aboutsummaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
-rw-r--r--Documentation/ABI/testing/debugfs-pktcdvd6
-rw-r--r--Documentation/block/biodoc.txt19
-rw-r--r--Documentation/cgroups/memory.txt55
-rw-r--r--Documentation/cgroups/resource_counter.txt27
-rw-r--r--Documentation/networking/bonding.txt2
-rw-r--r--Documentation/powerpc/dts-bindings/fsl/i2c.txt46
-rw-r--r--Documentation/sound/alsa/HD-Audio.txt4
-rw-r--r--Documentation/sysctl/net.txt2
-rw-r--r--Documentation/tomoyo.txt55
-rw-r--r--Documentation/vm/00-INDEX2
-rw-r--r--Documentation/vm/active_mm.txt83
-rw-r--r--Documentation/vm/unevictable-lru.txt1041
-rw-r--r--MAINTAINERS46
-rw-r--r--Makefile9
-rw-r--r--arch/arm/mach-omap2/usb-musb.c8
-rw-r--r--arch/ia64/kernel/pci-swiotlb.c2
-rw-r--r--arch/microblaze/include/asm/auxvec.h1
-rw-r--r--arch/microblaze/include/asm/cputable.h1
-rw-r--r--arch/microblaze/include/asm/ftrace.h1
-rw-r--r--arch/microblaze/include/asm/hw_irq.h1
-rw-r--r--arch/microblaze/include/asm/io.h1
-rw-r--r--arch/microblaze/include/asm/socket.h3
-rw-r--r--arch/microblaze/include/asm/user.h1
-rw-r--r--arch/microblaze/include/asm/vga.h1
-rw-r--r--arch/microblaze/kernel/of_device.c10
-rw-r--r--arch/microblaze/kernel/process.c3
-rw-r--r--arch/microblaze/kernel/prom.c1
-rw-r--r--arch/microblaze/kernel/ptrace.c1
-rw-r--r--arch/microblaze/kernel/signal.c1
-rw-r--r--arch/microblaze/kernel/sys_microblaze.c2
-rw-r--r--arch/mn10300/include/asm/bug.h10
-rw-r--r--arch/mn10300/include/asm/unistd.h2
-rw-r--r--arch/mn10300/kernel/entry.S2
-rw-r--r--arch/mn10300/kernel/setup.c4
-rw-r--r--arch/powerpc/Kconfig2
-rw-r--r--arch/powerpc/boot/dts/tqm8540.dts4
-rw-r--r--arch/powerpc/boot/dts/tqm8541.dts4
-rw-r--r--arch/powerpc/boot/dts/tqm8548-bigflash.dts8
-rw-r--r--arch/powerpc/boot/dts/tqm8548.dts8
-rw-r--r--arch/powerpc/boot/dts/tqm8555.dts4
-rw-r--r--arch/powerpc/boot/dts/tqm8560.dts4
-rw-r--r--arch/powerpc/configs/85xx/tqm8548_defconfig164
-rw-r--r--arch/powerpc/include/asm/futex.h12
-rw-r--r--arch/powerpc/include/asm/mmu.h6
-rw-r--r--arch/powerpc/include/asm/parport.h2
-rw-r--r--arch/powerpc/include/asm/ppc-opcode.h13
-rw-r--r--arch/powerpc/kernel/cputable.c2
-rw-r--r--arch/powerpc/mm/tlb_nohash.c1
-rw-r--r--arch/powerpc/mm/tlb_nohash_low.S14
-rw-r--r--arch/powerpc/platforms/pseries/dtl.c1
-rw-r--r--arch/powerpc/platforms/pseries/eeh_driver.c2
-rw-r--r--arch/s390/appldata/appldata_base.c2
-rw-r--r--arch/s390/include/asm/cpuid.h25
-rw-r--r--arch/s390/include/asm/kvm_host.h1
-rw-r--r--arch/s390/include/asm/lowcore.h12
-rw-r--r--arch/s390/include/asm/processor.h17
-rw-r--r--arch/s390/include/asm/ptrace.h2
-rw-r--r--arch/s390/include/asm/setup.h24
-rw-r--r--arch/s390/include/asm/thread_info.h3
-rw-r--r--arch/s390/include/asm/timer.h1
-rw-r--r--arch/s390/include/asm/timex.h5
-rw-r--r--arch/s390/include/asm/unistd.h4
-rw-r--r--arch/s390/kernel/asm-offsets.c2
-rw-r--r--arch/s390/kernel/compat_wrapper.S18
-rw-r--r--arch/s390/kernel/early.c19
-rw-r--r--arch/s390/kernel/entry.S13
-rw-r--r--arch/s390/kernel/entry64.S13
-rw-r--r--arch/s390/kernel/head.S15
-rw-r--r--arch/s390/kernel/nmi.c5
-rw-r--r--arch/s390/kernel/setup.c12
-rw-r--r--arch/s390/kernel/smp.c4
-rw-r--r--arch/s390/kernel/syscalls.S2
-rw-r--r--arch/s390/kernel/time.c77
-rw-r--r--arch/s390/kernel/vtime.c67
-rw-r--r--arch/sh/Kconfig18
-rw-r--r--arch/sh/boards/board-ap325rxa.c1
-rw-r--r--arch/sh/boards/board-urquell.c30
-rw-r--r--arch/sh/drivers/pci/ops-sh7785lcr.c5
-rw-r--r--arch/sh/drivers/pci/pci-sh7780.h2
-rw-r--r--arch/sh/drivers/pci/pci.c3
-rw-r--r--arch/sh/include/asm/dma-mapping.h36
-rw-r--r--arch/sh/include/asm/scatterlist.h11
-rw-r--r--arch/sh/include/asm/topology.h7
-rw-r--r--arch/sh/include/asm/unistd_32.h4
-rw-r--r--arch/sh/include/asm/unistd_64.h4
-rw-r--r--arch/sh/kernel/cpu/sh4a/setup-sh7786.c14
-rw-r--r--arch/sh/kernel/syscalls_32.S2
-rw-r--r--arch/sh/kernel/syscalls_64.S2
-rw-r--r--arch/sh/mm/consistent.c31
-rw-r--r--arch/sparc/include/asm/atomic_32.h2
-rw-r--r--arch/sparc/include/asm/parport.h5
-rw-r--r--arch/sparc/kernel/ldc.c6
-rw-r--r--arch/sparc/kernel/smp_64.c4
-rw-r--r--arch/x86/Kconfig1
-rw-r--r--arch/x86/include/asm/required-features.h2
-rw-r--r--arch/x86/include/asm/uv/uv_mmrs.h5
-rw-r--r--arch/x86/include/asm/xen/page.h3
-rw-r--r--arch/x86/kernel/apic/x2apic_uv_x.c16
-rw-r--r--arch/x86/kernel/bios_uv.c3
-rw-r--r--arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c8
-rw-r--r--arch/x86/kernel/microcode_core.c33
-rw-r--r--arch/x86/kernel/tlb_uv.c189
-rw-r--r--arch/x86/kernel/uv_sysfs.c4
-rw-r--r--arch/x86/xen/enlighten.c89
-rw-r--r--arch/x86/xen/mmu.c116
-rw-r--r--arch/x86/xen/mmu.h3
-rw-r--r--arch/x86/xen/smp.c4
-rw-r--r--arch/x86/xen/xen-ops.h2
-rw-r--r--block/as-iosched.c116
-rw-r--r--block/blk-barrier.c3
-rw-r--r--block/blk-sysfs.c4
-rw-r--r--block/blk.h4
-rw-r--r--block/cfq-iosched.c270
-rw-r--r--block/elevator.c8
-rw-r--r--block/ioctl.c2
-rw-r--r--block/scsi_ioctl.c6
-rw-r--r--drivers/ata/libata-core.c25
-rw-r--r--drivers/ata/libata-scsi.c30
-rw-r--r--drivers/ata/libata-sff.c27
-rw-r--r--drivers/ata/pata_hpt37x.c22
-rw-r--r--drivers/ata/pata_legacy.c34
-rw-r--r--drivers/ata/pata_ninja32.c4
-rw-r--r--drivers/atm/solos-pci.c2
-rw-r--r--drivers/block/brd.c5
-rw-r--r--drivers/block/cciss.c2
-rw-r--r--drivers/char/agp/intel-agp.c3
-rw-r--r--drivers/char/sysrq.c1
-rw-r--r--drivers/edac/edac_core.h12
-rw-r--r--drivers/edac/edac_device.c2
-rw-r--r--drivers/edac/edac_mc.c2
-rw-r--r--drivers/edac/edac_pci.c2
-rw-r--r--drivers/gpu/drm/i915/i915_drv.h7
-rw-r--r--drivers/gpu/drm/i915/i915_gem.c187
-rw-r--r--drivers/gpu/drm/i915/i915_gem_debugfs.c93
-rw-r--r--drivers/gpu/drm/i915/i915_gem_tiling.c111
-rw-r--r--drivers/gpu/drm/i915/intel_display.c2
-rw-r--r--drivers/gpu/drm/i915/intel_fb.c4
-rw-r--r--drivers/gpu/drm/i915/intel_hdmi.c23
-rw-r--r--drivers/gpu/drm/i915/intel_sdvo.c22
-rw-r--r--drivers/hwmon/Kconfig10
-rw-r--r--drivers/hwmon/Makefile1
-rw-r--r--drivers/hwmon/hp_accel.c1
-rw-r--r--drivers/hwmon/sht15.c692
-rw-r--r--drivers/md/dm-bio-list.h117
-rw-r--r--drivers/md/dm-delay.c2
-rw-r--r--drivers/md/dm-mpath.c1
-rw-r--r--drivers/md/dm-raid1.c1
-rw-r--r--drivers/md/dm-region-hash.c1
-rw-r--r--drivers/md/dm-snap.c1
-rw-r--r--drivers/md/dm.c1
-rw-r--r--drivers/md/raid1.c1
-rw-r--r--drivers/md/raid10.c1
-rw-r--r--drivers/misc/eeprom/at24.c8
-rw-r--r--drivers/misc/eeprom/at25.c5
-rw-r--r--drivers/misc/sgi-xp/xpc.h254
-rw-r--r--drivers/misc/sgi-xp/xpc_channel.c138
-rw-r--r--drivers/misc/sgi-xp/xpc_main.c128
-rw-r--r--drivers/misc/sgi-xp/xpc_partition.c20
-rw-r--r--drivers/misc/sgi-xp/xpc_sn2.c164
-rw-r--r--drivers/misc/sgi-xp/xpc_uv.c257
-rw-r--r--drivers/net/a2065.c17
-rw-r--r--drivers/net/ariadne.c19
-rw-r--r--drivers/net/arm/am79c961a.c24
-rw-r--r--drivers/net/arm/at91_ether.c32
-rw-r--r--drivers/net/arm/ep93xx_eth.c19
-rw-r--r--drivers/net/arm/ether1.c19
-rw-r--r--drivers/net/arm/ether3.c19
-rw-r--r--drivers/net/atarilance.c18
-rw-r--r--drivers/net/atl1c/atl1c_main.c4
-rw-r--r--drivers/net/au1000_eth.c23
-rw-r--r--drivers/net/benet/be_ethtool.c4
-rw-r--r--drivers/net/benet/be_main.c4
-rw-r--r--drivers/net/bfin_mac.c24
-rw-r--r--drivers/net/bonding/bond_alb.c2
-rw-r--r--drivers/net/bonding/bond_main.c5
-rw-r--r--drivers/net/bonding/bond_sysfs.c14
-rw-r--r--drivers/net/cris/eth_v10.c30
-rw-r--r--drivers/net/declance.c17
-rw-r--r--drivers/net/e1000/e1000_main.c48
-rw-r--r--drivers/net/e1000e/netdev.c61
-rw-r--r--drivers/net/ehea/ehea_main.c4
-rw-r--r--drivers/net/forcedeth.c4
-rw-r--r--drivers/net/fs_enet/fs_enet-main.c27
-rw-r--r--drivers/net/gianfar.c2
-rw-r--r--drivers/net/ibm_newemac/core.c41
-rw-r--r--drivers/net/igb/e1000_mac.c2
-rw-r--r--drivers/net/igb/e1000_mac.h1
-rw-r--r--drivers/net/igb/e1000_mbx.c17
-rw-r--r--drivers/net/igb/e1000_mbx.h2
-rw-r--r--drivers/net/igb/igb_main.c37
-rw-r--r--drivers/net/igbvf/igbvf.h3
-rw-r--r--drivers/net/igbvf/netdev.c9
-rw-r--r--drivers/net/igbvf/vf.c2
-rw-r--r--drivers/net/igbvf/vf.h1
-rw-r--r--drivers/net/ioc3-eth.c22
-rw-r--r--drivers/net/isa-skeleton.c20
-rw-r--r--drivers/net/ixgbe/ixgbe_82598.c59
-rw-r--r--drivers/net/ixgbe/ixgbe_82599.c40
-rw-r--r--drivers/net/ixgbe/ixgbe_common.c55
-rw-r--r--drivers/net/ixgbe/ixgbe_common.h3
-rw-r--r--drivers/net/ixgbe/ixgbe_ethtool.c24
-rw-r--r--drivers/net/ixgbe/ixgbe_main.c57
-rw-r--r--drivers/net/jme.c8
-rw-r--r--drivers/net/mac89x0.c18
-rw-r--r--drivers/net/macb.c19
-rw-r--r--drivers/net/macsonic.c19
-rw-r--r--drivers/net/myri10ge/myri10ge.c1
-rw-r--r--drivers/net/phy/fixed.c5
-rw-r--r--drivers/net/phy/marvell.c54
-rw-r--r--drivers/net/phy/phy.c14
-rw-r--r--drivers/net/sfc/efx.c7
-rw-r--r--drivers/net/sfc/falcon.c4
-rw-r--r--drivers/net/sh_eth.c21
-rw-r--r--drivers/net/skge.c4
-rw-r--r--drivers/net/sun3_82586.c19
-rw-r--r--drivers/net/tc35815.c27
-rw-r--r--drivers/net/tg3.c9
-rw-r--r--drivers/net/tsi108_eth.c20
-rw-r--r--drivers/net/tun.c5
-rw-r--r--drivers/net/via-velocity.c4
-rw-r--r--drivers/net/wireless/ath9k/pci.c4
-rw-r--r--drivers/net/wireless/p54/p54pci.c4
-rw-r--r--drivers/net/xtsonic.c19
-rw-r--r--drivers/parisc/superio.c3
-rw-r--r--drivers/s390/block/dasd.c37
-rw-r--r--drivers/s390/block/dasd_eckd.c16
-rw-r--r--drivers/s390/cio/qdio_main.c43
-rw-r--r--drivers/sbus/char/jsflash.c15
-rw-r--r--drivers/sbus/char/uctrl.c7
-rw-r--r--drivers/scsi/3w-9xxx.c8
-rw-r--r--drivers/scsi/aacraid/aachba.c2
-rw-r--r--drivers/scsi/lpfc/lpfc_scsi.c2
-rw-r--r--drivers/scsi/mpt2sas/mpt2sas_base.c10
-rw-r--r--drivers/sh/intc.c35
-rw-r--r--drivers/spi/spi.c22
-rw-r--r--drivers/staging/b3dfg/b3dfg.c2
-rw-r--r--drivers/usb/class/cdc-acm.c4
-rw-r--r--drivers/usb/otg/nop-usb-xceiv.c4
-rw-r--r--drivers/usb/serial/ti_usb_3410_5052.c26
-rw-r--r--drivers/video/aty/radeon_base.c4
-rw-r--r--drivers/video/backlight/backlight.c3
-rw-r--r--drivers/video/backlight/lcd.c3
-rw-r--r--drivers/video/cirrusfb.c4
-rw-r--r--drivers/video/console/fbcon.c55
-rw-r--r--drivers/video/efifb.c7
-rw-r--r--drivers/video/fbmem.c19
-rw-r--r--drivers/video/intelfb/intelfb.h2
-rw-r--r--drivers/video/intelfb/intelfb_i2c.c1
-rw-r--r--drivers/video/intelfb/intelfbdrv.c1
-rw-r--r--drivers/video/intelfb/intelfbhw.c5
-rw-r--r--drivers/video/s3fb.c6
-rw-r--r--drivers/video/sa1100fb.c15
-rw-r--r--drivers/video/sa1100fb.h7
-rw-r--r--drivers/video/sis/sis_main.c2
-rw-r--r--drivers/video/skeletonfb.c8
-rw-r--r--drivers/video/uvesafb.c35
-rw-r--r--drivers/video/vfb.c11
-rw-r--r--drivers/xen/cpu_hotplug.c40
-rw-r--r--drivers/xen/manage.c5
-rw-r--r--fs/bio.c18
-rw-r--r--fs/buffer.c56
-rw-r--r--fs/direct-io.c2
-rw-r--r--fs/ext2/inode.c44
-rw-r--r--fs/ext4/extents.c2
-rw-r--r--fs/fuse/file.c8
-rw-r--r--fs/gfs2/glock.c10
-rw-r--r--fs/gfs2/inode.c8
-rw-r--r--fs/gfs2/inode.h14
-rw-r--r--fs/gfs2/ops_file.c8
-rw-r--r--fs/gfs2/ops_fstype.c5
-rw-r--r--fs/gfs2/ops_inode.c1
-rw-r--r--fs/gfs2/quota.c4
-rw-r--r--fs/hfs/inode.c4
-rw-r--r--fs/hfs/mdb.c1
-rw-r--r--fs/inode.c36
-rw-r--r--fs/jbd/revoke.c24
-rw-r--r--fs/nilfs2/bmap.c5
-rw-r--r--fs/nilfs2/nilfs.h5
-rw-r--r--fs/nilfs2/recovery.c20
-rw-r--r--fs/nilfs2/sufile.c290
-rw-r--r--fs/nilfs2/sufile.h79
-rw-r--r--fs/nilfs2/super.c7
-rw-r--r--fs/nilfs2/the_nilfs.c4
-rw-r--r--fs/ocfs2/file.c94
-rw-r--r--fs/pipe.c42
-rw-r--r--fs/splice.c371
-rw-r--r--fs/xfs/linux-2.6/xfs_aops.c38
-rw-r--r--fs/xfs/linux-2.6/xfs_aops.h1
-rw-r--r--fs/xfs/linux-2.6/xfs_buf.c9
-rw-r--r--fs/xfs/linux-2.6/xfs_fs_subr.c14
-rw-r--r--fs/xfs/linux-2.6/xfs_lrw.c18
-rw-r--r--fs/xfs/linux-2.6/xfs_sync.c78
-rw-r--r--fs/xfs/linux-2.6/xfs_sync.h9
-rw-r--r--fs/xfs/xfs_iget.c23
-rw-r--r--fs/xfs/xfs_iomap.c61
-rw-r--r--fs/xfs/xfs_iomap.h3
-rw-r--r--fs/xfs/xfs_log.c78
-rw-r--r--fs/xfs/xfs_mount.h2
-rw-r--r--fs/xfs/xfs_vnodeops.c7
-rw-r--r--include/asm-generic/bug.h2
-rw-r--r--include/asm-generic/siginfo.h2
-rw-r--r--include/drm/drm_pciids.h2
-rw-r--r--include/drm/i915_drm.h3
-rw-r--r--include/linux/bio.h109
-rw-r--r--include/linux/buffer_head.h3
-rw-r--r--include/linux/debug_locks.h1
-rw-r--r--include/linux/fb.h8
-rw-r--r--include/linux/fiemap.h2
-rw-r--r--include/linux/fs.h64
-rw-r--r--include/linux/fsl_devices.h4
-rw-r--r--include/linux/init_task.h13
-rw-r--r--include/linux/libata.h8
-rw-r--r--include/linux/pci_ids.h2
-rw-r--r--include/linux/phy.h6
-rw-r--r--include/linux/pipe_fs_i.h5
-rw-r--r--include/linux/rcutree.h3
-rw-r--r--include/linux/sht15.h24
-rw-r--r--include/linux/splice.h12
-rw-r--r--include/linux/usb/serial.h7
-rw-r--r--include/net/udp.h2
-rw-r--r--include/sound/jack.h2
-rw-r--r--include/sound/pcm.h3
-rw-r--r--include/video/cyblafb.h175
-rw-r--r--init/initramfs.c5
-rw-r--r--ipc/mq_sysctl.c2
-rw-r--r--kernel/power/swap.c2
-rw-r--r--kernel/ptrace.c7
-rw-r--r--kernel/rcupdate.c18
-rw-r--r--kernel/rcutree.c19
-rw-r--r--kernel/rcutree_trace.c14
-rw-r--r--kernel/sched.c10
-rw-r--r--kernel/softirq.c4
-rw-r--r--kernel/sys.c24
-rw-r--r--kernel/sysctl.c20
-rw-r--r--lib/kobject_uevent.c2
-rw-r--r--mm/Kconfig2
-rw-r--r--mm/filemap.c5
-rw-r--r--mm/memcontrol.c2
-rw-r--r--mm/mmap.c2
-rw-r--r--mm/shmem.c27
-rw-r--r--mm/util.c16
-rw-r--r--net/802/tr.c3
-rw-r--r--net/8021q/vlan_core.c4
-rw-r--r--net/core/dev.c9
-rw-r--r--net/ipv4/tcp_input.c3
-rw-r--r--net/ipv4/udp.c3
-rw-r--r--net/ipv6/ipv6_sockglue.c4
-rw-r--r--net/ipv6/udp.c6
-rw-r--r--net/packet/af_packet.c5
-rw-r--r--net/rose/af_rose.c10
-rw-r--r--net/sched/em_meta.c6
-rw-r--r--security/tomoyo/common.c6
-rw-r--r--security/tomoyo/common.h2
-rw-r--r--security/tomoyo/domain.c2
-rw-r--r--security/tomoyo/file.c2
-rw-r--r--security/tomoyo/realpath.c2
-rw-r--r--security/tomoyo/realpath.h2
-rw-r--r--security/tomoyo/tomoyo.c2
-rw-r--r--security/tomoyo/tomoyo.h2
-rw-r--r--sound/core/control.c35
-rw-r--r--sound/core/jack.c3
-rw-r--r--sound/core/pcm_compat.c11
-rw-r--r--sound/core/pcm_lib.c47
-rw-r--r--sound/core/pcm_native.c93
-rw-r--r--sound/core/seq/seq_compat.c9
-rw-r--r--sound/core/timer.c11
-rw-r--r--sound/isa/sb/sb16_csp.c19
-rw-r--r--sound/isa/wavefront/wavefront_fx.c14
-rw-r--r--sound/isa/wavefront/wavefront_synth.c11
-rw-r--r--sound/pci/emu10k1/emufx.c41
-rw-r--r--sound/pci/hda/hda_codec.c6
-rw-r--r--sound/pci/hda/hda_intel.c47
-rw-r--r--sound/pci/hda/patch_conexant.c21
-rw-r--r--sound/pci/hda/patch_realtek.c5
-rw-r--r--sound/pci/hda/patch_sigmatel.c27
-rw-r--r--sound/pci/intel8x0.c91
-rw-r--r--sound/soc/pxa/magician.c2
-rw-r--r--sound/soc/s3c24xx/Kconfig6
-rw-r--r--sound/usb/caiaq/Makefile4
-rw-r--r--sound/usb/caiaq/audio.c (renamed from sound/usb/caiaq/caiaq-audio.c)12
-rw-r--r--sound/usb/caiaq/audio.h (renamed from sound/usb/caiaq/caiaq-audio.h)0
-rw-r--r--sound/usb/caiaq/control.c (renamed from sound/usb/caiaq/caiaq-control.c)10
-rw-r--r--sound/usb/caiaq/control.h (renamed from sound/usb/caiaq/caiaq-control.h)0
-rw-r--r--sound/usb/caiaq/device.c (renamed from sound/usb/caiaq/caiaq-device.c)23
-rw-r--r--sound/usb/caiaq/device.h (renamed from sound/usb/caiaq/caiaq-device.h)0
-rw-r--r--sound/usb/caiaq/input.c (renamed from sound/usb/caiaq/caiaq-input.c)11
-rw-r--r--sound/usb/caiaq/input.h (renamed from sound/usb/caiaq/caiaq-input.h)0
-rw-r--r--sound/usb/caiaq/midi.c (renamed from sound/usb/caiaq/caiaq-midi.c)13
-rw-r--r--sound/usb/caiaq/midi.h (renamed from sound/usb/caiaq/caiaq-midi.h)0
-rw-r--r--sound/usb/usx2y/us122l.c10
-rw-r--r--sound/usb/usx2y/usX2Yhwdep.c13
391 files changed, 6180 insertions, 3774 deletions
diff --git a/Documentation/ABI/testing/debugfs-pktcdvd b/Documentation/ABI/testing/debugfs-pktcdvd
index bf9c16b64c34..cf11736acb76 100644
--- a/Documentation/ABI/testing/debugfs-pktcdvd
+++ b/Documentation/ABI/testing/debugfs-pktcdvd
@@ -1,4 +1,4 @@
1What: /debug/pktcdvd/pktcdvd[0-7] 1What: /sys/kernel/debug/pktcdvd/pktcdvd[0-7]
2Date: Oct. 2006 2Date: Oct. 2006
3KernelVersion: 2.6.20 3KernelVersion: 2.6.20
4Contact: Thomas Maier <balagi@justmail.de> 4Contact: Thomas Maier <balagi@justmail.de>
@@ -10,10 +10,10 @@ debugfs interface
10The pktcdvd module (packet writing driver) creates 10The pktcdvd module (packet writing driver) creates
11these files in debugfs: 11these files in debugfs:
12 12
13/debug/pktcdvd/pktcdvd[0-7]/ 13/sys/kernel/debug/pktcdvd/pktcdvd[0-7]/
14 info (0444) Lots of driver statistics and infos. 14 info (0444) Lots of driver statistics and infos.
15 15
16Example: 16Example:
17------- 17-------
18 18
19cat /debug/pktcdvd/pktcdvd0/info 19cat /sys/kernel/debug/pktcdvd/pktcdvd0/info
diff --git a/Documentation/block/biodoc.txt b/Documentation/block/biodoc.txt
index ecad6ee75705..6fab97ea7e6b 100644
--- a/Documentation/block/biodoc.txt
+++ b/Documentation/block/biodoc.txt
@@ -1040,23 +1040,21 @@ Front merges are handled by the binary trees in AS and deadline schedulers.
1040iii. Plugging the queue to batch requests in anticipation of opportunities for 1040iii. Plugging the queue to batch requests in anticipation of opportunities for
1041 merge/sort optimizations 1041 merge/sort optimizations
1042 1042
1043This is just the same as in 2.4 so far, though per-device unplugging
1044support is anticipated for 2.5. Also with a priority-based i/o scheduler,
1045such decisions could be based on request priorities.
1046
1047Plugging is an approach that the current i/o scheduling algorithm resorts to so 1043Plugging is an approach that the current i/o scheduling algorithm resorts to so
1048that it collects up enough requests in the queue to be able to take 1044that it collects up enough requests in the queue to be able to take
1049advantage of the sorting/merging logic in the elevator. If the 1045advantage of the sorting/merging logic in the elevator. If the
1050queue is empty when a request comes in, then it plugs the request queue 1046queue is empty when a request comes in, then it plugs the request queue
1051(sort of like plugging the bottom of a vessel to get fluid to build up) 1047(sort of like plugging the bath tub of a vessel to get fluid to build up)
1052till it fills up with a few more requests, before starting to service 1048till it fills up with a few more requests, before starting to service
1053the requests. This provides an opportunity to merge/sort the requests before 1049the requests. This provides an opportunity to merge/sort the requests before
1054passing them down to the device. There are various conditions when the queue is 1050passing them down to the device. There are various conditions when the queue is
1055unplugged (to open up the flow again), either through a scheduled task or 1051unplugged (to open up the flow again), either through a scheduled task or
1056could be on demand. For example wait_on_buffer sets the unplugging going 1052could be on demand. For example wait_on_buffer sets the unplugging going
1057(by running tq_disk) so the read gets satisfied soon. So in the read case, 1053through sync_buffer() running blk_run_address_space(mapping). Or the caller
1058the queue gets explicitly unplugged as part of waiting for completion, 1054can do it explicity through blk_unplug(bdev). So in the read case,
1059in fact all queues get unplugged as a side-effect. 1055the queue gets explicitly unplugged as part of waiting for completion on that
1056buffer. For page driven IO, the address space ->sync_page() takes care of
1057doing the blk_run_address_space().
1060 1058
1061Aside: 1059Aside:
1062 This is kind of controversial territory, as it's not clear if plugging is 1060 This is kind of controversial territory, as it's not clear if plugging is
@@ -1067,11 +1065,6 @@ Aside:
1067 multi-page bios being queued in one shot, we may not need to wait to merge 1065 multi-page bios being queued in one shot, we may not need to wait to merge
1068 a big request from the broken up pieces coming by. 1066 a big request from the broken up pieces coming by.
1069 1067
1070 Per-queue granularity unplugging (still a Todo) may help reduce some of the
1071 concerns with just a single tq_disk flush approach. Something like
1072 blk_kick_queue() to unplug a specific queue (right away ?)
1073 or optionally, all queues, is in the plan.
1074
10754.4 I/O contexts 10684.4 I/O contexts
1076I/O contexts provide a dynamically allocated per process data area. They may 1069I/O contexts provide a dynamically allocated per process data area. They may
1077be used in I/O schedulers, and in the block layer (could be used for IO statis, 1070be used in I/O schedulers, and in the block layer (could be used for IO statis,
diff --git a/Documentation/cgroups/memory.txt b/Documentation/cgroups/memory.txt
index a98a7fe7aabb..1a608877b14e 100644
--- a/Documentation/cgroups/memory.txt
+++ b/Documentation/cgroups/memory.txt
@@ -6,15 +6,14 @@ used here with the memory controller that is used in hardware.
6 6
7Salient features 7Salient features
8 8
9a. Enable control of both RSS (mapped) and Page Cache (unmapped) pages 9a. Enable control of Anonymous, Page Cache (mapped and unmapped) and
10 Swap Cache memory pages.
10b. The infrastructure allows easy addition of other types of memory to control 11b. The infrastructure allows easy addition of other types of memory to control
11c. Provides *zero overhead* for non memory controller users 12c. Provides *zero overhead* for non memory controller users
12d. Provides a double LRU: global memory pressure causes reclaim from the 13d. Provides a double LRU: global memory pressure causes reclaim from the
13 global LRU; a cgroup on hitting a limit, reclaims from the per 14 global LRU; a cgroup on hitting a limit, reclaims from the per
14 cgroup LRU 15 cgroup LRU
15 16
16NOTE: Swap Cache (unmapped) is not accounted now.
17
18Benefits and Purpose of the memory controller 17Benefits and Purpose of the memory controller
19 18
20The memory controller isolates the memory behaviour of a group of tasks 19The memory controller isolates the memory behaviour of a group of tasks
@@ -290,34 +289,44 @@ will be charged as a new owner of it.
290 moved to the parent. If you want to avoid that, force_empty will be useful. 289 moved to the parent. If you want to avoid that, force_empty will be useful.
291 290
2925.2 stat file 2915.2 stat file
293 memory.stat file includes following statistics (now) 292
294 cache - # of pages from page-cache and shmem. 293memory.stat file includes following statistics
295 rss - # of pages from anonymous memory. 294
296 pgpgin - # of event of charging 295cache - # of bytes of page cache memory.
297 pgpgout - # of event of uncharging 296rss - # of bytes of anonymous and swap cache memory.
298 active_anon - # of pages on active lru of anon, shmem. 297pgpgin - # of pages paged in (equivalent to # of charging events).
299 inactive_anon - # of pages on active lru of anon, shmem 298pgpgout - # of pages paged out (equivalent to # of uncharging events).
300 active_file - # of pages on active lru of file-cache 299active_anon - # of bytes of anonymous and swap cache memory on active
301 inactive_file - # of pages on inactive lru of file cache 300 lru list.
302 unevictable - # of pages cannot be reclaimed.(mlocked etc) 301inactive_anon - # of bytes of anonymous memory and swap cache memory on
303 302 inactive lru list.
304 Below is depend on CONFIG_DEBUG_VM. 303active_file - # of bytes of file-backed memory on active lru list.
305 inactive_ratio - VM internal parameter. (see mm/page_alloc.c) 304inactive_file - # of bytes of file-backed memory on inactive lru list.
306 recent_rotated_anon - VM internal parameter. (see mm/vmscan.c) 305unevictable - # of bytes of memory that cannot be reclaimed (mlocked etc).
307 recent_rotated_file - VM internal parameter. (see mm/vmscan.c) 306
308 recent_scanned_anon - VM internal parameter. (see mm/vmscan.c) 307The following additional stats are dependent on CONFIG_DEBUG_VM.
309 recent_scanned_file - VM internal parameter. (see mm/vmscan.c) 308
310 309inactive_ratio - VM internal parameter. (see mm/page_alloc.c)
311 Memo: 310recent_rotated_anon - VM internal parameter. (see mm/vmscan.c)
311recent_rotated_file - VM internal parameter. (see mm/vmscan.c)
312recent_scanned_anon - VM internal parameter. (see mm/vmscan.c)
313recent_scanned_file - VM internal parameter. (see mm/vmscan.c)
314
315Memo:
312 recent_rotated means recent frequency of lru rotation. 316 recent_rotated means recent frequency of lru rotation.
313 recent_scanned means recent # of scans to lru. 317 recent_scanned means recent # of scans to lru.
314 showing for better debug please see the code for meanings. 318 showing for better debug please see the code for meanings.
315 319
320Note:
321 Only anonymous and swap cache memory is listed as part of 'rss' stat.
322 This should not be confused with the true 'resident set size' or the
323 amount of physical memory used by the cgroup. Per-cgroup rss
324 accounting is not done yet.
316 325
3175.3 swappiness 3265.3 swappiness
318 Similar to /proc/sys/vm/swappiness, but affecting a hierarchy of groups only. 327 Similar to /proc/sys/vm/swappiness, but affecting a hierarchy of groups only.
319 328
320 Following cgroup's swapiness can't be changed. 329 Following cgroups' swapiness can't be changed.
321 - root cgroup (uses /proc/sys/vm/swappiness). 330 - root cgroup (uses /proc/sys/vm/swappiness).
322 - a cgroup which uses hierarchy and it has child cgroup. 331 - a cgroup which uses hierarchy and it has child cgroup.
323 - a cgroup which uses hierarchy and not the root of hierarchy. 332 - a cgroup which uses hierarchy and not the root of hierarchy.
diff --git a/Documentation/cgroups/resource_counter.txt b/Documentation/cgroups/resource_counter.txt
index f196ac1d7d25..95b24d766eab 100644
--- a/Documentation/cgroups/resource_counter.txt
+++ b/Documentation/cgroups/resource_counter.txt
@@ -47,13 +47,18 @@ to work with it.
47 47
482. Basic accounting routines 482. Basic accounting routines
49 49
50 a. void res_counter_init(struct res_counter *rc) 50 a. void res_counter_init(struct res_counter *rc,
51 struct res_counter *rc_parent)
51 52
52 Initializes the resource counter. As usual, should be the first 53 Initializes the resource counter. As usual, should be the first
53 routine called for a new counter. 54 routine called for a new counter.
54 55
55 b. int res_counter_charge[_locked] 56 The struct res_counter *parent can be used to define a hierarchical
56 (struct res_counter *rc, unsigned long val) 57 child -> parent relationship directly in the res_counter structure,
58 NULL can be used to define no relationship.
59
60 c. int res_counter_charge(struct res_counter *rc, unsigned long val,
61 struct res_counter **limit_fail_at)
57 62
58 When a resource is about to be allocated it has to be accounted 63 When a resource is about to be allocated it has to be accounted
59 with the appropriate resource counter (controller should determine 64 with the appropriate resource counter (controller should determine
@@ -67,15 +72,25 @@ to work with it.
67 * if the charging is performed first, then it should be uncharged 72 * if the charging is performed first, then it should be uncharged
68 on error path (if the one is called). 73 on error path (if the one is called).
69 74
70 c. void res_counter_uncharge[_locked] 75 If the charging fails and a hierarchical dependency exists, the
76 limit_fail_at parameter is set to the particular res_counter element
77 where the charging failed.
78
79 d. int res_counter_charge_locked
80 (struct res_counter *rc, unsigned long val)
81
82 The same as res_counter_charge(), but it must not acquire/release the
83 res_counter->lock internally (it must be called with res_counter->lock
84 held).
85
86 e. void res_counter_uncharge[_locked]
71 (struct res_counter *rc, unsigned long val) 87 (struct res_counter *rc, unsigned long val)
72 88
73 When a resource is released (freed) it should be de-accounted 89 When a resource is released (freed) it should be de-accounted
74 from the resource counter it was accounted to. This is called 90 from the resource counter it was accounted to. This is called
75 "uncharging". 91 "uncharging".
76 92
77 The _locked routines imply that the res_counter->lock is taken. 93 The _locked routines imply that the res_counter->lock is taken.
78
79 94
80 2.1 Other accounting routines 95 2.1 Other accounting routines
81 96
diff --git a/Documentation/networking/bonding.txt b/Documentation/networking/bonding.txt
index 5ede7473b425..08762750f121 100644
--- a/Documentation/networking/bonding.txt
+++ b/Documentation/networking/bonding.txt
@@ -1242,7 +1242,7 @@ monitoring is enabled, and vice-versa.
1242To add ARP targets: 1242To add ARP targets:
1243# echo +192.168.0.100 > /sys/class/net/bond0/bonding/arp_ip_target 1243# echo +192.168.0.100 > /sys/class/net/bond0/bonding/arp_ip_target
1244# echo +192.168.0.101 > /sys/class/net/bond0/bonding/arp_ip_target 1244# echo +192.168.0.101 > /sys/class/net/bond0/bonding/arp_ip_target
1245 NOTE: up to 10 target addresses may be specified. 1245 NOTE: up to 16 target addresses may be specified.
1246 1246
1247To remove an ARP target: 1247To remove an ARP target:
1248# echo -192.168.0.100 > /sys/class/net/bond0/bonding/arp_ip_target 1248# echo -192.168.0.100 > /sys/class/net/bond0/bonding/arp_ip_target
diff --git a/Documentation/powerpc/dts-bindings/fsl/i2c.txt b/Documentation/powerpc/dts-bindings/fsl/i2c.txt
index d0ab33e21fe6..b6d2e21474f9 100644
--- a/Documentation/powerpc/dts-bindings/fsl/i2c.txt
+++ b/Documentation/powerpc/dts-bindings/fsl/i2c.txt
@@ -7,8 +7,10 @@ Required properties :
7 7
8Recommended properties : 8Recommended properties :
9 9
10 - compatible : Should be "fsl-i2c" for parts compatible with 10 - compatible : compatibility list with 2 entries, the first should
11 Freescale I2C specifications. 11 be "fsl,CHIP-i2c" where CHIP is the name of a compatible processor,
12 e.g. mpc8313, mpc8543, mpc8544, mpc5200 or mpc5200b. The second one
13 should be "fsl-i2c".
12 - interrupts : <a b> where a is the interrupt number and b is a 14 - interrupts : <a b> where a is the interrupt number and b is a
13 field that represents an encoding of the sense and level 15 field that represents an encoding of the sense and level
14 information for the interrupt. This should be encoded based on 16 information for the interrupt. This should be encoded based on
@@ -16,17 +18,31 @@ Recommended properties :
16 controller you have. 18 controller you have.
17 - interrupt-parent : the phandle for the interrupt controller that 19 - interrupt-parent : the phandle for the interrupt controller that
18 services interrupts for this device. 20 services interrupts for this device.
19 - dfsrr : boolean; if defined, indicates that this I2C device has 21 - fsl,preserve-clocking : boolean; if defined, the clock settings
20 a digital filter sampling rate register 22 from the bootloader are preserved (not touched).
21 - fsl5200-clocking : boolean; if defined, indicated that this device 23 - clock-frequency : desired I2C bus clock frequency in Hz.
22 uses the FSL 5200 clocking mechanism. 24
23 25Examples :
24Example : 26
25 i2c@3000 { 27 i2c@3d00 {
26 interrupt-parent = <40000>; 28 #address-cells = <1>;
27 interrupts = <1b 3>; 29 #size-cells = <0>;
28 reg = <3000 18>; 30 compatible = "fsl,mpc5200b-i2c","fsl,mpc5200-i2c","fsl-i2c";
29 device_type = "i2c"; 31 cell-index = <0>;
30 compatible = "fsl-i2c"; 32 reg = <0x3d00 0x40>;
31 dfsrr; 33 interrupts = <2 15 0>;
34 interrupt-parent = <&mpc5200_pic>;
35 fsl,preserve-clocking;
32 }; 36 };
37
38 i2c@3100 {
39 #address-cells = <1>;
40 #size-cells = <0>;
41 cell-index = <1>;
42 compatible = "fsl,mpc8544-i2c", "fsl-i2c";
43 reg = <0x3100 0x100>;
44 interrupts = <43 2>;
45 interrupt-parent = <&mpic>;
46 clock-frequency = <400000>;
47 };
48
diff --git a/Documentation/sound/alsa/HD-Audio.txt b/Documentation/sound/alsa/HD-Audio.txt
index c5948f2f9a25..88b7433d2f11 100644
--- a/Documentation/sound/alsa/HD-Audio.txt
+++ b/Documentation/sound/alsa/HD-Audio.txt
@@ -169,7 +169,7 @@ PCI SSID look-up.
