfs-rootfs-initramfs.txt
+        - info on the 'in memory' filesystems ramfs, rootfs and initramfs.
+reiser4.txt
+        - info on the Reiser4 filesystem based on dancing tree algorithms.
+relayfs.txt
+        - info on relayfs, for efficient streaming from kernel to user space.
 romfs.txt
-        - Description of the ROMFS filesystem.
+        - description of the ROMFS filesystem.
 smbfs.txt
-        - info on using filesystems with the SMB protocol (Windows 3.11 and NT)
+        - info on using filesystems with the SMB protocol (Win 3.11 and NT).
+spufs.txt
+        - info and mount options for the SPU filesystem used on Cell.
+sysfs-pci.txt
+        - info on accessing PCI device resources through sysfs.
+sysfs.txt
+        - info on sysfs, a ram-based filesystem for exporting kernel objects.
 sysv-fs.txt
         - info on the SystemV/V7/Xenix/Coherent filesystem.
+tmpfs.txt
+        - info on tmpfs, a filesystem that holds all files in virtual memory.
 udf.txt
         - info and mount options for the UDF filesystem.
 ufs.txt
         - info on the ufs filesystem.
+v9fs.txt
+        - v9fs is a Unix implementation of the Plan 9 9p remote fs protocol.
 vfat.txt
         - info on using the VFAT filesystem used in Windows NT and Windows 95
 vfs.txt
-        - Overview of the Virtual File System
+        - overview of the Virtual File System
 xfs.txt
         - info and mount options for the XFS filesystem.
+xip.txt
+        - info on execute-in-place for file mappings.
diff --git a/Documentation/ioctl-number.txt b/Documentation/ioctl-number.txt
index aa7ba00ec082..171a44ebd939 100644
--- a/Documentation/ioctl-number.txt
+++ b/Documentation/ioctl-number.txt
@@ -78,8 +78,6 @@ Code	Seq#	Include File		Comments
 '#'     00-3F   IEEE 1394 Subsystem     Block for the entire subsystem
 '1'     00-1F   <linux/timepps.h>       PPS kit from Ulrich Windl
                                         <ftp://ftp.de.kernel.org/pub/linux/daemons/ntp/PPS/>
-'6'     00-10   <asm-i386/processor.h>  Intel IA32 microcode update driver
-                                        <mailto:tigran@veritas.com>
 '8'     all                             SNP8023 advanced NIC card
                                         <mailto:mcr@solidum.com>
 'A'     00-1F   linux/apm_bios.h
diff --git a/Documentation/m68k/README.buddha b/Documentation/m68k/README.buddha
index bf802ffc98ad..ef484a719bb9 100644
--- a/Documentation/m68k/README.buddha
+++ b/Documentation/m68k/README.buddha
@@ -29,7 +29,7 @@ address is written to $4a, then the whole Byte is written to
 $48, while it doesn't matter how often you're writing to $4a
 as  long as $48 is not touched.  After $48 has been written,
 the  whole card disappears from $e8 and is mapped to the new
-address just written.  Make shure $4a is written before $48,
+address just written.  Make sure $4a is written before $48,
 otherwise your chance is only 1:16 to find the board :-).
 
 The local memory-map is even active when mapped to $e8:
diff --git a/Documentation/networking/ifenslave.c b/Documentation/networking/ifenslave.c
index 545447ac503a..a12059886755 100644
--- a/Documentation/networking/ifenslave.c
+++ b/Documentation/networking/ifenslave.c
@@ -87,7 +87,7 @@
  *         would fail and generate an error message in the system log.
  *       - For opt_c: slave should not be set to the master's setting
  *         while it is running. It was already set during enslave. To
- *         simplify things, it is now handeled separately.
+ *         simplify things, it is now handled separately.
  *
  *    - 2003/12/01 - Shmulik Hen <shmulik.hen at intel dot com>
  *       - Code cleanup and style changes
diff --git a/Documentation/pnp.txt b/Documentation/pnp.txt
index af0f6eabfa1c..9529c9c9fd59 100644
--- a/Documentation/pnp.txt
+++ b/Documentation/pnp.txt
@@ -115,6 +115,9 @@ pnp_unregister_protocol
 pnp_register_driver
 - adds a PnP driver to the Plug and Play Layer
 - this includes driver model integration
+- returns zero for success or a negative error number for failure; count
+  calls to the .add() method if you need to know how many devices bind to
+  the driver
 