169What `model` option values are available depends on the codec chip. 169What `model` option values are available depends on the codec chip.
170Check your codec chip from the codec proc file (see "Codec Proc-File" 170Check your codec chip from the codec proc file (see "Codec Proc-File"
171section below). It will show the vendor/product name of your codec 171section below). It will show the vendor/product name of your codec
172chip. Then, see Documentation/sound/alsa/HD-Audio-Modelstxt file, 172chip. Then, see Documentation/sound/alsa/HD-Audio-Models.txt file,
173the section of HD-audio driver. You can find a list of codecs 173the section of HD-audio driver. You can find a list of codecs
174and `model` options belonging to each codec. For example, for Realtek 174and `model` options belonging to each codec. For example, for Realtek
175ALC262 codec chip, pass `model=ultra` for devices that are compatible 175ALC262 codec chip, pass `model=ultra` for devices that are compatible
@@ -177,7 +177,7 @@ with Samsung Q1 Ultra.
177 177
178Thus, the first thing you can do for any brand-new, unsupported and 178Thus, the first thing you can do for any brand-new, unsupported and
179non-working HD-audio hardware is to check HD-audio codec and several 179non-working HD-audio hardware is to check HD-audio codec and several
180different `model` option values. If you have a luck, some of them 180different `model` option values. If you have any luck, some of them
181might suit with your device well. 181might suit with your device well.
182 182
183Some codecs such as ALC880 have a special model option `model=test`. 183Some codecs such as ALC880 have a special model option `model=test`.
diff --git a/Documentation/sysctl/net.txt b/Documentation/sysctl/net.txt
index a34d55b65441..df38ef046f8d 100644
--- a/Documentation/sysctl/net.txt
+++ b/Documentation/sysctl/net.txt
@@ -95,7 +95,7 @@ of struct cmsghdr structures with appended data.
95 95
96There is only one file in this directory. 96There is only one file in this directory.
97unix_dgram_qlen limits the max number of datagrams queued in Unix domain 97unix_dgram_qlen limits the max number of datagrams queued in Unix domain
98socket's buffer. It will not take effect unless PF_UNIX flag is spicified. 98socket's buffer. It will not take effect unless PF_UNIX flag is specified.
99 99
100 100
1013. /proc/sys/net/ipv4 - IPV4 settings 1013. /proc/sys/net/ipv4 - IPV4 settings
diff --git a/Documentation/tomoyo.txt b/Documentation/tomoyo.txt
new file mode 100644
index 000000000000..b3a232cae7f8
--- /dev/null
+++ b/Documentation/tomoyo.txt
@@ -0,0 +1,55 @@
1--- What is TOMOYO? ---
2
3TOMOYO is a name-based MAC extension (LSM module) for the Linux kernel.
4
5LiveCD-based tutorials are available at
6http://tomoyo.sourceforge.jp/en/1.6.x/1st-step/ubuntu8.04-live/
7http://tomoyo.sourceforge.jp/en/1.6.x/1st-step/centos5-live/ .
8Though these tutorials use non-LSM version of TOMOYO, they are useful for you
9to know what TOMOYO is.
10
11--- How to enable TOMOYO? ---
12
13Build the kernel with CONFIG_SECURITY_TOMOYO=y and pass "security=tomoyo" on
14kernel's command line.
15
16Please see http://tomoyo.sourceforge.jp/en/2.2.x/ for details.
17
18--- Where is documentation? ---
19
20User <-> Kernel interface documentation is available at
21http://tomoyo.sourceforge.jp/en/2.2.x/policy-reference.html .
22
23Materials we prepared for seminars and symposiums are available at
24http://sourceforge.jp/projects/tomoyo/docs/?category_id=532&language_id=1 .
25Below lists are chosen from three aspects.
26
27What is TOMOYO?
28 TOMOYO Linux Overview
29 http://sourceforge.jp/projects/tomoyo/docs/lca2009-takeda.pdf
30 TOMOYO Linux: pragmatic and manageable security for Linux
31 http://sourceforge.jp/projects/tomoyo/docs/freedomhectaipei-tomoyo.pdf
32 TOMOYO Linux: A Practical Method to Understand and Protect Your Own Linux Box
33 http://sourceforge.jp/projects/tomoyo/docs/PacSec2007-en-no-demo.pdf
34
35What can TOMOYO do?
36 Deep inside TOMOYO Linux
37 http://sourceforge.jp/projects/tomoyo/docs/lca2009-kumaneko.pdf
38 The role of "pathname based access control" in security.
39 http://sourceforge.jp/projects/tomoyo/docs/lfj2008-bof.pdf
40
41History of TOMOYO?
42 Realities of Mainlining
43 http://sourceforge.jp/projects/tomoyo/docs/lfj2008.pdf
44
45--- What is future plan? ---
46
47We believe that inode based security and name based security are complementary
48and both should be used together. But unfortunately, so far, we cannot enable
49multiple LSM modules at the same time. We feel sorry that you have to give up
50SELinux/SMACK/AppArmor etc. when you want to use TOMOYO.
51
52We hope that LSM becomes stackable in future. Meanwhile, you can use non-LSM
53version of TOMOYO, available at http://tomoyo.sourceforge.jp/en/1.6.x/ .
54LSM version of TOMOYO is a subset of non-LSM version of TOMOYO. We are planning
55to port non-LSM version's functionalities to LSM versions.
diff --git a/Documentation/vm/00-INDEX b/Documentation/vm/00-INDEX
index 2131b00b63f6..2f77ced35df7 100644
--- a/Documentation/vm/00-INDEX
+++ b/Documentation/vm/00-INDEX
@@ -1,5 +1,7 @@
100-INDEX 100-INDEX
2 - this file. 2 - this file.
3active_mm.txt
4 - An explanation from Linus about tsk->active_mm vs tsk->mm.
3balance 5balance
4 - various information on memory balancing. 6 - various information on memory balancing.
5hugetlbpage.txt 7hugetlbpage.txt
diff --git a/Documentation/vm/active_mm.txt b/Documentation/vm/active_mm.txt
new file mode 100644
index 000000000000..4ee1f643d897
--- /dev/null
+++ b/Documentation/vm/active_mm.txt
@@ -0,0 +1,83 @@
1List: linux-kernel
2Subject: Re: active_mm
3From: Linus Torvalds <torvalds () transmeta ! com>
4Date: 1999-07-30 21:36:24
5
6Cc'd to linux-kernel, because I don't write explanations all that often,
7and when I do I feel better about more people reading them.
8
9On Fri, 30 Jul 1999, David Mosberger wrote:
10>
11> Is there a brief description someplace on how "mm" vs. "active_mm" in
12> the task_struct are supposed to be used? (My apologies if this was
13> discussed on the mailing lists---I just returned from vacation and
14> wasn't able to follow linux-kernel for a while).
15
16Basically, the new setup is:
17
18 - we have "real address spaces" and "anonymous address spaces". The
19 difference is that an anonymous address space doesn't care about the
20 user-level page tables at all, so when we do a context switch into an
21 anonymous address space we just leave the previous address space
22 active.
23
24 The obvious use for a "anonymous address space" is any thread that
25 doesn't need any user mappings - all kernel threads basically fall into
26 this category, but even "real" threads can temporarily say that for
27 some amount of time they are not going to be interested in user space,
28 and that the scheduler might as well try to avoid wasting time on
29 switching the VM state around. Currently only the old-style bdflush
30 sync does that.
31
32 - "tsk->mm" points to the "real address space". For an anonymous process,
33 tsk->mm will be NULL, for the logical reason that an anonymous process
34 really doesn't _have_ a real address space at all.
35
36 - however, we obviously need to keep track of which address space we
37 "stole" for such an anonymous user. For that, we have "tsk->active_mm",
38 which shows what the currently active address space is.
39
40 The rule is that for a process with a real address space (ie tsk->mm is
41 non-NULL) the active_mm obviously always has to be the same as the real
42 one.
43
44 For a anonymous process, tsk->mm == NULL, and tsk->active_mm is the
45 "borrowed" mm while the anonymous process is running. When the
46 anonymous process gets scheduled away, the borrowed address space is
47 returned and cleared.
48
49To support all that, the "struct mm_struct" now has two counters: a
50"mm_users" counter that is how many "real address space users" there are,
51and a "mm_count" counter that is the number of "lazy" users (ie anonymous
52users) plus one if there are any real users.
53
54Usually there is at least one real user, but it could be that the real
55user exited on another CPU while a lazy user was still active, so you do
56actually get cases where you have a address space that is _only_ used by
57lazy users. That is often a short-lived state, because once that thread
58gets scheduled away in favour of a real thread, the "zombie" mm gets
59released because "mm_users" becomes zero.
60
61Also, a new rule is that _nobody_ ever has "init_mm" as a real MM any
62more. "init_mm" should be considered just a "lazy context when no other
63context is available", and in fact it is mainly used just at bootup when
64no real VM has yet been created. So code that used to check
65
66 if (current->mm == &init_mm)
67
68should generally just do
69
70 if (!current->mm)
71
72instead (which makes more sense anyway - the test is basically one of "do
73we have a user context", and is generally done by the page fault handler
74and things like that).
75
76Anyway, I put a pre-patch-2.3.13-1 on ftp.kernel.org just a moment ago,
77because it slightly changes the interfaces to accomodate the alpha (who
78would have thought it, but the alpha actually ends up having one of the
79ugliest context switch codes - unlike the other architectures where the MM
80and register state is separate, the alpha PALcode joins the two, and you
81need to switch both together).
82
83(From http://marc.info/?l=linux-kernel&m=93337278602211&w=2)
diff --git a/Documentation/vm/unevictable-lru.txt b/Documentation/vm/unevictable-lru.txt
index 0706a7282a8c..2d70d0d95108 100644
--- a/Documentation/vm/unevictable-lru.txt
+++ b/Documentation/vm/unevictable-lru.txt
@@ -1,588 +1,691 @@
1 1 ==============================
2This document describes the Linux memory management "Unevictable LRU" 2 UNEVICTABLE LRU INFRASTRUCTURE
3infrastructure and the use of this infrastructure to manage several types 3 ==============================
4of "unevictable" pages. The document attempts to provide the overall 4
5rationale behind this mechanism and the rationale for some of the design 5========
6decisions that drove the implementation. The latter design rationale is 6CONTENTS
7discussed in the context of an implementation description. Admittedly, one 7========
8can obtain the implementation details--the "what does it do?"--by reading the 8
9code. One hopes that the descriptions below add value by provide the answer 9 (*) The Unevictable LRU
10to "why does it do that?". 10
11 11 - The unevictable page list.
12Unevictable LRU Infrastructure: 12 - Memory control group interaction.
13 13 - Marking address spaces unevictable.
14The Unevictable LRU adds an additional LRU list to track unevictable pages 14 - Detecting Unevictable Pages.
15and to hide these pages from vmscan. This mechanism is based on a patch by 15 - vmscan's handling of unevictable pages.
16Larry Woodman of Red Hat to address several scalability problems with page 16
17 (*) mlock()'d pages.
18
19 - History.
20 - Basic management.
21 - mlock()/mlockall() system call handling.
22 - Filtering special vmas.
23 - munlock()/munlockall() system call handling.
24 - Migrating mlocked pages.
25 - mmap(MAP_LOCKED) system call handling.
26 - munmap()/exit()/exec() system call handling.
27 - try_to_unmap().
28 - try_to_munlock() reverse map scan.
29 - Page reclaim in shrink_*_list().
30
31
32============
33INTRODUCTION
34============
35
36This document describes the Linux memory manager's "Unevictable LRU"
37infrastructure and the use of this to manage several types of "unevictable"
38pages.
39
40The document attempts to provide the overall rationale behind this mechanism
41and the rationale for some of the design decisions that drove the
42implementation. The latter design rationale is discussed in the context of an
43implementation description. Admittedly, one can obtain the implementation
44details - the "what does it do?" - by reading the code. One hopes that the
45descriptions below add value by provide the answer to "why does it do that?".
46
47
48===================
49THE UNEVICTABLE LRU
50===================
51
52The Unevictable LRU facility adds an additional LRU list to track unevictable
53pages and to hide these pages from vmscan. This mechanism is based on a patch
54by Larry Woodman of Red Hat to address several scalability problems with page
17reclaim in Linux. The problems have been observed at customer sites on large 55reclaim in Linux. The problems have been observed at customer sites on large
18memory x86_64 systems. For example, a non-numal x86_64 platform with 128GB 56memory x86_64 systems.
19of main memory will have over 32 million 4k pages in a single zone. When a 57
20large fraction of these pages are not evictable for any reason [see below], 58To illustrate this with an example, a non-NUMA x86_64 platform with 128GB of
21vmscan will spend a lot of time scanning the LRU lists looking for the small 59main memory will have over 32 million 4k pages in a single zone. When a large
22fraction of pages that are evictable. This can result in a situation where 60fraction of these pages are not evictable for any reason [see below], vmscan
23all cpus are spending 100% of their time in vmscan for hours or days on end, 61will spend a lot of time scanning the LRU lists looking for the small fraction
24with the system completely unresponsive. 62of pages that are evictable. This can result in a situation where all CPUs are
25 63spending 100% of their time in vmscan for hours or days on end, with the system
26The Unevictable LRU infrastructure addresses the following classes of 64completely unresponsive.
27unevictable pages: 65
28 66The unevictable list addresses the following classes of unevictable pages:
29+ page owned by ramfs 67
30+ page mapped into SHM_LOCKed shared memory regions 68 (*) Those owned by ramfs.
31+ page mapped into VM_LOCKED [mlock()ed] vmas 69
32 70 (*) Those mapped into SHM_LOCK'd shared memory regions.
33The infrastructure might be able to handle other conditions that make pages 71
72 (*) Those mapped into VM_LOCKED [mlock()ed] VMAs.
73
74The infrastructure may also be able to handle other conditions that make pages
34unevictable, either by definition or by circumstance, in the future. 75unevictable, either by definition or by circumstance, in the future.
35 76
36 77
37The Unevictable LRU List 78THE UNEVICTABLE PAGE LIST
79-------------------------
38 80
39The Unevictable LRU infrastructure consists of an additional, per-zone, LRU list 81The Unevictable LRU infrastructure consists of an additional, per-zone, LRU list
40called the "unevictable" list and an associated page flag, PG_unevictable, to 82called the "unevictable" list and an associated page flag, PG_unevictable, to
41indicate that the page is being managed on the unevictable list. The 83indicate that the page is being managed on the unevictable list.
42PG_unevictable flag is analogous to, and mutually exclusive with, the PG_active 84
43flag in that it indicates on which LRU list a page resides when PG_lru is set. 85The PG_unevictable flag is analogous to, and mutually exclusive with, the
44The unevictable LRU list is source configurable based on the UNEVICTABLE_LRU 86PG_active flag in that it indicates on which LRU list a page resides when
45Kconfig option. 87PG_lru is set. The unevictable list is compile-time configurable based on the
88UNEVICTABLE_LRU Kconfig option.
46 89
47The Unevictable LRU infrastructure maintains unevictable pages on an additional 90The Unevictable LRU infrastructure maintains unevictable pages on an additional
48LRU list for a few reasons: 91LRU list for a few reasons:
49 92
501) We get to "treat unevictable pages just like we treat other pages in the 93 (1) We get to "treat unevictable pages just like we treat other pages in the
51 system, which means we get to use the same code to manipulate them, the 94 system - which means we get to use the same code to manipulate them, the
52 same code to isolate them (for migrate, etc.), the same code to keep track 95 same code to isolate them (for migrate, etc.), the same code to keep track
53 of the statistics, etc..." [Rik van Riel] 96 of the statistics, etc..." [Rik van Riel]
97
98 (2) We want to be able to migrate unevictable pages between nodes for memory
99 defragmentation, workload management and memory hotplug. The linux kernel
100 can only migrate pages that it can successfully isolate from the LRU
101 lists. If we were to maintain pages elsewhere than on an LRU-like list,
102 where they can be found by isolate_lru_page(), we would prevent their
103 migration, unless we reworked migration code to find the unevictable pages
104 itself.
54 105
552) We want to be able to migrate unevictable pages between nodes--for memory
56 defragmentation, workload management and memory hotplug. The linux kernel
57 can only migrate pages that it can successfully isolate from the lru lists.
58 If we were to maintain pages elsewise than on an lru-like list, where they
59 can be found by isolate_lru_page(), we would prevent their migration, unless
60 we reworked migration code to find the unevictable pages.
61 106
107The unevictable list does not differentiate between file-backed and anonymous,
108swap-backed pages. This differentiation is only important while the pages are,
109in fact, evictable.
62 110
63The unevictable LRU list does not differentiate between file backed and swap 111The unevictable list benefits from the "arrayification" of the per-zone LRU
64backed [anon] pages. This differentiation is only important while the pages 112lists and statistics originally proposed and posted by Christoph Lameter.
65are, in fact, evictable.
66 113
67The unevictable LRU list benefits from the "arrayification" of the per-zone 114The unevictable list does not use the LRU pagevec mechanism. Rather,
68LRU lists and statistics originally proposed and posted by Christoph Lameter. 115unevictable pages are placed directly on the page's zone's unevictable list
116under the zone lru_lock. This allows us to prevent the stranding of pages on
117the unevictable list when one task has the page isolated from the LRU and other
118tasks are changing the "evictability" state of the page.
69 119
70The unevictable list does not use the lru pagevec mechanism. Rather,
71unevictable pages are placed directly on the page's zone's unevictable
72list under the zone lru_lock. The reason for this is to prevent stranding
73of pages on the unevictable list when one task has the page isolated from the
74lru and other tasks are changing the "evictability" state of the page.
75 120
121MEMORY CONTROL GROUP INTERACTION
122--------------------------------
76 123
77Unevictable LRU and Memory Controller Interaction 124The unevictable LRU facility interacts with the memory control group [aka
125memory controller; see Documentation/cgroups/memory.txt] by extending the
126lru_list enum.
127
128The memory controller data structure automatically gets a per-zone unevictable
129list as a result of the "arrayification" of the per-zone LRU lists (one per
130lru_list enum element). The memory controller tracks the movement of pages to
131and from the unevictable list.
78 132
79The memory controller data structure automatically gets a per zone unevictable
80lru list as a result of the "arrayification" of the per-zone LRU lists. The
81memory controller tracks the movement of pages to and from the unevictable list.
82When a memory control group comes under memory pressure, the controller will 133When a memory control group comes under memory pressure, the controller will
83not attempt to reclaim pages on the unevictable list. This has a couple of 134not attempt to reclaim pages on the unevictable list. This has a couple of
84effects. Because the pages are "hidden" from reclaim on the unevictable list, 135effects:
85the reclaim process can be more efficient, dealing only with pages that have 136
86a chance of being reclaimed. On the other hand, if too many of the pages 137 (1) Because the pages are "hidden" from reclaim on the unevictable list, the
87charged to the control group are unevictable, the evictable portion of the 138 reclaim process can be more efficient, dealing only with pages that have a
88working set of the tasks in the control group may not fit into the available 139 chance of being reclaimed.
89memory. This can cause the control group to thrash or to oom-kill tasks. 140
90 141 (2) On the other hand, if too many of the pages charged to the control group
91 142 are unevictable, the evictable portion of the working set of the tasks in
92Unevictable LRU: Detecting Unevictable Pages 143 the control group may not fit into the available memory. This can cause
93 144 the control group to thrash or to OOM-kill tasks.
94The function page_evictable(page, vma) in vmscan.c determines whether a 145
95page is evictable or not. For ramfs pages and pages in SHM_LOCKed regions, 146
96page_evictable() tests a new address space flag, AS_UNEVICTABLE, in the page's 147MARKING ADDRESS SPACES UNEVICTABLE
97address space using a wrapper function. Wrapper functions are used to set, 148----------------------------------
98clear and test the flag to reduce the requirement for #ifdef's throughout the 149
99source code. AS_UNEVICTABLE is set on ramfs inode/mapping when it is created. 150For facilities such as ramfs none of the pages attached to the address space
100This flag remains for the life of the inode. 151may be evicted. To prevent eviction of any such pages, the AS_UNEVICTABLE
101 152address space flag is provided, and this can be manipulated by a filesystem
102For shared memory regions, AS_UNEVICTABLE is set when an application 153using a number of wrapper functions:
103successfully SHM_LOCKs the region and is removed when the region is 154
104SHM_UNLOCKed. Note that shmctl(SHM_LOCK, ...) does not populate the page 155 (*) void mapping_set_unevictable(struct address_space *mapping);
105tables for the region as does, for example, mlock(). So, we make no special 156
106effort to push any pages in the SHM_LOCKed region to the unevictable list. 157 Mark the address space as being completely unevictable.
107Vmscan will do this when/if it encounters the pages during reclaim. On 158
108SHM_UNLOCK, shmctl() scans the pages in the region and "rescues" them from the 159 (*) void mapping_clear_unevictable(struct address_space *mapping);
109unevictable list if no other condition keeps them unevictable. If a SHM_LOCKed 160
110region is destroyed, the pages are also "rescued" from the unevictable list in 161 Mark the address space as being evictable.
111the process of freeing them. 162
112 163 (*) int mapping_unevictable(struct address_space *mapping);
113page_evictable() detects mlock()ed pages by testing an additional page flag, 164
114PG_mlocked via the PageMlocked() wrapper. If the page is NOT mlocked, and a 165 Query the address space, and return true if it is completely
115non-NULL vma is supplied, page_evictable() will check whether the vma is 166 unevictable.
167
168These are currently used in two places in the kernel:
169
170 (1) By ramfs to mark the address spaces of its inodes when they are created,
171 and this mark remains for the life of the inode.
172
173 (2) By SYSV SHM to mark SHM_LOCK'd address spaces until SHM_UNLOCK is called.
174
175 Note that SHM_LOCK is not required to page in the locked pages if they're
176 swapped out; the application must touch the pages manually if it wants to
177 ensure they're in memory.
178
179
180DETECTING UNEVICTABLE PAGES
181---------------------------
182
183The function page_evictable() in vmscan.c determines whether a page is
184evictable or not using the query function outlined above [see section "Marking
185address spaces unevictable"] to check the AS_UNEVICTABLE flag.
186
187For address spaces that are so marked after being populated (as SHM regions
188might be), the lock action (eg: SHM_LOCK) can be lazy, and need not populate
189the page tables for the region as does, for example, mlock(), nor need it make
190any special effort to push any pages in the SHM_LOCK'd area to the unevictable
191list. Instead, vmscan will do this if and when it encounters the pages during
192a reclamation scan.
193
194On an unlock action (such as SHM_UNLOCK), the unlocker (eg: shmctl()) must scan
195the pages in the region and "rescue" them from the unevictable list if no other
196condition is keeping them unevictable. If an unevictable region is destroyed,
197the pages are also "rescued" from the unevictable list in the process of
198freeing them.
199
200page_evictable() also checks for mlocked pages by testing an additional page
201flag, PG_mlocked (as wrapped by PageMlocked()). If the page is NOT mlocked,
202and a non-NULL VMA is supplied, page_evictable() will check whether the VMA is
116VM_LOCKED via is_mlocked_vma(). is_mlocked_vma() will SetPageMlocked() and 203VM_LOCKED via is_mlocked_vma(). is_mlocked_vma() will SetPageMlocked() and
117update the appropriate statistics if the vma is VM_LOCKED. This method allows 204update the appropriate statistics if the vma is VM_LOCKED. This method allows
118efficient "culling" of pages in the fault path that are being faulted in to 205efficient "culling" of pages in the fault path that are being faulted in to
119VM_LOCKED vmas. 206VM_LOCKED VMAs.
120 207
121 208
122Unevictable Pages and Vmscan [shrink_*_list()] 209VMSCAN'S HANDLING OF UNEVICTABLE PAGES
210--------------------------------------
123 211
124If unevictable pages are culled in the fault path, or moved to the unevictable 212If unevictable pages are culled in the fault path, or moved to the unevictable
125list at mlock() or mmap() time, vmscan will never encounter the pages until 213list at mlock() or mmap() time, vmscan will not encounter the pages until they
126they have become evictable again, for example, via munlock() and have been 214have become evictable again (via munlock() for example) and have been "rescued"
127"rescued" from the unevictable list. However, there may be situations where we 215from the unevictable list. However, there may be situations where we decide,
128decide, for the sake of expediency, to leave a unevictable page on one of the 216for the sake of expediency, to leave a unevictable page on one of the regular
129regular active/inactive LRU lists for vmscan to deal with. Vmscan checks for 217active/inactive LRU lists for vmscan to deal with. vmscan checks for such
130such pages in all of the shrink_{active|inactive|page}_list() functions and 218pages in all of the shrink_{active|inactive|page}_list() functions and will
131will "cull" such pages that it encounters--that is, it diverts those pages to 219"cull" such pages that it encounters: that is, it diverts those pages to the
132the unevictable list for the zone being scanned. 220unevictable list for the zone being scanned.
133 221
134There may be situations where a page is mapped into a VM_LOCKED vma, but the 222There may be situations where a page is mapped into a VM_LOCKED VMA, but the
135page is not marked as PageMlocked. Such pages will make it all the way to 223page is not marked as PG_mlocked. Such pages will make it all the way to
136shrink_page_list() where they will be detected when vmscan walks the reverse 224shrink_page_list() where they will be detected when vmscan walks the reverse
137map in try_to_unmap(). If try_to_unmap() returns SWAP_MLOCK, shrink_page_list() 225map in try_to_unmap(). If try_to_unmap() returns SWAP_MLOCK,
138will cull the page at that point. 226shrink_page_list() will cull the page at that point.
139 227
140To "cull" an unevictable page, vmscan simply puts the page back on the lru 228To "cull" an unevictable page, vmscan simply puts the page back on the LRU list
141list using putback_lru_page()--the inverse operation to isolate_lru_page()-- 229using putback_lru_page() - the inverse operation to isolate_lru_page() - after
142after dropping the page lock. Because the condition which makes the page 230dropping the page lock. Because the condition which makes the page unevictable
143unevictable may change once the page is unlocked, putback_lru_page() will 231may change once the page is unlocked, putback_lru_page() will recheck the
144recheck the unevictable state of a page that it places on the unevictable lru 232unevictable state of a page that it places on the unevictable list. If the
145list. If the page has become unevictable, putback_lru_page() removes it from 233page has become unevictable, putback_lru_page() removes it from the list and
146the list and retries, including the page_unevictable() test. Because such a 234retries, including the page_unevictable() test. Because such a race is a rare
147race is a rare event and movement of pages onto the unevictable list should be 235event and movement of pages onto the unevictable list should be rare, these
148rare, these extra evictabilty checks should not occur in the majority of calls 236extra evictabilty checks should not occur in the majority of calls to
149to putback_lru_page(). 237putback_lru_page().
150 238
151 239
152Mlocked Page: Prior Work 240=============
241MLOCKED PAGES
242=============
153 243
154The "Unevictable Mlocked Pages" infrastructure is based on work originally 244The unevictable page list is also useful for mlock(), in addition to ramfs and
245SYSV SHM. Note that mlock() is only available in CONFIG_MMU=y situations; in
246NOMMU situations, all mappings are effectively mlocked.
247
248
249HISTORY
250-------
251
252The "Unevictable mlocked Pages" infrastructure is based on work originally
155posted by Nick Piggin in an RFC patch entitled "mm: mlocked pages off LRU". 253posted by Nick Piggin in an RFC patch entitled "mm: mlocked pages off LRU".
156Nick posted his patch as an alternative to a patch posted by Christoph 254Nick posted his patch as an alternative to a patch posted by Christoph Lameter
157Lameter to achieve the same objective--hiding mlocked pages from vmscan. 255to achieve the same objective: hiding mlocked pages from vmscan.
158In Nick's patch, he used one of the struct page lru list link fields as a count 256
159of VM_LOCKED vmas that map the page. This use of the link field for a count 257In Nick's patch, he used one of the struct page LRU list link fields as a count
160prevented the management of the pages on an LRU list. Thus, mlocked pages were 258of VM_LOCKED VMAs that map the page. This use of the link field for a count
161not migratable as isolate_lru_page() could not find them and the lru list link 259prevented the management of the pages on an LRU list, and thus mlocked pages
162field was not available to the migration subsystem. Nick resolved this by 260were not migratable as isolate_lru_page() could not find them, and the LRU list
163putting mlocked pages back on the lru list before attempting to isolate them, 261link field was not available to the migration subsystem.
164thus abandoning the count of VM_LOCKED vmas. When Nick's patch was integrated 262
165with the Unevictable LRU work, the count was replaced by walking the reverse 263Nick resolved this by putting mlocked pages back on the lru list before
166map to determine whether any VM_LOCKED vmas mapped the page. More on this 264attempting to isolate them, thus abandoning the count of VM_LOCKED VMAs. When
167below. 265Nick's patch was integrated with the Unevictable LRU work, the count was
168 266replaced by walking the reverse map to determine whether any VM_LOCKED VMAs
169 267mapped the page. More on this below.
170Mlocked Pages: Basic Management 268
171 269
172Mlocked pages--pages mapped into a VM_LOCKED vma--represent one class of 270BASIC MANAGEMENT
173unevictable pages. When such a page has been "noticed" by the memory 271----------------
174management subsystem, the page is marked with the PG_mlocked [PageMlocked()] 272
175flag. A PageMlocked() page will be placed on the unevictable LRU list when 273mlocked pages - pages mapped into a VM_LOCKED VMA - are a class of unevictable
176it is added to the LRU. Pages can be "noticed" by memory management in 274pages. When such a page has been "noticed" by the memory management subsystem,
177several places: 275the page is marked with the PG_mlocked flag. This can be manipulated using the
178 276PageMlocked() functions.
1791) in the mlock()/mlockall() system call handlers. 277
1802) in the mmap() system call handler when mmap()ing a region with the 278A PG_mlocked page will be placed on the unevictable list when it is added to
181 MAP_LOCKED flag, or mmap()ing a region in a task that has called 279the LRU. Such pages can be "noticed" by memory management in several places:
182 mlockall() with the MCL_FUTURE flag. Both of these conditions result 280
183 in the VM_LOCKED flag being set for the vma. 281 (1) in the mlock()/mlockall() system call handlers;
1843) in the fault path, if mlocked pages are "culled" in the fault path, 282
185 and when a VM_LOCKED stack segment is expanded. 283 (2) in the mmap() system call handler when mmapping a region with the
1864) as mentioned above, in vmscan:shrink_page_list() when attempting to 284 MAP_LOCKED flag;
187 reclaim a page in a VM_LOCKED vma via try_to_unmap(). 285
188 286 (3) mmapping a region in a task that has called mlockall() with the MCL_FUTURE
189Mlocked pages become unlocked and rescued from the unevictable list when: 287 flag
190 288
1911) mapped in a range unlocked via the munlock()/munlockall() system calls. 289 (4) in the fault path, if mlocked pages are "culled" in the fault path,
1922) munmapped() out of the last VM_LOCKED vma that maps the page, including 290 and when a VM_LOCKED stack segment is expanded; or
193 unmapping at task exit. 291
1943) when the page is truncated from the last VM_LOCKED vma of an mmap()ed file. 292 (5) as mentioned above, in vmscan:shrink_page_list() when attempting to
1954) before a page is COWed in a VM_LOCKED vma. 293 reclaim a page in a VM_LOCKED VMA via try_to_unmap()
196 294
197 295all of which result in the VM_LOCKED flag being set for the VMA if it doesn't
198Mlocked Pages: mlock()/mlockall() System Call Handling 296already have it set.
297
298mlocked pages become unlocked and rescued from the unevictable list when:
299
300 (1) mapped in a range unlocked via the munlock()/munlockall() system calls;
301
302 (2) munmap()'d out of the last VM_LOCKED VMA that maps the page, including
303 unmapping at task exit;
304
305 (3) when the page is truncated from the last VM_LOCKED VMA of an mmapped file;
306 or
307
308 (4) before a page is COW'd in a VM_LOCKED VMA.
309
310
311mlock()/mlockall() SYSTEM CALL HANDLING
312---------------------------------------
199 313
200Both [do_]mlock() and [do_]mlockall() system call handlers call mlock_fixup() 314Both [do_]mlock() and [do_]mlockall() system call handlers call mlock_fixup()
201for each vma in the range specified by the call. In the case of mlockall(), 315for each VMA in the range specified by the call. In the case of mlockall(),
202this is the entire active address space of the task. Note that mlock_fixup() 316this is the entire active address space of the task. Note that mlock_fixup()
203is used for both mlock()ing and munlock()ing a range of memory. A call to 317is used for both mlocking and munlocking a range of memory. A call to mlock()
204mlock() an already VM_LOCKED vma, or to munlock() a vma that is not VM_LOCKED 318an already VM_LOCKED VMA, or to munlock() a VMA that is not VM_LOCKED is
205is treated as a no-op--mlock_fixup() simply returns. 319treated as a no-op, and mlock_fixup() simply returns.
206 320
207If the vma passes some filtering described in "Mlocked Pages: Filtering Vmas" 321If the VMA passes some filtering as described in "Filtering Special Vmas"
208below, mlock_fixup() will attempt to merge the vma with its neighbors or split 322below, mlock_fixup() will attempt to merge the VMA with its neighbors or split
209off a subset of the vma if the range does not cover the entire vma. Once the 323off a subset of the VMA if the range does not cover the entire VMA. Once the
210vma has been merged or split or neither, mlock_fixup() will call 324VMA has been merged or split or neither, mlock_fixup() will call
211__mlock_vma_pages_range() to fault in the pages via get_user_pages() and 325__mlock_vma_pages_range() to fault in the pages via get_user_pages() and to
212to mark the pages as mlocked via mlock_vma_page(). 326mark the pages as mlocked via mlock_vma_page().
213 327
214Note that the vma being mlocked might be mapped with PROT_NONE. In this case, 328Note that the VMA being mlocked might be mapped with PROT_NONE. In this case,
215get_user_pages() will be unable to fault in the pages. That's OK. If pages 329get_user_pages() will be unable to fault in the pages. That's okay. If pages
216do end up getting faulted into this VM_LOCKED vma, we'll handle them in the 330do end up getting faulted into this VM_LOCKED VMA, we'll handle them in the
217fault path or in vmscan. 331fault path or in vmscan.
218 332
219Also note that a page returned by get_user_pages() could be truncated or 333Also note that a page returned by get_user_pages() could be truncated or
220migrated out from under us, while we're trying to mlock it. To detect 334migrated out from under us, while we're trying to mlock it. To detect this,
221this, __mlock_vma_pages_range() tests the page_mapping after acquiring 335__mlock_vma_pages_range() checks page_mapping() after acquiring the page lock.
222the page lock. If the page is still associated with its mapping, we'll 336If the page is still associated with its mapping, we'll go ahead and call
223go ahead and call mlock_vma_page(). If the mapping is gone, we just 337mlock_vma_page(). If the mapping is gone, we just unlock the page and move on.
224unlock the page and move on. Worse case, this results in page mapped 338In the worst case, this will result in a page mapped in a VM_LOCKED VMA
225in a VM_LOCKED vma remaining on a normal LRU list without being 339remaining on a normal LRU list without being PageMlocked(). Again, vmscan will
226PageMlocked(). Again, vmscan will detect and cull such pages. 340detect and cull such pages.
227 341
228mlock_vma_page(), called with the page locked [N.B., not "mlocked"], will 342mlock_vma_page() will call TestSetPageMlocked() for each page returned by
229TestSetPageMlocked() for each page returned by get_user_pages(). We use 343get_user_pages(). We use TestSetPageMlocked() because the page might already
230TestSetPageMlocked() because the page might already be mlocked by another 344be mlocked by another task/VMA and we don't want to do extra work. We
231task/vma and we don't want to do extra work. We especially do not want to 345especially do not want to count an mlocked page more than once in the
232count an mlocked page more than once in the statistics. If the page was 346statistics. If the page was already mlocked, mlock_vma_page() need do nothing
233already mlocked, mlock_vma_page() is done. 347more.
234 348
235If the page was NOT already mlocked, mlock_vma_page() attempts to isolate the 349If the page was NOT already mlocked, mlock_vma_page() attempts to isolate the
236page from the LRU, as it is likely on the appropriate active or inactive list 350page from the LRU, as it is likely on the appropriate active or inactive list
237at that time. If the isolate_lru_page() succeeds, mlock_vma_page() will 351at that time. If the isolate_lru_page() succeeds, mlock_vma_page() will put
238putback the page--putback_lru_page()--which will notice that the page is now 352back the page - by calling putback_lru_page() - which will notice that the page
239mlocked and divert the page to the zone's unevictable LRU list. If 353is now mlocked and divert the page to the zone's unevictable list. If
240mlock_vma_page() is unable to isolate the page from the LRU, vmscan will handle 354mlock_vma_page() is unable to isolate the page from the LRU, vmscan will handle
241it later if/when it attempts to reclaim the page. 355it later if and when it attempts to reclaim the page.
242 356
243 357
244Mlocked Pages: Filtering Special Vmas 358FILTERING SPECIAL VMAS
359----------------------
245 360
246mlock_fixup() filters several classes of "special" vmas: 361mlock_fixup() filters several classes of "special" VMAs:
247 362
2481) vmas with VM_IO|VM_PFNMAP set are skipped entirely. The pages behind 3631) VMAs with VM_IO or VM_PFNMAP set are skipped entirely. The pages behind
249 these mappings are inherently pinned, so we don't need to mark them as 364 these mappings are inherently pinned, so we don't need to mark them as
250 mlocked. In any case, most of the pages have no struct page in which to 365 mlocked. In any case, most of the pages have no struct page in which to so
251 so mark the page. Because of this, get_user_pages() will fail for these 366 mark the page. Because of this, get_user_pages() will fail for these VMAs,
252 vmas, so there is no sense in attempting to visit them. 367 so there is no sense in attempting to visit them.
253 368
2542) vmas mapping hugetlbfs page are already effectively pinned into memory. 3692) VMAs mapping hugetlbfs page are already effectively pinned into memory. We
255 We don't need nor want to mlock() these pages. However, to preserve the 370 neither need nor want to mlock() these pages. However, to preserve the
256 prior behavior of mlock()--before the unevictable/mlock changes-- 371 prior behavior of mlock() - before the unevictable/mlock changes -
257 mlock_fixup() will call make_pages_present() in the hugetlbfs vma range 372 mlock_fixup() will call make_pages_present() in the hugetlbfs VMA range to
258 to allocate the huge pages and populate the ptes. 373 allocate the huge pages and populate the ptes.
259 374
2603) vmas with VM_DONTEXPAND|VM_RESERVED are generally user space mappings of 3753) VMAs with VM_DONTEXPAND or VM_RESERVED are generally userspace mappings of
261 kernel pages, such as the vdso page, relay channel pages, etc. These pages 376 kernel pages, such as the VDSO page, relay channel pages, etc. These pages
262 are inherently unevictable and are not managed on the LRU lists. 377 are inherently unevictable and are not managed on the LRU lists.
263 mlock_fixup() treats these vmas the same as hugetlbfs vmas. It calls 378 mlock_fixup() treats these VMAs the same as hugetlbfs VMAs. It calls
264 make_pages_present() to populate the ptes. 379 make_pages_present() to populate the ptes.
265 380
266Note that for all of these special vmas, mlock_fixup() does not set the 381Note that for all of these special VMAs, mlock_fixup() does not set the
267VM_LOCKED flag. Therefore, we won't have to deal with them later during 382VM_LOCKED flag. Therefore, we won't have to deal with them later during
268munlock() or munmap()--for example, at task exit. Neither does mlock_fixup() 383munlock(), munmap() or task exit. Neither does mlock_fixup() account these
269account these vmas against the task's "locked_vm". 384VMAs against the task's "locked_vm".
270 385
271Mlocked Pages: Downgrading the Mmap Semaphore. 386
272 387munlock()/munlockall() SYSTEM CALL HANDLING
273mlock_fixup() must be called with the mmap semaphore held for write, because 388-------------------------------------------
274it may have to merge or split vmas. However, mlocking a large region of 389
275memory can take a long time--especially if vmscan must reclaim pages to 390The munlock() and munlockall() system calls are handled by the same functions -
276satisfy the regions requirements. Faulting in a large region with the mmap 391do_mlock[all]() - as the mlock() and mlockall() system calls with the unlock vs
277semaphore held for write can hold off other faults on the address space, in 392lock operation indicated by an argument. So, these system calls are also
278the case of a multi-threaded task. It can also hold off scans of the task's 393handled by mlock_fixup(). Again, if called for an already munlocked VMA,
279address space via /proc. While testing under heavy load, it was observed that 394mlock_fixup() simply returns. Because of the VMA filtering discussed above,
280the ps(1) command could be held off for many minutes while a large segment was 395VM_LOCKED will not be set in any "special" VMAs. So, these VMAs will be
281mlock()ed down.