 pnp_unregister_driver
 - removes a PnP driver from the Plug and Play Layer
diff --git a/Documentation/robust-futex-ABI.txt b/Documentation/robust-futex-ABI.txt
new file mode 100644
index 000000000000..8529a17ffaa1
--- /dev/null
+++ b/Documentation/robust-futex-ABI.txt
@@ -0,0 +1,182 @@
+Started by Paul Jackson <pj@sgi.com>
+
+The robust futex ABI
+--------------------
+
+Robust_futexes provide a mechanism that is used in addition to normal
+futexes, for kernel assist of cleanup of held locks on task exit.
+
+The interesting data as to what futexes a thread is holding is kept on a
+linked list in user space, where it can be updated efficiently as locks
+are taken and dropped, without kernel intervention.  The only additional
+kernel intervention required for robust_futexes above and beyond what is
+required for futexes is:
+
+ 1) a one time call, per thread, to tell the kernel where its list of
+    held robust_futexes begins, and
+ 2) internal kernel code at exit, to handle any listed locks held
+    by the exiting thread.
+
+The existing normal futexes already provide a "Fast Userspace Locking"
+mechanism, which handles uncontested locking without needing a system
+call, and handles contested locking by maintaining a list of waiting
+threads in the kernel.  Options on the sys_futex(2) system call support
+waiting on a particular futex, and waking up the next waiter on a
+particular futex.
+
+For robust_futexes to work, the user code (typically in a library such
+as glibc linked with the application) has to manage and place the
+necessary list elements exactly as the kernel expects them.  If it fails
+to do so, then improperly listed locks will not be cleaned up on exit,
+probably causing deadlock or other such failure of the other threads
+waiting on the same locks.
+
+A thread that anticipates possibly using robust_futexes should first
+issue the system call:
+
+    asmlinkage long
+    sys_set_robust_list(struct robust_list_head __user *head, size_t len);
+
+The pointer 'head' points to a structure in the threads address space
+consisting of three words.  Each word is 32 bits on 32 bit arch's, or 64
+bits on 64 bit arch's, and local byte order.  Each thread should have
+its own thread private 'head'.
+
+If a thread is running in 32 bit compatibility mode on a 64 native arch
+kernel, then it can actually have two such structures - one using 32 bit
+words for 32 bit compatibility mode, and one using 64 bit words for 64
+bit native mode.  The kernel, if it is a 64 bit kernel supporting 32 bit
+compatibility mode, will attempt to process both lists on each task
+exit, if the corresponding sys_set_robust_list() call has been made to
+setup that list.
+
+  The first word in the memory structure at 'head' contains a
+  pointer to a single linked list of 'lock entries', one per lock,
+  as described below.  If the list is empty, the pointer will point
+  to itself, 'head'.  The last 'lock entry' points back to the 'head'.
+
+  The second word, called 'offset', specifies the offset from the
+  address of the associated 'lock entry', plus or minus, of what will
+  be called the 'lock word', from that 'lock entry'.  The 'lock word'
+  is always a 32 bit word, unlike the other words above.  The 'lock
+  word' holds 3 flag bits in the upper 3 bits, and the thread id (TID)
+  of the thread holding the lock in the bottom 29 bits.  See further
+  below for a description of the flag bits.
+
+  The third word, called 'list_op_pending', contains transient copy of
+  the address of the 'lock entry', during list insertion and removal,
+  and is needed to correctly resolve races should a thread exit while
+  in the middle of a locking or unlocking operation.
+
+Each 'lock entry' on the single linked list starting at 'head' consists
+of just a single word, pointing to the next 'lock entry', or back to
+'head' if there are no more entries.  In addition, nearby to each 'lock
+entry', at an offset from the 'lock entry' specified by the 'offset'
+word, is one 'lock word'.
+
+The 'lock word' is always 32 bits, and is intended to be the same 32 bit
+lock variable used by the futex mechanism, in conjunction with
+robust_futexes.  The kernel will only be able to wakeup the next thread
+waiting for a lock on a threads exit if that next thread used the futex
+mechanism to register the address of that 'lock word' with the kernel.
+
+For each futex lock currently held by a thread, if it wants this
+robust_futex support for exit cleanup of that lock, it should have one
+'lock entry' on this list, with its associated 'lock word' at the
+specified 'offset'.  Should a thread die while holding any such locks,
+the kernel will walk this list, mark any such locks with a bit
+indicating their holder died, and wakeup the next thread waiting for
+that lock using the futex mechanism.
+
+When a thread has invoked the above system call to indicate it
+anticipates using robust_futexes, the kernel stores the passed in 'head'