282
283To address this issue, and to make the system more responsive during mlock()ing
284of large segments, mlock_fixup() downgrades the mmap semaphore to read mode
285during the call to __mlock_vma_pages_range(). This works fine. However, the
286callers of mlock_fixup() expect the semaphore to be returned in write mode.
287So, mlock_fixup() "upgrades" the semphore to write mode. Linux does not
288support an atomic upgrade_sem() call, so mlock_fixup() must drop the semaphore
289and reacquire it in write mode. In a multi-threaded task, it is possible for
290the task memory map to change while the semaphore is dropped. Therefore,
291mlock_fixup() looks up the vma at the range start address after reacquiring
292the semaphore in write mode and verifies that it still covers the original
293range. If not, mlock_fixup() returns an error [-EAGAIN]. All callers of
294mlock_fixup() have been changed to deal with this new error condition.
295
296Note: when munlocking a region, all of the pages should already be resident--
297unless we have racing threads mlocking() and munlocking() regions. So,
298unlocking should not have to wait for page allocations nor faults of any kind.
299Therefore mlock_fixup() does not downgrade the semaphore for munlock().
300
301
302Mlocked Pages: munlock()/munlockall() System Call Handling
303
304The munlock() and munlockall() system calls are handled by the same functions--
305do_mlock[all]()--as the mlock() and mlockall() system calls with the unlock
306vs lock operation indicated by an argument. So, these system calls are also
307handled by mlock_fixup(). Again, if called for an already munlock()ed vma,
308mlock_fixup() simply returns. Because of the vma filtering discussed above,
309VM_LOCKED will not be set in any "special" vmas. So, these vmas will be
310ignored for munlock. 396ignored for munlock.
311 397
312If the vma is VM_LOCKED, mlock_fixup() again attempts to merge or split off 398If the VMA is VM_LOCKED, mlock_fixup() again attempts to merge or split off the
313the specified range. The range is then munlocked via the function 399specified range. The range is then munlocked via the function
314__mlock_vma_pages_range()--the same function used to mlock a vma range-- 400__mlock_vma_pages_range() - the same function used to mlock a VMA range -
315passing a flag to indicate that munlock() is being performed. 401passing a flag to indicate that munlock() is being performed.
316 402
317Because the vma access protections could have been changed to PROT_NONE after 403Because the VMA access protections could have been changed to PROT_NONE after
318faulting in and mlocking pages, get_user_pages() was unreliable for visiting 404faulting in and mlocking pages, get_user_pages() was unreliable for visiting
319these pages for munlocking. Because we don't want to leave pages mlocked(), 405these pages for munlocking. Because we don't want to leave pages mlocked,
320get_user_pages() was enhanced to accept a flag to ignore the permissions when 406get_user_pages() was enhanced to accept a flag to ignore the permissions when
321fetching the pages--all of which should be resident as a result of previous 407fetching the pages - all of which should be resident as a result of previous
322mlock()ing. 408mlocking.
323 409
324For munlock(), __mlock_vma_pages_range() unlocks individual pages by calling 410For munlock(), __mlock_vma_pages_range() unlocks individual pages by calling
325munlock_vma_page(). munlock_vma_page() unconditionally clears the PG_mlocked 411munlock_vma_page(). munlock_vma_page() unconditionally clears the PG_mlocked
326flag using TestClearPageMlocked(). As with mlock_vma_page(), munlock_vma_page() 412flag using TestClearPageMlocked(). As with mlock_vma_page(),
327use the Test*PageMlocked() function to handle the case where the page might 413munlock_vma_page() use the Test*PageMlocked() function to handle the case where
328have already been unlocked by another task. If the page was mlocked, 414the page might have already been unlocked by another task. If the page was
329munlock_vma_page() updates that zone statistics for the number of mlocked 415mlocked, munlock_vma_page() updates that zone statistics for the number of
330pages. Note, however, that at this point we haven't checked whether the page 416mlocked pages. Note, however, that at this point we haven't checked whether
331is mapped by other VM_LOCKED vmas. 417the page is mapped by other VM_LOCKED VMAs.
332 418
333We can't call try_to_munlock(), the function that walks the reverse map to check 419We can't call try_to_munlock(), the function that walks the reverse map to
334for other VM_LOCKED vmas, without first isolating the page from the LRU. 420check for other VM_LOCKED VMAs, without first isolating the page from the LRU.
335try_to_munlock() is a variant of try_to_unmap() and thus requires that the page 421try_to_munlock() is a variant of try_to_unmap() and thus requires that the page
336not be on an lru list. [More on these below.] However, the call to 422not be on an LRU list [more on these below]. However, the call to
337isolate_lru_page() could fail, in which case we couldn't try_to_munlock(). 423isolate_lru_page() could fail, in which case we couldn't try_to_munlock(). So,
338So, we go ahead and clear PG_mlocked up front, as this might be the only chance 424we go ahead and clear PG_mlocked up front, as this might be the only chance we
339we have. If we can successfully isolate the page, we go ahead and 425have. If we can successfully isolate the page, we go ahead and
340try_to_munlock(), which will restore the PG_mlocked flag and update the zone 426try_to_munlock(), which will restore the PG_mlocked flag and update the zone
341page statistics if it finds another vma holding the page mlocked. If we fail 427page statistics if it finds another VMA holding the page mlocked. If we fail
342to isolate the page, we'll have left a potentially mlocked page on the LRU. 428to isolate the page, we'll have left a potentially mlocked page on the LRU.
343This is fine, because we'll catch it later when/if vmscan tries to reclaim the 429This is fine, because we'll catch it later if and if vmscan tries to reclaim
344page. This should be relatively rare. 430the page. This should be relatively rare.
345 431
346Mlocked Pages: Migrating Them... 432
347 433MIGRATING MLOCKED PAGES
348A page that is being migrated has been isolated from the lru lists and is 434-----------------------
349held locked across unmapping of the page, updating the page's mapping 435
350[address_space] entry and copying the contents and state, until the 436A page that is being migrated has been isolated from the LRU lists and is held
351page table entry has been replaced with an entry that refers to the new 437locked across unmapping of the page, updating the page's address space entry
352page. Linux supports migration of mlocked pages and other unevictable 438and copying the contents and state, until the page table entry has been
353pages. This involves simply moving the PageMlocked and PageUnevictable states 439replaced with an entry that refers to the new page. Linux supports migration
354from the old page to the new page. 440of mlocked pages and other unevictable pages. This involves simply moving the
355 441PG_mlocked and PG_unevictable states from the old page to the new page.
356Note that page migration can race with mlocking or munlocking of the same 442
357page. This has been discussed from the mlock/munlock perspective in the 443Note that page migration can race with mlocking or munlocking of the same page.
358respective sections above. Both processes [migration, m[un]locking], hold 444This has been discussed from the mlock/munlock perspective in the respective
359the page locked. This provides the first level of synchronization. Page 445sections above. Both processes (migration and m[un]locking) hold the page
360migration zeros out the page_mapping of the old page before unlocking it, 446locked. This provides the first level of synchronization. Page migration
361so m[un]lock can skip these pages by testing the page mapping under page 447zeros out the page_mapping of the old page before unlocking it, so m[un]lock
362lock. 448can skip these pages by testing the page mapping under page lock.
363 449
364When completing page migration, we place the new and old pages back onto the 450To complete page migration, we place the new and old pages back onto the LRU
365lru after dropping the page lock. The "unneeded" page--old page on success, 451after dropping the page lock. The "unneeded" page - old page on success, new
366new page on failure--will be freed when the reference count held by the 452page on failure - will be freed when the reference count held by the migration
367migration process is released. To ensure that we don't strand pages on the 453process is released. To ensure that we don't strand pages on the unevictable
368unevictable list because of a race between munlock and migration, page 454list because of a race between munlock and migration, page migration uses the
369migration uses the putback_lru_page() function to add migrated pages back to 455putback_lru_page() function to add migrated pages back to the LRU.
370the lru. 456
371 457
372 458mmap(MAP_LOCKED) SYSTEM CALL HANDLING
373Mlocked Pages: mmap(MAP_LOCKED) System Call Handling 459-------------------------------------
374 460
375In addition the the mlock()/mlockall() system calls, an application can request 461In addition the the mlock()/mlockall() system calls, an application can request
376that a region of memory be mlocked using the MAP_LOCKED flag with the mmap() 462that a region of memory be mlocked supplying the MAP_LOCKED flag to the mmap()
377call. Furthermore, any mmap() call or brk() call that expands the heap by a 463call. Furthermore, any mmap() call or brk() call that expands the heap by a
378task that has previously called mlockall() with the MCL_FUTURE flag will result 464task that has previously called mlockall() with the MCL_FUTURE flag will result
379in the newly mapped memory being mlocked. Before the unevictable/mlock changes, 465in the newly mapped memory being mlocked. Before the unevictable/mlock
380the kernel simply called make_pages_present() to allocate pages and populate 466changes, the kernel simply called make_pages_present() to allocate pages and
381the page table. 467populate the page table.
382 468
383To mlock a range of memory under the unevictable/mlock infrastructure, the 469To mlock a range of memory under the unevictable/mlock infrastructure, the
384mmap() handler and task address space expansion functions call 470mmap() handler and task address space expansion functions call
385mlock_vma_pages_range() specifying the vma and the address range to mlock. 471mlock_vma_pages_range() specifying the vma and the address range to mlock.
386mlock_vma_pages_range() filters vmas like mlock_fixup(), as described above in 472mlock_vma_pages_range() filters VMAs like mlock_fixup(), as described above in
387"Mlocked Pages: Filtering Vmas". It will clear the VM_LOCKED flag, which will 473"Filtering Special VMAs". It will clear the VM_LOCKED flag, which will have
388have already been set by the caller, in filtered vmas. Thus these vma's need 474already been set by the caller, in filtered VMAs. Thus these VMA's need not be
389not be visited for munlock when the region is unmapped. 475visited for munlock when the region is unmapped.
390 476
391For "normal" vmas, mlock_vma_pages_range() calls __mlock_vma_pages_range() to 477For "normal" VMAs, mlock_vma_pages_range() calls __mlock_vma_pages_range() to
392fault/allocate the pages and mlock them. Again, like mlock_fixup(), 478fault/allocate the pages and mlock them. Again, like mlock_fixup(),
393mlock_vma_pages_range() downgrades the mmap semaphore to read mode before 479mlock_vma_pages_range() downgrades the mmap semaphore to read mode before
394attempting to fault/allocate and mlock the pages; and "upgrades" the semaphore 480attempting to fault/allocate and mlock the pages and "upgrades" the semaphore
395back to write mode before returning. 481back to write mode before returning.
396 482
397The callers of mlock_vma_pages_range() will have already added the memory 483The callers of mlock_vma_pages_range() will have already added the memory range
398range to be mlocked to the task's "locked_vm". To account for filtered vmas, 484to be mlocked to the task's "locked_vm". To account for filtered VMAs,
399mlock_vma_pages_range() returns the number of pages NOT mlocked. All of the 485mlock_vma_pages_range() returns the number of pages NOT mlocked. All of the
400callers then subtract a non-negative return value from the task's locked_vm. 486callers then subtract a non-negative return value from the task's locked_vm. A
401A negative return value represent an error--for example, from get_user_pages() 487negative return value represent an error - for example, from get_user_pages()
402attempting to fault in a vma with PROT_NONE access. In this case, we leave 488attempting to fault in a VMA with PROT_NONE access. In this case, we leave the
403the memory range accounted as locked_vm, as the protections could be changed 489memory range accounted as locked_vm, as the protections could be changed later
404later and pages allocated into that region. 490and pages allocated into that region.
405 491
406 492
407Mlocked Pages: munmap()/exit()/exec() System Call Handling 493munmap()/exit()/exec() SYSTEM CALL HANDLING
494-------------------------------------------
408 495
409When unmapping an mlocked region of memory, whether by an explicit call to 496When unmapping an mlocked region of memory, whether by an explicit call to
410munmap() or via an internal unmap from exit() or exec() processing, we must 497munmap() or via an internal unmap from exit() or exec() processing, we must
411munlock the pages if we're removing the last VM_LOCKED vma that maps the pages. 498munlock the pages if we're removing the last VM_LOCKED VMA that maps the pages.
412Before the unevictable/mlock changes, mlocking did not mark the pages in any 499Before the unevictable/mlock changes, mlocking did not mark the pages in any
413way, so unmapping them required no processing. 500way, so unmapping them required no processing.
414 501
415To munlock a range of memory under the unevictable/mlock infrastructure, the 502To munlock a range of memory under the unevictable/mlock infrastructure, the
416munmap() hander and task address space tear down function call 503munmap() handler and task address space call tear down function
417munlock_vma_pages_all(). The name reflects the observation that one always 504munlock_vma_pages_all(). The name reflects the observation that one always
418specifies the entire vma range when munlock()ing during unmap of a region. 505specifies the entire VMA range when munlock()ing during unmap of a region.
419Because of the vma filtering when mlocking() regions, only "normal" vmas that 506Because of the VMA filtering when mlocking() regions, only "normal" VMAs that
420actually contain mlocked pages will be passed to munlock_vma_pages_all(). 507actually contain mlocked pages will be passed to munlock_vma_pages_all().
421 508
422munlock_vma_pages_all() clears the VM_LOCKED vma flag and, like mlock_fixup() 509munlock_vma_pages_all() clears the VM_LOCKED VMA flag and, like mlock_fixup()
423for the munlock case, calls __munlock_vma_pages_range() to walk the page table 510for the munlock case, calls __munlock_vma_pages_range() to walk the page table
424for the vma's memory range and munlock_vma_page() each resident page mapped by 511for the VMA's memory range and munlock_vma_page() each resident page mapped by
425the vma. This effectively munlocks the page, only if this is the last 512the VMA. This effectively munlocks the page, only if this is the last
426VM_LOCKED vma that maps the page. 513VM_LOCKED VMA that maps the page.
427
428 514
429Mlocked Page: try_to_unmap()
430 515
431[Note: the code changes represented by this section are really quite small 516try_to_unmap()
432compared to the text to describe what happening and why, and to discuss the 517--------------
433implications.]
434 518
435Pages can, of course, be mapped into multiple vmas. Some of these vmas may 519Pages can, of course, be mapped into multiple VMAs. Some of these VMAs may
436have VM_LOCKED flag set. It is possible for a page mapped into one or more 520have VM_LOCKED flag set. It is possible for a page mapped into one or more
437VM_LOCKED vmas not to have the PG_mlocked flag set and therefore reside on one 521VM_LOCKED VMAs not to have the PG_mlocked flag set and therefore reside on one
438of the active or inactive LRU lists. This could happen if, for example, a 522of the active or inactive LRU lists. This could happen if, for example, a task
439task in the process of munlock()ing the page could not isolate the page from 523in the process of munlocking the page could not isolate the page from the LRU.
440the LRU. As a result, vmscan/shrink_page_list() might encounter such a page 524As a result, vmscan/shrink_page_list() might encounter such a page as described
441as described in "Unevictable Pages and Vmscan [shrink_*_list()]". To 525in section "vmscan's handling of unevictable pages". To handle this situation,
442handle this situation, try_to_unmap() has been enhanced to check for VM_LOCKED 526try_to_unmap() checks for VM_LOCKED VMAs while it is walking a page's reverse
443vmas while it is walking a page's reverse map. 527map.
444 528
445try_to_unmap() is always called, by either vmscan for reclaim or for page 529try_to_unmap() is always called, by either vmscan for reclaim or for page
446migration, with the argument page locked and isolated from the LRU. BUG_ON() 530migration, with the argument page locked and isolated from the LRU. Separate
447assertions enforce this requirement. Separate functions handle anonymous and 531functions handle anonymous and mapped file pages, as these types of pages have
448mapped file pages, as these types of pages have different reverse map 532different reverse map mechanisms.
449mechanisms. 533
450 534 (*) try_to_unmap_anon()
451 try_to_unmap_anon() 535
452 536 To unmap anonymous pages, each VMA in the list anchored in the anon_vma
453To unmap anonymous pages, each vma in the list anchored in the anon_vma must be 537 must be visited - at least until a VM_LOCKED VMA is encountered. If the
454visited--at least until a VM_LOCKED vma is encountered. If the page is being 538 page is being unmapped for migration, VM_LOCKED VMAs do not stop the
455unmapped for migration, VM_LOCKED vmas do not stop the process because mlocked 539 process because mlocked pages are migratable. However, for reclaim, if
456pages are migratable. However, for reclaim, if the page is mapped into a 540 the page is mapped into a VM_LOCKED VMA, the scan stops.
457VM_LOCKED vma, the scan stops. try_to_unmap() attempts to acquire the mmap 541
458semphore of the mm_struct to which the vma belongs in read mode. If this is 542 try_to_unmap_anon() attempts to acquire in read mode the mmap semphore of
459successful, try_to_unmap() will mlock the page via mlock_vma_page()--we 543 the mm_struct to which the VMA belongs. If this is successful, it will
460wouldn't have gotten to try_to_unmap() if the page were already mlocked--and 544 mlock the page via mlock_vma_page() - we wouldn't have gotten to
461will return SWAP_MLOCK, indicating that the page is unevictable. If the 545 try_to_unmap_anon() if the page were already mlocked - and will return
462mmap semaphore cannot be acquired, we are not sure whether the page is really 546 SWAP_MLOCK, indicating that the page is unevictable.
463unevictable or not. In this case, try_to_unmap() will return SWAP_AGAIN. 547
464 548 If the mmap semaphore cannot be acquired, we are not sure whether the page
465 try_to_unmap_file() -- linear mappings 549 is really unevictable or not. In this case, try_to_unmap_anon() will
466 550 return SWAP_AGAIN.
467Unmapping of a mapped file page works the same, except that the scan visits 551
468all vmas that maps the page's index/page offset in the page's mapping's 552 (*) try_to_unmap_file() - linear mappings
469reverse map priority search tree. It must also visit each vma in the page's 553
470mapping's non-linear list, if the list is non-empty. As for anonymous pages, 554 Unmapping of a mapped file page works the same as for anonymous mappings,
471on encountering a VM_LOCKED vma for a mapped file page, try_to_unmap() will 555 except that the scan visits all VMAs that map the page's index/page offset
472attempt to acquire the associated mm_struct's mmap semaphore to mlock the page, 556 in the page's mapping's reverse map priority search tree. It also visits
473returning SWAP_MLOCK if this is successful, and SWAP_AGAIN, if not. 557 each VMA in the page's mapping's non-linear list, if the list is
474 558 non-empty.
475 try_to_unmap_file() -- non-linear mappings 559
476 560 As for anonymous pages, on encountering a VM_LOCKED VMA for a mapped file
477If a page's mapping contains a non-empty non-linear mapping vma list, then 561 page, try_to_unmap_file() will attempt to acquire the associated
478try_to_un{map|lock}() must also visit each vma in that list to determine 562 mm_struct's mmap semaphore to mlock the page, returning SWAP_MLOCK if this
479whether the page is mapped in a VM_LOCKED vma. Again, the scan must visit 563 is successful, and SWAP_AGAIN, if not.
480all vmas in the non-linear list to ensure that the pages is not/should not be 564
481mlocked. If a VM_LOCKED vma is found in the list, the scan could terminate. 565 (*) try_to_unmap_file() - non-linear mappings
482However, there is no easy way to determine whether the page is actually mapped 566
483in a given vma--either for unmapping or testing whether the VM_LOCKED vma 567 If a page's mapping contains a non-empty non-linear mapping VMA list, then
484actually pins the page. 568 try_to_un{map|lock}() must also visit each VMA in that list to determine
485 569 whether the page is mapped in a VM_LOCKED VMA. Again, the scan must visit
486So, try_to_unmap_file() handles non-linear mappings by scanning a certain 570 all VMAs in the non-linear list to ensure that the pages is not/should not
487number of pages--a "cluster"--in each non-linear vma associated with the page's 571 be mlocked.
488mapping, for each file mapped page that vmscan tries to unmap. If this happens 572
489to unmap the page we're trying to unmap, try_to_unmap() will notice this on 573 If a VM_LOCKED VMA is found in the list, the scan could terminate.
490return--(page_mapcount(page) == 0)--and return SWAP_SUCCESS. Otherwise, it 574 However, there is no easy way to determine whether the page is actually
491will return SWAP_AGAIN, causing vmscan to recirculate this page. We take 575 mapped in a given VMA - either for unmapping or testing whether the
492advantage of the cluster scan in try_to_unmap_cluster() as follows: 576 VM_LOCKED VMA actually pins the page.
493 577
494For each non-linear vma, try_to_unmap_cluster() attempts to acquire the mmap 578 try_to_unmap_file() handles non-linear mappings by scanning a certain
495semaphore of the associated mm_struct for read without blocking. If this 579 number of pages - a "cluster" - in each non-linear VMA associated with the
496attempt is successful and the vma is VM_LOCKED, try_to_unmap_cluster() will 580 page's mapping, for each file mapped page that vmscan tries to unmap. If
497retain the mmap semaphore for the scan; otherwise it drops it here. Then, 581 this happens to unmap the page we're trying to unmap, try_to_unmap() will
498for each page in the cluster, if we're holding the mmap semaphore for a locked 582 notice this on return (page_mapcount(page) will be 0) and return
499vma, try_to_unmap_cluster() calls mlock_vma_page() to mlock the page. This 583 SWAP_SUCCESS. Otherwise, it will return SWAP_AGAIN, causing vmscan to
500call is a no-op if the page is already locked, but will mlock any pages in 584 recirculate this page. We take advantage of the cluster scan in
501the non-linear mapping that happen to be unlocked. If one of the pages so 585 try_to_unmap_cluster() as follows:
502mlocked is the page passed in to try_to_unmap(), try_to_unmap_cluster() will 586
503return SWAP_MLOCK, rather than the default SWAP_AGAIN. This will allow vmscan 587 For each non-linear VMA, try_to_unmap_cluster() attempts to acquire the
504to cull the page, rather than recirculating it on the inactive list. Again, 588 mmap semaphore of the associated mm_struct for read without blocking.
505if try_to_unmap_cluster() cannot acquire the vma's mmap sem, it returns 589
506SWAP_AGAIN, indicating that the page is mapped by a VM_LOCKED vma, but 590 If this attempt is successful and the VMA is VM_LOCKED,
507couldn't be mlocked. 591 try_to_unmap_cluster() will retain the mmap semaphore for the scan;
508 592 otherwise it drops it here.
509 593
510Mlocked pages: try_to_munlock() Reverse Map Scan 594 Then, for each page in the cluster, if we're holding the mmap semaphore
511 595 for a locked VMA, try_to_unmap_cluster() calls mlock_vma_page() to
512TODO/FIXME: a better name might be page_mlocked()--analogous to the 596 mlock the page. This call is a no-op if the page is already locked,
513page_referenced() reverse map walker. 597 but will mlock any pages in the non-linear mapping that happen to be
514 598 unlocked.
515When munlock_vma_page()--see "Mlocked Pages: munlock()/munlockall() 599
516System Call Handling" above--tries to munlock a page, it needs to 600 If one of the pages so mlocked is the page passed in to try_to_unmap(),
517determine whether or not the page is mapped by any VM_LOCKED vma, without 601 try_to_unmap_cluster() will return SWAP_MLOCK, rather than the default
518actually attempting to unmap all ptes from the page. For this purpose, the 602 SWAP_AGAIN. This will allow vmscan to cull the page, rather than
519unevictable/mlock infrastructure introduced a variant of try_to_unmap() called 603 recirculating it on the inactive list.
520try_to_munlock(). 604
605 Again, if try_to_unmap_cluster() cannot acquire the VMA's mmap sem, it
606 returns SWAP_AGAIN, indicating that the page is mapped by a VM_LOCKED
607 VMA, but couldn't be mlocked.
608
609
610try_to_munlock() REVERSE MAP SCAN
611---------------------------------
612
613 [!] TODO/FIXME: a better name might be page_mlocked() - analogous to the
614 page_referenced() reverse map walker.
615
616When munlock_vma_page() [see section "munlock()/munlockall() System Call
617Handling" above] tries to munlock a page, it needs to determine whether or not
618the page is mapped by any VM_LOCKED VMA without actually attempting to unmap
619all PTEs from the page. For this purpose, the unevictable/mlock infrastructure
620introduced a variant of try_to_unmap() called try_to_munlock().
521 621
522try_to_munlock() calls the same functions as try_to_unmap() for anonymous and 622try_to_munlock() calls the same functions as try_to_unmap() for anonymous and
523mapped file pages with an additional argument specifing unlock versus unmap 623mapped file pages with an additional argument specifing unlock versus unmap
524processing. Again, these functions walk the respective reverse maps looking 624processing. Again, these functions walk the respective reverse maps looking
525for VM_LOCKED vmas. When such a vma is found for anonymous pages and file 625for VM_LOCKED VMAs. When such a VMA is found for anonymous pages and file
526pages mapped in linear VMAs, as in the try_to_unmap() case, the functions 626pages mapped in linear VMAs, as in the try_to_unmap() case, the functions
527attempt to acquire the associated mmap semphore, mlock the page via 627attempt to acquire the associated mmap semphore, mlock the page via
528mlock_vma_page() and return SWAP_MLOCK. This effectively undoes the 628mlock_vma_page() and return SWAP_MLOCK. This effectively undoes the
529pre-clearing of the page's PG_mlocked done by munlock_vma_page. 629pre-clearing of the page's PG_mlocked done by munlock_vma_page.
530 630
531If try_to_unmap() is unable to acquire a VM_LOCKED vma's associated mmap 631If try_to_unmap() is unable to acquire a VM_LOCKED VMA's associated mmap
532semaphore, it will return SWAP_AGAIN. This will allow shrink_page_list() 632semaphore, it will return SWAP_AGAIN. This will allow shrink_page_list() to
533to recycle the page on the inactive list and hope that it has better luck 633recycle the page on the inactive list and hope that it has better luck with the
534with the page next time. 634page next time.
535 635
536For file pages mapped into non-linear vmas, the try_to_munlock() logic works 636For file pages mapped into non-linear VMAs, the try_to_munlock() logic works
537slightly differently. On encountering a VM_LOCKED non-linear vma that might 637slightly differently. On encountering a VM_LOCKED non-linear VMA that might
538map the page, try_to_munlock() returns SWAP_AGAIN without actually mlocking 638map the page, try_to_munlock() returns SWAP_AGAIN without actually mlocking the
539the page. munlock_vma_page() will just leave the page unlocked and let 639page. munlock_vma_page() will just leave the page unlocked and let vmscan deal
540vmscan deal with it--the usual fallback position. 640with it - the usual fallback position.
541 641
542Note that try_to_munlock()'s reverse map walk must visit every vma in a pages' 642Note that try_to_munlock()'s reverse map walk must visit every VMA in a page's
543reverse map to determine that a page is NOT mapped into any VM_LOCKED vma. 643reverse map to determine that a page is NOT mapped into any VM_LOCKED VMA.
544However, the scan can terminate when it encounters a VM_LOCKED vma and can 644However, the scan can terminate when it encounters a VM_LOCKED VMA and can
545successfully acquire the vma's mmap semphore for read and mlock the page. 645successfully acquire the VMA's mmap semphore for read and mlock the page.
546Although try_to_munlock() can be called many [very many!] times when 646Although try_to_munlock() might be called a great many times when munlocking a
547munlock()ing a large region or tearing down a large address space that has been 647large region or tearing down a large address space that has been mlocked via
548mlocked via mlockall(), overall this is a fairly rare event. 648mlockall(), overall this is a fairly rare event.
549 649
550Mlocked Page: Page Reclaim in shrink_*_list() 650
551 651PAGE RECLAIM IN shrink_*_list()
552shrink_active_list() culls any obviously unevictable pages--i.e., 652-------------------------------
553!page_evictable(page, NULL)--diverting these to the unevictable lru 653
554list. However, shrink_active_list() only sees unevictable pages that 654shrink_active_list() culls any obviously unevictable pages - i.e.
555made it onto the active/inactive lru lists. Note that these pages do not 655!page_evictable(page, NULL) - diverting these to the unevictable list.
556have PageUnevictable set--otherwise, they would be on the unevictable list and 656However, shrink_active_list() only sees unevictable pages that made it onto the
557shrink_active_list would never see them. 657active/inactive lru lists. Note that these pages do not have PageUnevictable
658set - otherwise they would be on the unevictable list and shrink_active_list
659would never see them.
558 660
559Some examples of these unevictable pages on the LRU lists are: 661Some examples of these unevictable pages on the LRU lists are:
560 662
5611) ramfs pages that have been placed on the lru lists when first allocated. 663 (1) ramfs pages that have been placed on the LRU lists when first allocated.
664
665 (2) SHM_LOCK'd shared memory pages. shmctl(SHM_LOCK) does not attempt to
666 allocate or fault in the pages in the shared memory region. This happens
667 when an application accesses the page the first time after SHM_LOCK'ing
668 the segment.
562 669
5632) SHM_LOCKed shared memory pages. shmctl(SHM_LOCK) does not attempt to 670 (3) mlocked pages that could not be isolated from the LRU and moved to the
564 allocate or fault in the pages in the shared memory region. This happens 671 unevictable list in mlock_vma_page().
565 when an application accesses the page the first time after SHM_LOCKing
566 the segment.
567 672
5683) Mlocked pages that could not be isolated from the lru and moved to the 673 (4) Pages mapped into multiple VM_LOCKED VMAs, but try_to_munlock() couldn't
569 unevictable list in mlock_vma_page(). 674 acquire the VMA's mmap semaphore to test the flags and set PageMlocked.
675 munlock_vma_page() was forced to let the page back on to the normal LRU
676 list for vmscan to handle.
570 677
5713) Pages mapped into multiple VM_LOCKED vmas, but try_to_munlock() couldn't 678shrink_inactive_list() also diverts any unevictable pages that it finds on the
572 acquire the vma's mmap semaphore to test the flags and set PageMlocked. 679inactive lists to the appropriate zone's unevictable list.
573 munlock_vma_page() was forced to let the page back on to the normal
574 LRU list for vmscan to handle.
575 680
576shrink_inactive_list() also culls any unevictable pages that it finds on 681shrink_inactive_list() should only see SHM_LOCK'd pages that became SHM_LOCK'd
577the inactive lists, again diverting them to the appropriate zone's unevictable 682after shrink_active_list() had moved them to the inactive list, or pages mapped
578lru list. shrink_inactive_list() should only see SHM_LOCKed pages that became 683into VM_LOCKED VMAs that munlock_vma_page() couldn't isolate from the LRU to
579SHM_LOCKed after shrink_active_list() had moved them to the inactive list, or 684recheck via try_to_munlock(). shrink_inactive_list() won't notice the latter,
580pages mapped into VM_LOCKED vmas that munlock_vma_page() couldn't isolate from 685but will pass on to shrink_page_list().
581the lru to recheck via try_to_munlock(). shrink_inactive_list() won't notice
582the latter, but will pass on to shrink_page_list().
583 686
584shrink_page_list() again culls obviously unevictable pages that it could 687shrink_page_list() again culls obviously unevictable pages that it could
585encounter for similar reason to shrink_inactive_list(). Pages mapped into 688encounter for similar reason to shrink_inactive_list(). Pages mapped into
586VM_LOCKED vmas but without PG_mlocked set will make it all the way to 689VM_LOCKED VMAs but without PG_mlocked set will make it all the way to
587try_to_unmap(). shrink_page_list() will divert them to the unevictable list 690try_to_unmap(). shrink_page_list() will divert them to the unevictable list
588when try_to_unmap() returns SWAP_MLOCK, as discussed above. 691when try_to_unmap() returns SWAP_MLOCK, as discussed above.
diff --git a/MAINTAINERS b/MAINTAINERS
index ca2997a45766..abbedb658c29 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -461,7 +461,7 @@ F: arch/x86/include/asm/amd_iommu*.h
461 461
462AMD MICROCODE UPDATE SUPPORT 462AMD MICROCODE UPDATE SUPPORT
463P: Andreas Herrmann 463P: Andreas Herrmann
464M: andeas.herrmann3@amd.com 464M: andreas.herrmann3@amd.com
465L: amd64-microcode@amd64.org 465L: amd64-microcode@amd64.org
466S: Supported 466S: Supported
467F: arch/x86/kernel/microcode_amd.c 467F: arch/x86/kernel/microcode_amd.c
@@ -1894,7 +1894,7 @@ F: fs/ecryptfs/
1894EDAC-CORE 1894EDAC-CORE
1895P: Doug Thompson 1895P: Doug Thompson
1896M: dougthompson@xmission.com 1896M: dougthompson@xmission.com
1897L: bluesmoke-devel@lists.sourceforge.net 1897L: bluesmoke-devel@lists.sourceforge.net (moderated for non-subscribers)
1898W: bluesmoke.sourceforge.net 1898W: bluesmoke.sourceforge.net
1899S: Supported 1899S: Supported
1900F: Documentation/edac.txt 1900F: Documentation/edac.txt
@@ -1906,7 +1906,7 @@ P: Mark Gross
1906P: Doug Thompson 1906P: Doug Thompson
1907M: mark.gross@intel.com 1907M: mark.gross@intel.com
1908M: dougthompson@xmission.com 1908M: dougthompson@xmission.com
1909L: bluesmoke-devel@lists.sourceforge.net 1909L: bluesmoke-devel@lists.sourceforge.net (moderated for non-subscribers)
1910W: bluesmoke.sourceforge.net 1910W: bluesmoke.sourceforge.net
1911S: Maintained 1911S: Maintained
1912F: drivers/edac/e752x_edac.c 1912F: drivers/edac/e752x_edac.c
@@ -1914,7 +1914,7 @@ F: drivers/edac/e752x_edac.c
1914EDAC-E7XXX 1914EDAC-E7XXX
1915P: Doug Thompson 1915P: Doug Thompson
1916M: dougthompson@xmission.com 1916M: dougthompson@xmission.com
1917L: bluesmoke-devel@lists.sourceforge.net 1917L: bluesmoke-devel@lists.sourceforge.net (moderated for non-subscribers)
1918W: bluesmoke.sourceforge.net 1918W: bluesmoke.sourceforge.net
1919S: Maintained 1919S: Maintained
1920F: drivers/edac/e7xxx_edac.c 1920F: drivers/edac/e7xxx_edac.c
@@ -1922,7 +1922,7 @@ F: drivers/edac/e7xxx_edac.c
1922EDAC-I82443BXGX 1922EDAC-I82443BXGX
1923P: Tim Small 1923P: Tim Small
1924M: tim@buttersideup.com 1924M: tim@buttersideup.com
1925L: bluesmoke-devel@lists.sourceforge.net 1925L: bluesmoke-devel@lists.sourceforge.net (moderated for non-subscribers)
1926W: bluesmoke.sourceforge.net 1926W: bluesmoke.sourceforge.net
1927S: Maintained 1927S: Maintained
1928F: drivers/edac/i82443bxgx_edac.c 1928F: drivers/edac/i82443bxgx_edac.c
@@ -1930,7 +1930,7 @@ F: drivers/edac/i82443bxgx_edac.c
1930EDAC-I3000 1930EDAC-I3000
1931P: Jason Uhlenkott 1931P: Jason Uhlenkott
1932M: juhlenko@akamai.com 1932M: juhlenko@akamai.com
1933L: bluesmoke-devel@lists.sourceforge.net 1933L: bluesmoke-devel@lists.sourceforge.net (moderated for non-subscribers)
1934W: bluesmoke.sourceforge.net 1934W: bluesmoke.sourceforge.net
1935S: Maintained 1935S: Maintained
1936F: drivers/edac/i3000_edac.c 1936F: drivers/edac/i3000_edac.c
@@ -1938,7 +1938,7 @@ F: drivers/edac/i3000_edac.c
1938EDAC-I5000 1938EDAC-I5000
1939P: Doug Thompson 1939P: Doug Thompson
1940M: dougthompson@xmission.com 1940M: dougthompson@xmission.com
1941L: bluesmoke-devel@lists.sourceforge.net 1941L: bluesmoke-devel@lists.sourceforge.net (moderated for non-subscribers)
1942W: bluesmoke.sourceforge.net 1942W: bluesmoke.sourceforge.net
1943S: Maintained 1943S: Maintained
1944F: drivers/edac/i5000_edac.c 1944F: drivers/edac/i5000_edac.c
@@ -1946,7 +1946,7 @@ F: drivers/edac/i5000_edac.c
1946EDAC-I5400 1946EDAC-I5400
1947P: Mauro Carvalho Chehab 1947P: Mauro Carvalho Chehab
1948M: mchehab@redhat.com 1948M: mchehab@redhat.com
1949L: bluesmoke-devel@lists.sourceforge.net 1949L: bluesmoke-devel@lists.sourceforge.net (moderated for non-subscribers)
1950W: bluesmoke.sourceforge.net 1950W: bluesmoke.sourceforge.net
1951S: Maintained 1951S: Maintained
1952F: drivers/edac/i5400_edac.c 1952F: drivers/edac/i5400_edac.c
@@ -1956,7 +1956,7 @@ P: Ranganathan Desikan
1956P: Arvind R. 1956P: Arvind R.
1957M: rdesikan@jetzbroadband.com 1957M: rdesikan@jetzbroadband.com
1958M: arvind@acarlab.com 1958M: arvind@acarlab.com
1959L: bluesmoke-devel@lists.sourceforge.net 1959L: bluesmoke-devel@lists.sourceforge.net (moderated for non-subscribers)
1960W: bluesmoke.sourceforge.net 1960W: bluesmoke.sourceforge.net
1961S: Maintained 1961S: Maintained
1962F: drivers/edac/i82975x_edac.c 1962F: drivers/edac/i82975x_edac.c
@@ -1964,7 +1964,7 @@ F: drivers/edac/i82975x_edac.c
1964EDAC-PASEMI 1964EDAC-PASEMI
1965P: Egor Martovetsky 1965P: Egor Martovetsky
1966M: egor@pasemi.com 1966M: egor@pasemi.com
1967L: bluesmoke-devel@lists.sourceforge.net 1967L: bluesmoke-devel@lists.sourceforge.net (moderated for non-subscribers)
1968W: bluesmoke.sourceforge.net 1968W: bluesmoke.sourceforge.net
1969S: Maintained 1969S: Maintained
1970F: drivers/edac/pasemi_edac.c 1970F: drivers/edac/pasemi_edac.c
@@ -1972,7 +1972,7 @@ F: drivers/edac/pasemi_edac.c
1972EDAC-R82600 1972EDAC-R82600
1973P: Tim Small 1973P: Tim Small
1974M: tim@buttersideup.com 1974M: tim@buttersideup.com
1975L: bluesmoke-devel@lists.sourceforge.net 1975L: bluesmoke-devel@lists.sourceforge.net (moderated for non-subscribers)
1976W: bluesmoke.sourceforge.net 1976W: bluesmoke.sourceforge.net
1977S: Maintained 1977S: Maintained
1978F: drivers/edac/r82600_edac.c 1978F: drivers/edac/r82600_edac.c
@@ -2592,8 +2592,8 @@ S: Maintained
2592F: fs/hpfs/ 2592F: fs/hpfs/
2593 2593
2594HSO 3G MODEM DRIVER 2594HSO 3G MODEM DRIVER
2595P: Denis Joseph Barrow 2595P: Jan Dumon
2596M: d.barow@option.com 2596M: j.dumon@option.com
2597W: http://www.pharscape.org 2597W: http://www.pharscape.org
2598S: Maintained 2598S: Maintained
2599F: drivers/net/usb/hso.c 2599F: drivers/net/usb/hso.c
@@ -3680,6 +3680,7 @@ L: microblaze-uclinux@itee.uq.edu.au
3680W: http://www.monstr.eu/fdt/ 3680W: http://www.monstr.eu/fdt/
3681T: git git://git.monstr.eu/linux-2.6-microblaze.git 3681T: git git://git.monstr.eu/linux-2.6-microblaze.git
3682S: Supported 3682S: Supported
3683F: arch/microblaze/
3683 3684
3684MICROTEK X6 SCANNER 3685MICROTEK X6 SCANNER
3685P: Oliver Neukum 3686P: Oliver Neukum
@@ -3916,19 +3917,12 @@ F: Documentation/blockdev/nbd.txt
3916F: drivers/block/nbd.c 3917F: drivers/block/nbd.c
3917F: include/linux/nbd.h 3918F: include/linux/nbd.h
3918 3919
3919NETWORK DEVICE DRIVERS
3920P: Jeff Garzik
3921M: jgarzik@pobox.com
3922L: netdev@vger.kernel.org
3923T: git git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/netdev-2.6.git
3924S: Maintained
3925F: drivers/net/
3926
3927NETWORKING [GENERAL] 3920NETWORKING [GENERAL]
3928P: Networking Team 3921P: David S. Miller
3929M: netdev@vger.kernel.org 3922M: davem@davemloft.net
3930L: netdev@vger.kernel.org 3923L: netdev@vger.kernel.org
3931W: http://linux-net.osdl.org/ 3924W: http://linux-net.osdl.org/
3925T: git kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6.git
3932S: Maintained 3926S: Maintained
3933F: net/ 3927F: net/
3934F: include/net/ 3928F: include/net/
@@ -4978,8 +4972,8 @@ S: Maintained for 2.6.