+pointer for that task.  The task may retrieve that value later on by
+using the system call:
+
+    asmlinkage long
+    sys_get_robust_list(int pid, struct robust_list_head __user **head_ptr,
+                        size_t __user *len_ptr);
+
+It is anticipated that threads will use robust_futexes embedded in
+larger, user level locking structures, one per lock.  The kernel
+robust_futex mechanism doesn't care what else is in that structure, so
+long as the 'offset' to the 'lock word' is the same for all
+robust_futexes used by that thread.  The thread should link those locks
+it currently holds using the 'lock entry' pointers.  It may also have
+other links between the locks, such as the reverse side of a double
+linked list, but that doesn't matter to the kernel.
+
+By keeping its locks linked this way, on a list starting with a 'head'
+pointer known to the kernel, the kernel can provide to a thread the
+essential service available for robust_futexes, which is to help clean
+up locks held at the time of (a perhaps unexpectedly) exit.
+
+Actual locking and unlocking, during normal operations, is handled
+entirely by user level code in the contending threads, and by the
+existing futex mechanism to wait for, and wakeup, locks.  The kernels
+only essential involvement in robust_futexes is to remember where the
+list 'head' is, and to walk the list on thread exit, handling locks
+still held by the departing thread, as described below.
+
+There may exist thousands of futex lock structures in a threads shared
+memory, on various data structures, at a given point in time. Only those
+lock structures for locks currently held by that thread should be on
+that thread's robust_futex linked lock list a given time.
+
+A given futex lock structure in a user shared memory region may be held
+at different times by any of the threads with access to that region. The
+thread currently holding such a lock, if any, is marked with the threads
+TID in the lower 29 bits of the 'lock word'.
+
+When adding or removing a lock from its list of held locks, in order for
+the kernel to correctly handle lock cleanup regardless of when the task
+exits (perhaps it gets an unexpected signal 9 in the middle of
+manipulating this list), the user code must observe the following
+protocol on 'lock entry' insertion and removal:
+
+On insertion:
+ 1) set the 'list_op_pending' word to the address of the 'lock word'
+    to be inserted,
+ 2) acquire the futex lock,
+ 3) add the lock entry, with its thread id (TID) in the bottom 29 bits
+    of the 'lock word', to the linked list starting at 'head', and
+ 4) clear the 'list_op_pending' word.
+
+On removal:
+ 1) set the 'list_op_pending' word to the address of the 'lock word'
+    to be removed,
+ 2) remove the lock entry for this lock from the 'head' list,
+ 2) release the futex lock, and
+ 2) clear the 'lock_op_pending' word.
+
+On exit, the kernel will consider the address stored in
+'list_op_pending' and the address of each 'lock word' found by walking
+the list starting at 'head'.  For each such address, if the bottom 29
+bits of the 'lock word' at offset 'offset' from that address equals the
+exiting threads TID, then the kernel will do two things:
+
+ 1) if bit 31 (0x80000000) is set in that word, then attempt a futex
+    wakeup on that address, which will waken the next thread that has
+    used to the futex mechanism to wait on that address, and
+ 2) atomically set  bit 30 (0x40000000) in the 'lock word'.
+
+In the above, bit 31 was set by futex waiters on that lock to indicate
+they were waiting, and bit 30 is set by the kernel to indicate that the
+lock owner died holding the lock.
+
+The kernel exit code will silently stop scanning the list further if at
+any point:
+
+ 1) the 'head' pointer or an subsequent linked list pointer
+    is not a valid address of a user space word
+ 2) the calculated location of the 'lock word' (address plus
+    'offset') is not the valud address of a 32 bit user space
+    word
+ 3) if the list contains more than 1 million (subject to
+    future kernel configuration changes) elements.
+
+When the kernel sees a list entry whose 'lock word' doesn't have the
+current threads TID in the lower 29 bits, it does nothing with that
+entry, and goes on to the next entry.
+
+Bit 29 (0x20000000) of the 'lock word' is reserved for future use.
diff --git a/Documentation/robust-futexes.txt b/Documentation/robust-futexes.txt
new file mode 100644
index 000000000000..df82d75245a0
--- /dev/null
+++ b/Documentation/robust-futexes.txt
@@ -0,0 +1,218 @@
+Started by: Ingo Molnar <mingo@redhat.com>
+
+Background
+----------
+
+what are robust futexes? To answer that, we first need to understand
+what futexes are: normal futexes are special types of locks that in the
+noncontended case can be acquired/released from userspace without having
+to enter the kernel.
+
+A futex is in essence a user-space address, e.g. a 32-bit lock variable
+field. If userspace notices contention (the lock is already owned and
+someone else wants to grab it too) then the lock is marked with a value
+that says "there's a waiter pending", and the sys_futex(FUTEX_WAIT)
+syscall is used to wait for the other guy to release it. The kernel
+creates a 'futex queue' internally, so that it can later on match up the
+waiter with the waker - without them having to know about each other.
+When the owner thread releases the futex, it notices (via the variable
+value) that there were waiter(s) pending, and does the
+sys_futex(FUTEX_WAKE) syscall to wake them up.  Once all waiters have
+taken and released the lock, the futex is again back to 'uncontended'
+state, and there's no in-kernel state associated with it. The kernel
+completely forgets that there ever was a futex at that address. This
+method makes futexes very lightweight and scalable.
+
+"Robustness" is about dealing with crashes while holding a lock: if a
+process exits prematurely while holding a pthread_mutex_t lock that is
+also shared with some other process (e.g. yum segfaults while holding a
+pthread_mutex_t, or yum is kill -9-ed), then waiters for that lock need
+to be notified that the last owner of the lock exited in some irregular
+way.
+
+To solve such types of problems, "robust mutex" userspace APIs were
+created: pthread_mutex_lock() returns an error value if the owner exits
+prematurely - and the new owner can decide whether the data protected by
+the lock can be recovered safely.
+
+There is a big conceptual problem with futex based mutexes though: it is
+the kernel that destroys the owner task (e.g. due to a SEGFAULT), but
+the kernel cannot help with the cleanup: if there is no 'futex queue'
+(and in most cases there is none, futexes being fast lightweight locks)
+then the kernel has no information to clean up after the held lock!
+Userspace has no chance to clean up after the lock either - userspace is
+the one that crashes, so it has no opportunity to clean up. Catch-22.
+
+In practice, when e.g. yum is kill -9-ed (or segfaults), a system reboot
+is needed to release that futex based lock. This is one of the leading
+bugreports against yum.
+
+To solve this problem, the traditional approach was to extend the vma
+(virtual memory area descriptor) concept to have a notion of 'pending
+robust futexes attached to this area'. This approach requires 3 new
+syscall variants to sys_futex(): FUTEX_REGISTER, FUTEX_DEREGISTER and
+FUTEX_RECOVER. At do_exit() time, all vmas are searched to see whether
+they have a robust_head set. This approach has two fundamental problems
+left:
+
+ - it has quite complex locking and race scenarios. The vma-based
+   approach had been pending for years, but they are still not completely
+   reliable.
+
+ - they have to scan _every_ vma at sys_exit() time, per thread!
+
+The second disadvantage is a real killer: pthread_exit() takes around 1
+microsecond on Linux, but with thousands (or tens of thousands) of vmas
+every pthread_exit() takes a millisecond or more, also totally
+destroying the CPU's L1 and L2 caches!
+
+This is very much noticeable even for normal process sys_exit_group()
+calls: the kernel has to do the vma scanning unconditionally! (this is
+because the kernel has no knowledge about how many robust futexes there
+are to be cleaned up, because a robust futex might have been registered
+in another task, and the futex variable might have been simply mmap()-ed
+into this process's address space).
+
+This huge overhead forced the creation of CONFIG_FUTEX_ROBUST so that
+normal kernels can turn it off, but worse than that: the overhead makes
+robust futexes impractical for any type of generic Linux distribution.
+
+So something had to be done.
+
+New approach to robust futexes
+------------------------------
+
+At the heart of this new approach there is a per-thread private list of
+robust locks that userspace is holding (maintained by glibc) - which
+userspace list is registered with the kernel via a new syscall [this
+registration happens at most once per thread lifetime]. At do_exit()
+time, the kernel checks this user-space list: are there any robust futex
+locks to be cleaned up?
+
+In the common case, at do_exit() time, there is no list registered, so
+the cost of robust futexes is just a simple current->robust_list != NULL
+comparison. If the thread has registered a list, then normally the list
+is empty. If the thread/process crashed or terminated in some incorrect
+way then the list might be non-empty: in this case the kernel carefully
+walks the list [not trusting it], and marks all locks that are owned by
+this thread with the FUTEX_OWNER_DEAD bit, and wakes up one waiter (if
+any).
+
+The list is guaranteed to be private and per-thread at do_exit() time,
+so it can be accessed by the kernel in a lockless way.
+
+There is one race possible though: since adding to and removing from the