4978F: Documentation/sgi-visws.txt 4972F: Documentation/sgi-visws.txt
4979 4973
4980SGI XP/XPC/XPNET DRIVER 4974SGI XP/XPC/XPNET DRIVER
4981P: Dean Nelson 4975P: Robin Holt
4982M: dcn@sgi.com 4976M: holt@sgi.com
4983S: Maintained 4977S: Maintained
4984F: drivers/misc/sgi-xp/ 4978F: drivers/misc/sgi-xp/
4985 4979
@@ -5313,7 +5307,9 @@ L: linux-sh@vger.kernel.org
5313W: http://www.linux-sh.org 5307W: http://www.linux-sh.org
5314T: git git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6.git 5308T: git git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6.git
5315S: Supported 5309S: Supported
5310F: Documentation/sh/
5316F: arch/sh/ 5311F: arch/sh/
5312F: drivers/sh/
5317 5313
5318SUSPEND TO RAM 5314SUSPEND TO RAM
5319P: Len Brown 5315P: Len Brown
diff --git a/Makefile b/Makefile
index ad830bd45a4b..bfdef56add34 100644
--- a/Makefile
+++ b/Makefile
@@ -1,7 +1,7 @@
1VERSION = 2 1VERSION = 2
2PATCHLEVEL = 6 2PATCHLEVEL = 6
3SUBLEVEL = 30 3SUBLEVEL = 30
4EXTRAVERSION = -rc1 4EXTRAVERSION = -rc2
5NAME = Temporary Tasmanian Devil 5NAME = Temporary Tasmanian Devil
6 6
7# *DOCUMENTATION* 7# *DOCUMENTATION*
@@ -169,7 +169,7 @@ SUBARCH := $(shell uname -m | sed -e s/i.86/i386/ -e s/sun4u/sparc64/ \
169 -e s/arm.*/arm/ -e s/sa110/arm/ \ 169 -e s/arm.*/arm/ -e s/sa110/arm/ \
170 -e s/s390x/s390/ -e s/parisc64/parisc/ \ 170 -e s/s390x/s390/ -e s/parisc64/parisc/ \
171 -e s/ppc.*/powerpc/ -e s/mips.*/mips/ \ 171 -e s/ppc.*/powerpc/ -e s/mips.*/mips/ \
172 -e s/sh.*/sh/ ) 172 -e s/sh[234].*/sh/ )
173 173
174# Cross compiling and selecting different set of gcc/bin-utils 174# Cross compiling and selecting different set of gcc/bin-utils
175# --------------------------------------------------------------------------- 175# ---------------------------------------------------------------------------
@@ -210,6 +210,11 @@ ifeq ($(ARCH),sparc64)
210 SRCARCH := sparc 210 SRCARCH := sparc
211endif 211endif
212 212
213# Additional ARCH settings for sh
214ifeq ($(ARCH),sh64)
215 SRCARCH := sh
216endif
217
213# Where to locate arch specific headers 218# Where to locate arch specific headers
214hdr-arch := $(SRCARCH) 219hdr-arch := $(SRCARCH)
215 220
diff --git a/arch/arm/mach-omap2/usb-musb.c b/arch/arm/mach-omap2/usb-musb.c
index fc74e913c415..34a56a136efd 100644
--- a/arch/arm/mach-omap2/usb-musb.c
+++ b/arch/arm/mach-omap2/usb-musb.c
@@ -131,14 +131,14 @@ static struct musb_hdrc_platform_data musb_plat = {
131 .power = 50, /* up to 100 mA */ 131 .power = 50, /* up to 100 mA */
132}; 132};
133 133
134static u64 musb_dmamask = DMA_32BIT_MASK; 134static u64 musb_dmamask = DMA_BIT_MASK(32);
135 135
136static struct platform_device musb_device = { 136static struct platform_device musb_device = {
137 .name = "musb_hdrc", 137 .name = "musb_hdrc",
138 .id = -1, 138 .id = -1,
139 .dev = { 139 .dev = {
140 .dma_mask = &musb_dmamask, 140 .dma_mask = &musb_dmamask,
141 .coherent_dma_mask = DMA_32BIT_MASK, 141 .coherent_dma_mask = DMA_BIT_MASK(32),
142 .platform_data = &musb_plat, 142 .platform_data = &musb_plat,
143 }, 143 },
144 .num_resources = ARRAY_SIZE(musb_resources), 144 .num_resources = ARRAY_SIZE(musb_resources),
@@ -146,14 +146,14 @@ static struct platform_device musb_device = {
146}; 146};
147 147
148#ifdef CONFIG_NOP_USB_XCEIV 148#ifdef CONFIG_NOP_USB_XCEIV
149static u64 nop_xceiv_dmamask = DMA_32BIT_MASK; 149static u64 nop_xceiv_dmamask = DMA_BIT_MASK(32);
150 150
151static struct platform_device nop_xceiv_device = { 151static struct platform_device nop_xceiv_device = {
152 .name = "nop_usb_xceiv", 152 .name = "nop_usb_xceiv",
153 .id = -1, 153 .id = -1,
154 .dev = { 154 .dev = {
155 .dma_mask = &nop_xceiv_dmamask, 155 .dma_mask = &nop_xceiv_dmamask,
156 .coherent_dma_mask = DMA_32BIT_MASK, 156 .coherent_dma_mask = DMA_BIT_MASK(32),
157 .platform_data = NULL, 157 .platform_data = NULL,
158 }, 158 },
159}; 159};
diff --git a/arch/ia64/kernel/pci-swiotlb.c b/arch/ia64/kernel/pci-swiotlb.c
index 573f02c39a00..285aae8431c6 100644
--- a/arch/ia64/kernel/pci-swiotlb.c
+++ b/arch/ia64/kernel/pci-swiotlb.c
@@ -16,7 +16,7 @@ EXPORT_SYMBOL(swiotlb);
16static void *ia64_swiotlb_alloc_coherent(struct device *dev, size_t size, 16static void *ia64_swiotlb_alloc_coherent(struct device *dev, size_t size,
17 dma_addr_t *dma_handle, gfp_t gfp) 17 dma_addr_t *dma_handle, gfp_t gfp)
18{ 18{
19 if (dev->coherent_dma_mask != DMA_64BIT_MASK) 19 if (dev->coherent_dma_mask != DMA_BIT_MASK(64))
20 gfp |= GFP_DMA; 20 gfp |= GFP_DMA;
21 return swiotlb_alloc_coherent(dev, size, dma_handle, gfp); 21 return swiotlb_alloc_coherent(dev, size, dma_handle, gfp);
22} 22}
diff --git a/arch/microblaze/include/asm/auxvec.h b/arch/microblaze/include/asm/auxvec.h
index e69de29bb2d1..8b137891791f 100644
--- a/arch/microblaze/include/asm/auxvec.h
+++ b/arch/microblaze/include/asm/auxvec.h
@@ -0,0 +1 @@
diff --git a/arch/microblaze/include/asm/cputable.h b/arch/microblaze/include/asm/cputable.h
index e69de29bb2d1..8b137891791f 100644
--- a/arch/microblaze/include/asm/cputable.h
+++ b/arch/microblaze/include/asm/cputable.h
@@ -0,0 +1 @@
diff --git a/arch/microblaze/include/asm/ftrace.h b/arch/microblaze/include/asm/ftrace.h
new file mode 100644
index 000000000000..8b137891791f
--- /dev/null
+++ b/arch/microblaze/include/asm/ftrace.h
@@ -0,0 +1 @@
diff --git a/arch/microblaze/include/asm/hw_irq.h b/arch/microblaze/include/asm/hw_irq.h
index e69de29bb2d1..8b137891791f 100644
--- a/arch/microblaze/include/asm/hw_irq.h
+++ b/arch/microblaze/include/asm/hw_irq.h
@@ -0,0 +1 @@
diff --git a/arch/microblaze/include/asm/io.h b/arch/microblaze/include/asm/io.h
index cfab0342588d..8b5853ee6b5c 100644
--- a/arch/microblaze/include/asm/io.h
+++ b/arch/microblaze/include/asm/io.h
@@ -12,7 +12,6 @@
12#include <asm/byteorder.h> 12#include <asm/byteorder.h>
13#include <asm/page.h> 13#include <asm/page.h>
14#include <linux/types.h> 14#include <linux/types.h>
15#include <asm/page.h>
16 15
17#define IO_SPACE_LIMIT (0xFFFFFFFF) 16#define IO_SPACE_LIMIT (0xFFFFFFFF)
18 17
diff --git a/arch/microblaze/include/asm/socket.h b/arch/microblaze/include/asm/socket.h
index f919b6b540ac..825936860314 100644
--- a/arch/microblaze/include/asm/socket.h
+++ b/arch/microblaze/include/asm/socket.h
@@ -63,4 +63,7 @@
63 63
64#define SO_MARK 36 64#define SO_MARK 36
65 65
66#define SO_TIMESTAMPING 37
67#define SCM_TIMESTAMPING SO_TIMESTAMPING
68
66#endif /* _ASM_MICROBLAZE_SOCKET_H */ 69#endif /* _ASM_MICROBLAZE_SOCKET_H */
diff --git a/arch/microblaze/include/asm/user.h b/arch/microblaze/include/asm/user.h
index e69de29bb2d1..8b137891791f 100644
--- a/arch/microblaze/include/asm/user.h
+++ b/arch/microblaze/include/asm/user.h
@@ -0,0 +1 @@
diff --git a/arch/microblaze/include/asm/vga.h b/arch/microblaze/include/asm/vga.h
index e69de29bb2d1..8b137891791f 100644
--- a/arch/microblaze/include/asm/vga.h
+++ b/arch/microblaze/include/asm/vga.h
@@ -0,0 +1 @@
diff --git a/arch/microblaze/kernel/of_device.c b/arch/microblaze/kernel/of_device.c
index 717edf4ad0b4..9a0f7632c47c 100644
--- a/arch/microblaze/kernel/of_device.c
+++ b/arch/microblaze/kernel/of_device.c
@@ -13,7 +13,6 @@ void of_device_make_bus_id(struct of_device *dev)
13{ 13{
14 static atomic_t bus_no_reg_magic; 14 static atomic_t bus_no_reg_magic;
15 struct device_node *node = dev->node; 15 struct device_node *node = dev->node;
16 char *name = dev->dev.bus_id;
17 const u32 *reg; 16 const u32 *reg;
18 u64 addr; 17 u64 addr;
19 int magic; 18 int magic;
@@ -25,9 +24,8 @@ void of_device_make_bus_id(struct of_device *dev)
25 if (reg) { 24 if (reg) {
26 addr = of_translate_address(node, reg); 25 addr = of_translate_address(node, reg);
27 if (addr != OF_BAD_ADDR) { 26 if (addr != OF_BAD_ADDR) {
28 snprintf(name, BUS_ID_SIZE, 27 dev_set_name(&dev->dev, "%llx.%s",
29 "%llx.%s", (unsigned long long)addr, 28 (unsigned long long)addr, node->name);
30 node->name);
31 return; 29 return;
32 } 30 }
33 } 31 }
@@ -37,7 +35,7 @@ void of_device_make_bus_id(struct of_device *dev)
37 * counter (and pray...) 35 * counter (and pray...)
38 */ 36 */
39 magic = atomic_add_return(1, &bus_no_reg_magic); 37 magic = atomic_add_return(1, &bus_no_reg_magic);
40 snprintf(name, BUS_ID_SIZE, "%s.%d", node->name, magic - 1); 38 dev_set_name(&dev->dev, "%s.%d", node->name, magic - 1);
41} 39}
42EXPORT_SYMBOL(of_device_make_bus_id); 40EXPORT_SYMBOL(of_device_make_bus_id);
43 41
@@ -58,7 +56,7 @@ struct of_device *of_device_alloc(struct device_node *np,
58 dev->dev.archdata.of_node = np; 56 dev->dev.archdata.of_node = np;
59 57
60 if (bus_id) 58 if (bus_id)
61 strlcpy(dev->dev.bus_id, bus_id, BUS_ID_SIZE); 59 dev_set_name(&dev->dev, bus_id);
62 else 60 else
63 of_device_make_bus_id(dev); 61 of_device_make_bus_id(dev);
64 62
diff --git a/arch/microblaze/kernel/process.c b/arch/microblaze/kernel/process.c
index 60e9ed7d3132..436f26ccbfa9 100644
--- a/arch/microblaze/kernel/process.c
+++ b/arch/microblaze/kernel/process.c
@@ -115,8 +115,7 @@ void flush_thread(void)
115{ 115{
116} 116}
117 117
118/* FIXME - here will be a proposed change -> remove nr parameter */ 118int copy_thread(unsigned long clone_flags, unsigned long usp,
119int copy_thread(int nr, unsigned long clone_flags, unsigned long usp,
120 unsigned long unused, 119 unsigned long unused,
121 struct task_struct *p, struct pt_regs *regs) 120 struct task_struct *p, struct pt_regs *regs)
122{ 121{
diff --git a/arch/microblaze/kernel/prom.c b/arch/microblaze/kernel/prom.c
index 475b1fac5cfd..34c48718061a 100644
--- a/arch/microblaze/kernel/prom.c
+++ b/arch/microblaze/kernel/prom.c
@@ -39,7 +39,6 @@
39#include <asm/system.h> 39#include <asm/system.h>
40#include <asm/mmu.h> 40#include <asm/mmu.h>
41#include <asm/pgtable.h> 41#include <asm/pgtable.h>
42#include <linux/pci.h>
43#include <asm/sections.h> 42#include <asm/sections.h>
44#include <asm/pci-bridge.h> 43#include <asm/pci-bridge.h>
45 44
diff --git a/arch/microblaze/kernel/ptrace.c b/arch/microblaze/kernel/ptrace.c
index 3171e39e3220..b86aa623e36d 100644
--- a/arch/microblaze/kernel/ptrace.c
+++ b/arch/microblaze/kernel/ptrace.c
@@ -32,7 +32,6 @@
32#include <linux/signal.h> 32#include <linux/signal.h>
33 33
34#include <linux/errno.h> 34#include <linux/errno.h>
35#include <linux/ptrace.h>
36#include <asm/processor.h> 35#include <asm/processor.h>
37#include <linux/uaccess.h> 36#include <linux/uaccess.h>
38#include <asm/asm-offsets.h> 37#include <asm/asm-offsets.h>
diff --git a/arch/microblaze/kernel/signal.c b/arch/microblaze/kernel/signal.c
index ff347b98863a..3889cf45fa71 100644
--- a/arch/microblaze/kernel/signal.c
+++ b/arch/microblaze/kernel/signal.c
@@ -37,7 +37,6 @@
37#include <linux/uaccess.h> 37#include <linux/uaccess.h>
38#include <asm/pgtable.h> 38#include <asm/pgtable.h>
39#include <asm/pgalloc.h> 39#include <asm/pgalloc.h>
40#include <linux/signal.h>
41#include <linux/syscalls.h> 40#include <linux/syscalls.h>
42#include <asm/cacheflush.h> 41#include <asm/cacheflush.h>
43#include <asm/syscalls.h> 42#include <asm/syscalls.h>
diff --git a/arch/microblaze/kernel/sys_microblaze.c b/arch/microblaze/kernel/sys_microblaze.c
index d90b548fb1bb..ba0568c2cc1c 100644
--- a/arch/microblaze/kernel/sys_microblaze.c
+++ b/arch/microblaze/kernel/sys_microblaze.c
@@ -29,9 +29,7 @@
29#include <linux/module.h> 29#include <linux/module.h>
30#include <linux/err.h> 30#include <linux/err.h>
31#include <linux/fs.h> 31#include <linux/fs.h>
32#include <linux/ipc.h>
33#include <linux/semaphore.h> 32#include <linux/semaphore.h>
34#include <linux/syscalls.h>
35#include <linux/uaccess.h> 33#include <linux/uaccess.h>
36#include <linux/unistd.h> 34#include <linux/unistd.h>
37 35
diff --git a/arch/mn10300/include/asm/bug.h b/arch/mn10300/include/asm/bug.h
index 4fcf3384e259..aa6a38886391 100644
--- a/arch/mn10300/include/asm/bug.h
+++ b/arch/mn10300/include/asm/bug.h
@@ -11,10 +11,12 @@
11#ifndef _ASM_BUG_H 11#ifndef _ASM_BUG_H
12#define _ASM_BUG_H 12#define _ASM_BUG_H
13 13
14#ifdef CONFIG_BUG
15
14/* 16/*
15 * Tell the user there is some problem. 17 * Tell the user there is some problem.
16 */ 18 */
17#define _debug_bug_trap() \ 19#define BUG() \
18do { \ 20do { \
19 asm volatile( \ 21 asm volatile( \
20 " syscall 15 \n" \ 22 " syscall 15 \n" \
@@ -25,11 +27,11 @@ do { \
25 : \ 27 : \
26 : "i"(__FILE__), "i"(__LINE__) \ 28 : "i"(__FILE__), "i"(__LINE__) \
27 ); \ 29 ); \
28} while (0) 30} while (1)
29
30#define BUG() _debug_bug_trap()
31 31
32#define HAVE_ARCH_BUG 32#define HAVE_ARCH_BUG
33#endif /* CONFIG_BUG */
34
33#include <asm-generic/bug.h> 35#include <asm-generic/bug.h>
34 36
35#endif /* _ASM_BUG_H */ 37#endif /* _ASM_BUG_H */
diff --git a/arch/mn10300/include/asm/unistd.h b/arch/mn10300/include/asm/unistd.h
index 543a4f98695d..fef5b434dadc 100644
--- a/arch/mn10300/include/asm/unistd.h
+++ b/arch/mn10300/include/asm/unistd.h
@@ -344,6 +344,8 @@
344#define __NR_dup3 331 344#define __NR_dup3 331
345#define __NR_pipe2 332 345#define __NR_pipe2 332
346#define __NR_inotify_init1 333 346#define __NR_inotify_init1 333
347#define __NR_preadv 334
348#define __NR_pwritev 335
347 349
348#ifdef __KERNEL__ 350#ifdef __KERNEL__
349 351
diff --git a/arch/mn10300/kernel/entry.S b/arch/mn10300/kernel/entry.S
index 34ab5a293153..3dc3e462f92a 100644
--- a/arch/mn10300/kernel/entry.S
+++ b/arch/mn10300/kernel/entry.S
@@ -723,6 +723,8 @@ ENTRY(sys_call_table)
723 .long sys_dup3 723 .long sys_dup3
724 .long sys_pipe2 724 .long sys_pipe2
725 .long sys_inotify_init1 725 .long sys_inotify_init1
726 .long sys_preadv
727 .long sys_pwritev /* 335 */
726 728
727 729
728nr_syscalls=(.-sys_call_table)/4 730nr_syscalls=(.-sys_call_table)/4
diff --git a/arch/mn10300/kernel/setup.c b/arch/mn10300/kernel/setup.c
index 71414e19fd16..79890edfd67a 100644
--- a/arch/mn10300/kernel/setup.c
+++ b/arch/mn10300/kernel/setup.c
@@ -136,10 +136,6 @@ void __init setup_arch(char **cmdline_p)
136 data_resource.start = virt_to_bus(&_etext); 136 data_resource.start = virt_to_bus(&_etext);
137 data_resource.end = virt_to_bus(&_edata)-1; 137 data_resource.end = virt_to_bus(&_edata)-1;
138 138
139#define PFN_UP(x) (((x) + PAGE_SIZE-1) >> PAGE_SHIFT)
140#define PFN_DOWN(x) ((x) >> PAGE_SHIFT)
141#define PFN_PHYS(x) ((x) << PAGE_SHIFT)
142
143 start_pfn = (CONFIG_KERNEL_RAM_BASE_ADDRESS >> PAGE_SHIFT); 139 start_pfn = (CONFIG_KERNEL_RAM_BASE_ADDRESS >> PAGE_SHIFT);
144 kstart_pfn = PFN_UP(__pa(&_text)); 140 kstart_pfn = PFN_UP(__pa(&_text));
145 free_pfn = PFN_UP(__pa(&_end)); 141 free_pfn = PFN_UP(__pa(&_end));
diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
index 5b50e1ac6179..4c7804551362 100644
--- a/arch/powerpc/Kconfig
+++ b/arch/powerpc/Kconfig
@@ -462,7 +462,7 @@ config PPC_64K_PAGES
462 462
463config PPC_256K_PAGES 463config PPC_256K_PAGES
464 bool "256k page size" if 44x 464 bool "256k page size" if 44x
465 depends on !STDBINUTILS && (!SHMEM || BROKEN) 465 depends on !STDBINUTILS
466 help 466 help
467 Make the page size 256k. 467 Make the page size 256k.
468 468
diff --git a/arch/powerpc/boot/dts/tqm8540.dts b/arch/powerpc/boot/dts/tqm8540.dts
index 231bae756637..b6f1fc6eb960 100644
--- a/arch/powerpc/boot/dts/tqm8540.dts
+++ b/arch/powerpc/boot/dts/tqm8540.dts
@@ -84,9 +84,9 @@
84 interrupt-parent = <&mpic>; 84 interrupt-parent = <&mpic>;
85 dfsrr; 85 dfsrr;
86 86
87 dtt@50 { 87 dtt@48 {
88 compatible = "national,lm75"; 88 compatible = "national,lm75";
89 reg = <0x50>; 89 reg = <0x48>;
90 }; 90 };
91 91
92 rtc@68 { 92 rtc@68 {
diff --git a/arch/powerpc/boot/dts/tqm8541.dts b/arch/powerpc/boot/dts/tqm8541.dts
index 4356a1f08295..fa6a3d54a8a5 100644
--- a/arch/powerpc/boot/dts/tqm8541.dts
+++ b/arch/powerpc/boot/dts/tqm8541.dts
@@ -83,9 +83,9 @@
83 interrupt-parent = <&mpic>; 83 interrupt-parent = <&mpic>;
84 dfsrr; 84 dfsrr;
85 85
86 dtt@50 { 86 dtt@48 {
87 compatible = "national,lm75"; 87 compatible = "national,lm75";
88 reg = <0x50>; 88 reg = <0x48>;
89 }; 89 };
90 90
91 rtc@68 { 91 rtc@68 {
diff --git a/arch/powerpc/boot/dts/tqm8548-bigflash.dts b/arch/powerpc/boot/dts/tqm8548-bigflash.dts
index 19aa72301c83..00f7ed7a2455 100644
--- a/arch/powerpc/boot/dts/tqm8548-bigflash.dts
+++ b/arch/powerpc/boot/dts/tqm8548-bigflash.dts
@@ -85,9 +85,9 @@
85 interrupt-parent = <&mpic>; 85 interrupt-parent = <&mpic>;
86 dfsrr; 86 dfsrr;
87 87
88 dtt@50 { 88 dtt@48 {
89 compatible = "national,lm75"; 89 compatible = "national,lm75";
90 reg = <0x50>; 90 reg = <0x48>;
91 }; 91 };
92 92
93 rtc@68 { 93 rtc@68 {
@@ -247,7 +247,7 @@
247 interrupts = <31 2 32 2 33 2>; 247 interrupts = <31 2 32 2 33 2>;
248 interrupt-parent = <&mpic>; 248 interrupt-parent = <&mpic>;
249 tbi-handle = <&tbi2>; 249 tbi-handle = <&tbi2>;
250 phy-handle = <&phy3>; 250 phy-handle = <&phy4>;
251 251
252 mdio@520 { 252 mdio@520 {
253 #address-cells = <1>; 253 #address-cells = <1>;
@@ -275,7 +275,7 @@
275 interrupts = <37 2 38 2 39 2>; 275 interrupts = <37 2 38 2 39 2>;
276 interrupt-parent = <&mpic>; 276 interrupt-parent = <&mpic>;
277 tbi-handle = <&tbi3>; 277 tbi-handle = <&tbi3>;
278 phy-handle = <&phy4>; 278 phy-handle = <&phy5>;
279 279
280 mdio@520 { 280 mdio@520 {
281 #address-cells = <1>; 281 #address-cells = <1>;
diff --git a/arch/powerpc/boot/dts/tqm8548.dts b/arch/powerpc/boot/dts/tqm8548.dts
index 49145a04fc6c..673e4a778ac8 100644
--- a/arch/powerpc/boot/dts/tqm8548.dts
+++ b/arch/powerpc/boot/dts/tqm8548.dts
@@ -85,9 +85,9 @@
85 interrupt-parent = <&mpic>; 85 interrupt-parent = <&mpic>;
86 dfsrr; 86 dfsrr;
87 87
88 dtt@50 { 88 dtt@48 {
89 compatible = "national,lm75"; 89 compatible = "national,lm75";
90 reg = <0x50>; 90 reg = <0x48>;
91 }; 91 };
92 92
93 rtc@68 { 93 rtc@68 {
@@ -247,7 +247,7 @@
247 interrupts = <31 2 32 2 33 2>; 247 interrupts = <31 2 32 2 33 2>;
248 interrupt-parent = <&mpic>; 248 interrupt-parent = <&mpic>;
249 tbi-handle = <&tbi2>; 249 tbi-handle = <&tbi2>;
250 phy-handle = <&phy3>; 250 phy-handle = <&phy4>;
251 251
252 mdio@520 { 252 mdio@520 {
253 #address-cells = <1>; 253 #address-cells = <1>;
@@ -275,7 +275,7 @@
275 interrupts = <37 2 38 2 39 2>; 275 interrupts = <37 2 38 2 39 2>;
276 interrupt-parent = <&mpic>; 276 interrupt-parent = <&mpic>;
277 tbi-handle = <&tbi3>; 277 tbi-handle = <&tbi3>;
278 phy-handle = <&phy4>; 278 phy-handle = <&phy5>;
279 279
280 mdio@520 { 280 mdio@520 {
281 #address-cells = <1>; 281 #address-cells = <1>;
diff --git a/arch/powerpc/boot/dts/tqm8555.dts b/arch/powerpc/boot/dts/tqm8555.dts
index 06d366ebbda3..6a99f1eef7ad 100644
--- a/arch/powerpc/boot/dts/tqm8555.dts
+++ b/arch/powerpc/boot/dts/tqm8555.dts
@@ -83,9 +83,9 @@
83 interrupt-parent = <&mpic>; 83 interrupt-parent = <&mpic>;
84 dfsrr; 84 dfsrr;
85 85
86 dtt@50 { 86 dtt@48 {
87 compatible = "national,lm75"; 87 compatible = "national,lm75";
88 reg = <0x50>; 88 reg = <0x48>;
89 }; 89 };
90 90
91 rtc@68 { 91 rtc@68 {
diff --git a/arch/powerpc/boot/dts/tqm8560.dts b/arch/powerpc/boot/dts/tqm8560.dts
index feff915e0492..b6c2d71defd3 100644
--- a/arch/powerpc/boot/dts/tqm8560.dts
+++ b/arch/powerpc/boot/dts/tqm8560.dts
@@ -85,9 +85,9 @@
85 interrupt-parent = <&mpic>; 85 interrupt-parent = <&mpic>;
86 dfsrr; 86 dfsrr;
87 87
88 dtt@50 { 88 dtt@48 {
89 compatible = "national,lm75"; 89 compatible = "national,lm75";
90 reg = <0x50>; 90 reg = <0x48>;
91 }; 91 };
92 92
93 rtc@68 { 93 rtc@68 {
diff --git a/arch/powerpc/configs/85xx/tqm8548_defconfig b/arch/powerpc/configs/85xx/tqm8548_defconfig
index 0bc45975911a..43030fea2eee 100644
--- a/arch/powerpc/configs/85xx/tqm8548_defconfig
+++ b/arch/powerpc/configs/85xx/tqm8548_defconfig
@@ -1,7 +1,7 @@
1# 1#
2# Automatically generated make config: don't edit 2# Automatically generated make config: don't edit
3# Linux kernel version: 2.6.29-rc2 3# Linux kernel version: 2.6.29-rc7
4# Mon Jan 26 15:36:20 2009 4# Mon Mar 16 09:03:28 2009
5# 5#
6# CONFIG_PPC64 is not set 6# CONFIG_PPC64 is not set
7 7
@@ -22,6 +22,7 @@ CONFIG_FSL_EMB_PERFMON=y
22# CONFIG_PHYS_64BIT is not set 22# CONFIG_PHYS_64BIT is not set
23CONFIG_SPE=y 23CONFIG_SPE=y
24CONFIG_PPC_MMU_NOHASH=y 24CONFIG_PPC_MMU_NOHASH=y
25CONFIG_PPC_BOOK3E_MMU=y
25# CONFIG_PPC_MM_SLICES is not set 26# CONFIG_PPC_MM_SLICES is not set
26# CONFIG_SMP is not set 27# CONFIG_SMP is not set
27CONFIG_PPC32=y 28CONFIG_PPC32=y
@@ -75,6 +76,15 @@ CONFIG_SYSVIPC_SYSCTL=y
75# CONFIG_BSD_PROCESS_ACCT is not set 76# CONFIG_BSD_PROCESS_ACCT is not set
76# CONFIG_TASKSTATS is not set 77# CONFIG_TASKSTATS is not set
77# CONFIG_AUDIT is not set 78# CONFIG_AUDIT is not set
79
80#
81# RCU Subsystem
82#
83CONFIG_CLASSIC_RCU=y
84# CONFIG_TREE_RCU is not set
85# CONFIG_PREEMPT_RCU is not set
86# CONFIG_TREE_RCU_TRACE is not set
87# CONFIG_PREEMPT_RCU_TRACE is not set
78# CONFIG_IKCONFIG is not set 88# CONFIG_IKCONFIG is not set
79CONFIG_LOG_BUF_SHIFT=14 89CONFIG_LOG_BUF_SHIFT=14
80CONFIG_GROUP_SCHED=y 90CONFIG_GROUP_SCHED=y
@@ -152,11 +162,6 @@ CONFIG_DEFAULT_AS=y
152# CONFIG_DEFAULT_CFQ is not set 162# CONFIG_DEFAULT_CFQ is not set
153# CONFIG_DEFAULT_NOOP is not set 163# CONFIG_DEFAULT_NOOP is not set
154CONFIG_DEFAULT_IOSCHED="anticipatory" 164CONFIG_DEFAULT_IOSCHED="anticipatory"
155CONFIG_CLASSIC_RCU=y
156# CONFIG_TREE_RCU is not set
157# CONFIG_PREEMPT_RCU is not set
158# CONFIG_TREE_RCU_TRACE is not set
159# CONFIG_PREEMPT_RCU_TRACE is not set
160# CONFIG_FREEZER is not set 165# CONFIG_FREEZER is not set
161 166
162# 167#
@@ -202,7 +207,7 @@ CONFIG_MPIC=y
202# 207#
203# Kernel options 208# Kernel options
204# 209#
205# CONFIG_HIGHMEM is not set 210CONFIG_HIGHMEM=y
206CONFIG_TICK_ONESHOT=y 211CONFIG_TICK_ONESHOT=y
207CONFIG_NO_HZ=y 212CONFIG_NO_HZ=y
208CONFIG_HIGH_RES_TIMERS=y 213CONFIG_HIGH_RES_TIMERS=y
@@ -244,6 +249,7 @@ CONFIG_UNEVICTABLE_LRU=y
244CONFIG_PPC_4K_PAGES=y 249CONFIG_PPC_4K_PAGES=y
245# CONFIG_PPC_16K_PAGES is not set 250# CONFIG_PPC_16K_PAGES is not set
246# CONFIG_PPC_64K_PAGES is not set 251# CONFIG_PPC_64K_PAGES is not set
252# CONFIG_PPC_256K_PAGES is not set
247CONFIG_FORCE_MAX_ZONEORDER=11 253CONFIG_FORCE_MAX_ZONEORDER=11
248CONFIG_PROC_DEVICETREE=y 254CONFIG_PROC_DEVICETREE=y
249# CONFIG_CMDLINE_BOOL is not set 255# CONFIG_CMDLINE_BOOL is not set
@@ -259,6 +265,7 @@ CONFIG_ZONE_DMA=y
259CONFIG_PPC_INDIRECT_PCI=y 265CONFIG_PPC_INDIRECT_PCI=y
260CONFIG_FSL_SOC=y 266CONFIG_FSL_SOC=y
261CONFIG_FSL_PCI=y 267CONFIG_FSL_PCI=y
268CONFIG_FSL_LBC=y
262CONFIG_PPC_PCI_CHOICE=y 269CONFIG_PPC_PCI_CHOICE=y
263CONFIG_PCI=y 270CONFIG_PCI=y
264CONFIG_PCI_DOMAINS=y 271CONFIG_PCI_DOMAINS=y
@@ -284,10 +291,11 @@ CONFIG_ARCH_SUPPORTS_MSI=y
284# Default settings for advanced configuration options are used 291# Default settings for advanced configuration options are used
285# 292#
286CONFIG_LOWMEM_SIZE=0x30000000 293CONFIG_LOWMEM_SIZE=0x30000000
294CONFIG_LOWMEM_CAM_NUM=3
287CONFIG_PAGE_OFFSET=0xc0000000 295CONFIG_PAGE_OFFSET=0xc0000000
288CONFIG_KERNEL_START=0xc0000000 296CONFIG_KERNEL_START=0xc0000000
289CONFIG_PHYSICAL_START=0x00000000 297CONFIG_PHYSICAL_START=0x00000000
290CONFIG_PHYSICAL_ALIGN=0x10000000 298CONFIG_PHYSICAL_ALIGN=0x04000000
291CONFIG_TASK_SIZE=0xc0000000 299CONFIG_TASK_SIZE=0xc0000000
292CONFIG_NET=y 300CONFIG_NET=y
293 301
@@ -363,12 +371,7 @@ CONFIG_DEFAULT_TCP_CONG="cubic"
363# CONFIG_BT is not set 371# CONFIG_BT is not set
364# CONFIG_AF_RXRPC is not set 372# CONFIG_AF_RXRPC is not set
365# CONFIG_PHONET is not set 373# CONFIG_PHONET is not set
366CONFIG_WIRELESS=y 374# CONFIG_WIRELESS is not set
367# CONFIG_CFG80211 is not set
368CONFIG_WIRELESS_OLD_REGULATORY=y
369# CONFIG_WIRELESS_EXT is not set
370# CONFIG_LIB80211 is not set
371# CONFIG_MAC80211 is not set
372# CONFIG_WIMAX is not set 375# CONFIG_WIMAX is not set
373# CONFIG_RFKILL is not set 376# CONFIG_RFKILL is not set
374# CONFIG_NET_9P is not set 377# CONFIG_NET_9P is not set
@@ -471,27 +474,18 @@ CONFIG_MTD_NAND_IDS=y
471# CONFIG_MTD_NAND_NANDSIM is not set 474# CONFIG_MTD_NAND_NANDSIM is not set
472# CONFIG_MTD_NAND_PLATFORM is not set 475# CONFIG_MTD_NAND_PLATFORM is not set
473# CONFIG_MTD_NAND_FSL_ELBC is not set 476# CONFIG_MTD_NAND_FSL_ELBC is not set
474# CONFIG_MTD_NAND_FSL_UPM is not set 477CONFIG_MTD_NAND_FSL_UPM=y
475# CONFIG_MTD_ONENAND is not set 478# CONFIG_MTD_ONENAND is not set
476 479
477# 480#
478# LPDDR flash memory drivers 481# LPDDR flash memory drivers
479# 482#
480# CONFIG_MTD_LPDDR is not set 483# CONFIG_MTD_LPDDR is not set
481# CONFIG_MTD_QINFO_PROBE is not set
482 484
483# 485#
484# UBI - Unsorted block images 486# UBI - Unsorted block images
485# 487#
486CONFIG_MTD_UBI=m 488# CONFIG_MTD_UBI is not set
487CONFIG_MTD_UBI_WL_THRESHOLD=4096
488CONFIG_MTD_UBI_BEB_RESERVE=1
489# CONFIG_MTD_UBI_GLUEBI is not set
490
491#
492# UBI debugging options
493#
494# CONFIG_MTD_UBI_DEBUG is not set
495CONFIG_OF_DEVICE=y 489CONFIG_OF_DEVICE=y
496CONFIG_OF_I2C=y 490CONFIG_OF_I2C=y
497# CONFIG_PARPORT is not set 491# CONFIG_PARPORT is not set
@@ -515,69 +509,21 @@ CONFIG_BLK_DEV_RAM_SIZE=32768
515# CONFIG_BLK_DEV_HD is not set 509# CONFIG_BLK_DEV_HD is not set
516CONFIG_MISC_DEVICES=y 510CONFIG_MISC_DEVICES=y
517# CONFIG_PHANTOM is not set 511# CONFIG_PHANTOM is not set
518# CONFIG_EEPROM_93CX6 is not set
519# CONFIG_SGI_IOC4 is not set 512# CONFIG_SGI_IOC4 is not set
520# CONFIG_TIFM_CORE is not set 513# CONFIG_TIFM_CORE is not set
521# CONFIG_ICS932S401 is not set 514# CONFIG_ICS932S401 is not set
522# CONFIG_ENCLOSURE_SERVICES is not set 515# CONFIG_ENCLOSURE_SERVICES is not set
523# CONFIG_HP_ILO is not set 516# CONFIG_HP_ILO is not set
524# CONFIG_C2PORT is not set 517# CONFIG_C2PORT is not set
518
519#
520# EEPROM support
521#
522# CONFIG_EEPROM_AT24 is not set
523# CONFIG_EEPROM_LEGACY is not set
524# CONFIG_EEPROM_93CX6 is not set
525CONFIG_HAVE_IDE=y 525CONFIG_HAVE_IDE=y
526CONFIG_IDE=y 526# CONFIG_IDE is not set
527
528#
529# Please see Documentation/ide/ide.txt for help/info on IDE drives
530#
531CONFIG_IDE_TIMINGS=y
532# CONFIG_BLK_DEV_IDE_SATA is not set
533CONFIG_IDE_GD=y
534CONFIG_IDE_GD_ATA=y
535# CONFIG_IDE_GD_ATAPI is not set
536# CONFIG_BLK_DEV_IDECD is not set
537# CONFIG_BLK_DEV_IDETAPE is not set
538# CONFIG_IDE_TASK_IOCTL is not set
539CONFIG_IDE_PROC_FS=y
540
541#
542# IDE chipset support/bugfixes
543#
544# CONFIG_BLK_DEV_PLATFORM is not set
545CONFIG_BLK_DEV_IDEDMA_SFF=y
546
547#
548# PCI IDE chipsets support
549#
550CONFIG_BLK_DEV_IDEPCI=y
551CONFIG_IDEPCI_PCIBUS_ORDER=y
552# CONFIG_BLK_DEV_OFFBOARD is not set
553CONFIG_BLK_DEV_GENERIC=y
554# CONFIG_BLK_DEV_OPTI621 is not set
555CONFIG_BLK_DEV_IDEDMA_PCI=y
556# CONFIG_BLK_DEV_AEC62XX is not set
557# CONFIG_BLK_DEV_ALI15X3 is not set
558# CONFIG_BLK_DEV_AMD74XX is not set
559# CONFIG_BLK_DEV_CMD64X is not set
560# CONFIG_BLK_DEV_TRIFLEX is not set
561# CONFIG_BLK_DEV_CS5520 is not set
562# CONFIG_BLK_DEV_CS5530 is not set
563# CONFIG_BLK_DEV_HPT366 is not set
564# CONFIG_BLK_DEV_JMICRON is not set
565# CONFIG_BLK_DEV_SC1200 is not set
566# CONFIG_BLK_DEV_PIIX is not set
567# CONFIG_BLK_DEV_IT8172 is not set
568# CONFIG_BLK_DEV_IT8213 is not set
569# CONFIG_BLK_DEV_IT821X is not set
570# CONFIG_BLK_DEV_NS87415 is not set
571# CONFIG_BLK_DEV_PDC202XX_OLD is not set
572# CONFIG_BLK_DEV_PDC202XX_NEW is not set
573# CONFIG_BLK_DEV_SVWKS is not set
574# CONFIG_BLK_DEV_SIIMAGE is not set
575# CONFIG_BLK_DEV_SL82C105 is not set
576# CONFIG_BLK_DEV_SLC90E66 is not set
577# CONFIG_BLK_DEV_TRM290 is not set
578CONFIG_BLK_DEV_VIA82CXXX=y
579# CONFIG_BLK_DEV_TC86C001 is not set
580CONFIG_BLK_DEV_IDEDMA=y
581 527
582# 528#
583# SCSI device support 529# SCSI device support
@@ -650,7 +596,7 @@ CONFIG_MII=y
650CONFIG_NETDEV_1000=y 596CONFIG_NETDEV_1000=y
651# CONFIG_ACENIC is not set 597# CONFIG_ACENIC is not set
652# CONFIG_DL2K is not set 598# CONFIG_DL2K is not set
653CONFIG_E1000=y 599# CONFIG_E1000 is not set
654# CONFIG_E1000E is not set 600# CONFIG_E1000E is not set
655# CONFIG_IP1000 is not set 601# CONFIG_IP1000 is not set
656# CONFIG_IGB is not set 602# CONFIG_IGB is not set
@@ -668,6 +614,7 @@ CONFIG_GIANFAR=y
668# CONFIG_QLA3XXX is not set 614# CONFIG_QLA3XXX is not set
669# CONFIG_ATL1 is not set 615# CONFIG_ATL1 is not set
670# CONFIG_ATL1E is not set 616# CONFIG_ATL1E is not set
617# CONFIG_ATL1C is not set
671# CONFIG_JME is not set 618# CONFIG_JME is not set
672CONFIG_NETDEV_10000=y 619CONFIG_NETDEV_10000=y
673# CONFIG_CHELSIO_T1 is not set 620# CONFIG_CHELSIO_T1 is not set
@@ -835,8 +782,6 @@ CONFIG_I2C_MPC=y
835# Miscellaneous I2C Chip support 782# Miscellaneous I2C Chip support
836# 783#
837# CONFIG_DS1682 is not set 784# CONFIG_DS1682 is not set
838# CONFIG_EEPROM_AT24 is not set
839# CONFIG_EEPROM_LEGACY is not set
840# CONFIG_SENSORS_PCF8574 is not set 785# CONFIG_SENSORS_PCF8574 is not set
841# CONFIG_PCF8575 is not set 786# CONFIG_PCF8575 is not set
842# CONFIG_SENSORS_PCA9539 is not set 787# CONFIG_SENSORS_PCA9539 is not set
@@ -975,26 +920,7 @@ CONFIG_HID=y
975# Special HID drivers 920# Special HID drivers
976# 921#
977CONFIG_HID_COMPAT=y 922CONFIG_HID_COMPAT=y
978CONFIG_USB_SUPPORT=y 923# CONFIG_USB_SUPPORT is not set
979CONFIG_USB_ARCH_HAS_HCD=y
980CONFIG_USB_ARCH_HAS_OHCI=y
981CONFIG_USB_ARCH_HAS_EHCI=y
982# CONFIG_USB is not set
983# CONFIG_USB_OTG_WHITELIST is not set
984# CONFIG_USB_OTG_BLACKLIST_HUB is not set
985
986#
987# Enable Host or Gadget support to see Inventra options
988#
989
990#
991# NOTE: USB_STORAGE depends on SCSI but BLK_DEV_SD may also be needed;
992#
993# CONFIG_USB_GADGET is not set
994
995#
996# OTG and related infrastructure
997#
998# CONFIG_UWB is not set 924# CONFIG_UWB is not set
999# CONFIG_MMC is not set 925# CONFIG_MMC is not set
1000# CONFIG_MEMSTICK is not set 926# CONFIG_MEMSTICK is not set
@@ -1064,16 +990,9 @@ CONFIG_RTC_DRV_DS1307=y
1064# 990#
1065# File systems 991# File systems
1066# 992#
1067CONFIG_EXT2_FS=y 993# CONFIG_EXT2_FS is not set
1068# CONFIG_EXT2_FS_XATTR is not set 994# CONFIG_EXT3_FS is not set
1069# CONFIG_EXT2_FS_XIP is not set
1070CONFIG_EXT3_FS=y
1071CONFIG_EXT3_FS_XATTR=y
1072# CONFIG_EXT3_FS_POSIX_ACL is not set
1073# CONFIG_EXT3_FS_SECURITY is not set
1074# CONFIG_EXT4_FS is not set 995# CONFIG_EXT4_FS is not set
1075CONFIG_JBD=y
1076CONFIG_FS_MBCACHE=y
1077# CONFIG_REISERFS_FS is not set 996# CONFIG_REISERFS_FS is not set
1078# CONFIG_JFS_FS is not set 997# CONFIG_JFS_FS is not set
1079# CONFIG_FS_POSIX_ACL is not set 998# CONFIG_FS_POSIX_ACL is not set
@@ -1122,8 +1041,17 @@ CONFIG_MISC_FILESYSTEMS=y
1122# CONFIG_BEFS_FS is not set 1041# CONFIG_BEFS_FS is not set
1123# CONFIG_BFS_FS is not set 1042# CONFIG_BFS_FS is not set
1124# CONFIG_EFS_FS is not set 1043# CONFIG_EFS_FS is not set
1125# CONFIG_JFFS2_FS is not set 1044CONFIG_JFFS2_FS=y
1126# CONFIG_UBIFS_FS is not set 1045CONFIG_JFFS2_FS_DEBUG=0
1046CONFIG_JFFS2_FS_WRITEBUFFER=y
1047# CONFIG_JFFS2_FS_WBUF_VERIFY is not set
1048# CONFIG_JFFS2_SUMMARY is not set
1049# CONFIG_JFFS2_FS_XATTR is not set
1050# CONFIG_JFFS2_COMPRESSION_OPTIONS is not set
1051CONFIG_JFFS2_ZLIB=y
1052# CONFIG_JFFS2_LZO is not set
1053CONFIG_JFFS2_RTIME=y
1054# CONFIG_JFFS2_RUBIN is not set
1127# CONFIG_CRAMFS is not set 1055# CONFIG_CRAMFS is not set
1128# CONFIG_SQUASHFS is not set 1056# CONFIG_SQUASHFS is not set
1129# CONFIG_VXFS_FS is not set 1057# CONFIG_VXFS_FS is not set
@@ -1184,6 +1112,8 @@ CONFIG_GENERIC_FIND_LAST_BIT=y
1184CONFIG_CRC32=y 1112CONFIG_CRC32=y
1185# CONFIG_CRC7 is not set 1113# CONFIG_CRC7 is not set
1186# CONFIG_LIBCRC32C is not set 1114# CONFIG_LIBCRC32C is not set
1115CONFIG_ZLIB_INFLATE=y
1116CONFIG_ZLIB_DEFLATE=y
1187CONFIG_PLIST=y 1117CONFIG_PLIST=y
1188CONFIG_HAS_IOMEM=y 1118CONFIG_HAS_IOMEM=y
1189CONFIG_HAS_IOPORT=y 1119CONFIG_HAS_IOPORT=y
@@ -1219,6 +1149,7 @@ CONFIG_DEBUG_MUTEXES=y
1219# CONFIG_DEBUG_SPINLOCK_SLEEP is not set 1149# CONFIG_DEBUG_SPINLOCK_SLEEP is not set
1220# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set 1150# CONFIG_DEBUG_LOCKING_API_SELFTESTS is not set
1221# CONFIG_DEBUG_KOBJECT is not set 1151# CONFIG_DEBUG_KOBJECT is not set
1152# CONFIG_DEBUG_HIGHMEM is not set
1222# CONFIG_DEBUG_BUGVERBOSE is not set 1153# CONFIG_DEBUG_BUGVERBOSE is not set
1223# CONFIG_DEBUG_INFO is not set 1154# CONFIG_DEBUG_INFO is not set
1224# CONFIG_DEBUG_VM is not set 1155# CONFIG_DEBUG_VM is not set
@@ -1236,6 +1167,7 @@ CONFIG_DEBUG_MUTEXES=y
1236# CONFIG_LATENCYTOP is not set 1167# CONFIG_LATENCYTOP is not set
1237CONFIG_SYSCTL_SYSCALL_CHECK=y 1168CONFIG_SYSCTL_SYSCALL_CHECK=y
1238CONFIG_HAVE_FUNCTION_TRACER=y 1169CONFIG_HAVE_FUNCTION_TRACER=y
1170CONFIG_HAVE_FUNCTION_GRAPH_TRACER=y
1239CONFIG_HAVE_DYNAMIC_FTRACE=y 1171CONFIG_HAVE_DYNAMIC_FTRACE=y
1240CONFIG_HAVE_FTRACE_MCOUNT_RECORD=y 1172CONFIG_HAVE_FTRACE_MCOUNT_RECORD=y
1241 1173
diff --git a/arch/powerpc/include/asm/futex.h b/arch/powerpc/include/asm/futex.h
index 6d406c5c5de4..9696cc36d2dc 100644
--- a/arch/powerpc/include/asm/futex.h
+++ b/arch/powerpc/include/asm/futex.h
@@ -27,7 +27,7 @@
27 PPC_LONG "1b,4b,2b,4b\n" \ 27 PPC_LONG "1b,4b,2b,4b\n" \
28 ".previous" \ 28 ".previous" \
29 : "=&r" (oldval), "=&r" (ret) \ 29 : "=&r" (oldval), "=&r" (ret) \
30 : "b" (uaddr), "i" (-EFAULT), "1" (oparg) \ 30 : "b" (uaddr), "i" (-EFAULT), "r" (oparg) \
31 : "cr0", "memory") 31 : "cr0", "memory")
32 32
33static inline int futex_atomic_op_inuser (int encoded_op, int __user *uaddr) 33static inline int futex_atomic_op_inuser (int encoded_op, int __user *uaddr)
@@ -47,19 +47,19 @@ static inline int futex_atomic_op_inuser (int encoded_op, int __user *uaddr)
47 47
48 switch (op) { 48 switch (op) {
49 case FUTEX_OP_SET: 49 case FUTEX_OP_SET:
50 __futex_atomic_op("", ret, oldval, uaddr, oparg); 50 __futex_atomic_op("mr %1,%4\n", ret, oldval, uaddr, oparg);
51 break; 51 break;
52 case FUTEX_OP_ADD: 52 case FUTEX_OP_ADD:
53 __futex_atomic_op("add %1,%0,%1\n", ret, oldval, uaddr, oparg); 53 __futex_atomic_op("add %1,%0,%4\n", ret, oldval, uaddr, oparg);
54 break; 54 break;
55 case FUTEX_OP_OR: 55 case FUTEX_OP_OR:
56 __futex_atomic_op("or %1,%0,%1\n", ret, oldval, uaddr, oparg); 56 __futex_atomic_op("or %1,%0,%4\n", ret, oldval, uaddr, oparg);
57 break; 57 break;
58 case FUTEX_OP_ANDN: 58 case FUTEX_OP_ANDN:
59 __futex_atomic_op("andc %1,%0,%1\n", ret, oldval, uaddr, oparg); 59 __futex_atomic_op("andc %1,%0,%4\n", ret, oldval, uaddr, oparg);
60 break; 60 break;
61 case FUTEX_OP_XOR: 61 case FUTEX_OP_XOR:
62 __futex_atomic_op("xor %1,%0,%1\n", ret, oldval, uaddr, oparg); 62 __futex_atomic_op("xor %1,%0,%4\n", ret, oldval, uaddr, oparg);
63 break; 63 break;
64 default: 64 default:
65 ret = -ENOSYS; 65 ret = -ENOSYS;
diff --git a/arch/powerpc/include/asm/mmu.h b/arch/powerpc/include/asm/mmu.h
index cbf154387091..86d2366ab6a1 100644
--- a/arch/powerpc/include/asm/mmu.h
+++ b/arch/powerpc/include/asm/mmu.h
@@ -52,6 +52,12 @@
52 */ 52 */
53#define MMU_FTR_NEED_DTLB_SW_LRU ASM_CONST(0x00200000) 53#define MMU_FTR_NEED_DTLB_SW_LRU ASM_CONST(0x00200000)
54 54
55/* This indicates that the processor uses the wrong opcode for tlbilx
56 * instructions. During the ISA 2.06 development the opcode for tlbilx
57 * changed and some early implementations used to old opcode
58 */
59#define MMU_FTR_TLBILX_EARLY_OPCODE ASM_CONST(0x00400000)
60
55#ifndef __ASSEMBLY__ 61#ifndef __ASSEMBLY__
56#include <asm/cputable.h> 62#include <asm/cputable.h>
57 63
diff --git a/arch/powerpc/include/asm/parport.h b/arch/powerpc/include/asm/parport.h
index 414c50e2e881..94942d60ddfd 100644
--- a/arch/powerpc/include/asm/parport.h
+++ b/arch/powerpc/include/asm/parport.h
@@ -29,7 +29,7 @@ static int __devinit parport_pc_find_nonpci_ports (int autoirq, int autodma)
29 prop = of_get_property(np, "interrupts", NULL); 29 prop = of_get_property(np, "interrupts", NULL);
30 if (!prop) 30 if (!prop)
31 continue; 31 continue;
32 if (parport_pc_probe_port(io1, io2, prop[0], autodma, NULL) != NULL) 32 if (parport_pc_probe_port(io1, io2, prop[0], autodma, NULL, 0) != NULL)
33 count++; 33 count++;
34 } 34 }
35 return count; 35 return count;
diff --git a/arch/powerpc/include/asm/ppc-opcode.h b/arch/powerpc/include/asm/ppc-opcode.h
index f4a4db8d5555..ef4da37f3c10 100644
--- a/arch/powerpc/include/asm/ppc-opcode.h
+++ b/arch/powerpc/include/asm/ppc-opcode.h
@@ -43,7 +43,8 @@
43 43
44#define PPC_INST_STSWI 0x7c0005aa 44#define PPC_INST_STSWI 0x7c0005aa
45#define PPC_INST_STSWX 0x7c00052a 45#define PPC_INST_STSWX 0x7c00052a
46#define PPC_INST_TLBILX 0x7c000626 46#define PPC_INST_TLBILX 0x7c000024
47#define PPC_INST_TLBILX_EARLY 0x7c000626
47#define PPC_INST_WAIT 0x7c00007c 48#define PPC_INST_WAIT 0x7c00007c
48 49
49/* macros to insert fields into opcodes */ 50/* macros to insert fields into opcodes */
@@ -63,10 +64,18 @@
63#define PPC_RFDI stringify_in_c(.long PPC_INST_RFDI) 64#define PPC_RFDI stringify_in_c(.long PPC_INST_RFDI)
64#define PPC_RFMCI stringify_in_c(.long PPC_INST_RFMCI) 65#define PPC_RFMCI stringify_in_c(.long PPC_INST_RFMCI)
65#define PPC_TLBILX(t, a, b) stringify_in_c(.long PPC_INST_TLBILX | \ 66#define PPC_TLBILX(t, a, b) stringify_in_c(.long PPC_INST_TLBILX | \
66 __PPC_T_TLB(t) | __PPC_RA(a) | __PPC_RB(b)) 67 __PPC_T_TLB(t) | \
68 __PPC_RA(a) | __PPC_RB(b))
67#define PPC_TLBILX_ALL(a, b) PPC_TLBILX(0, a, b) 69#define PPC_TLBILX_ALL(a, b) PPC_TLBILX(0, a, b)
68#define PPC_TLBILX_PID(a, b) PPC_TLBILX(1, a, b) 70#define PPC_TLBILX_PID(a, b) PPC_TLBILX(1, a, b)
69#define PPC_TLBILX_VA(a, b) PPC_TLBILX(3, a, b) 71#define PPC_TLBILX_VA(a, b) PPC_TLBILX(3, a, b)
72
73#define PPC_TLBILX_EARLY(t, a, b) stringify_in_c(.long PPC_INST_TLBILX_EARLY | \
74 __PPC_T_TLB(t) | \
75 __PPC_RA(a) | __PPC_RB(b))
76#define PPC_TLBILX_ALL_EARLY(a, b) PPC_TLBILX_EARLY(0, a, b)
77#define PPC_TLBILX_PID_EARLY(a, b) PPC_TLBILX_EARLY(1, a, b)
78#define PPC_TLBILX_VA_EARLY(a, b) PPC_TLBILX_EARLY(3, a, b)
70#define PPC_WAIT(w) stringify_in_c(.long PPC_INST_WAIT | \ 79#define PPC_WAIT(w) stringify_in_c(.long PPC_INST_WAIT | \
71 __PPC_WC(w)) 80 __PPC_WC(w))
72 81
diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c
index cd1b687544f3..57db50f40289 100644
--- a/arch/powerpc/kernel/cputable.c
+++ b/arch/powerpc/kernel/cputable.c
@@ -1766,7 +1766,7 @@ static struct cpu_spec __initdata cpu_specs[] = {
1766 .cpu_features = CPU_FTRS_E500MC, 1766 .cpu_features = CPU_FTRS_E500MC,
1767 .cpu_user_features = COMMON_USER_BOOKE | PPC_FEATURE_HAS_FPU, 1767 .cpu_user_features = COMMON_USER_BOOKE | PPC_FEATURE_HAS_FPU,
1768 .mmu_features = MMU_FTR_TYPE_FSL_E | MMU_FTR_BIG_PHYS | 1768 .mmu_features = MMU_FTR_TYPE_FSL_E | MMU_FTR_BIG_PHYS |
1769 MMU_FTR_USE_TLBILX, 1769 MMU_FTR_USE_TLBILX | MMU_FTR_TLBILX_EARLY_OPCODE,
1770 .icache_bsize = 64, 1770 .icache_bsize = 64,
1771 .dcache_bsize = 64, 1771 .dcache_bsize = 64,
1772 .num_pmcs = 4, 1772 .num_pmcs = 4,
diff --git a/arch/powerpc/mm/tlb_nohash.c b/arch/powerpc/mm/tlb_nohash.c
index 7af72970faed..ad2eb4d34dd4 100644
--- a/arch/powerpc/mm/tlb_nohash.c
+++ b/arch/powerpc/mm/tlb_nohash.c
@@ -125,7 +125,6 @@ static void do_flush_tlb_page_ipi(void *param)
125 125
126void flush_tlb_mm(struct mm_struct *mm) 126void flush_tlb_mm(struct mm_struct *mm)
127{ 127{
128 cpumask_t cpu_mask;
129 unsigned int pid; 128 unsigned int pid;
130 129
131 preempt_disable(); 130 preempt_disable();
diff --git a/arch/powerpc/mm/tlb_nohash_low.S b/arch/powerpc/mm/tlb_nohash_low.S
index 788b87c36f77..45fed3698349 100644
--- a/arch/powerpc/mm/tlb_nohash_low.S
+++ b/arch/powerpc/mm/tlb_nohash_low.S
@@ -138,7 +138,11 @@ BEGIN_MMU_FTR_SECTION
138 andi. r3,r3,MMUCSR0_TLBFI@l 138 andi. r3,r3,MMUCSR0_TLBFI@l
139 bne 1b 139 bne 1b
140MMU_FTR_SECTION_ELSE 140MMU_FTR_SECTION_ELSE
141 PPC_TLBILX_ALL(0,0) 141 BEGIN_MMU_FTR_SECTION_NESTED(96)
142 PPC_TLBILX_ALL(0,r3)
143 MMU_FTR_SECTION_ELSE_NESTED(96)
144 PPC_TLBILX_ALL_EARLY(0,r3)
145 ALT_MMU_FTR_SECTION_END_NESTED_IFCLR(MMU_FTR_TLBILX_EARLY_OPCODE, 96)
142ALT_MMU_FTR_SECTION_END_IFCLR(MMU_FTR_USE_TLBILX) 146ALT_MMU_FTR_SECTION_END_IFCLR(MMU_FTR_USE_TLBILX)
143 msync 147 msync
144 isync 148 isync
@@ -151,7 +155,11 @@ BEGIN_MMU_FTR_SECTION
151 wrteei 0 155 wrteei 0
152 mfspr r4,SPRN_MAS6 /* save MAS6 */ 156 mfspr r4,SPRN_MAS6 /* save MAS6 */
153 mtspr SPRN_MAS6,r3 157 mtspr SPRN_MAS6,r3
158 BEGIN_MMU_FTR_SECTION_NESTED(96)
154 PPC_TLBILX_PID(0,0) 159 PPC_TLBILX_PID(0,0)
160 MMU_FTR_SECTION_ELSE_NESTED(96)
161 PPC_TLBILX_PID_EARLY(0,0)
162 ALT_MMU_FTR_SECTION_END_NESTED_IFCLR(MMU_FTR_TLBILX_EARLY_OPCODE, 96)
155 mtspr SPRN_MAS6,r4 /* restore MAS6 */ 163 mtspr SPRN_MAS6,r4 /* restore MAS6 */
156 wrtee r10 164 wrtee r10
157MMU_FTR_SECTION_ELSE 165MMU_FTR_SECTION_ELSE
@@ -185,7 +193,11 @@ BEGIN_MMU_FTR_SECTION
185 mtspr SPRN_MAS1,r4 193 mtspr SPRN_MAS1,r4
186 tlbwe 194 tlbwe
187MMU_FTR_SECTION_ELSE 195MMU_FTR_SECTION_ELSE
196 BEGIN_MMU_FTR_SECTION_NESTED(96)
188 PPC_TLBILX_VA(0,r3) 197 PPC_TLBILX_VA(0,r3)
198 MMU_FTR_SECTION_ELSE_NESTED(96)
199 PPC_TLBILX_VA_EARLY(0,r3)
200 ALT_MMU_FTR_SECTION_END_NESTED_IFCLR(MMU_FTR_TLBILX_EARLY_OPCODE, 96)
189ALT_MMU_FTR_SECTION_END_IFCLR(MMU_FTR_USE_TLBILX) 201ALT_MMU_FTR_SECTION_END_IFCLR(MMU_FTR_USE_TLBILX)
190 msync 202 msync
191 isync 203 isync
diff --git a/arch/powerpc/platforms/pseries/dtl.c b/arch/powerpc/platforms/pseries/dtl.c
index fafcaa0e81ef..ab69925d579b 100644
--- a/arch/powerpc/platforms/pseries/dtl.c
+++ b/arch/powerpc/platforms/pseries/dtl.c
@@ -25,6 +25,7 @@
25#include <asm/smp.h> 25#include <asm/smp.h>
26#include <asm/system.h> 26#include <asm/system.h>
27#include <asm/uaccess.h> 27#include <asm/uaccess.h>
28#include <asm/firmware.h>
28 29
29#include "plpar_wrappers.h" 30#include "plpar_wrappers.h"
30 31
diff --git a/arch/powerpc/platforms/pseries/eeh_driver.c b/arch/powerpc/platforms/pseries/eeh_driver.c
index 380420f8c400..9a2a6e32f00f 100644
--- a/arch/powerpc/platforms/pseries/eeh_driver.c
+++ b/arch/powerpc/platforms/pseries/eeh_driver.c
@@ -182,6 +182,8 @@ static void eeh_report_reset(struct pci_dev *dev, void *userdata)
182 if (!driver) 182 if (!driver)
183 return; 183 return;
184 184
185 dev->error_state = pci_channel_io_normal;
186
185 eeh_enable_irq(dev); 187 eeh_enable_irq(dev);
186 188
187 if (!driver->err_handler || 189 if (!driver->err_handler ||
diff --git a/arch/s390/appldata/appldata_base.c b/arch/s390/appldata/appldata_base.c
index 27b70d8a359c..aeb3cff95f63 100644
--- a/arch/s390/appldata/appldata_base.c
+++ b/arch/s390/appldata/appldata_base.c
@@ -176,7 +176,7 @@ static void __appldata_mod_vtimer_wrap(void *p) {
176 struct vtimer_list *timer; 176 struct vtimer_list *timer;
177 u64 expires; 177 u64 expires;
178 } *args = p; 178 } *args = p;
179 mod_virt_timer(args->timer, args->expires); 179 mod_virt_timer_periodic(args->timer, args->expires);
180} 180}
181 181
182#define APPLDATA_ADD_TIMER 0 182#define APPLDATA_ADD_TIMER 0
diff --git a/arch/s390/include/asm/cpuid.h b/arch/s390/include/asm/cpuid.h
new file mode 100644
index 000000000000..07836a2e5222
--- /dev/null
+++ b/arch/s390/include/asm/cpuid.h
@@ -0,0 +1,25 @@
1/*
2 * Copyright IBM Corp. 2000,2009
3 * Author(s): Hartmut Penner <hp@de.ibm.com>,
4 * Martin Schwidefsky <schwidefsky@de.ibm.com>
5 * Christian Ehrhardt <ehrhardt@de.ibm.com>
6 */
7
8#ifndef _ASM_S390_CPUID_H_
9#define _ASM_S390_CPUID_H_
10
11/*
12 * CPU type and hardware bug flags. Kept separately for each CPU.
13 * Members of this structure are referenced in head.S, so think twice
14 * before touching them. [mj]
15 */
16
17typedef struct
18{
19 unsigned int version : 8;
20 unsigned int ident : 24;
21 unsigned int machine : 16;
22 unsigned int unused : 16;
23} __attribute__ ((packed)) cpuid_t;
24
25#endif /* _ASM_S390_CPUID_H_ */
diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h
index c6e674f5fca9..54ea39f96ecd 100644
--- a/arch/s390/include/asm/kvm_host.h
+++ b/arch/s390/include/asm/kvm_host.h
@@ -15,6 +15,7 @@
15#define ASM_KVM_HOST_H 15#define ASM_KVM_HOST_H
16#include <linux/kvm_host.h> 16#include <linux/kvm_host.h>
17#include <asm/debug.h> 17#include <asm/debug.h>
18#include <asm/cpuid.h>
18 19
19#define KVM_MAX_VCPUS 64 20#define KVM_MAX_VCPUS 64
20#define KVM_MEMORY_SLOTS 32 21#define KVM_MEMORY_SLOTS 32
diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h
index b349f1c7fdfa..3aeca492b147 100644
--- a/arch/s390/include/asm/lowcore.h
+++ b/arch/s390/include/asm/lowcore.h
@@ -66,6 +66,7 @@
66#define __LC_USER_EXEC_ASCE 0x02ac 66#define __LC_USER_EXEC_ASCE 0x02ac
67#define __LC_CPUID 0x02b0 67#define __LC_CPUID 0x02b0
68#define __LC_INT_CLOCK 0x02c8 68#define __LC_INT_CLOCK 0x02c8
69#define __LC_MACHINE_FLAGS 0x02d8
69#define __LC_IRB 0x0300 70#define __LC_IRB 0x0300
70#define __LC_PFAULT_INTPARM 0x0080 71#define __LC_PFAULT_INTPARM 0x0080
71#define __LC_CPU_TIMER_SAVE_AREA 0x00d8 72#define __LC_CPU_TIMER_SAVE_AREA 0x00d8
@@ -110,6 +111,7 @@
110#define __LC_CPUID 0x0320 111#define __LC_CPUID 0x0320
111#define __LC_INT_CLOCK 0x0340 112#define __LC_INT_CLOCK 0x0340
112#define __LC_VDSO_PER_CPU 0x0350 113#define __LC_VDSO_PER_CPU 0x0350
114#define __LC_MACHINE_FLAGS 0x0358
113#define __LC_IRB 0x0380 115#define __LC_IRB 0x0380
114#define __LC_PASTE 0x03c0 116#define __LC_PASTE 0x03c0
115#define __LC_PFAULT_INTPARM 0x11b8 117#define __LC_PFAULT_INTPARM 0x11b8
@@ -127,9 +129,9 @@
127 129
128#ifndef __ASSEMBLY__ 130#ifndef __ASSEMBLY__
129 131
130#include <asm/processor.h> 132#include <asm/cpuid.h>
133#include <asm/ptrace.h>
131#include <linux/types.h> 134#include <linux/types.h>
132#include <asm/sigp.h>
133 135
134void restart_int_handler(void); 136void restart_int_handler(void);
135void ext_int_handler(void); 137void ext_int_handler(void);
@@ -277,7 +279,8 @@ struct _lowcore
277 __u32 ext_call_fast; /* 0x02c4 */ 279 __u32 ext_call_fast; /* 0x02c4 */
278 __u64 int_clock; /* 0x02c8 */ 280 __u64 int_clock; /* 0x02c8 */
279 __u64 clock_comparator; /* 0x02d0 */ 281 __u64 clock_comparator; /* 0x02d0 */
280 __u8 pad_0x02d8[0x0300-0x02d8]; /* 0x02d8 */ 282 __u32 machine_flags; /* 0x02d8 */
283 __u8 pad_0x02dc[0x0300-0x02dc]; /* 0x02dc */
281 284
282 /* Interrupt response block */ 285 /* Interrupt response block */
283 __u8 irb[64]; /* 0x0300 */ 286 __u8 irb[64]; /* 0x0300 */
@@ -381,7 +384,8 @@ struct _lowcore
381 __u64 int_clock; /* 0x0340 */ 384 __u64 int_clock; /* 0x0340 */
382 __u64 clock_comparator; /* 0x0348 */ 385 __u64 clock_comparator; /* 0x0348 */
383 __u64 vdso_per_cpu_data; /* 0x0350 */ 386 __u64 vdso_per_cpu_data; /* 0x0350 */
384 __u8 pad_0x0358[0x0380-0x0358]; /* 0x0358 */ 387 __u64 machine_flags; /* 0x0358 */
388 __u8 pad_0x0360[0x0380-0x0360]; /* 0x0360 */
385 389
386 /* Interrupt response block. */ 390 /* Interrupt response block. */
387 __u8 irb[64]; /* 0x0380 */ 391 __u8 irb[64]; /* 0x0380 */
diff --git a/arch/s390/include/asm/processor.h b/arch/s390/include/asm/processor.h
index 61862b3ac794..c139fa7b8e89 100644
--- a/arch/s390/include/asm/processor.h
+++ b/arch/s390/include/asm/processor.h
@@ -14,7 +14,10 @@
14#define __ASM_S390_PROCESSOR_H 14#define __ASM_S390_PROCESSOR_H
15 15
16#include <linux/linkage.h> 16#include <linux/linkage.h>
17#include <asm/cpuid.h>
18#include <asm/page.h>
17#include <asm/ptrace.h> 19#include <asm/ptrace.h>
20#include <asm/setup.h>
18 21
19#ifdef __KERNEL__ 22#ifdef __KERNEL__
20/* 23/*
@@ -23,20 +26,6 @@
23 */ 26 */
24#define current_text_addr() ({ void *pc; asm("basr %0,0" : "=a" (pc)); pc; }) 27#define current_text_addr() ({ void *pc; asm("basr %0,0" : "=a" (pc)); pc; })
25 28
26/*
27 * CPU type and hardware bug flags. Kept separately for each CPU.
28 * Members of this structure are referenced in head.S, so think twice
29 * before touching them. [mj]
30 */
31
32typedef struct
33{
34 unsigned int version : 8;
35 unsigned int ident : 24;
36 unsigned int machine : 16;
37 unsigned int unused : 16;
38} __attribute__ ((packed)) cpuid_t;
39
40static inline void get_cpu_id(cpuid_t *ptr) 29static inline void get_cpu_id(cpuid_t *ptr)
41{ 30{
42 asm volatile("stidp 0(%1)" : "=m" (*ptr) : "a" (ptr)); 31 asm volatile("stidp 0(%1)" : "=m" (*ptr) : "a" (ptr));
diff --git a/arch/s390/include/asm/ptrace.h b/arch/s390/include/asm/ptrace.h
index f1b051630c50..539263fc9ab9 100644
--- a/arch/s390/include/asm/ptrace.h
+++ b/arch/s390/include/asm/ptrace.h
@@ -313,8 +313,6 @@ typedef struct
313 313
314 314
315#ifdef __KERNEL__ 315#ifdef __KERNEL__
316#include <asm/setup.h>
317#include <asm/page.h>
318 316
319/* 317/*
320 * The pt_regs struct defines the way the registers are stored on 318 * The pt_regs struct defines the way the registers are stored on
diff --git a/arch/s390/include/asm/setup.h b/arch/s390/include/asm/setup.h
index e8bd6ac22c99..38b0fc221ed7 100644
--- a/arch/s390/include/asm/setup.h
+++ b/arch/s390/include/asm/setup.h
@@ -14,6 +14,7 @@
14 14
15#ifdef __KERNEL__ 15#ifdef __KERNEL__
16 16
17#include <asm/lowcore.h>
17#include <asm/types.h> 18#include <asm/types.h>
18 19
19#define PARMAREA 0x10400 20#define PARMAREA 0x10400
@@ -63,7 +64,6 @@ extern unsigned int s390_noexec;
63/* 64/*
64 * Machine features detected in head.S 65 * Machine features detected in head.S
65 */ 66 */
66extern unsigned long machine_flags;
67 67
68#define MACHINE_FLAG_VM (1UL << 0) 68#define MACHINE_FLAG_VM (1UL << 0)
69#define MACHINE_FLAG_IEEE (1UL << 1) 69#define MACHINE_FLAG_IEEE (1UL << 1)
@@ -77,28 +77,28 @@ extern unsigned long machine_flags;
77#define MACHINE_FLAG_HPAGE (1UL << 10) 77#define MACHINE_FLAG_HPAGE (1UL << 10)
78#define MACHINE_FLAG_PFMF (1UL << 11) 78#define MACHINE_FLAG_PFMF (1UL << 11)
79 79
80#define MACHINE_IS_VM (machine_flags & MACHINE_FLAG_VM) 80#define MACHINE_IS_VM (S390_lowcore.machine_flags & MACHINE_FLAG_VM)
81#define MACHINE_IS_KVM (machine_flags & MACHINE_FLAG_KVM) 81#define MACHINE_IS_KVM (S390_lowcore.machine_flags & MACHINE_FLAG_KVM)
82#define MACHINE_HAS_DIAG9C (machine_flags & MACHINE_FLAG_DIAG9C) 82#define MACHINE_HAS_DIAG9C (S390_lowcore.machine_flags & MACHINE_FLAG_DIAG9C)
83 83
84#ifndef __s390x__ 84#ifndef __s390x__
85#define MACHINE_HAS_IEEE (machine_flags & MACHINE_FLAG_IEEE) 85#define MACHINE_HAS_IEEE (S390_lowcore.machine_flags & MACHINE_FLAG_IEEE)
86#define MACHINE_HAS_CSP (machine_flags & MACHINE_FLAG_CSP) 86#define MACHINE_HAS_CSP (S390_lowcore.machine_flags & MACHINE_FLAG_CSP)
87#define MACHINE_HAS_IDTE (0) 87#define MACHINE_HAS_IDTE (0)
88#define MACHINE_HAS_DIAG44 (1) 88#define MACHINE_HAS_DIAG44 (1)
89#define MACHINE_HAS_MVPG (machine_flags & MACHINE_FLAG_MVPG) 89#define MACHINE_HAS_MVPG (S390_lowcore.machine_flags & MACHINE_FLAG_MVPG)
90#define MACHINE_HAS_MVCOS (0) 90#define MACHINE_HAS_MVCOS (0)
91#define MACHINE_HAS_HPAGE (0) 91#define MACHINE_HAS_HPAGE (0)
92#define MACHINE_HAS_PFMF (0) 92#define MACHINE_HAS_PFMF (0)
93#else /* __s390x__ */ 93#else /* __s390x__ */
94#define MACHINE_HAS_IEEE (1) 94#define MACHINE_HAS_IEEE (1)
95#define MACHINE_HAS_CSP (1) 95#define MACHINE_HAS_CSP (1)
96#define MACHINE_HAS_IDTE (machine_flags & MACHINE_FLAG_IDTE) 96#define MACHINE_HAS_IDTE (S390_lowcore.machine_flags & MACHINE_FLAG_IDTE)
97#define MACHINE_HAS_DIAG44 (machine_flags & MACHINE_FLAG_DIAG44) 97#define MACHINE_HAS_DIAG44 (S390_lowcore.machine_flags & MACHINE_FLAG_DIAG44)
98#define MACHINE_HAS_MVPG (1) 98#define MACHINE_HAS_MVPG (1)
99#define MACHINE_HAS_MVCOS (machine_flags & MACHINE_FLAG_MVCOS) 99#define MACHINE_HAS_MVCOS (S390_lowcore.machine_flags & MACHINE_FLAG_MVCOS)
100#define MACHINE_HAS_HPAGE (machine_flags & MACHINE_FLAG_HPAGE) 100#define MACHINE_HAS_HPAGE (S390_lowcore.machine_flags & MACHINE_FLAG_HPAGE)
101#define MACHINE_HAS_PFMF (machine_flags & MACHINE_FLAG_PFMF) 101#define MACHINE_HAS_PFMF (S390_lowcore.machine_flags & MACHINE_FLAG_PFMF)
102#endif /* __s390x__ */ 102#endif /* __s390x__ */
103 103
104#define ZFCPDUMP_HSA_SIZE (32UL<<20) 104#define ZFCPDUMP_HSA_SIZE (32UL<<20)
diff --git a/arch/s390/include/asm/thread_info.h b/arch/s390/include/asm/thread_info.h
index c544aa524535..461f2abd2e6f 100644
--- a/arch/s390/include/asm/thread_info.h
+++ b/arch/s390/include/asm/thread_info.h
@@ -31,8 +31,9 @@
31#define ASYNC_SIZE (PAGE_SIZE << ASYNC_ORDER) 31#define ASYNC_SIZE (PAGE_SIZE << ASYNC_ORDER)
32 32
33#ifndef __ASSEMBLY__ 33#ifndef __ASSEMBLY__
34#include <asm/processor.h>
35#include <asm/lowcore.h> 34#include <asm/lowcore.h>
35#include <asm/page.h>
36#include <asm/processor.h>
36 37
37/* 38/*
38 * low level task data that entry.S needs immediate access to 39 * low level task data that entry.S needs immediate access to
diff --git a/arch/s390/include/asm/timer.h b/arch/s390/include/asm/timer.h
index e4bcab739c19..814243cafdfe 100644
--- a/arch/s390/include/asm/timer.h
+++ b/arch/s390/include/asm/timer.h
@@ -41,6 +41,7 @@ extern void init_virt_timer(struct vtimer_list *timer);
41extern void add_virt_timer(void *new); 41extern void add_virt_timer(void *new);
42extern void add_virt_timer_periodic(void *new); 42extern void add_virt_timer_periodic(void *new);
43extern int mod_virt_timer(struct vtimer_list *timer, __u64 expires); 43extern int mod_virt_timer(struct vtimer_list *timer, __u64 expires);
44extern int mod_virt_timer_periodic(struct vtimer_list *timer, __u64 expires);
44extern int del_virt_timer(struct vtimer_list *timer); 45extern int del_virt_timer(struct vtimer_list *timer);
45 46
46extern void init_cpu_vtimer(void); 47extern void init_cpu_vtimer(void);
diff --git a/arch/s390/include/asm/timex.h b/arch/s390/include/asm/timex.h
index d744c3d62de5..cc21e3e20fd7 100644
--- a/arch/s390/include/asm/timex.h
+++ b/arch/s390/include/asm/timex.h
@@ -11,6 +11,9 @@
11#ifndef _ASM_S390_TIMEX_H 11#ifndef _ASM_S390_TIMEX_H
12#define _ASM_S390_TIMEX_H 12#define _ASM_S390_TIMEX_H
13 13
14/* The value of the TOD clock for 1.1.1970. */
15#define TOD_UNIX_EPOCH 0x7d91048bca000000ULL
16
14/* Inline functions for clock register access. */ 17/* Inline functions for clock register access. */
15static inline int set_clock(__u64 time) 18static inline int set_clock(__u64 time)
16{ 19{
@@ -85,4 +88,6 @@ int get_sync_clock(unsigned long long *clock);
85void init_cpu_timer(void); 88void init_cpu_timer(void);
86unsigned long long monotonic_clock(void); 89unsigned long long monotonic_clock(void);
87 90
91extern u64 sched_clock_base_cc;
92
88#endif 93#endif
diff --git a/arch/s390/include/asm/unistd.h b/arch/s390/include/asm/unistd.h
index c8ad350d1444..f0f19e6ace6c 100644
--- a/arch/s390/include/asm/unistd.h
+++ b/arch/s390/include/asm/unistd.h
@@ -265,7 +265,9 @@
265#define __NR_pipe2 325 265#define __NR_pipe2 325
266#define __NR_dup3 326 266#define __NR_dup3 326
267#define __NR_epoll_create1 327 267#define __NR_epoll_create1 327
268#define NR_syscalls 328 268#define __NR_preadv 328
269#define __NR_pwritev 329
270#define NR_syscalls 330
269 271
270/* 272/*
271 * There are some system calls that are not present on 64 bit, some 273 * There are some system calls that are not present on 64 bit, some
diff --git a/arch/s390/kernel/asm-offsets.c b/arch/s390/kernel/asm-offsets.c
index 67a60016babb..fa9905ce7d0b 100644
--- a/arch/s390/kernel/asm-offsets.c
+++ b/arch/s390/kernel/asm-offsets.c
@@ -27,6 +27,8 @@ int main(void)
27 DEFINE(__TI_flags, offsetof(struct thread_info, flags)); 27 DEFINE(__TI_flags, offsetof(struct thread_info, flags));
28 DEFINE(__TI_cpu, offsetof(struct thread_info, cpu)); 28 DEFINE(__TI_cpu, offsetof(struct thread_info, cpu));
29 DEFINE(__TI_precount, offsetof(struct thread_info, preempt_count)); 29 DEFINE(__TI_precount, offsetof(struct thread_info, preempt_count));
30 DEFINE(__TI_user_timer, offsetof(struct thread_info, user_timer));
31 DEFINE(__TI_system_timer, offsetof(struct thread_info, system_timer));
30 BLANK(); 32 BLANK();
31 DEFINE(__PT_ARGS, offsetof(struct pt_regs, args)); 33 DEFINE(__PT_ARGS, offsetof(struct pt_regs, args));
32 DEFINE(__PT_PSW, offsetof(struct pt_regs, psw)); 34 DEFINE(__PT_PSW, offsetof(struct pt_regs, psw));
diff --git a/arch/s390/kernel/compat_wrapper.S b/arch/s390/kernel/compat_wrapper.S
index 87cf5a79a351..fb38af6316bb 100644
--- a/arch/s390/kernel/compat_wrapper.S
+++ b/arch/s390/kernel/compat_wrapper.S
@@ -1805,3 +1805,21 @@ compat_sys_keyctl_wrapper:
1805 llgfr %r5,%r5 # u32 1805 llgfr %r5,%r5 # u32
1806 llgfr %r6,%r6 # u32 1806 llgfr %r6,%r6 # u32
1807 jg compat_sys_keyctl # branch to system call 1807 jg compat_sys_keyctl # branch to system call
1808
1809 .globl compat_sys_preadv_wrapper
1810compat_sys_preadv_wrapper:
1811 llgfr %r2,%r2 # unsigned long
1812 llgtr %r3,%r3 # compat_iovec *
1813 llgfr %r4,%r4 # unsigned long
1814 llgfr %r5,%r5 # u32
1815 llgfr %r6,%r6 # u32
1816 jg compat_sys_preadv # branch to system call
1817
1818 .globl compat_sys_pwritev_wrapper
1819compat_sys_pwritev_wrapper:
1820 llgfr %r2,%r2 # unsigned long
1821 llgtr %r3,%r3 # compat_iovec *
1822 llgfr %r4,%r4 # unsigned long
1823 llgfr %r5,%r5 # u32
1824 llgfr %r6,%r6 # u32
1825 jg compat_sys_pwritev # branch to system call
diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c
index 4d221c81c849..cf09948faad6 100644
--- a/arch/s390/kernel/early.c
+++ b/arch/s390/kernel/early.c
@@ -34,8 +34,25 @@
34 34
35char kernel_nss_name[NSS_NAME_SIZE + 1]; 35char kernel_nss_name[NSS_NAME_SIZE + 1];
36 36
37static unsigned long machine_flags;
38
37static void __init setup_boot_command_line(void); 39static void __init setup_boot_command_line(void);
38 40
41/*
42 * Get the TOD clock running.