+list is done after the futex is acquired by glibc, there is a few
+instructions window for the thread (or process) to die there, leaving
+the futex hung. To protect against this possibility, userspace (glibc)
+also maintains a simple per-thread 'list_op_pending' field, to allow the
+kernel to clean up if the thread dies after acquiring the lock, but just
+before it could have added itself to the list. Glibc sets this
+list_op_pending field before it tries to acquire the futex, and clears
+it after the list-add (or list-remove) has finished.
+
+That's all that is needed - all the rest of robust-futex cleanup is done
+in userspace [just like with the previous patches].
+
+Ulrich Drepper has implemented the necessary glibc support for this new
+mechanism, which fully enables robust mutexes.
+
+Key differences of this userspace-list based approach, compared to the
+vma based method:
+
+ - it's much, much faster: at thread exit time, there's no need to loop
+   over every vma (!), which the VM-based method has to do. Only a very
+   simple 'is the list empty' op is done.
+
+ - no VM changes are needed - 'struct address_space' is left alone.
+
+ - no registration of individual locks is needed: robust mutexes dont
+   need any extra per-lock syscalls. Robust mutexes thus become a very
+   lightweight primitive - so they dont force the application designer
+   to do a hard choice between performance and robustness - robust
+   mutexes are just as fast.
+
+ - no per-lock kernel allocation happens.
+
+ - no resource limits are needed.
+
+ - no kernel-space recovery call (FUTEX_RECOVER) is needed.
+
+ - the implementation and the locking is "obvious", and there are no
+   interactions with the VM.
+
+Performance
+-----------
+
+I have benchmarked the time needed for the kernel to process a list of 1
+million (!) held locks, using the new method [on a 2GHz CPU]:
+
+ - with FUTEX_WAIT set [contended mutex]: 130 msecs
+ - without FUTEX_WAIT set [uncontended mutex]: 30 msecs
+
+I have also measured an approach where glibc does the lock notification
+[which it currently does for !pshared robust mutexes], and that took 256
+msecs - clearly slower, due to the 1 million FUTEX_WAKE syscalls
+userspace had to do.
+
+(1 million held locks are unheard of - we expect at most a handful of
+locks to be held at a time. Nevertheless it's nice to know that this
+approach scales nicely.)
+
+Implementation details
+----------------------
+
+The patch adds two new syscalls: one to register the userspace list, and
+one to query the registered list pointer:
+
+ asmlinkage long
+ sys_set_robust_list(struct robust_list_head __user *head,
+                     size_t len);
+
+ asmlinkage long
+ sys_get_robust_list(int pid, struct robust_list_head __user **head_ptr,
+                     size_t __user *len_ptr);
+
+List registration is very fast: the pointer is simply stored in
+current->robust_list. [Note that in the future, if robust futexes become
+widespread, we could extend sys_clone() to register a robust-list head
+for new threads, without the need of another syscall.]
+
+So there is virtually zero overhead for tasks not using robust futexes,
+and even for robust futex users, there is only one extra syscall per
+thread lifetime, and the cleanup operation, if it happens, is fast and
+straightforward. The kernel doesnt have any internal distinction between
+robust and normal futexes.
+
+If a futex is found to be held at exit time, the kernel sets the
+following bit of the futex word:
+
+        #define FUTEX_OWNER_DIED        0x40000000
+
+and wakes up the next futex waiter (if any). User-space does the rest of
+the cleanup.
+
+Otherwise, robust futexes are acquired by glibc by putting the TID into
+the futex field atomically. Waiters set the FUTEX_WAITERS bit:
+
+        #define FUTEX_WAITERS           0x80000000
+
+and the remaining bits are for the TID.
+
+Testing, architecture support
+-----------------------------
+
+i've tested the new syscalls on x86 and x86_64, and have made sure the
+parsing of the userspace list is robust [ ;-) ] even if the list is
+deliberately corrupted.
+
+i386 and x86_64 syscalls are wired up at the moment, and Ulrich has
+tested the new glibc code (on x86_64 and i386), and it works for his
+robust-mutex testcases.
+
+All other architectures should build just fine too - but they wont have
+the new syscalls yet.
+
+Architectures need to implement the new futex_atomic_cmpxchg_inatomic()
+inline function before writing up the syscalls (that function returns
+-ENOSYS right now).
diff --git a/Documentation/rpc-cache.txt b/Documentation/rpc-cache.txt
index 2b5d4434fa5a..5f757c8cf979 100644
--- a/Documentation/rpc-cache.txt
+++ b/Documentation/rpc-cache.txt
@@ -1,4 +1,4 @@
-This document gives a brief introduction to the caching
+        This document gives a brief introduction to the caching
 mechanisms in the sunrpc layer that is used, in particular,
 for NFS authentication.
 