43 */
44static void __init reset_tod_clock(void)
45{
46 u64 time;
47
48 if (store_clock(&time) == 0)
49 return;
50 /* TOD clock not running. Set the clock to Unix Epoch. */
51 if (set_clock(TOD_UNIX_EPOCH) != 0 || store_clock(&time) != 0)
52 disabled_wait(0);
53
54 sched_clock_base_cc = TOD_UNIX_EPOCH;
55}
39 56
40#ifdef CONFIG_SHARED_KERNEL 57#ifdef CONFIG_SHARED_KERNEL
41int __init savesys_ipl_nss(char *cmd, const int cmdlen); 58int __init savesys_ipl_nss(char *cmd, const int cmdlen);
@@ -370,6 +387,7 @@ static void __init setup_boot_command_line(void)
370 */ 387 */
371void __init startup_init(void) 388void __init startup_init(void)
372{ 389{
390 reset_tod_clock();
373 ipl_save_parameters(); 391 ipl_save_parameters();
374 rescue_initrd(); 392 rescue_initrd();
375 clear_bss_section(); 393 clear_bss_section();
@@ -391,5 +409,6 @@ void __init startup_init(void)
391 setup_hpage(); 409 setup_hpage();
392 sclp_facilities_detect(); 410 sclp_facilities_detect();
393 detect_memory_layout(memory_chunk); 411 detect_memory_layout(memory_chunk);
412 S390_lowcore.machine_flags = machine_flags;
394 lockdep_on(); 413 lockdep_on();
395} 414}
diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S
index 1268aa2991bf..f3e275934213 100644
--- a/arch/s390/kernel/entry.S
+++ b/arch/s390/kernel/entry.S
@@ -837,16 +837,29 @@ mcck_return:
837 __CPUINIT 837 __CPUINIT
838 .globl restart_int_handler 838 .globl restart_int_handler
839restart_int_handler: 839restart_int_handler:
840 basr %r1,0
841restart_base:
842 spt restart_vtime-restart_base(%r1)
843 stck __LC_LAST_UPDATE_CLOCK
844 mvc __LC_LAST_UPDATE_TIMER(8),restart_vtime-restart_base(%r1)
845 mvc __LC_EXIT_TIMER(8),restart_vtime-restart_base(%r1)
840 l %r15,__LC_SAVE_AREA+60 # load ksp 846 l %r15,__LC_SAVE_AREA+60 # load ksp
841 lctl %c0,%c15,__LC_CREGS_SAVE_AREA # get new ctl regs 847 lctl %c0,%c15,__LC_CREGS_SAVE_AREA # get new ctl regs
842 lam %a0,%a15,__LC_AREGS_SAVE_AREA 848 lam %a0,%a15,__LC_AREGS_SAVE_AREA
843 lm %r6,%r15,__SF_GPRS(%r15) # load registers from clone 849 lm %r6,%r15,__SF_GPRS(%r15) # load registers from clone
850 l %r1,__LC_THREAD_INFO
851 mvc __LC_USER_TIMER(8),__TI_user_timer(%r1)
852 mvc __LC_SYSTEM_TIMER(8),__TI_system_timer(%r1)
853 xc __LC_STEAL_TIMER(8),__LC_STEAL_TIMER
844 stosm __SF_EMPTY(%r15),0x04 # now we can turn dat on 854 stosm __SF_EMPTY(%r15),0x04 # now we can turn dat on
845 basr %r14,0 855 basr %r14,0
846 l %r14,restart_addr-.(%r14) 856 l %r14,restart_addr-.(%r14)
847 br %r14 # branch to start_secondary 857 br %r14 # branch to start_secondary
848restart_addr: 858restart_addr:
849 .long start_secondary 859 .long start_secondary
860 .align 8
861restart_vtime:
862 .long 0x7fffffff,0xffffffff
850 .previous 863 .previous
851#else 864#else
852/* 865/*
diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S
index c6fbde13971a..84a105838e03 100644
--- a/arch/s390/kernel/entry64.S
+++ b/arch/s390/kernel/entry64.S
@@ -831,14 +831,27 @@ mcck_return:
831 __CPUINIT 831 __CPUINIT
832 .globl restart_int_handler 832 .globl restart_int_handler
833restart_int_handler: 833restart_int_handler:
834 basr %r1,0
835restart_base:
836 spt restart_vtime-restart_base(%r1)
837 stck __LC_LAST_UPDATE_CLOCK
838 mvc __LC_LAST_UPDATE_TIMER(8),restart_vtime-restart_base(%r1)
839 mvc __LC_EXIT_TIMER(8),restart_vtime-restart_base(%r1)
834 lg %r15,__LC_SAVE_AREA+120 # load ksp 840 lg %r15,__LC_SAVE_AREA+120 # load ksp
835 lghi %r10,__LC_CREGS_SAVE_AREA 841 lghi %r10,__LC_CREGS_SAVE_AREA
836 lctlg %c0,%c15,0(%r10) # get new ctl regs 842 lctlg %c0,%c15,0(%r10) # get new ctl regs
837 lghi %r10,__LC_AREGS_SAVE_AREA 843 lghi %r10,__LC_AREGS_SAVE_AREA
838 lam %a0,%a15,0(%r10) 844 lam %a0,%a15,0(%r10)
839 lmg %r6,%r15,__SF_GPRS(%r15) # load registers from clone 845 lmg %r6,%r15,__SF_GPRS(%r15) # load registers from clone
846 lg %r1,__LC_THREAD_INFO
847 mvc __LC_USER_TIMER(8),__TI_user_timer(%r1)
848 mvc __LC_SYSTEM_TIMER(8),__TI_system_timer(%r1)
849 xc __LC_STEAL_TIMER(8),__LC_STEAL_TIMER
840 stosm __SF_EMPTY(%r15),0x04 # now we can turn dat on 850 stosm __SF_EMPTY(%r15),0x04 # now we can turn dat on
841 jg start_secondary 851 jg start_secondary
852 .align 8
853restart_vtime:
854 .long 0x7fffffff,0xffffffff
842 .previous 855 .previous
843#else 856#else
844/* 857/*
diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S
index 1046c2c9f8d1..bba14494ee00 100644
--- a/arch/s390/kernel/head.S
+++ b/arch/s390/kernel/head.S
@@ -471,7 +471,12 @@ startup:basr %r13,0 # get base
471.LPG0: 471.LPG0:
472 xc 0x200(256),0x200 # partially clear lowcore 472 xc 0x200(256),0x200 # partially clear lowcore
473 xc 0x300(256),0x300 473 xc 0x300(256),0x300
474 474 l %r1,5f-.LPG0(%r13)
475 stck 0(%r1)
476 spt 6f-.LPG0(%r13)
477 mvc __LC_LAST_UPDATE_CLOCK(8),0(%r1)
478 mvc __LC_LAST_UPDATE_TIMER(8),6f-.LPG0(%r13)
479 mvc __LC_EXIT_TIMER(8),5f-.LPG0(%r13)
475#ifndef CONFIG_MARCH_G5 480#ifndef CONFIG_MARCH_G5
476 # check processor version against MARCH_{G5,Z900,Z990,Z9_109,Z10} 481 # check processor version against MARCH_{G5,Z900,Z990,Z9_109,Z10}
477 stidp __LC_CPUID # store cpuid 482 stidp __LC_CPUID # store cpuid
@@ -496,9 +501,13 @@ startup:basr %r13,0 # get base
496 brct %r0,0b 501 brct %r0,0b
497#endif 502#endif
498 503
499 l %r13,0f-.LPG0(%r13) 504 l %r13,4f-.LPG0(%r13)
500 b 0(%r13) 505 b 0(%r13)
5010: .long startup_continue 506 .align 4
5074: .long startup_continue
5085: .long sched_clock_base_cc
509 .align 8
5106: .long 0x7fffffff,0xffffffff
502 511
503# 512#
504# params at 10400 (setup.h) 513# params at 10400 (setup.h)
diff --git a/arch/s390/kernel/nmi.c b/arch/s390/kernel/nmi.c
index 4bfdc421d7e9..28cf196ba775 100644
--- a/arch/s390/kernel/nmi.c
+++ b/arch/s390/kernel/nmi.c
@@ -10,6 +10,7 @@
10 10
11#include <linux/init.h> 11#include <linux/init.h>
12#include <linux/errno.h> 12#include <linux/errno.h>
13#include <linux/hardirq.h>
13#include <linux/time.h> 14#include <linux/time.h>
14#include <linux/module.h> 15#include <linux/module.h>
15#include <asm/lowcore.h> 16#include <asm/lowcore.h>
@@ -253,7 +254,7 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
253 struct mci *mci; 254 struct mci *mci;
254 int umode; 255 int umode;
255 256
256 lockdep_off(); 257 nmi_enter();
257 s390_idle_check(); 258 s390_idle_check();
258 259
259 mci = (struct mci *) &S390_lowcore.mcck_interruption_code; 260 mci = (struct mci *) &S390_lowcore.mcck_interruption_code;
@@ -363,7 +364,7 @@ void notrace s390_do_machine_check(struct pt_regs *regs)
363 mcck->warning = 1; 364 mcck->warning = 1;
364 set_thread_flag(TIF_MCCK_PENDING); 365 set_thread_flag(TIF_MCCK_PENDING);
365 } 366 }
366 lockdep_on(); 367 nmi_exit();
367} 368}
368 369
369static int __init machine_check_init(void) 370static int __init machine_check_init(void)
diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index 06201b93cbbf..7402b6a39ead 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -82,9 +82,6 @@ EXPORT_SYMBOL(console_devno);
82unsigned int console_irq = -1; 82unsigned int console_irq = -1;
83EXPORT_SYMBOL(console_irq); 83EXPORT_SYMBOL(console_irq);
84 84
85unsigned long machine_flags;
86EXPORT_SYMBOL(machine_flags);
87
88unsigned long elf_hwcap = 0; 85unsigned long elf_hwcap = 0;
89char elf_platform[ELF_PLATFORM_SIZE]; 86char elf_platform[ELF_PLATFORM_SIZE];
90 87
@@ -426,6 +423,7 @@ setup_lowcore(void)
426 __alloc_bootmem(PAGE_SIZE, PAGE_SIZE, 0) + PAGE_SIZE; 423 __alloc_bootmem(PAGE_SIZE, PAGE_SIZE, 0) + PAGE_SIZE;
427 lc->current_task = (unsigned long) init_thread_union.thread_info.task; 424 lc->current_task = (unsigned long) init_thread_union.thread_info.task;
428 lc->thread_info = (unsigned long) &init_thread_union; 425 lc->thread_info = (unsigned long) &init_thread_union;
426 lc->machine_flags = S390_lowcore.machine_flags;
429#ifndef CONFIG_64BIT 427#ifndef CONFIG_64BIT
430 if (MACHINE_HAS_IEEE) { 428 if (MACHINE_HAS_IEEE) {
431 lc->extended_save_area_addr = (__u32) 429 lc->extended_save_area_addr = (__u32)
@@ -436,6 +434,14 @@ setup_lowcore(void)
436#else 434#else
437 lc->vdso_per_cpu_data = (unsigned long) &lc->paste[0]; 435 lc->vdso_per_cpu_data = (unsigned long) &lc->paste[0];
438#endif 436#endif
437 lc->sync_enter_timer = S390_lowcore.sync_enter_timer;
438 lc->async_enter_timer = S390_lowcore.async_enter_timer;
439 lc->exit_timer = S390_lowcore.exit_timer;
440 lc->user_timer = S390_lowcore.user_timer;
441 lc->system_timer = S390_lowcore.system_timer;
442 lc->steal_timer = S390_lowcore.steal_timer;
443 lc->last_update_timer = S390_lowcore.last_update_timer;
444 lc->last_update_clock = S390_lowcore.last_update_clock;
439 set_prefix((u32)(unsigned long) lc); 445 set_prefix((u32)(unsigned long) lc);
440 lowcore_ptr[0] = lc; 446 lowcore_ptr[0] = lc;
441} 447}
diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index 006ed5016eb4..a985a3ba4401 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -571,6 +571,7 @@ int __cpuinit __cpu_up(unsigned int cpu)
571 cpu_lowcore->current_task = (unsigned long) idle; 571 cpu_lowcore->current_task = (unsigned long) idle;
572 cpu_lowcore->cpu_nr = cpu; 572 cpu_lowcore->cpu_nr = cpu;
573 cpu_lowcore->kernel_asce = S390_lowcore.kernel_asce; 573 cpu_lowcore->kernel_asce = S390_lowcore.kernel_asce;
574 cpu_lowcore->machine_flags = S390_lowcore.machine_flags;
574 eieio(); 575 eieio();
575 576
576 while (signal_processor(cpu, sigp_restart) == sigp_busy) 577 while (signal_processor(cpu, sigp_restart) == sigp_busy)
@@ -590,7 +591,8 @@ static int __init setup_possible_cpus(char *s)
590 int pcpus, cpu; 591 int pcpus, cpu;
591 592
592 pcpus = simple_strtoul(s, NULL, 0); 593 pcpus = simple_strtoul(s, NULL, 0);
593 for (cpu = 0; cpu < pcpus && cpu < nr_cpu_ids; cpu++) 594 init_cpu_possible(cpumask_of(0));
595 for (cpu = 1; cpu < pcpus && cpu < nr_cpu_ids; cpu++)
594 set_cpu_possible(cpu, true); 596 set_cpu_possible(cpu, true);
595 return 0; 597 return 0;
596} 598}
diff --git a/arch/s390/kernel/syscalls.S b/arch/s390/kernel/syscalls.S
index fe5b25a988ab..2c7739fe70b1 100644
--- a/arch/s390/kernel/syscalls.S
+++ b/arch/s390/kernel/syscalls.S
@@ -336,3 +336,5 @@ SYSCALL(sys_inotify_init1,sys_inotify_init1,sys_inotify_init1_wrapper)
336SYSCALL(sys_pipe2,sys_pipe2,sys_pipe2_wrapper) /* 325 */ 336SYSCALL(sys_pipe2,sys_pipe2,sys_pipe2_wrapper) /* 325 */
337SYSCALL(sys_dup3,sys_dup3,sys_dup3_wrapper) 337SYSCALL(sys_dup3,sys_dup3,sys_dup3_wrapper)
338SYSCALL(sys_epoll_create1,sys_epoll_create1,sys_epoll_create1_wrapper) 338SYSCALL(sys_epoll_create1,sys_epoll_create1,sys_epoll_create1_wrapper)
339SYSCALL(sys_preadv,sys_preadv,compat_sys_preadv_wrapper)
340SYSCALL(sys_pwritev,sys_pwritev,compat_sys_pwritev_wrapper)
diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c
index f72d41068dc2..6ded50dfa75a 100644
--- a/arch/s390/kernel/time.c
+++ b/arch/s390/kernel/time.c
@@ -52,9 +52,6 @@
52#define USECS_PER_JIFFY ((unsigned long) 1000000/HZ) 52#define USECS_PER_JIFFY ((unsigned long) 1000000/HZ)
53#define CLK_TICKS_PER_JIFFY ((unsigned long) USECS_PER_JIFFY << 12) 53#define CLK_TICKS_PER_JIFFY ((unsigned long) USECS_PER_JIFFY << 12)
54 54
55/* The value of the TOD clock for 1.1.1970. */
56#define TOD_UNIX_EPOCH 0x7d91048bca000000ULL
57
58/* 55/*
59 * Create a small time difference between the timer interrupts 56 * Create a small time difference between the timer interrupts
60 * on the different cpus to avoid lock contention. 57 * on the different cpus to avoid lock contention.
@@ -63,9 +60,10 @@
63 60
64#define TICK_SIZE tick 61#define TICK_SIZE tick
65 62
63u64 sched_clock_base_cc = -1; /* Force to data section. */
64
66static ext_int_info_t ext_int_info_cc; 65static ext_int_info_t ext_int_info_cc;
67static ext_int_info_t ext_int_etr_cc; 66static ext_int_info_t ext_int_etr_cc;
68static u64 sched_clock_base_cc;
69 67
70static DEFINE_PER_CPU(struct clock_event_device, comparators); 68static DEFINE_PER_CPU(struct clock_event_device, comparators);
71 69
@@ -195,22 +193,12 @@ static void timing_alert_interrupt(__u16 code)
195static void etr_reset(void); 193static void etr_reset(void);
196static void stp_reset(void); 194static void stp_reset(void);
197 195
198/* 196unsigned long read_persistent_clock(void)
199 * Get the TOD clock running.
200 */
201static u64 __init reset_tod_clock(void)
202{ 197{
203 u64 time; 198 struct timespec ts;
204
205 etr_reset();
206 stp_reset();
207 if (store_clock(&time) == 0)
208 return time;
209 /* TOD clock not running. Set the clock to Unix Epoch. */
210 if (set_clock(TOD_UNIX_EPOCH) != 0 || store_clock(&time) != 0)
211 panic("TOD clock not operational.");
212 199
213 return TOD_UNIX_EPOCH; 200 tod_to_timeval(get_clock() - TOD_UNIX_EPOCH, &ts);
201 return ts.tv_sec;
214} 202}
215 203
216static cycle_t read_tod_clock(void) 204static cycle_t read_tod_clock(void)
@@ -265,12 +253,13 @@ void update_vsyscall_tz(void)
265 */ 253 */
266void __init time_init(void) 254void __init time_init(void)
267{ 255{
268 sched_clock_base_cc = reset_tod_clock(); 256 struct timespec ts;
257 unsigned long flags;
258 cycle_t now;
269 259
270 /* set xtime */ 260 /* Reset time synchronization interfaces. */
271 tod_to_timeval(sched_clock_base_cc - TOD_UNIX_EPOCH, &xtime); 261 etr_reset();
272 set_normalized_timespec(&wall_to_monotonic, 262 stp_reset();
273 -xtime.tv_sec, -xtime.tv_nsec);
274 263
275 /* request the clock comparator external interrupt */ 264 /* request the clock comparator external interrupt */
276 if (register_early_external_interrupt(0x1004, 265 if (register_early_external_interrupt(0x1004,
@@ -278,17 +267,38 @@ void __init time_init(void)
278 &ext_int_info_cc) != 0) 267 &ext_int_info_cc) != 0)
279 panic("Couldn't request external interrupt 0x1004"); 268 panic("Couldn't request external interrupt 0x1004");
280 269
281 if (clocksource_register(&clocksource_tod) != 0)
282 panic("Could not register TOD clock source");
283
284 /* request the timing alert external interrupt */ 270 /* request the timing alert external interrupt */
285 if (register_early_external_interrupt(0x1406, 271 if (register_early_external_interrupt(0x1406,
286 timing_alert_interrupt, 272 timing_alert_interrupt,
287 &ext_int_etr_cc) != 0) 273 &ext_int_etr_cc) != 0)
288 panic("Couldn't request external interrupt 0x1406"); 274 panic("Couldn't request external interrupt 0x1406");
289 275
276 if (clocksource_register(&clocksource_tod) != 0)
277 panic("Could not register TOD clock source");
278
279 /*
280 * The TOD clock is an accurate clock. The xtime should be
281 * initialized in a way that the difference between TOD and
282 * xtime is reasonably small. Too bad that timekeeping_init
283 * sets xtime.tv_nsec to zero. In addition the clock source
284 * change from the jiffies clock source to the TOD clock
285 * source add another error of up to 1/HZ second. The same
286 * function sets wall_to_monotonic to a value that is too
287 * small for /proc/uptime to be accurate.
288 * Reset xtime and wall_to_monotonic to sane values.
289 */
290 write_seqlock_irqsave(&xtime_lock, flags);
291 now = get_clock();
292 tod_to_timeval(now - TOD_UNIX_EPOCH, &xtime);
293 clocksource_tod.cycle_last = now;
294 clocksource_tod.raw_time = xtime;
295 tod_to_timeval(sched_clock_base_cc - TOD_UNIX_EPOCH, &ts);
296 set_normalized_timespec(&wall_to_monotonic, -ts.tv_sec, -ts.tv_nsec);
297 write_sequnlock_irqrestore(&xtime_lock, flags);
298
290 /* Enable TOD clock interrupts on the boot cpu. */ 299 /* Enable TOD clock interrupts on the boot cpu. */
291 init_cpu_timer(); 300 init_cpu_timer();
301
292 /* Enable cpu timer interrupts on the boot cpu. */ 302 /* Enable cpu timer interrupts on the boot cpu. */
293 vtime_init(); 303 vtime_init();
294} 304}
@@ -1423,6 +1433,7 @@ static void *stp_page;
1423static void stp_work_fn(struct work_struct *work); 1433static void stp_work_fn(struct work_struct *work);
1424static DEFINE_MUTEX(stp_work_mutex); 1434static DEFINE_MUTEX(stp_work_mutex);
1425static DECLARE_WORK(stp_work, stp_work_fn); 1435static DECLARE_WORK(stp_work, stp_work_fn);
1436static struct timer_list stp_timer;
1426 1437
1427static int __init early_parse_stp(char *p) 1438static int __init early_parse_stp(char *p)
1428{ 1439{
@@ -1454,10 +1465,16 @@ static void __init stp_reset(void)
1454 } 1465 }
1455} 1466}
1456 1467
1468static void stp_timeout(unsigned long dummy)
1469{
1470 queue_work(time_sync_wq, &stp_work);
1471}
1472
1457static int __init stp_init(void) 1473static int __init stp_init(void)
1458{ 1474{
1459 if (!test_bit(CLOCK_SYNC_HAS_STP, &clock_sync_flags)) 1475 if (!test_bit(CLOCK_SYNC_HAS_STP, &clock_sync_flags))
1460 return 0; 1476 return 0;
1477 setup_timer(&stp_timer, stp_timeout, 0UL);
1461 time_init_wq(); 1478 time_init_wq();
1462 if (!stp_online) 1479 if (!stp_online)
1463 return 0; 1480 return 0;
@@ -1565,6 +1582,7 @@ static void stp_work_fn(struct work_struct *work)
1565 1582
1566 if (!stp_online) { 1583 if (!stp_online) {
1567 chsc_sstpc(stp_page, STP_OP_CTRL, 0x0000); 1584 chsc_sstpc(stp_page, STP_OP_CTRL, 0x0000);
1585 del_timer_sync(&stp_timer);
1568 goto out_unlock; 1586 goto out_unlock;
1569 } 1587 }
1570 1588
@@ -1586,6 +1604,13 @@ static void stp_work_fn(struct work_struct *work)
1586 stop_machine(stp_sync_clock, &stp_sync, &cpu_online_map); 1604 stop_machine(stp_sync_clock, &stp_sync, &cpu_online_map);
1587 put_online_cpus(); 1605 put_online_cpus();
1588 1606
1607 if (!check_sync_clock())
1608 /*
1609 * There is a usable clock but the synchonization failed.
1610 * Retry after a second.
1611 */
1612 mod_timer(&stp_timer, jiffies + HZ);
1613
1589out_unlock: 1614out_unlock:
1590 mutex_unlock(&stp_work_mutex); 1615 mutex_unlock(&stp_work_mutex);
1591} 1616}
diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c
index ecf0304e61c1..38ea92ff04f9 100644
--- a/arch/s390/kernel/vtime.c
+++ b/arch/s390/kernel/vtime.c
@@ -134,6 +134,8 @@ void vtime_start_cpu(void)
134 /* Account time spent with enabled wait psw loaded as idle time. */ 134 /* Account time spent with enabled wait psw loaded as idle time. */
135 idle_time = S390_lowcore.int_clock - idle->idle_enter; 135 idle_time = S390_lowcore.int_clock - idle->idle_enter;
136 account_idle_time(idle_time); 136 account_idle_time(idle_time);
137 S390_lowcore.steal_timer +=
138 idle->idle_enter - S390_lowcore.last_update_clock;
137 S390_lowcore.last_update_clock = S390_lowcore.int_clock; 139 S390_lowcore.last_update_clock = S390_lowcore.int_clock;
138 140
139 /* Account system time spent going idle. */ 141 /* Account system time spent going idle. */
@@ -425,17 +427,7 @@ void add_virt_timer_periodic(void *new)
425} 427}
426EXPORT_SYMBOL(add_virt_timer_periodic); 428EXPORT_SYMBOL(add_virt_timer_periodic);
427 429
428/* 430int __mod_vtimer(struct vtimer_list *timer, __u64 expires, int periodic)
429 * If we change a pending timer the function must be called on the CPU
430 * where the timer is running on, e.g. by smp_call_function_single()
431 *
432 * The original mod_timer adds the timer if it is not pending. For
433 * compatibility we do the same. The timer will be added on the current
434 * CPU as a oneshot timer.
435 *
436 * returns whether it has modified a pending timer (1) or not (0)
437 */
438int mod_virt_timer(struct vtimer_list *timer, __u64 expires)
439{ 431{
440 struct vtimer_queue *vq; 432 struct vtimer_queue *vq;
441 unsigned long flags; 433 unsigned long flags;
@@ -444,39 +436,35 @@ int mod_virt_timer(struct vtimer_list *timer, __u64 expires)
444 BUG_ON(!timer->function); 436 BUG_ON(!timer->function);
445 BUG_ON(!expires || expires > VTIMER_MAX_SLICE); 437 BUG_ON(!expires || expires > VTIMER_MAX_SLICE);
446 438
447 /*
448 * This is a common optimization triggered by the
449 * networking code - if the timer is re-modified
450 * to be the same thing then just return:
451 */
452 if (timer->expires == expires && vtimer_pending(timer)) 439 if (timer->expires == expires && vtimer_pending(timer))
453 return 1; 440 return 1;
454 441
455 cpu = get_cpu(); 442 cpu = get_cpu();
456 vq = &per_cpu(virt_cpu_timer, cpu); 443 vq = &per_cpu(virt_cpu_timer, cpu);
457 444
458 /* check if we run on the right CPU */
459 BUG_ON(timer->cpu != cpu);
460
461 /* disable interrupts before test if timer is pending */ 445 /* disable interrupts before test if timer is pending */
462 spin_lock_irqsave(&vq->lock, flags); 446 spin_lock_irqsave(&vq->lock, flags);
463 447
464 /* if timer isn't pending add it on the current CPU */ 448 /* if timer isn't pending add it on the current CPU */
465 if (!vtimer_pending(timer)) { 449 if (!vtimer_pending(timer)) {
466 spin_unlock_irqrestore(&vq->lock, flags); 450 spin_unlock_irqrestore(&vq->lock, flags);
467 /* we do not activate an interval timer with mod_virt_timer */ 451
468 timer->interval = 0; 452 if (periodic)
453 timer->interval = expires;
454 else
455 timer->interval = 0;
469 timer->expires = expires; 456 timer->expires = expires;
470 timer->cpu = cpu; 457 timer->cpu = cpu;
471 internal_add_vtimer(timer); 458 internal_add_vtimer(timer);
472 return 0; 459 return 0;
473 } 460 }
474 461
462 /* check if we run on the right CPU */
463 BUG_ON(timer->cpu != cpu);
464
475 list_del_init(&timer->entry); 465 list_del_init(&timer->entry);
476 timer->expires = expires; 466 timer->expires = expires;
477 467 if (periodic)
478 /* also change the interval if we have an interval timer */
479 if (timer->interval)
480 timer->interval = expires; 468 timer->interval = expires;
481 469
482 /* the timer can't expire anymore so we can release the lock */ 470 /* the timer can't expire anymore so we can release the lock */
@@ -484,9 +472,32 @@ int mod_virt_timer(struct vtimer_list *timer, __u64 expires)
484 internal_add_vtimer(timer); 472 internal_add_vtimer(timer);
485 return 1; 473 return 1;
486} 474}
475
476/*
477 * If we change a pending timer the function must be called on the CPU
478 * where the timer is running on.
479 *
480 * returns whether it has modified a pending timer (1) or not (0)
481 */
482int mod_virt_timer(struct vtimer_list *timer, __u64 expires)
483{
484 return __mod_vtimer(timer, expires, 0);
485}
487EXPORT_SYMBOL(mod_virt_timer); 486EXPORT_SYMBOL(mod_virt_timer);
488 487
489/* 488/*
489 * If we change a pending timer the function must be called on the CPU
490 * where the timer is running on.
491 *
492 * returns whether it has modified a pending timer (1) or not (0)
493 */
494int mod_virt_timer_periodic(struct vtimer_list *timer, __u64 expires)
495{
496 return __mod_vtimer(timer, expires, 1);
497}
498EXPORT_SYMBOL(mod_virt_timer_periodic);
499
500/*
490 * delete a virtual timer 501 * delete a virtual timer
491 * 502 *
492 * returns whether the deleted timer was pending (1) or not (0) 503 * returns whether the deleted timer was pending (1) or not (0)
@@ -516,16 +527,8 @@ EXPORT_SYMBOL(del_virt_timer);
516 */ 527 */
517void init_cpu_vtimer(void) 528void init_cpu_vtimer(void)
518{ 529{
519 struct thread_info *ti = current_thread_info();
520 struct vtimer_queue *vq; 530 struct vtimer_queue *vq;
521 531
522 S390_lowcore.user_timer = ti->user_timer;
523 S390_lowcore.system_timer = ti->system_timer;
524
525 /* kick the virtual timer */
526 asm volatile ("STCK %0" : "=m" (S390_lowcore.last_update_clock));
527 asm volatile ("STPT %0" : "=m" (S390_lowcore.last_update_timer));
528
529 /* initialize per cpu vtimer structure */ 532 /* initialize per cpu vtimer structure */
530 vq = &__get_cpu_var(virt_cpu_timer); 533 vq = &__get_cpu_var(virt_cpu_timer);
531 INIT_LIST_HEAD(&vq->list); 534 INIT_LIST_HEAD(&vq->list);
diff --git a/arch/sh/Kconfig b/arch/sh/Kconfig
index 5e4babecf934..e7390dd0283d 100644
--- a/arch/sh/Kconfig
+++ b/arch/sh/Kconfig
@@ -14,6 +14,7 @@ config SUPERH
14 select HAVE_GENERIC_DMA_COHERENT 14 select HAVE_GENERIC_DMA_COHERENT
15 select HAVE_IOREMAP_PROT if MMU 15 select HAVE_IOREMAP_PROT if MMU
16 select HAVE_ARCH_TRACEHOOK 16 select HAVE_ARCH_TRACEHOOK
17 select HAVE_DMA_API_DEBUG
17 help 18 help
18 The SuperH is a RISC processor targeted for use in embedded systems 19 The SuperH is a RISC processor targeted for use in embedded systems
19 and consumer electronics; it was also used in the Sega Dreamcast 20 and consumer electronics; it was also used in the Sega Dreamcast
@@ -21,7 +22,7 @@ config SUPERH
21 <http://www.linux-sh.org/>. 22 <http://www.linux-sh.org/>.
22 23
23config SUPERH32 24config SUPERH32
24 def_bool !SUPERH64 25 def_bool ARCH = "sh"
25 select HAVE_KPROBES 26 select HAVE_KPROBES
26 select HAVE_KRETPROBES 27 select HAVE_KRETPROBES
27 select HAVE_FUNCTION_TRACER 28 select HAVE_FUNCTION_TRACER
@@ -31,7 +32,7 @@ config SUPERH32
31 select ARCH_HIBERNATION_POSSIBLE if MMU 32 select ARCH_HIBERNATION_POSSIBLE if MMU
32 33
33config SUPERH64 34config SUPERH64
34 def_bool y if CPU_SH5 35 def_bool ARCH = "sh64"
35 36
36config ARCH_DEFCONFIG 37config ARCH_DEFCONFIG
37 string 38 string
@@ -187,6 +188,8 @@ config ARCH_SHMOBILE
187 bool 188 bool
188 select ARCH_SUSPEND_POSSIBLE 189 select ARCH_SUSPEND_POSSIBLE
189 190
191if SUPERH32
192
190choice 193choice
191 prompt "Processor sub-type selection" 194 prompt "Processor sub-type selection"
192 195
@@ -408,6 +411,15 @@ config CPU_SUBTYPE_SH7366
408 select SYS_SUPPORTS_NUMA 411 select SYS_SUPPORTS_NUMA
409 select SYS_SUPPORTS_CMT 412 select SYS_SUPPORTS_CMT
410 413
414endchoice
415
416endif
417
418if SUPERH64
419
420choice
421 prompt "Processor sub-type selection"
422
411# SH-5 Processor Support 423# SH-5 Processor Support
412 424
413config CPU_SUBTYPE_SH5_101 425config CPU_SUBTYPE_SH5_101
@@ -420,6 +432,8 @@ config CPU_SUBTYPE_SH5_103
420 432
421endchoice 433endchoice
422 434
435endif
436
423source "arch/sh/mm/Kconfig" 437source "arch/sh/mm/Kconfig"
424 438
425source "arch/sh/Kconfig.cpu" 439source "arch/sh/Kconfig.cpu"
diff --git a/arch/sh/boards/board-ap325rxa.c b/arch/sh/boards/board-ap325rxa.c
index 912458f666eb..39e46919df14 100644
--- a/arch/sh/boards/board-ap325rxa.c
+++ b/arch/sh/boards/board-ap325rxa.c
@@ -349,6 +349,7 @@ static int ov7725_power(struct device *dev, int mode)
349static struct ov772x_camera_info ov7725_info = { 349static struct ov772x_camera_info ov7725_info = {
350 .buswidth = SOCAM_DATAWIDTH_8, 350 .buswidth = SOCAM_DATAWIDTH_8,
351 .flags = OV772X_FLAG_VFLIP | OV772X_FLAG_HFLIP, 351 .flags = OV772X_FLAG_VFLIP | OV772X_FLAG_HFLIP,
352 .edgectrl = OV772X_AUTO_EDGECTRL(0xf, 0),
352 .link = { 353 .link = {
353 .power = ov7725_power, 354 .power = ov7725_power,
354 }, 355 },
diff --git a/arch/sh/boards/board-urquell.c b/arch/sh/boards/board-urquell.c
index 8367d1d789c3..beb88c4da2c1 100644
--- a/arch/sh/boards/board-urquell.c
+++ b/arch/sh/boards/board-urquell.c
@@ -2,6 +2,8 @@
2 * Renesas Technology Corp. SH7786 Urquell Support. 2 * Renesas Technology Corp. SH7786 Urquell Support.
3 * 3 *
4 * Copyright (C) 2008 Kuninori Morimoto <morimoto.kuninori@renesas.com> 4 * Copyright (C) 2008 Kuninori Morimoto <morimoto.kuninori@renesas.com>
5 *
6 * Based on board-sh7785lcr.c
5 * Copyright (C) 2008 Yoshihiro Shimoda 7 * Copyright (C) 2008 Yoshihiro Shimoda
6 * 8 *
7 * This file is subject to the terms and conditions of the GNU General Public 9 * This file is subject to the terms and conditions of the GNU General Public
@@ -21,6 +23,32 @@
21#include <asm/heartbeat.h> 23#include <asm/heartbeat.h>
22#include <asm/sizes.h> 24#include <asm/sizes.h>
23 25
26/*
27 * bit 1234 5678
28 *----------------------------
29 * SW1 0101 0010 -> Pck 33MHz version
30 * (1101 0010) Pck 66MHz version
31 * SW2 0x1x xxxx -> little endian
32 * 29bit mode
33 * SW47 0001 1000 -> CS0 : on-board flash
34 * CS1 : SRAM, registers, LAN, PCMCIA
35 * 38400 bps for SCIF1
36 *
37 * Address
38 * 0x00000000 - 0x04000000 (CS0) Nor Flash
39 * 0x04000000 - 0x04200000 (CS1) SRAM
40 * 0x05000000 - 0x05800000 (CS1) on board register
41 * 0x05800000 - 0x06000000 (CS1) LAN91C111
42 * 0x06000000 - 0x06400000 (CS1) PCMCIA
43 * 0x08000000 - 0x10000000 (CS2-CS3) DDR3
44 * 0x10000000 - 0x14000000 (CS4) PCIe
45 * 0x14000000 - 0x14800000 (CS5) Core0 LRAM/URAM
46 * 0x14800000 - 0x15000000 (CS5) Core1 LRAM/URAM
47 * 0x18000000 - 0x1C000000 (CS6) ATA/NAND-Flash
48 * 0x1C000000 - (CS7) SH7786 Control register
49 */
50
51/* HeartBeat */
24static struct resource heartbeat_resources[] = { 52static struct resource heartbeat_resources[] = {
25 [0] = { 53 [0] = {
26 .start = BOARDREG(SLEDR), 54 .start = BOARDREG(SLEDR),
@@ -43,6 +71,7 @@ static struct platform_device heartbeat_device = {
43 .resource = heartbeat_resources, 71 .resource = heartbeat_resources,
44}; 72};
45 73
74/* LAN91C111 */
46static struct smc91x_platdata smc91x_info = { 75static struct smc91x_platdata smc91x_info = {
47 .flags = SMC91X_USE_16BIT | SMC91X_NOWAIT, 76 .flags = SMC91X_USE_16BIT | SMC91X_NOWAIT,
48}; 77};
@@ -69,6 +98,7 @@ static struct platform_device smc91x_eth_device = {
69 }, 98 },
70}; 99};
71 100
101/* Nor Flash */
72static struct mtd_partition nor_flash_partitions[] = { 102static struct mtd_partition nor_flash_partitions[] = {
73 { 103 {
74 .name = "loader", 104 .name = "loader",
diff --git a/arch/sh/drivers/pci/ops-sh7785lcr.c b/arch/sh/drivers/pci/ops-sh7785lcr.c
index e8b7446a7c2b..fb0869f0bef8 100644
--- a/arch/sh/drivers/pci/ops-sh7785lcr.c
+++ b/arch/sh/drivers/pci/ops-sh7785lcr.c
@@ -48,8 +48,13 @@ EXPORT_SYMBOL(board_pci_channels);
48 48
49static struct sh4_pci_address_map sh7785_pci_map = { 49static struct sh4_pci_address_map sh7785_pci_map = {
50 .window0 = { 50 .window0 = {
51#if defined(CONFIG_32BIT)
52 .base = SH7780_32BIT_DDR_BASE_ADDR,
53 .size = 0x40000000,
54#else
51 .base = SH7780_CS0_BASE_ADDR, 55 .base = SH7780_CS0_BASE_ADDR,
52 .size = 0x20000000, 56 .size = 0x20000000,
57#endif
53 }, 58 },
54 59
55 .flags = SH4_PCIC_NO_RESET, 60 .flags = SH4_PCIC_NO_RESET,
diff --git a/arch/sh/drivers/pci/pci-sh7780.h b/arch/sh/drivers/pci/pci-sh7780.h
index 97b2c98f05c4..93adc7119b79 100644
--- a/arch/sh/drivers/pci/pci-sh7780.h
+++ b/arch/sh/drivers/pci/pci-sh7780.h
@@ -104,6 +104,8 @@
104#define SH7780_CS5_BASE_ADDR (SH7780_CS4_BASE_ADDR + SH7780_MEM_REGION_SIZE) 104#define SH7780_CS5_BASE_ADDR (SH7780_CS4_BASE_ADDR + SH7780_MEM_REGION_SIZE)
105#define SH7780_CS6_BASE_ADDR (SH7780_CS5_BASE_ADDR + SH7780_MEM_REGION_SIZE) 105#define SH7780_CS6_BASE_ADDR (SH7780_CS5_BASE_ADDR + SH7780_MEM_REGION_SIZE)
106 106
107#define SH7780_32BIT_DDR_BASE_ADDR 0x40000000
108
107struct sh4_pci_address_map; 109struct sh4_pci_address_map;
108 110
109/* arch/sh/drivers/pci/pci-sh7780.c */ 111/* arch/sh/drivers/pci/pci-sh7780.c */
diff --git a/arch/sh/drivers/pci/pci.c b/arch/sh/drivers/pci/pci.c
index e36c7b870861..0d6ac7a1db49 100644
--- a/arch/sh/drivers/pci/pci.c
+++ b/arch/sh/drivers/pci/pci.c
@@ -19,6 +19,7 @@
19#include <linux/kernel.h> 19#include <linux/kernel.h>
20#include <linux/pci.h> 20#include <linux/pci.h>
21#include <linux/init.h> 21#include <linux/init.h>
22#include <linux/dma-debug.h>
22#include <asm/io.h> 23#include <asm/io.h>
23 24
24static int __init pcibios_init(void) 25static int __init pcibios_init(void)
@@ -43,6 +44,8 @@ static int __init pcibios_init(void)
43 44
44 pci_fixup_irqs(pci_common_swizzle, pcibios_map_platform_irq); 45 pci_fixup_irqs(pci_common_swizzle, pcibios_map_platform_irq);
45 46
47 dma_debug_add_bus(&pci_bus_type);
48
46 return 0; 49 return 0;
47} 50}
48subsys_initcall(pcibios_init); 51subsys_initcall(pcibios_init);
diff --git a/arch/sh/include/asm/dma-mapping.h b/arch/sh/include/asm/dma-mapping.h
index 627315ecdb52..ea9d4f41c9d2 100644
--- a/arch/sh/include/asm/dma-mapping.h
+++ b/arch/sh/include/asm/dma-mapping.h
@@ -3,6 +3,7 @@
3 3
4#include <linux/mm.h> 4#include <linux/mm.h>
5#include <linux/scatterlist.h> 5#include <linux/scatterlist.h>
6#include <linux/dma-debug.h>
6#include <asm/cacheflush.h> 7#include <asm/cacheflush.h>
7#include <asm/io.h> 8#include <asm/io.h>
8#include <asm-generic/dma-coherent.h> 9#include <asm-generic/dma-coherent.h>
@@ -38,16 +39,26 @@ static inline dma_addr_t dma_map_single(struct device *dev,
38 void *ptr, size_t size, 39 void *ptr, size_t size,
39 enum dma_data_direction dir) 40 enum dma_data_direction dir)
40{ 41{
42 dma_addr_t addr = virt_to_phys(ptr);
43
41#if defined(CONFIG_PCI) && !defined(CONFIG_SH_PCIDMA_NONCOHERENT) 44#if defined(CONFIG_PCI) && !defined(CONFIG_SH_PCIDMA_NONCOHERENT)
42 if (dev->bus == &pci_bus_type) 45 if (dev->bus == &pci_bus_type)
43 return virt_to_phys(ptr); 46 return addr;
44#endif 47#endif
45 dma_cache_sync(dev, ptr, size, dir); 48 dma_cache_sync(dev, ptr, size, dir);
46 49
47 return virt_to_phys(ptr); 50 debug_dma_map_page(dev, virt_to_page(ptr),
51 (unsigned long)ptr & ~PAGE_MASK, size,
52 dir, addr, true);
53
54 return addr;
48} 55}
49 56
50#define dma_unmap_single(dev, addr, size, dir) do { } while (0) 57static inline void dma_unmap_single(struct device *dev, dma_addr_t addr,
58 size_t size, enum dma_data_direction dir)
59{
60 debug_dma_unmap_page(dev, addr, size, dir, true);
61}
51 62
52static inline int dma_map_sg(struct device *dev, struct scatterlist *sg, 63static inline int dma_map_sg(struct device *dev, struct scatterlist *sg,
53 int nents, enum dma_data_direction dir) 64 int nents, enum dma_data_direction dir)
@@ -59,12 +70,19 @@ static inline int dma_map_sg(struct device *dev, struct scatterlist *sg,
59 dma_cache_sync(dev, sg_virt(&sg[i]), sg[i].length, dir); 70 dma_cache_sync(dev, sg_virt(&sg[i]), sg[i].length, dir);
60#endif 71#endif
61 sg[i].dma_address = sg_phys(&sg[i]); 72 sg[i].dma_address = sg_phys(&sg[i]);
73 sg[i].dma_length = sg[i].length;
62 } 74 }
63 75
76 debug_dma_map_sg(dev, sg, nents, i, dir);
77
64 return nents; 78 return nents;
65} 79}
66 80
67#define dma_unmap_sg(dev, sg, nents, dir) do { } while (0) 81static inline void dma_unmap_sg(struct device *dev, struct scatterlist *sg,
82 int nents, enum dma_data_direction dir)
83{
84 debug_dma_unmap_sg(dev, sg, nents, dir);
85}
68 86
69static inline dma_addr_t dma_map_page(struct device *dev, struct page *page, 87static inline dma_addr_t dma_map_page(struct device *dev, struct page *page,
70 unsigned long offset, size_t size, 88 unsigned long offset, size_t size,
@@ -111,6 +129,7 @@ static inline void dma_sync_sg(struct device *dev, struct scatterlist *sg,
111 dma_cache_sync(dev, sg_virt(&sg[i]), sg[i].length, dir); 129 dma_cache_sync(dev, sg_virt(&sg[i]), sg[i].length, dir);
112#endif 130#endif
113 sg[i].dma_address = sg_phys(&sg[i]); 131 sg[i].dma_address = sg_phys(&sg[i]);
132 sg[i].dma_length = sg[i].length;
114 } 133 }
115} 134}
116 135
@@ -119,6 +138,7 @@ static inline void dma_sync_single_for_cpu(struct device *dev,
119 enum dma_data_direction dir) 138 enum dma_data_direction dir)
120{ 139{
121 dma_sync_single(dev, dma_handle, size, dir); 140 dma_sync_single(dev, dma_handle, size, dir);
141 debug_dma_sync_single_for_cpu(dev, dma_handle, size, dir);
122} 142}
123 143
124static inline void dma_sync_single_for_device(struct device *dev, 144static inline void dma_sync_single_for_device(struct device *dev,
@@ -127,6 +147,7 @@ static inline void dma_sync_single_for_device(struct device *dev,
127 enum dma_data_direction dir) 147 enum dma_data_direction dir)
128{ 148{
129 dma_sync_single(dev, dma_handle, size, dir); 149 dma_sync_single(dev, dma_handle, size, dir);
150 debug_dma_sync_single_for_device(dev, dma_handle, size, dir);
130} 151}
131 152
132static inline void dma_sync_single_range_for_cpu(struct device *dev, 153static inline void dma_sync_single_range_for_cpu(struct device *dev,
@@ -136,6 +157,8 @@ static inline void dma_sync_single_range_for_cpu(struct device *dev,
136 enum dma_data_direction direction) 157 enum dma_data_direction direction)
137{ 158{
138 dma_sync_single_for_cpu(dev, dma_handle+offset, size, direction); 159 dma_sync_single_for_cpu(dev, dma_handle+offset, size, direction);
160 debug_dma_sync_single_range_for_cpu(dev, dma_handle,
161 offset, size, direction);
139} 162}
140 163
141static inline void dma_sync_single_range_for_device(struct device *dev, 164static inline void dma_sync_single_range_for_device(struct device *dev,
@@ -145,6 +168,8 @@ static inline void dma_sync_single_range_for_device(struct device *dev,
145 enum dma_data_direction direction) 168 enum dma_data_direction direction)
146{ 169{
147 dma_sync_single_for_device(dev, dma_handle+offset, size, direction); 170 dma_sync_single_for_device(dev, dma_handle+offset, size, direction);
171 debug_dma_sync_single_range_for_device(dev, dma_handle,
172 offset, size, direction);
148} 173}
149 174
150 175
@@ -153,6 +178,7 @@ static inline void dma_sync_sg_for_cpu(struct device *dev,
153 enum dma_data_direction dir) 178 enum dma_data_direction dir)
154{ 179{
155 dma_sync_sg(dev, sg, nelems, dir); 180 dma_sync_sg(dev, sg, nelems, dir);
181 debug_dma_sync_sg_for_cpu(dev, sg, nelems, dir);
156} 182}
157 183
158static inline void dma_sync_sg_for_device(struct device *dev, 184static inline void dma_sync_sg_for_device(struct device *dev,
@@ -160,9 +186,9 @@ static inline void dma_sync_sg_for_device(struct device *dev,
160 enum dma_data_direction dir) 186 enum dma_data_direction dir)
161{ 187{
162 dma_sync_sg(dev, sg, nelems, dir); 188 dma_sync_sg(dev, sg, nelems, dir);
189 debug_dma_sync_sg_for_device(dev, sg, nelems, dir);
163} 190}
164 191
165
166static inline int dma_get_cache_alignment(void) 192static inline int dma_get_cache_alignment(void)
167{ 193{
168 /* 194 /*
diff --git a/arch/sh/include/asm/scatterlist.h b/arch/sh/include/asm/scatterlist.h
index 2084d0373693..c693d268a413 100644
--- a/arch/sh/include/asm/scatterlist.h
+++ b/arch/sh/include/asm/scatterlist.h
@@ -5,12 +5,13 @@
5 5
6struct scatterlist { 6struct scatterlist {
7#ifdef CONFIG_DEBUG_SG 7#ifdef CONFIG_DEBUG_SG
8 unsigned long sg_magic; 8 unsigned long sg_magic;
9#endif 9#endif
10 unsigned long page_link; 10 unsigned long page_link;
11 unsigned int offset;/* for highmem, page offset */ 11 unsigned int offset; /* for highmem, page offset */
12 dma_addr_t dma_address; 12 unsigned int length;
13 unsigned int length; 13 dma_addr_t dma_address;
14 unsigned int dma_length;
14}; 15};
15 16
16#define ISA_DMA_THRESHOLD PHYS_ADDR_MASK 17#define ISA_DMA_THRESHOLD PHYS_ADDR_MASK
diff --git a/arch/sh/include/asm/topology.h b/arch/sh/include/asm/topology.h
index a3f239545897..8489a0905a87 100644
--- a/arch/sh/include/asm/topology.h
+++ b/arch/sh/include/asm/topology.h
@@ -37,8 +37,11 @@
37#define pcibus_to_node(bus) ((void)(bus), -1) 37#define pcibus_to_node(bus) ((void)(bus), -1)
38#define pcibus_to_cpumask(bus) (pcibus_to_node(bus) == -1 ? \ 38#define pcibus_to_cpumask(bus) (pcibus_to_node(bus) == -1 ? \
39 CPU_MASK_ALL : \ 39 CPU_MASK_ALL : \
40 node_to_cpumask(pcibus_to_node(bus)) \ 40 node_to_cpumask(pcibus_to_node(bus)))
41 ) 41#define cpumask_of_pcibus(bus) (pcibus_to_node(bus) == -1 ? \
42 CPU_MASK_ALL_PTR : \
43 cpumask_of_node(pcibus_to_node(bus)))
44
42#endif 45#endif
43 46
44#include <asm-generic/topology.h> 47#include <asm-generic/topology.h>
diff --git a/arch/sh/include/asm/unistd_32.h b/arch/sh/include/asm/unistd_32.h
index d52c000cf924..2efb819e2db3 100644
--- a/arch/sh/include/asm/unistd_32.h
+++ b/arch/sh/include/asm/unistd_32.h
@@ -341,8 +341,10 @@
341#define __NR_dup3 330 341#define __NR_dup3 330
342#define __NR_pipe2 331 342#define __NR_pipe2 331
343#define __NR_inotify_init1 332 343#define __NR_inotify_init1 332
344#define __NR_preadv 333
345#define __NR_pwritev 334
344 346
345#define NR_syscalls 333 347#define NR_syscalls 335
346 348
347#ifdef __KERNEL__ 349#ifdef __KERNEL__
348 350
diff --git a/arch/sh/include/asm/unistd_64.h b/arch/sh/include/asm/unistd_64.h
index 7c54e91753c1..6eb9d2934c0f 100644
--- a/arch/sh/include/asm/unistd_64.h
+++ b/arch/sh/include/asm/unistd_64.h
@@ -381,10 +381,12 @@
381#define __NR_dup3 358 381#define __NR_dup3 358
382#define __NR_pipe2 359 382#define __NR_pipe2 359
383#define __NR_inotify_init1 360 383#define __NR_inotify_init1 360
384#define __NR_preadv 361
385#define __NR_pwritev 362
384 386
385#ifdef __KERNEL__ 387#ifdef __KERNEL__
386 388
387#define NR_syscalls 361 389#define NR_syscalls 363
388 390
389#define __ARCH_WANT_IPC_PARSE_VERSION 391#define __ARCH_WANT_IPC_PARSE_VERSION
390#define __ARCH_WANT_OLD_READDIR 392#define __ARCH_WANT_OLD_READDIR
diff --git a/arch/sh/kernel/cpu/sh4a/setup-sh7786.c b/arch/sh/kernel/cpu/sh4a/setup-sh7786.c
index 5a47e1cf442e..90e8cfff55fd 100644
--- a/arch/sh/kernel/cpu/sh4a/setup-sh7786.c
+++ b/arch/sh/kernel/cpu/sh4a/setup-sh7786.c
@@ -143,14 +143,14 @@ static void __init sh7786_usb_setup(void)
143 * Set the PHY and PLL enable bit 143 * Set the PHY and PLL enable bit
144 */ 144 */
145 __raw_writel(PHY_ENB | PLL_ENB, USBPCTL1); 145 __raw_writel(PHY_ENB | PLL_ENB, USBPCTL1);
146 while (i-- && 146 while (i--) {
147 ((__raw_readl(USBST) & ACT_PLL_STATUS) != ACT_PLL_STATUS)) 147 if (ACT_PLL_STATUS == (__raw_readl(USBST) & ACT_PLL_STATUS)) {
148 /* Set the PHY RST bit */
149 __raw_writel(PHY_ENB | PLL_ENB | PHY_RST, USBPCTL1);
150 printk(KERN_INFO "sh7786 usb setup done\n");
151 break;
152 }
148 cpu_relax(); 153 cpu_relax();
149
150 if (i) {
151 /* Set the PHY RST bit */
152 __raw_writel(PHY_ENB | PLL_ENB | PHY_RST, USBPCTL1);
153 printk(KERN_INFO "sh7786 usb setup done\n");
154 } 154 }
155} 155}
156 156
diff --git a/arch/sh/kernel/syscalls_32.S b/arch/sh/kernel/syscalls_32.S
index e67c1733e1b9..05202edd8e21 100644
--- a/arch/sh/kernel/syscalls_32.S
+++ b/arch/sh/kernel/syscalls_32.S
@@ -349,3 +349,5 @@ ENTRY(sys_call_table)
349 .long sys_dup3 /* 330 */ 349 .long sys_dup3 /* 330 */
350 .long sys_pipe2 350 .long sys_pipe2
351 .long sys_inotify_init1 351 .long sys_inotify_init1
352 .long sys_preadv
353 .long sys_writev
diff --git a/arch/sh/kernel/syscalls_64.S b/arch/sh/kernel/syscalls_64.S
index 557cb91f5caf..a083609f9284 100644
--- a/arch/sh/kernel/syscalls_64.S
+++ b/arch/sh/kernel/syscalls_64.S
@@ -387,3 +387,5 @@ sys_call_table:
387 .long sys_dup3 387 .long sys_dup3
388 .long sys_pipe2 388 .long sys_pipe2
389 .long sys_inotify_init1 /* 360 */ 389 .long sys_inotify_init1 /* 360 */
390 .long sys_preadv
391 .long sys_pwritev
diff --git a/arch/sh/mm/consistent.c b/arch/sh/mm/consistent.c
index edcd5fbf9651..e098ec158ddb 100644
--- a/arch/sh/mm/consistent.c
+++ b/arch/sh/mm/consistent.c
@@ -10,11 +10,22 @@
10 * for more details. 10 * for more details.
11 */ 11 */
12#include <linux/mm.h> 12#include <linux/mm.h>
13#include <linux/init.h>
13#include <linux/platform_device.h> 14#include <linux/platform_device.h>
14#include <linux/dma-mapping.h> 15#include <linux/dma-mapping.h>
16#include <linux/dma-debug.h>
17#include <linux/io.h>
15#include <asm/cacheflush.h> 18#include <asm/cacheflush.h>
16#include <asm/addrspace.h> 19#include <asm/addrspace.h>
17#include <asm/io.h> 20
21#define PREALLOC_DMA_DEBUG_ENTRIES 4096
22
23static int __init dma_init(void)
24{
25 dma_debug_init(PREALLOC_DMA_DEBUG_ENTRIES);
26 return 0;
27}
28fs_initcall(dma_init);
18 29
19void *dma_alloc_coherent(struct device *dev, size_t size, 30void *dma_alloc_coherent(struct device *dev, size_t size,
20 dma_addr_t *dma_handle, gfp_t gfp) 31 dma_addr_t *dma_handle, gfp_t gfp)
@@ -45,6 +56,9 @@ void *dma_alloc_coherent(struct device *dev, size_t size,
45 split_page(pfn_to_page(virt_to_phys(ret) >> PAGE_SHIFT), order); 56 split_page(pfn_to_page(virt_to_phys(ret) >> PAGE_SHIFT), order);
46 57
47 *dma_handle = virt_to_phys(ret); 58 *dma_handle = virt_to_phys(ret);
59
60 debug_dma_alloc_coherent(dev, size, *dma_handle, ret_nocache);
61
48 return ret_nocache; 62 return ret_nocache;
49} 63}
50EXPORT_SYMBOL(dma_alloc_coherent); 64EXPORT_SYMBOL(dma_alloc_coherent);
@@ -56,12 +70,15 @@ void dma_free_coherent(struct device *dev, size_t size,
56 unsigned long pfn = dma_handle >> PAGE_SHIFT; 70 unsigned long pfn = dma_handle >> PAGE_SHIFT;
57 int k; 71 int k;
58 72
59 if (!dma_release_from_coherent(dev, order, vaddr)) { 73 WARN_ON(irqs_disabled()); /* for portability */
60 WARN_ON(irqs_disabled()); /* for portability */ 74
61 for (k = 0; k < (1 << order); k++) 75 if (dma_release_from_coherent(dev, order, vaddr))
62 __free_pages(pfn_to_page(pfn + k), 0); 76 return;
63 iounmap(vaddr); 77
64 } 78 debug_dma_free_coherent(dev, size, vaddr, dma_handle);
79 for (k = 0; k < (1 << order); k++)
80 __free_pages(pfn_to_page(pfn + k), 0);
81 iounmap(vaddr);
65} 82}
66EXPORT_SYMBOL(dma_free_coherent); 83EXPORT_SYMBOL(dma_free_coherent);
67 84
diff --git a/arch/sparc/include/asm/atomic_32.h b/arch/sparc/include/asm/atomic_32.h
index ce465975a6a5..bb91b1248cd1 100644
--- a/arch/sparc/include/asm/atomic_32.h
+++ b/arch/sparc/include/asm/atomic_32.h
@@ -15,6 +15,8 @@
15 15
16#ifdef __KERNEL__ 16#ifdef __KERNEL__
17 17
18#include <asm/system.h>
19
18#define ATOMIC_INIT(i) { (i) } 20#define ATOMIC_INIT(i) { (i) }
19 21
20extern int __atomic_add_return(int, atomic_t *); 22extern int __atomic_add_return(int, atomic_t *);
diff --git a/arch/sparc/include/asm/parport.h b/arch/sparc/include/asm/parport.h
index dff3f0253aa8..ff9ead640c4a 100644
--- a/arch/sparc/include/asm/parport.h
+++ b/arch/sparc/include/asm/parport.h
@@ -117,7 +117,7 @@ static int __devinit ecpp_probe(struct of_device *op, const struct of_device_id
117 if (!strcmp(parent->name, "dma")) { 117 if (!strcmp(parent->name, "dma")) {
118 p = parport_pc_probe_port(base, base + 0x400, 118 p = parport_pc_probe_port(base, base + 0x400,
119 op->irqs[0], PARPORT_DMA_NOFIFO, 119 op->irqs[0], PARPORT_DMA_NOFIFO,
120 op->dev.parent->parent); 120 op->dev.parent->parent, 0);
121 if (!p) 121 if (!p)
122 return -ENOMEM; 122 return -ENOMEM;
123 dev_set_drvdata(&op->dev, p); 123 dev_set_drvdata(&op->dev, p);
@@ -168,7 +168,8 @@ static int __devinit ecpp_probe(struct of_device *op, const struct of_device_id
168 p = parport_pc_probe_port(base, base + 0x400, 168 p = parport_pc_probe_port(base, base + 0x400,
169 op->irqs[0], 169 op->irqs[0],
170 slot, 170 slot,
171 op->dev.parent); 171 op->dev.parent,
172 0);
172 err = -ENOMEM; 173 err = -ENOMEM;
173 if (!p) 174 if (!p)
174 goto out_disable_irq; 175 goto out_disable_irq;
diff --git a/arch/sparc/kernel/ldc.c b/arch/sparc/kernel/ldc.c
index 6ce5d2598a09..adf5f273868a 100644
--- a/arch/sparc/kernel/ldc.c
+++ b/arch/sparc/kernel/ldc.c
@@ -1183,8 +1183,7 @@ out_free_txq:
1183 free_queue(lp->tx_num_entries, lp->tx_base); 1183 free_queue(lp->tx_num_entries, lp->tx_base);
1184 1184
1185out_free_mssbuf: 1185out_free_mssbuf:
1186 if (mssbuf) 1186 kfree(mssbuf);
1187 kfree(mssbuf);
1188 1187
1189out_free_iommu: 1188out_free_iommu:
1190 ldc_iommu_release(lp); 1189 ldc_iommu_release(lp);
@@ -1217,8 +1216,7 @@ void ldc_free(struct ldc_channel *lp)
1217 1216
1218 hlist_del(&lp->list); 1217 hlist_del(&lp->list);
1219 1218
1220 if (lp->mssbuf) 1219 kfree(lp->mssbuf);
1221 kfree(lp->mssbuf);
1222 1220
1223 ldc_iommu_release(lp); 1221 ldc_iommu_release(lp);
1224 1222
diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c
index 708e12a26b05..f7642e5a94db 100644
--- a/arch/sparc/kernel/smp_64.c
+++ b/arch/sparc/kernel/smp_64.c
@@ -118,9 +118,9 @@ void __cpuinit smp_callin(void)
118 while (!cpu_isset(cpuid, smp_commenced_mask)) 118 while (!cpu_isset(cpuid, smp_commenced_mask))
119 rmb(); 119 rmb();
120 120
121 ipi_call_lock(); 121 ipi_call_lock_irq();
122 cpu_set(cpuid, cpu_online_map); 122 cpu_set(cpuid, cpu_online_map);
123 ipi_call_unlock(); 123 ipi_call_unlock_irq();
124 124
125 /* idle thread is expected to have preempt disabled */ 125 /* idle thread is expected to have preempt disabled */
126 preempt_disable(); 126 preempt_disable();
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
index bc25b9f5e4cd..c9086e6307a5 100644
--- a/arch/x86/Kconfig
+++ b/arch/x86/Kconfig
@@ -353,6 +353,7 @@ config X86_UV
353 bool "SGI Ultraviolet" 353 bool "SGI Ultraviolet"
354 depends on X86_64 354 depends on X86_64
355 depends on X86_EXTENDED_PLATFORM 355 depends on X86_EXTENDED_PLATFORM
356 depends on NUMA
356 select X86_X2APIC 357 select X86_X2APIC
357 ---help--- 358 ---help---
358 This option is needed in order to support SGI Ultraviolet systems. 359 This option is needed in order to support SGI Ultraviolet systems.
diff --git a/arch/x86/include/asm/required-features.h b/arch/x86/include/asm/required-features.h
index d5cd6c586881..a4737dddfd58 100644
--- a/arch/x86/include/asm/required-features.h
+++ b/arch/x86/include/asm/required-features.h
@@ -50,7 +50,7 @@
50#ifdef CONFIG_X86_64 50#ifdef CONFIG_X86_64
51#define NEED_PSE 0 51#define NEED_PSE 0
52#define NEED_MSR (1<<(X86_FEATURE_MSR & 31)) 52#define NEED_MSR (1<<(X86_FEATURE_MSR & 31))
53#define NEED_PGE (1<<(X86_FEATURE_PGE & 31)) 53#define NEED_PGE 0
54#define NEED_FXSR (1<<(X86_FEATURE_FXSR & 31)) 54#define NEED_FXSR (1<<(X86_FEATURE_FXSR & 31))
55#define NEED_XMM (1<<(X86_FEATURE_XMM & 31)) 55#define NEED_XMM (1<<(X86_FEATURE_XMM & 31))
56#define NEED_XMM2 (1<<(X86_FEATURE_XMM2 & 31)) 56#define NEED_XMM2 (1<<(X86_FEATURE_XMM2 & 31))
diff --git a/arch/x86/include/asm/uv/uv_mmrs.h b/arch/x86/include/asm/uv/uv_mmrs.h
index db68ac8a5ac2..2cae46c7c8a2 100644
--- a/arch/x86/include/asm/uv/uv_mmrs.h
+++ b/arch/x86/include/asm/uv/uv_mmrs.h
@@ -17,6 +17,11 @@
17/* ========================================================================= */ 17/* ========================================================================= */
18/* UVH_BAU_DATA_CONFIG */ 18/* UVH_BAU_DATA_CONFIG */
19/* ========================================================================= */ 19/* ========================================================================= */
20#define UVH_LB_BAU_MISC_CONTROL 0x320170UL
21#define UV_ENABLE_INTD_SOFT_ACK_MODE_SHIFT 15
22#define UV_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHIFT 16
23#define UV_INTD_SOFT_ACK_TIMEOUT_PERIOD 0x000000000bUL
24/* 1011 timebase 7 (168millisec) * 3 ticks -> 500ms */
20#define UVH_BAU_DATA_CONFIG 0x61680UL 25#define UVH_BAU_DATA_CONFIG 0x61680UL
21#define UVH_BAU_DATA_CONFIG_32 0x0438 26#define UVH_BAU_DATA_CONFIG_32 0x0438
22 27
diff --git a/arch/x86/include/asm/xen/page.h b/arch/x86/include/asm/xen/page.h
index 1a918dde46b5..018a0a400799 100644
--- a/arch/x86/include/asm/xen/page.h
+++ b/arch/x86/include/asm/xen/page.h
@@ -124,7 +124,8 @@ static inline unsigned long mfn_to_local_pfn(unsigned long mfn)
124 124
125/* VIRT <-> MACHINE conversion */ 125/* VIRT <-> MACHINE conversion */
126#define virt_to_machine(v) (phys_to_machine(XPADDR(__pa(v)))) 126#define virt_to_machine(v) (phys_to_machine(XPADDR(__pa(v))))
127#define virt_to_mfn(v) (pfn_to_mfn(PFN_DOWN(__pa(v)))) 127#define virt_to_pfn(v) (PFN_DOWN(__pa(v)))
128#define virt_to_mfn(v) (pfn_to_mfn(virt_to_pfn(v)))
128#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT)) 129#define mfn_to_virt(m) (__va(mfn_to_pfn(m) << PAGE_SHIFT))
129 130
130static inline unsigned long pte_mfn(pte_t pte) 131static inline unsigned long pte_mfn(pte_t pte)
diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c
index 1248318436e8..de1a50af807b 100644
--- a/arch/x86/kernel/apic/x2apic_uv_x.c
+++ b/arch/x86/kernel/apic/x2apic_uv_x.c
@@ -549,7 +549,8 @@ void __init uv_system_init(void)
549 unsigned long gnode_upper, lowmem_redir_base, lowmem_redir_size; 549 unsigned long gnode_upper, lowmem_redir_base, lowmem_redir_size;
550 int bytes, nid, cpu, lcpu, pnode, blade, i, j, m_val, n_val; 550 int bytes, nid, cpu, lcpu, pnode, blade, i, j, m_val, n_val;
551 int max_pnode = 0; 551 int max_pnode = 0;
552 unsigned long mmr_base, present; 552 unsigned long mmr_base, present, paddr;
553 unsigned short pnode_mask;
553 554
554 map_low_mmrs(); 555 map_low_mmrs();
555 556
@@ -592,6 +593,7 @@ void __init uv_system_init(void)
592 } 593 }
593 } 594 }
594 595
596 pnode_mask = (1 << n_val) - 1;
595 node_id.v = uv_read_local_mmr(UVH_NODE_ID); 597 node_id.v = uv_read_local_mmr(UVH_NODE_ID);
596 gnode_upper = (((unsigned long)node_id.s.node_id) & 598 gnode_upper = (((unsigned long)node_id.s.node_id) &
597 ~((1 << n_val) - 1)) << m_val; 599 ~((1 << n_val) - 1)) << m_val;
@@ -615,7 +617,7 @@ void __init uv_system_init(void)
615 uv_cpu_hub_info(cpu)->numa_blade_id = blade; 617 uv_cpu_hub_info(cpu)->numa_blade_id = blade;
616 uv_cpu_hub_info(cpu)->blade_processor_id = lcpu; 618 uv_cpu_hub_info(cpu)->blade_processor_id = lcpu;
617 uv_cpu_hub_info(cpu)->pnode = pnode; 619 uv_cpu_hub_info(cpu)->pnode = pnode;
618 uv_cpu_hub_info(cpu)->pnode_mask = (1 << n_val) - 1; 620 uv_cpu_hub_info(cpu)->pnode_mask = pnode_mask;
619 uv_cpu_hub_info(cpu)->gpa_mask = (1 << (m_val + n_val)) - 1; 621 uv_cpu_hub_info(cpu)->gpa_mask = (1 << (m_val + n_val)) - 1;
620 uv_cpu_hub_info(cpu)->gnode_upper = gnode_upper; 622 uv_cpu_hub_info(cpu)->gnode_upper = gnode_upper;
621 uv_cpu_hub_info(cpu)->global_mmr_base = mmr_base; 623 uv_cpu_hub_info(cpu)->global_mmr_base = mmr_base;
@@ -631,6 +633,16 @@ void __init uv_system_init(void)
631 lcpu, blade); 633 lcpu, blade);
632 } 634 }
633 635
636 /* Add blade/pnode info for nodes without cpus */
637 for_each_online_node(nid) {
638 if (uv_node_to_blade[nid] >= 0)
639 continue;
640 paddr = node_start_pfn(nid) << PAGE_SHIFT;
641 pnode = (paddr >> m_val) & pnode_mask;
642 blade = boot_pnode_to_blade(pnode);
643 uv_node_to_blade[nid] = blade;
644 }
645
634 map_gru_high(max_pnode); 646 map_gru_high(max_pnode);
635 map_mmr_high(max_pnode); 647 map_mmr_high(max_pnode);
636 map_config_high(max_pnode); 648 map_config_high(max_pnode);
diff --git a/arch/x86/kernel/bios_uv.c b/arch/x86/kernel/bios_uv.c
index f63882728d91..63a88e1f987d 100644
--- a/arch/x86/kernel/bios_uv.c
+++ b/arch/x86/kernel/bios_uv.c
@@ -182,7 +182,8 @@ void uv_bios_init(void)
182 memcpy(&uv_systab, tab, sizeof(struct uv_systab)); 182 memcpy(&uv_systab, tab, sizeof(struct uv_systab));
183 iounmap(tab); 183 iounmap(tab);
184 184
185 printk(KERN_INFO "EFI UV System Table Revision %d\n", tab->revision); 185 printk(KERN_INFO "EFI UV System Table Revision %d\n",
186 uv_systab.revision);
186} 187}
187#else /* !CONFIG_EFI */ 188#else /* !CONFIG_EFI */
188 189
diff --git a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
index 3e3cd3db7a0c..ecdb682ab516 100644
--- a/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
+++ b/arch/x86/kernel/cpu/cpufreq/acpi-cpufreq.c
@@ -204,7 +204,13 @@ static void drv_read(struct drv_cmd *cmd)
204 204
205static void drv_write(struct drv_cmd *cmd) 205static void drv_write(struct drv_cmd *cmd)
206{ 206{
207 int this_cpu;
208
209 this_cpu = get_cpu();
210 if (cpumask_test_cpu(this_cpu, cmd->mask))
211 do_drv_write(cmd);
207 smp_call_function_many(cmd->mask, do_drv_write, cmd, 1); 212 smp_call_function_many(cmd->mask, do_drv_write, cmd, 1);
213 put_cpu();
208} 214}
209 215
210static u32 get_cur_val(const struct cpumask *mask) 216static u32 get_cur_val(const struct cpumask *mask)
@@ -277,7 +283,7 @@ static unsigned int get_measured_perf(struct cpufreq_policy *policy,
277 unsigned int perf_percent; 283 unsigned int perf_percent;
278 unsigned int retval; 284 unsigned int retval;
279 285
280 if (smp_call_function_single(cpu, read_measured_perf_ctrs, &cur, 1)) 286 if (smp_call_function_single(cpu, read_measured_perf_ctrs, &readin, 1))
281 return 0; 287 return 0;
282 288
283 cur.aperf.whole = readin.aperf.whole - 289 cur.aperf.whole = readin.aperf.whole -
diff --git a/arch/x86/kernel/microcode_core.c b/arch/x86/kernel/microcode_core.c
index a0f3851ef310..2e0eb4140951 100644
--- a/arch/x86/kernel/microcode_core.c
+++ b/arch/x86/kernel/microcode_core.c
@@ -108,40 +108,29 @@ struct ucode_cpu_info ucode_cpu_info[NR_CPUS];
108EXPORT_SYMBOL_GPL(ucode_cpu_info); 108EXPORT_SYMBOL_GPL(ucode_cpu_info);
109 109
110#ifdef CONFIG_MICROCODE_OLD_INTERFACE 110#ifdef CONFIG_MICROCODE_OLD_INTERFACE
111struct update_for_cpu {
112 const void __user *buf;
113 size_t size;
114};
115
116static long update_for_cpu(void *_ufc)
117{
118 struct update_for_cpu *ufc = _ufc;
119 int error;
120
121 error = microcode_ops->request_microcode_user(smp_processor_id(),
122 ufc->buf, ufc->size);
123 if (error < 0)
124 return error;
125 if (!error)
126 microcode_ops->apply_microcode(smp_processor_id());
127 return error;
128}
129
130static int do_microcode_update(const void __user *buf, size_t size) 111static int do_microcode_update(const void __user *buf, size_t size)
131{ 112{
113 cpumask_t old;
132 int error = 0; 114 int error = 0;
133 int cpu; 115 int cpu;
134 struct update_for_cpu ufc = { .buf = buf, .size = size }; 116
117 old = current->cpus_allowed;
135 118
136 for_each_online_cpu(cpu) { 119 for_each_online_cpu(cpu) {
137 struct ucode_cpu_info *uci = ucode_cpu_info + cpu; 120 struct ucode_cpu_info *uci = ucode_cpu_info + cpu;
138 121
139 if (!uci->valid) 122 if (!uci->valid)
140 continue; 123 continue;
141 error = work_on_cpu(cpu, update_for_cpu, &ufc); 124
125 set_cpus_allowed_ptr(current, &cpumask_of_cpu(cpu));
126 error = microcode_ops->request_microcode_user(cpu, buf, size);
142 if (error < 0) 127 if (error < 0)
143 break; 128 goto out;
129 if (!error)
130 microcode_ops->apply_microcode(cpu);
144 } 131 }
132out:
133 set_cpus_allowed_ptr(current, &old);
145 return error; 134 return error;
146} 135}
147 136
diff --git a/arch/x86/kernel/tlb_uv.c b/arch/x86/kernel/tlb_uv.c
index deb5ebb32c3b..ed0c33761e6d 100644
--- a/arch/x86/kernel/tlb_uv.c
+++ b/arch/x86/kernel/tlb_uv.c
@@ -25,6 +25,8 @@ static int uv_bau_retry_limit __read_mostly;
25 25
26/* position of pnode (which is nasid>>1): */ 26/* position of pnode (which is nasid>>1): */
27static int uv_nshift __read_mostly; 27static int uv_nshift __read_mostly;
28/* base pnode in this partition */
29static int uv_partition_base_pnode __read_mostly;
28 30
29static unsigned long uv_mmask __read_mostly; 31static unsigned long uv_mmask __read_mostly;
30 32
@@ -32,6 +34,34 @@ static DEFINE_PER_CPU(struct ptc_stats, ptcstats);
32static DEFINE_PER_CPU(struct bau_control, bau_control); 34static DEFINE_PER_CPU(struct bau_control, bau_control);
33 35
34/* 36/*
37 * Determine the first node on a blade.
38 */
39static int __init blade_to_first_node(int blade)
40{
41 int node, b;
42
43 for_each_online_node(node) {
44 b = uv_node_to_blade_id(node);
45 if (blade == b)
46 return node;
47 }
48 return -1; /* shouldn't happen */
49}
50
51/*
52 * Determine the apicid of the first cpu on a blade.
53 */
54static int __init blade_to_first_apicid(int blade)
55{
56 int cpu;
57
58 for_each_present_cpu(cpu)
59 if (blade == uv_cpu_to_blade_id(cpu))
60 return per_cpu(x86_cpu_to_apicid, cpu);
61 return -1;
62}
63
64/*
35 * Free a software acknowledge hardware resource by clearing its Pending 65 * Free a software acknowledge hardware resource by clearing its Pending
36 * bit. This will return a reply to the sender. 66 * bit. This will return a reply to the sender.
37 * If the message has timed out, a reply has already been sent by the 67 * If the message has timed out, a reply has already been sent by the
@@ -67,7 +97,7 @@ static void uv_bau_process_message(struct bau_payload_queue_entry *msg,
67 msp = __get_cpu_var(bau_control).msg_statuses + msg_slot; 97 msp = __get_cpu_var(bau_control).msg_statuses + msg_slot;
68 cpu = uv_blade_processor_id(); 98 cpu = uv_blade_processor_id();
69 msg->number_of_cpus = 99 msg->number_of_cpus =
70 uv_blade_nr_online_cpus(uv_node_to_blade_id(numa_node_id())); 100 uv_blade_nr_online_cpus(uv_node_to_blade_id(numa_node_id()));
71 this_cpu_mask = 1UL << cpu; 101 this_cpu_mask = 1UL << cpu;
72 if (msp->seen_by.bits & this_cpu_mask) 102 if (msp->seen_by.bits & this_cpu_mask)
73 return; 103 return;
@@ -215,14 +245,14 @@ static int uv_wait_completion(struct bau_desc *bau_desc,
215 * Returns @flush_mask if some remote flushing remains to be done. The 245 * Returns @flush_mask if some remote flushing remains to be done. The
216 * mask will have some bits still set. 246 * mask will have some bits still set.
217 */ 247 */
218const struct cpumask *uv_flush_send_and_wait(int cpu, int this_blade, 248const struct cpumask *uv_flush_send_and_wait(int cpu, int this_pnode,
219 struct bau_desc *bau_desc, 249 struct bau_desc *bau_desc,
220 struct cpumask *flush_mask) 250 struct cpumask *flush_mask)
221{ 251{
222 int completion_status = 0; 252 int completion_status = 0;
223 int right_shift; 253 int right_shift;
224 int tries = 0; 254 int tries = 0;
225 int blade; 255 int pnode;
226 int bit; 256 int bit;
227 unsigned long mmr_offset; 257 unsigned long mmr_offset;
228 unsigned long index; 258 unsigned long index;
@@ -265,8 +295,8 @@ const struct cpumask *uv_flush_send_and_wait(int cpu, int this_blade,
265 * use the IPI method of shootdown on them. 295 * use the IPI method of shootdown on them.