@@ -25,25 +25,17 @@ The common code handles such things as:
    - supporting 'NEGATIVE' as well as positive entries
    - allowing an EXPIRED time on cache items, and removing
      items after they expire, and are no longe in-use.
-
-   Future code extensions are expect to handle
    - making requests to user-space to fill in cache entries
    - allowing user-space to directly set entries in the cache
    - delaying RPC requests that depend on as-yet incomplete
      cache entries, and replaying those requests when the cache entry
      is complete.
-   - maintaining last-access times on cache entries
-   - clean out old entries when the caches become full
-
-The code for performing a cache lookup is also common, but in the form
-of a template.  i.e. a #define.
-Each cache defines a lookup function by using the DefineCacheLookup
-macro, or the simpler DefineSimpleCacheLookup macro
+   - clean out old entries as they expire.
 
 Creating a Cache
 ----------------
 
-1/ A cache needs a datum to cache.  This is in the form of a
+1/ A cache needs a datum to store.  This is in the form of a
    structure definition that must contain a
      struct cache_head
    as an element, usually the first.
@@ -51,35 +43,69 @@ Creating a Cache
    Each cache element is reference counted and contains
    expiry and update times for use in cache management.
 2/ A cache needs a "cache_detail" structure that
-   describes the cache.  This stores the hash table, and some
-   parameters for cache management.
-3/ A cache needs a lookup function.  This is created using
-   the DefineCacheLookup macro.  This lookup function is used both
-   to find entries and to update entries.  The normal mode for
-   updating an entry is to replace the old entry with a new
-   entry.  However it is possible to allow update-in-place
-   for those caches where it makes sense (no atomicity issues
-   or indirect reference counting issue)
-4/ A cache needs to be registered using cache_register().  This
-   includes in on a list of caches that will be regularly
-   cleaned to discard old data.  For this to work, some
-   thread must periodically call cache_clean
-   
+   describes the cache.  This stores the hash table, some
+   parameters for cache management, and some operations detailing how
+   to work with particular cache items.
+   The operations requires are:
+        struct cache_head *alloc(void)
+                This simply allocates appropriate memory and returns
+                a pointer to the cache_detail embedded within the
+                structure
+        void cache_put(struct kref *)
+                This is called when the last reference to an item is
+                is dropped.  The pointer passed is to the 'ref' field
+                in the cache_head.  cache_put should release any
+                references create by 'cache_init' and, if CACHE_VALID
+                is set, any references created by cache_update.
+                It should then release the memory allocated by
+                'alloc'.
+        int match(struct cache_head *orig, struct cache_head *new)
+                test if the keys in the two structures match.  Return
+                1 if they do, 0 if they don't.
+        void init(struct cache_head *orig, struct cache_head *new)
+                Set the 'key' fields in 'new' from 'orig'.  This may
+                include taking references to shared objects.
+        void update(struct cache_head *orig, struct cache_head *new)
+                Set the 'content' fileds in 'new' from 'orig'.
+        int cache_show(struct seq_file *m, struct cache_detail *cd,
+                        struct cache_head *h)
+                Optional.  Used to provide a /proc file that lists the
+                contents of a cache.  This should show one item,
+                usually on just one line.
+        int cache_request(struct cache_detail *cd, struct cache_head *h,
+                char **bpp, int *blen)
+                Format a request to be send to user-space for an item
+                to be instantiated.  *bpp is a buffer of size *blen.
+                bpp should be moved forward over the encoded message,
+                and  *blen should be reduced to show how much free
+                space remains.  Return 0 on success or <0 if not
+                enough room or other problem.
+        int cache_parse(struct cache_detail *cd, char *buf, int len)
+                A message from user space has arrived to fill out a
+                cache entry.  It is in 'buf' of length 'len'.
+                cache_parse should parse this, find the item in the
+                cache with sunrpc_cache_lookup, and update the item
+                with sunrpc_cache_update.
+
+
+3/ A cache needs to be registered using cache_register().  This
+   includes it on a list of caches that will be regularly
+   cleaned to discard old data.
+
 Using a cache
 -------------
 
-To find a value in a cache, call the lookup function passing it a the
-datum which contains key, and possibly content, and a flag saying
-whether to update the cache with new data from the datum.   Depending
-on how the cache lookup function was defined, it may take an extra
-argument to identify the particular cache in question.
+To find a value in a cache, call sunrpc_cache_lookup passing a pointer
+to the cache_head in a sample item with the 'key' fields filled in.
+This will be passed to ->match to identify the target entry.  If no
+entry is found, a new entry will be create, added to the cache, and
+marked as not containing valid data.
 
-Except in cases of kmalloc failure, the lookup function
-will return a new datum which will store the key and
-may contain valid content, or may not.
-This datum is typically passed to cache_check which determines the
-validity of the datum and may later initiate an upcall to fill
-in the data.
+The item returned is typically passed to cache_check which will check
+if the data is valid, and may initiate an up-call to get fresh data.
+cache_check will return -ENOENT in the entry is negative or if an up
+call is needed but not possible, -EAGAIN if an upcall is pending,
+or 0 if the data is valid;
 
 cache_check can be passed a "struct cache_req *".  This structure is
 typically embedded in the actual request and can be used to create a
@@ -90,6 +116,13 @@ item does become valid, the deferred copy of the request will be
 revisited (->revisit).  It is expected that this method will
 reschedule the request for processing.
 
+The value returned by sunrpc_cache_lookup can also be passed to
+sunrpc_cache_update to set the content for the item.  A second item is
+passed which should hold the content.  If the item found by _lookup
+has valid data, then it is discarded and a new item is created.  This
+saves any user of an item from worrying about content changing while
+it is being inspected.  If the item found by _lookup does not contain
+valid data, then the content is copied across and CACHE_VALID is set.
 
 Populating a cache
 ------------------
@@ -114,8 +147,8 @@ should be create or updated to have the given content, and the
 expiry time should be set on that item.
 
 Reading from a channel is a bit more interesting.  When a cache
-lookup fail, or when it suceeds but finds an entry that may soon
-expiry, a request is lodged for that cache item to be updated by
+lookup fails, or when it succeeds but finds an entry that may soon
+expire, a request is lodged for that cache item to be updated by
 user-space.  These requests appear in the channel file.
 
 Successive reads will return successive requests.
@@ -130,7 +163,7 @@ Thus a user-space helper is likely to:
     write a response
   loop.
 
-If it dies and needs to be restarted, any requests that have not be
+If it dies and needs to be restarted, any requests that have not been
 answered will still appear in the file and will be read by the new
 instance of the helper.
 
@@ -142,10 +175,9 @@ Each cache should also define a "cache_request" method which
 takes a cache item and encodes a request into the buffer
 provided.
 
-
 Note: If a cache has no active readers on the channel, and has had not
 active readers for more than 60 seconds, further requests will not be
-added to the channel but instead all looks that do not find a valid
+added to the channel but instead all lookups that do not find a valid
 entry will fail.  This is partly for backward compatibility: The
 previous nfs exports table was deemed to be authoritative and a
 failed lookup meant a definite 'no'.
@@ -154,18 +186,17 @@ request/response format
 -----------------------
 
 While each cache is free to use it's own format for requests
-and responses over channel, the following is recommended are
+and responses over channel, the following is recommended as
 appropriate and support routines are available to help:
 Each request or response record should be printable ASCII
 with precisely one newline character which should be at the end.
 Fields within the record should be separated by spaces, normally one.
 If spaces, newlines, or nul characters are needed in a field they
-much be quotes.  two mechanisms are available:
+much be quoted.  two mechanisms are available:
 1/ If a field begins '\x' then it must contain an even number of
    hex digits, and pairs of these digits provide the bytes in the
    field.
 2/ otherwise a \ in the field must be followed by 3 octal digits
    which give the code for a byte.  Other characters are treated
-   as them selves.  At the very least, space, newlines nul, and
+   as them selves.  At the very least, space, newline, nul, and
    '\' must be quoted in this way.
-