266 */ 296 */
267 for_each_cpu(bit, flush_mask) { 297 for_each_cpu(bit, flush_mask) {
268 blade = uv_cpu_to_blade_id(bit); 298 pnode = uv_cpu_to_pnode(bit);
269 if (blade == this_blade) 299 if (pnode == this_pnode)
270 continue; 300 continue;
271 cpumask_clear_cpu(bit, flush_mask); 301 cpumask_clear_cpu(bit, flush_mask);
272 } 302 }
@@ -309,16 +339,16 @@ const struct cpumask *uv_flush_tlb_others(const struct cpumask *cpumask,
309 struct cpumask *flush_mask = __get_cpu_var(uv_flush_tlb_mask); 339 struct cpumask *flush_mask = __get_cpu_var(uv_flush_tlb_mask);
310 int i; 340 int i;
311 int bit; 341 int bit;
312 int blade; 342 int pnode;
313 int uv_cpu; 343 int uv_cpu;
314 int this_blade; 344 int this_pnode;
315 int locals = 0; 345 int locals = 0;
316 struct bau_desc *bau_desc; 346 struct bau_desc *bau_desc;
317 347
318 cpumask_andnot(flush_mask, cpumask, cpumask_of(cpu)); 348 cpumask_andnot(flush_mask, cpumask, cpumask_of(cpu));
319 349
320 uv_cpu = uv_blade_processor_id(); 350 uv_cpu = uv_blade_processor_id();
321 this_blade = uv_numa_blade_id(); 351 this_pnode = uv_hub_info->pnode;
322 bau_desc = __get_cpu_var(bau_control).descriptor_base; 352 bau_desc = __get_cpu_var(bau_control).descriptor_base;
323 bau_desc += UV_ITEMS_PER_DESCRIPTOR * uv_cpu; 353 bau_desc += UV_ITEMS_PER_DESCRIPTOR * uv_cpu;
324 354
@@ -326,13 +356,14 @@ const struct cpumask *uv_flush_tlb_others(const struct cpumask *cpumask,
326 356
327 i = 0; 357 i = 0;
328 for_each_cpu(bit, flush_mask) { 358 for_each_cpu(bit, flush_mask) {
329 blade = uv_cpu_to_blade_id(bit); 359 pnode = uv_cpu_to_pnode(bit);
330 BUG_ON(blade > (UV_DISTRIBUTION_SIZE - 1)); 360 BUG_ON(pnode > (UV_DISTRIBUTION_SIZE - 1));
331 if (blade == this_blade) { 361 if (pnode == this_pnode) {
332 locals++; 362 locals++;
333 continue; 363 continue;
334 } 364 }
335 bau_node_set(blade, &bau_desc->distribution); 365 bau_node_set(pnode - uv_partition_base_pnode,
366 &bau_desc->distribution);
336 i++; 367 i++;
337 } 368 }
338 if (i == 0) { 369 if (i == 0) {
@@ -350,7 +381,7 @@ const struct cpumask *uv_flush_tlb_others(const struct cpumask *cpumask,
350 bau_desc->payload.address = va; 381 bau_desc->payload.address = va;
351 bau_desc->payload.sending_cpu = cpu; 382 bau_desc->payload.sending_cpu = cpu;
352 383
353 return uv_flush_send_and_wait(uv_cpu, this_blade, bau_desc, flush_mask); 384 return uv_flush_send_and_wait(uv_cpu, this_pnode, bau_desc, flush_mask);
354} 385}
355 386
356/* 387/*
@@ -418,24 +449,58 @@ void uv_bau_message_interrupt(struct pt_regs *regs)
418 set_irq_regs(old_regs); 449 set_irq_regs(old_regs);
419} 450}
420 451
452/*
453 * uv_enable_timeouts
454 *
455 * Each target blade (i.e. blades that have cpu's) needs to have
456 * shootdown message timeouts enabled. The timeout does not cause
457 * an interrupt, but causes an error message to be returned to
458 * the sender.
459 */
421static void uv_enable_timeouts(void) 460static void uv_enable_timeouts(void)
422{ 461{
423 int i;
424 int blade; 462 int blade;
425 int last_blade; 463 int nblades;
426 int pnode; 464 int pnode;
427 int cur_cpu = 0; 465 unsigned long mmr_image;
428 unsigned long apicid;
429 466
430 last_blade = -1; 467 nblades = uv_num_possible_blades();
431 for_each_online_node(i) { 468
432 blade = uv_node_to_blade_id(i); 469 for (blade = 0; blade < nblades; blade++) {
433 if (blade == last_blade) 470 if (!uv_blade_nr_possible_cpus(blade))
434 continue; 471 continue;
435 last_blade = blade; 472
436 apicid = per_cpu(x86_cpu_to_apicid, cur_cpu);
437 pnode = uv_blade_to_pnode(blade); 473 pnode = uv_blade_to_pnode(blade);
438 cur_cpu += uv_blade_nr_possible_cpus(i); 474 mmr_image =
475 uv_read_global_mmr64(pnode, UVH_LB_BAU_MISC_CONTROL);
476 /*
477 * Set the timeout period and then lock it in, in three
478 * steps; captures and locks in the period.
479 *
480 * To program the period, the SOFT_ACK_MODE must be off.
481 */
482 mmr_image &= ~((unsigned long)1 <<
483 UV_ENABLE_INTD_SOFT_ACK_MODE_SHIFT);
484 uv_write_global_mmr64
485 (pnode, UVH_LB_BAU_MISC_CONTROL, mmr_image);
486 /*
487 * Set the 4-bit period.
488 */
489 mmr_image &= ~((unsigned long)0xf <<
490 UV_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHIFT);
491 mmr_image |= (UV_INTD_SOFT_ACK_TIMEOUT_PERIOD <<
492 UV_INTD_SOFT_ACK_TIMEOUT_PERIOD_SHIFT);
493 uv_write_global_mmr64
494 (pnode, UVH_LB_BAU_MISC_CONTROL, mmr_image);
495 /*
496 * Subsequent reversals of the timebase bit (3) cause an
497 * immediate timeout of one or all INTD resources as
498 * indicated in bits 2:0 (7 causes all of them to timeout).
499 */
500 mmr_image |= ((unsigned long)1 <<
501 UV_ENABLE_INTD_SOFT_ACK_MODE_SHIFT);
502 uv_write_global_mmr64
503 (pnode, UVH_LB_BAU_MISC_CONTROL, mmr_image);
439 } 504 }
440} 505}
441 506
@@ -482,8 +547,7 @@ static int uv_ptc_seq_show(struct seq_file *file, void *data)
482 stat->requestee, stat->onetlb, stat->alltlb, 547 stat->requestee, stat->onetlb, stat->alltlb,
483 stat->s_retry, stat->d_retry, stat->ptc_i); 548 stat->s_retry, stat->d_retry, stat->ptc_i);
484 seq_printf(file, "%lx %ld %ld %ld %ld %ld %ld\n", 549 seq_printf(file, "%lx %ld %ld %ld %ld %ld %ld\n",
485 uv_read_global_mmr64(uv_blade_to_pnode 550 uv_read_global_mmr64(uv_cpu_to_pnode(cpu),
486 (uv_cpu_to_blade_id(cpu)),
487 UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE), 551 UVH_LB_BAU_INTD_SOFTWARE_ACKNOWLEDGE),
488 stat->sflush, stat->dflush, 552 stat->sflush, stat->dflush,
489 stat->retriesok, stat->nomsg, 553 stat->retriesok, stat->nomsg,
@@ -617,16 +681,18 @@ static struct bau_control * __init uv_table_bases_init(int blade, int node)
617 * finish the initialization of the per-blade control structures 681 * finish the initialization of the per-blade control structures
618 */ 682 */
619static void __init 683static void __init
620uv_table_bases_finish(int blade, int node, int cur_cpu, 684uv_table_bases_finish(int blade,
621 struct bau_control *bau_tablesp, 685 struct bau_control *bau_tablesp,
622 struct bau_desc *adp) 686 struct bau_desc *adp)
623{ 687{
624 struct bau_control *bcp; 688 struct bau_control *bcp;
625 int i; 689 int cpu;
626 690
627 for (i = cur_cpu; i < cur_cpu + uv_blade_nr_possible_cpus(blade); i++) { 691 for_each_present_cpu(cpu) {
628 bcp = (struct bau_control *)&per_cpu(bau_control, i); 692 if (blade != uv_cpu_to_blade_id(cpu))
693 continue;
629 694
695 bcp = (struct bau_control *)&per_cpu(bau_control, cpu);
630 bcp->bau_msg_head = bau_tablesp->va_queue_first; 696 bcp->bau_msg_head = bau_tablesp->va_queue_first;
631 bcp->va_queue_first = bau_tablesp->va_queue_first; 697 bcp->va_queue_first = bau_tablesp->va_queue_first;
632 bcp->va_queue_last = bau_tablesp->va_queue_last; 698 bcp->va_queue_last = bau_tablesp->va_queue_last;
@@ -649,11 +715,10 @@ uv_activation_descriptor_init(int node, int pnode)
649 struct bau_desc *adp; 715 struct bau_desc *adp;
650 struct bau_desc *ad2; 716 struct bau_desc *ad2;
651 717
652 adp = (struct bau_desc *) 718 adp = (struct bau_desc *)kmalloc_node(16384, GFP_KERNEL, node);
653 kmalloc_node(16384, GFP_KERNEL, node);
654 BUG_ON(!adp); 719 BUG_ON(!adp);
655 720
656 pa = __pa((unsigned long)adp); 721 pa = uv_gpa(adp); /* need the real nasid*/
657 n = pa >> uv_nshift; 722 n = pa >> uv_nshift;
658 m = pa & uv_mmask; 723 m = pa & uv_mmask;
659 724
@@ -667,8 +732,12 @@ uv_activation_descriptor_init(int node, int pnode)
667 for (i = 0, ad2 = adp; i < UV_ACTIVATION_DESCRIPTOR_SIZE; i++, ad2++) { 732 for (i = 0, ad2 = adp; i < UV_ACTIVATION_DESCRIPTOR_SIZE; i++, ad2++) {
668 memset(ad2, 0, sizeof(struct bau_desc)); 733 memset(ad2, 0, sizeof(struct bau_desc));
669 ad2->header.sw_ack_flag = 1; 734 ad2->header.sw_ack_flag = 1;
670 ad2->header.base_dest_nodeid = 735 /*
671 uv_blade_to_pnode(uv_cpu_to_blade_id(0)); 736 * base_dest_nodeid is the first node in the partition, so
737 * the bit map will indicate partition-relative node numbers.
738 * note that base_dest_nodeid is actually a nasid.
739 */
740 ad2->header.base_dest_nodeid = uv_partition_base_pnode << 1;
672 ad2->header.command = UV_NET_ENDPOINT_INTD; 741 ad2->header.command = UV_NET_ENDPOINT_INTD;
673 ad2->header.int_both = 1; 742 ad2->header.int_both = 1;
674 /* 743 /*
@@ -686,6 +755,8 @@ static struct bau_payload_queue_entry * __init
686uv_payload_queue_init(int node, int pnode, struct bau_control *bau_tablesp) 755uv_payload_queue_init(int node, int pnode, struct bau_control *bau_tablesp)
687{ 756{
688 struct bau_payload_queue_entry *pqp; 757 struct bau_payload_queue_entry *pqp;
758 unsigned long pa;
759 int pn;
689 char *cp; 760 char *cp;
690 761
691 pqp = (struct bau_payload_queue_entry *) kmalloc_node( 762 pqp = (struct bau_payload_queue_entry *) kmalloc_node(
@@ -696,10 +767,14 @@ uv_payload_queue_init(int node, int pnode, struct bau_control *bau_tablesp)
696 cp = (char *)pqp + 31; 767 cp = (char *)pqp + 31;
697 pqp = (struct bau_payload_queue_entry *)(((unsigned long)cp >> 5) << 5); 768 pqp = (struct bau_payload_queue_entry *)(((unsigned long)cp >> 5) << 5);
698 bau_tablesp->va_queue_first = pqp; 769 bau_tablesp->va_queue_first = pqp;
770 /*
771 * need the pnode of where the memory was really allocated
772 */
773 pa = uv_gpa(pqp);
774 pn = pa >> uv_nshift;
699 uv_write_global_mmr64(pnode, 775 uv_write_global_mmr64(pnode,
700 UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST, 776 UVH_LB_BAU_INTD_PAYLOAD_QUEUE_FIRST,
701 ((unsigned long)pnode << 777 ((unsigned long)pn << UV_PAYLOADQ_PNODE_SHIFT) |
702 UV_PAYLOADQ_PNODE_SHIFT) |
703 uv_physnodeaddr(pqp)); 778 uv_physnodeaddr(pqp));
704 uv_write_global_mmr64(pnode, UVH_LB_BAU_INTD_PAYLOAD_QUEUE_TAIL, 779 uv_write_global_mmr64(pnode, UVH_LB_BAU_INTD_PAYLOAD_QUEUE_TAIL,
705 uv_physnodeaddr(pqp)); 780 uv_physnodeaddr(pqp));
@@ -715,8 +790,9 @@ uv_payload_queue_init(int node, int pnode, struct bau_control *bau_tablesp)
715/* 790/*
716 * Initialization of each UV blade's structures 791 * Initialization of each UV blade's structures
717 */ 792 */
718static int __init uv_init_blade(int blade, int node, int cur_cpu) 793static int __init uv_init_blade(int blade)
719{ 794{
795 int node;
720 int pnode; 796 int pnode;
721 unsigned long pa; 797 unsigned long pa;
722 unsigned long apicid; 798 unsigned long apicid;
@@ -724,16 +800,17 @@ static int __init uv_init_blade(int blade, int node, int cur_cpu)
724 struct bau_payload_queue_entry *pqp; 800 struct bau_payload_queue_entry *pqp;
725 struct bau_control *bau_tablesp; 801 struct bau_control *bau_tablesp;
726 802
803 node = blade_to_first_node(blade);
727 bau_tablesp = uv_table_bases_init(blade, node); 804 bau_tablesp = uv_table_bases_init(blade, node);
728 pnode = uv_blade_to_pnode(blade); 805 pnode = uv_blade_to_pnode(blade);
729 adp = uv_activation_descriptor_init(node, pnode); 806 adp = uv_activation_descriptor_init(node, pnode);
730 pqp = uv_payload_queue_init(node, pnode, bau_tablesp); 807 pqp = uv_payload_queue_init(node, pnode, bau_tablesp);
731 uv_table_bases_finish(blade, node, cur_cpu, bau_tablesp, adp); 808 uv_table_bases_finish(blade, bau_tablesp, adp);
732 /* 809 /*
733 * the below initialization can't be in firmware because the 810 * the below initialization can't be in firmware because the
734 * messaging IRQ will be determined by the OS 811 * messaging IRQ will be determined by the OS
735 */ 812 */
736 apicid = per_cpu(x86_cpu_to_apicid, cur_cpu); 813 apicid = blade_to_first_apicid(blade);
737 pa = uv_read_global_mmr64(pnode, UVH_BAU_DATA_CONFIG); 814 pa = uv_read_global_mmr64(pnode, UVH_BAU_DATA_CONFIG);
738 if ((pa & 0xff) != UV_BAU_MESSAGE) { 815 if ((pa & 0xff) != UV_BAU_MESSAGE) {
739 uv_write_global_mmr64(pnode, UVH_BAU_DATA_CONFIG, 816 uv_write_global_mmr64(pnode, UVH_BAU_DATA_CONFIG,
@@ -748,9 +825,7 @@ static int __init uv_init_blade(int blade, int node, int cur_cpu)
748static int __init uv_bau_init(void) 825static int __init uv_bau_init(void)
749{ 826{
750 int blade; 827 int blade;
751 int node;
752 int nblades; 828 int nblades;
753 int last_blade;
754 int cur_cpu; 829 int cur_cpu;
755 830
756 if (!is_uv_system()) 831 if (!is_uv_system())
@@ -763,29 +838,21 @@ static int __init uv_bau_init(void)
763 uv_bau_retry_limit = 1; 838 uv_bau_retry_limit = 1;
764 uv_nshift = uv_hub_info->n_val; 839 uv_nshift = uv_hub_info->n_val;
765 uv_mmask = (1UL << uv_hub_info->n_val) - 1; 840 uv_mmask = (1UL << uv_hub_info->n_val) - 1;
766 nblades = 0; 841 nblades = uv_num_possible_blades();
767 last_blade = -1; 842
768 cur_cpu = 0;
769 for_each_online_node(node) {
770 blade = uv_node_to_blade_id(node);
771 if (blade == last_blade)
772 continue;
773 last_blade = blade;
774 nblades++;
775 }
776 uv_bau_table_bases = (struct bau_control **) 843 uv_bau_table_bases = (struct bau_control **)
777 kmalloc(nblades * sizeof(struct bau_control *), GFP_KERNEL); 844 kmalloc(nblades * sizeof(struct bau_control *), GFP_KERNEL);
778 BUG_ON(!uv_bau_table_bases); 845 BUG_ON(!uv_bau_table_bases);
779 846
780 last_blade = -1; 847 uv_partition_base_pnode = 0x7fffffff;
781 for_each_online_node(node) { 848 for (blade = 0; blade < nblades; blade++)
782 blade = uv_node_to_blade_id(node); 849 if (uv_blade_nr_possible_cpus(blade) &&
783 if (blade == last_blade) 850 (uv_blade_to_pnode(blade) < uv_partition_base_pnode))
784 continue; 851 uv_partition_base_pnode = uv_blade_to_pnode(blade);
785 last_blade = blade; 852 for (blade = 0; blade < nblades; blade++)
786 uv_init_blade(blade, node, cur_cpu); 853 if (uv_blade_nr_possible_cpus(blade))
787 cur_cpu += uv_blade_nr_possible_cpus(blade); 854 uv_init_blade(blade);
788 } 855
789 alloc_intr_gate(UV_BAU_MESSAGE, uv_bau_message_intr1); 856 alloc_intr_gate(UV_BAU_MESSAGE, uv_bau_message_intr1);
790 uv_enable_timeouts(); 857 uv_enable_timeouts();
791 858
diff --git a/arch/x86/kernel/uv_sysfs.c b/arch/x86/kernel/uv_sysfs.c
index 67f9b9dbf800..36afb98675a4 100644
--- a/arch/x86/kernel/uv_sysfs.c
+++ b/arch/x86/kernel/uv_sysfs.c
@@ -21,6 +21,7 @@
21 21
22#include <linux/sysdev.h> 22#include <linux/sysdev.h>
23#include <asm/uv/bios.h> 23#include <asm/uv/bios.h>
24#include <asm/uv/uv.h>
24 25
25struct kobject *sgi_uv_kobj; 26struct kobject *sgi_uv_kobj;
26 27
@@ -47,6 +48,9 @@ static int __init sgi_uv_sysfs_init(void)
47{ 48{
48 unsigned long ret; 49 unsigned long ret;
49 50
51 if (!is_uv_system())
52 return -ENODEV;
53
50 if (!sgi_uv_kobj) 54 if (!sgi_uv_kobj)
51 sgi_uv_kobj = kobject_create_and_add("sgi_uv", firmware_kobj); 55 sgi_uv_kobj = kobject_create_and_add("sgi_uv", firmware_kobj);
52 if (!sgi_uv_kobj) { 56 if (!sgi_uv_kobj) {
diff --git a/arch/x86/xen/enlighten.c b/arch/x86/xen/enlighten.c
index 82cd39a6cbd3..f09e8c36ee80 100644
--- a/arch/x86/xen/enlighten.c
+++ b/arch/x86/xen/enlighten.c
@@ -42,6 +42,7 @@
42#include <asm/xen/hypervisor.h> 42#include <asm/xen/hypervisor.h>
43#include <asm/fixmap.h> 43#include <asm/fixmap.h>
44#include <asm/processor.h> 44#include <asm/processor.h>
45#include <asm/proto.h>
45#include <asm/msr-index.h> 46#include <asm/msr-index.h>
46#include <asm/setup.h> 47#include <asm/setup.h>
47#include <asm/desc.h> 48#include <asm/desc.h>
@@ -168,21 +169,23 @@ static void __init xen_banner(void)
168 xen_feature(XENFEAT_mmu_pt_update_preserve_ad) ? " (preserve-AD)" : ""); 169 xen_feature(XENFEAT_mmu_pt_update_preserve_ad) ? " (preserve-AD)" : "");
169} 170}
170 171
172static __read_mostly unsigned int cpuid_leaf1_edx_mask = ~0;
173static __read_mostly unsigned int cpuid_leaf1_ecx_mask = ~0;
174
171static void xen_cpuid(unsigned int *ax, unsigned int *bx, 175static void xen_cpuid(unsigned int *ax, unsigned int *bx,
172 unsigned int *cx, unsigned int *dx) 176 unsigned int *cx, unsigned int *dx)
173{ 177{
178 unsigned maskecx = ~0;
174 unsigned maskedx = ~0; 179 unsigned maskedx = ~0;
175 180
176 /* 181 /*
177 * Mask out inconvenient features, to try and disable as many 182 * Mask out inconvenient features, to try and disable as many
178 * unsupported kernel subsystems as possible. 183 * unsupported kernel subsystems as possible.
179 */ 184 */
180 if (*ax == 1) 185 if (*ax == 1) {
181 maskedx = ~((1 << X86_FEATURE_APIC) | /* disable APIC */ 186 maskecx = cpuid_leaf1_ecx_mask;
182 (1 << X86_FEATURE_ACPI) | /* disable ACPI */ 187 maskedx = cpuid_leaf1_edx_mask;
183 (1 << X86_FEATURE_MCE) | /* disable MCE */ 188 }
184 (1 << X86_FEATURE_MCA) | /* disable MCA */
185 (1 << X86_FEATURE_ACC)); /* thermal monitoring */
186 189
187 asm(XEN_EMULATE_PREFIX "cpuid" 190 asm(XEN_EMULATE_PREFIX "cpuid"
188 : "=a" (*ax), 191 : "=a" (*ax),
@@ -190,9 +193,43 @@ static void xen_cpuid(unsigned int *ax, unsigned int *bx,
190 "=c" (*cx), 193 "=c" (*cx),
191 "=d" (*dx) 194 "=d" (*dx)
192 : "0" (*ax), "2" (*cx)); 195 : "0" (*ax), "2" (*cx));
196
197 *cx &= maskecx;
193 *dx &= maskedx; 198 *dx &= maskedx;
194} 199}
195 200
201static __init void xen_init_cpuid_mask(void)
202{
203 unsigned int ax, bx, cx, dx;
204
205 cpuid_leaf1_edx_mask =
206 ~((1 << X86_FEATURE_MCE) | /* disable MCE */
207 (1 << X86_FEATURE_MCA) | /* disable MCA */
208 (1 << X86_FEATURE_ACC)); /* thermal monitoring */
209
210 if (!xen_initial_domain())
211 cpuid_leaf1_edx_mask &=
212 ~((1 << X86_FEATURE_APIC) | /* disable local APIC */
213 (1 << X86_FEATURE_ACPI)); /* disable ACPI */
214
215 ax = 1;
216 xen_cpuid(&ax, &bx, &cx, &dx);
217
218 /* cpuid claims we support xsave; try enabling it to see what happens */
219 if (cx & (1 << (X86_FEATURE_XSAVE % 32))) {
220 unsigned long cr4;
221
222 set_in_cr4(X86_CR4_OSXSAVE);
223
224 cr4 = read_cr4();
225
226 if ((cr4 & X86_CR4_OSXSAVE) == 0)
227 cpuid_leaf1_ecx_mask &= ~(1 << (X86_FEATURE_XSAVE % 32));
228
229 clear_in_cr4(X86_CR4_OSXSAVE);
230 }
231}
232
196static void xen_set_debugreg(int reg, unsigned long val) 233static void xen_set_debugreg(int reg, unsigned long val)
197{ 234{
198 HYPERVISOR_set_debugreg(reg, val); 235 HYPERVISOR_set_debugreg(reg, val);
@@ -284,12 +321,11 @@ static void xen_set_ldt(const void *addr, unsigned entries)
284 321
285static void xen_load_gdt(const struct desc_ptr *dtr) 322static void xen_load_gdt(const struct desc_ptr *dtr)
286{ 323{
287 unsigned long *frames;
288 unsigned long va = dtr->address; 324 unsigned long va = dtr->address;
289 unsigned int size = dtr->size + 1; 325 unsigned int size = dtr->size + 1;
290 unsigned pages = (size + PAGE_SIZE - 1) / PAGE_SIZE; 326 unsigned pages = (size + PAGE_SIZE - 1) / PAGE_SIZE;
327 unsigned long frames[pages];
291 int f; 328 int f;
292 struct multicall_space mcs;
293 329
294 /* A GDT can be up to 64k in size, which corresponds to 8192 330 /* A GDT can be up to 64k in size, which corresponds to 8192
295 8-byte entries, or 16 4k pages.. */ 331 8-byte entries, or 16 4k pages.. */
@@ -297,19 +333,26 @@ static void xen_load_gdt(const struct desc_ptr *dtr)
297 BUG_ON(size > 65536); 333 BUG_ON(size > 65536);
298 BUG_ON(va & ~PAGE_MASK); 334 BUG_ON(va & ~PAGE_MASK);
299 335
300 mcs = xen_mc_entry(sizeof(*frames) * pages);
301 frames = mcs.args;
302
303 for (f = 0; va < dtr->address + size; va += PAGE_SIZE, f++) { 336 for (f = 0; va < dtr->address + size; va += PAGE_SIZE, f++) {
304 frames[f] = arbitrary_virt_to_mfn((void *)va); 337 int level;
338 pte_t *ptep = lookup_address(va, &level);
339 unsigned long pfn, mfn;
340 void *virt;
341
342 BUG_ON(ptep == NULL);
343
344 pfn = pte_pfn(*ptep);
345 mfn = pfn_to_mfn(pfn);
346 virt = __va(PFN_PHYS(pfn));
347
348 frames[f] = mfn;
305 349
306 make_lowmem_page_readonly((void *)va); 350 make_lowmem_page_readonly((void *)va);
307 make_lowmem_page_readonly(mfn_to_virt(frames[f])); 351 make_lowmem_page_readonly(virt);
308 } 352 }
309 353
310 MULTI_set_gdt(mcs.mc, frames, size / sizeof(struct desc_struct)); 354 if (HYPERVISOR_set_gdt(frames, size / sizeof(struct desc_struct)))
311 355 BUG();
312 xen_mc_issue(PARAVIRT_LAZY_CPU);
313} 356}
314 357
315static void load_TLS_descriptor(struct thread_struct *t, 358static void load_TLS_descriptor(struct thread_struct *t,
@@ -385,7 +428,7 @@ static void xen_write_ldt_entry(struct desc_struct *dt, int entrynum,
385static int cvt_gate_to_trap(int vector, const gate_desc *val, 428static int cvt_gate_to_trap(int vector, const gate_desc *val,
386 struct trap_info *info) 429 struct trap_info *info)
387{ 430{
388 if (val->type != 0xf && val->type != 0xe) 431 if (val->type != GATE_TRAP && val->type != GATE_INTERRUPT)
389 return 0; 432 return 0;
390 433
391 info->vector = vector; 434 info->vector = vector;
@@ -393,8 +436,8 @@ static int cvt_gate_to_trap(int vector, const gate_desc *val,
393 info->cs = gate_segment(*val); 436 info->cs = gate_segment(*val);
394 info->flags = val->dpl; 437 info->flags = val->dpl;
395 /* interrupt gates clear IF */ 438 /* interrupt gates clear IF */
396 if (val->type == 0xe) 439 if (val->type == GATE_INTERRUPT)
397 info->flags |= 4; 440 info->flags |= 1 << 2;
398 441
399 return 1; 442 return 1;
400} 443}
@@ -872,7 +915,6 @@ static const struct machine_ops __initdata xen_machine_ops = {
872 .emergency_restart = xen_emergency_restart, 915 .emergency_restart = xen_emergency_restart,
873}; 916};
874 917
875
876/* First C function to be called on Xen boot */ 918/* First C function to be called on Xen boot */
877asmlinkage void __init xen_start_kernel(void) 919asmlinkage void __init xen_start_kernel(void)
878{ 920{
@@ -897,6 +939,8 @@ asmlinkage void __init xen_start_kernel(void)
897 939
898 xen_init_irq_ops(); 940 xen_init_irq_ops();
899 941
942 xen_init_cpuid_mask();
943
900#ifdef CONFIG_X86_LOCAL_APIC 944#ifdef CONFIG_X86_LOCAL_APIC
901 /* 945 /*
902 * set up the basic apic ops. 946 * set up the basic apic ops.
@@ -938,6 +982,11 @@ asmlinkage void __init xen_start_kernel(void)
938 if (!xen_initial_domain()) 982 if (!xen_initial_domain())
939 __supported_pte_mask &= ~(_PAGE_PWT | _PAGE_PCD); 983 __supported_pte_mask &= ~(_PAGE_PWT | _PAGE_PCD);
940 984
985#ifdef CONFIG_X86_64
986 /* Work out if we support NX */
987 check_efer();
988#endif
989
941 /* Don't do the full vcpu_info placement stuff until we have a 990 /* Don't do the full vcpu_info placement stuff until we have a
942 possible map and a non-dummy shared_info. */ 991 possible map and a non-dummy shared_info. */
943 per_cpu(xen_vcpu, 0) = &HYPERVISOR_shared_info->vcpu_info[0]; 992 per_cpu(xen_vcpu, 0) = &HYPERVISOR_shared_info->vcpu_info[0];
diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c
index 2a81838a9ab7..9842b1212407 100644
--- a/arch/x86/xen/mmu.c
+++ b/arch/x86/xen/mmu.c
@@ -184,7 +184,7 @@ static inline unsigned p2m_index(unsigned long pfn)
184} 184}
185 185
186/* Build the parallel p2m_top_mfn structures */ 186/* Build the parallel p2m_top_mfn structures */
187void xen_setup_mfn_list_list(void) 187static void __init xen_build_mfn_list_list(void)
188{ 188{
189 unsigned pfn, idx; 189 unsigned pfn, idx;
190 190
@@ -198,7 +198,10 @@ void xen_setup_mfn_list_list(void)
198 unsigned topidx = idx * P2M_ENTRIES_PER_PAGE; 198 unsigned topidx = idx * P2M_ENTRIES_PER_PAGE;
199 p2m_top_mfn_list[idx] = virt_to_mfn(&p2m_top_mfn[topidx]); 199 p2m_top_mfn_list[idx] = virt_to_mfn(&p2m_top_mfn[topidx]);
200 } 200 }
201}
201 202
203void xen_setup_mfn_list_list(void)
204{
202 BUG_ON(HYPERVISOR_shared_info == &xen_dummy_shared_info); 205 BUG_ON(HYPERVISOR_shared_info == &xen_dummy_shared_info);
203 206
204 HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list = 207 HYPERVISOR_shared_info->arch.pfn_to_mfn_frame_list_list =
@@ -218,6 +221,8 @@ void __init xen_build_dynamic_phys_to_machine(void)
218 221
219 p2m_top[topidx] = &mfn_list[pfn]; 222 p2m_top[topidx] = &mfn_list[pfn];
220 } 223 }
224
225 xen_build_mfn_list_list();
221} 226}
222 227
223unsigned long get_phys_to_machine(unsigned long pfn) 228unsigned long get_phys_to_machine(unsigned long pfn)
@@ -233,47 +238,74 @@ unsigned long get_phys_to_machine(unsigned long pfn)
233} 238}
234EXPORT_SYMBOL_GPL(get_phys_to_machine); 239EXPORT_SYMBOL_GPL(get_phys_to_machine);
235 240
236static void alloc_p2m(unsigned long **pp, unsigned long *mfnp) 241/* install a new p2m_top page */
242bool install_p2mtop_page(unsigned long pfn, unsigned long *p)
237{ 243{
238 unsigned long *p; 244 unsigned topidx = p2m_top_index(pfn);
245 unsigned long **pfnp, *mfnp;
239 unsigned i; 246 unsigned i;
240 247
241 p = (void *)__get_free_page(GFP_KERNEL | __GFP_NOFAIL); 248 pfnp = &p2m_top[topidx];
242 BUG_ON(p == NULL); 249 mfnp = &p2m_top_mfn[topidx];
243 250
244 for (i = 0; i < P2M_ENTRIES_PER_PAGE; i++) 251 for (i = 0; i < P2M_ENTRIES_PER_PAGE; i++)
245 p[i] = INVALID_P2M_ENTRY; 252 p[i] = INVALID_P2M_ENTRY;
246 253
247 if (cmpxchg(pp, p2m_missing, p) != p2m_missing) 254 if (cmpxchg(pfnp, p2m_missing, p) == p2m_missing) {
248 free_page((unsigned long)p);
249 else
250 *mfnp = virt_to_mfn(p); 255 *mfnp = virt_to_mfn(p);
256 return true;
257 }
258
259 return false;
251} 260}
252 261
253void set_phys_to_machine(unsigned long pfn, unsigned long mfn) 262static void alloc_p2m(unsigned long pfn)
254{ 263{
255 unsigned topidx, idx; 264 unsigned long *p;
256 265
257 if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) { 266 p = (void *)__get_free_page(GFP_KERNEL | __GFP_NOFAIL);
258 BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY); 267 BUG_ON(p == NULL);
259 return; 268
260 } 269 if (!install_p2mtop_page(pfn, p))
270 free_page((unsigned long)p);
271}
272
273/* Try to install p2m mapping; fail if intermediate bits missing */
274bool __set_phys_to_machine(unsigned long pfn, unsigned long mfn)
275{
276 unsigned topidx, idx;
261 277
262 if (unlikely(pfn >= MAX_DOMAIN_PAGES)) { 278 if (unlikely(pfn >= MAX_DOMAIN_PAGES)) {
263 BUG_ON(mfn != INVALID_P2M_ENTRY); 279 BUG_ON(mfn != INVALID_P2M_ENTRY);
264 return; 280 return true;
265 } 281 }
266 282
267 topidx = p2m_top_index(pfn); 283 topidx = p2m_top_index(pfn);
268 if (p2m_top[topidx] == p2m_missing) { 284 if (p2m_top[topidx] == p2m_missing) {
269 /* no need to allocate a page to store an invalid entry */
270 if (mfn == INVALID_P2M_ENTRY) 285 if (mfn == INVALID_P2M_ENTRY)
271 return; 286 return true;
272 alloc_p2m(&p2m_top[topidx], &p2m_top_mfn[topidx]); 287 return false;
273 } 288 }
274 289
275 idx = p2m_index(pfn); 290 idx = p2m_index(pfn);
276 p2m_top[topidx][idx] = mfn; 291 p2m_top[topidx][idx] = mfn;
292
293 return true;
294}
295
296void set_phys_to_machine(unsigned long pfn, unsigned long mfn)
297{
298 if (unlikely(xen_feature(XENFEAT_auto_translated_physmap))) {
299 BUG_ON(pfn != mfn && mfn != INVALID_P2M_ENTRY);
300 return;
301 }
302
303 if (unlikely(!__set_phys_to_machine(pfn, mfn))) {
304 alloc_p2m(pfn);
305
306 if (!__set_phys_to_machine(pfn, mfn))
307 BUG();
308 }
277} 309}
278 310
279unsigned long arbitrary_virt_to_mfn(void *vaddr) 311unsigned long arbitrary_virt_to_mfn(void *vaddr)
@@ -987,7 +1019,7 @@ static __init int xen_mark_pinned(struct mm_struct *mm, struct page *page,
987 return 0; 1019 return 0;
988} 1020}
989 1021
990void __init xen_mark_init_mm_pinned(void) 1022static void __init xen_mark_init_mm_pinned(void)
991{ 1023{
992 xen_pgd_walk(&init_mm, xen_mark_pinned, FIXADDR_TOP); 1024 xen_pgd_walk(&init_mm, xen_mark_pinned, FIXADDR_TOP);
993} 1025}
@@ -1270,8 +1302,8 @@ static void xen_flush_tlb_others(const struct cpumask *cpus,
1270 } *args; 1302 } *args;
1271 struct multicall_space mcs; 1303 struct multicall_space mcs;
1272 1304
1273 BUG_ON(cpumask_empty(cpus)); 1305 if (cpumask_empty(cpus))
1274 BUG_ON(!mm); 1306 return; /* nothing to do */
1275 1307
1276 mcs = xen_mc_entry(sizeof(*args)); 1308 mcs = xen_mc_entry(sizeof(*args));
1277 args = mcs.args; 1309 args = mcs.args;
@@ -1438,6 +1470,15 @@ static __init void xen_set_pte_init(pte_t *ptep, pte_t pte)
1438} 1470}
1439#endif 1471#endif
1440 1472
1473static void pin_pagetable_pfn(unsigned cmd, unsigned long pfn)
1474{
1475 struct mmuext_op op;
1476 op.cmd = cmd;
1477 op.arg1.mfn = pfn_to_mfn(pfn);
1478 if (HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF))
1479 BUG();
1480}
1481
1441/* Early in boot, while setting up the initial pagetable, assume 1482/* Early in boot, while setting up the initial pagetable, assume
1442 everything is pinned. */ 1483 everything is pinned. */
1443static __init void xen_alloc_pte_init(struct mm_struct *mm, unsigned long pfn) 1484static __init void xen_alloc_pte_init(struct mm_struct *mm, unsigned long pfn)
@@ -1446,22 +1487,29 @@ static __init void xen_alloc_pte_init(struct mm_struct *mm, unsigned long pfn)
1446 BUG_ON(mem_map); /* should only be used early */ 1487 BUG_ON(mem_map); /* should only be used early */
1447#endif 1488#endif
1448 make_lowmem_page_readonly(__va(PFN_PHYS(pfn))); 1489 make_lowmem_page_readonly(__va(PFN_PHYS(pfn)));
1490 pin_pagetable_pfn(MMUEXT_PIN_L1_TABLE, pfn);
1491}
1492
1493/* Used for pmd and pud */
1494static __init void xen_alloc_pmd_init(struct mm_struct *mm, unsigned long pfn)
1495{
1496#ifdef CONFIG_FLATMEM
1497 BUG_ON(mem_map); /* should only be used early */
1498#endif
1499 make_lowmem_page_readonly(__va(PFN_PHYS(pfn)));
1449} 1500}
1450 1501
1451/* Early release_pte assumes that all pts are pinned, since there's 1502/* Early release_pte assumes that all pts are pinned, since there's
1452 only init_mm and anything attached to that is pinned. */ 1503 only init_mm and anything attached to that is pinned. */
1453static void xen_release_pte_init(unsigned long pfn) 1504static __init void xen_release_pte_init(unsigned long pfn)
1454{ 1505{
1506 pin_pagetable_pfn(MMUEXT_UNPIN_TABLE, pfn);
1455 make_lowmem_page_readwrite(__va(PFN_PHYS(pfn))); 1507 make_lowmem_page_readwrite(__va(PFN_PHYS(pfn)));
1456} 1508}
1457 1509
1458static void pin_pagetable_pfn(unsigned cmd, unsigned long pfn) 1510static __init void xen_release_pmd_init(unsigned long pfn)
1459{ 1511{
1460 struct mmuext_op op; 1512 make_lowmem_page_readwrite(__va(PFN_PHYS(pfn)));
1461 op.cmd = cmd;
1462 op.arg1.mfn = pfn_to_mfn(pfn);
1463 if (HYPERVISOR_mmuext_op(&op, 1, NULL, DOMID_SELF))
1464 BUG();
1465} 1513}
1466 1514
1467/* This needs to make sure the new pte page is pinned iff its being 1515/* This needs to make sure the new pte page is pinned iff its being
@@ -1773,6 +1821,9 @@ static void xen_set_fixmap(unsigned idx, phys_addr_t phys, pgprot_t prot)
1773#ifdef CONFIG_X86_LOCAL_APIC 1821#ifdef CONFIG_X86_LOCAL_APIC
1774 case FIX_APIC_BASE: /* maps dummy local APIC */ 1822 case FIX_APIC_BASE: /* maps dummy local APIC */
1775#endif 1823#endif
1824 case FIX_TEXT_POKE0:
1825 case FIX_TEXT_POKE1:
1826 /* All local page mappings */
1776 pte = pfn_pte(phys, prot); 1827 pte = pfn_pte(phys, prot);
1777 break; 1828 break;
1778 1829
@@ -1819,7 +1870,6 @@ __init void xen_post_allocator_init(void)
1819 xen_mark_init_mm_pinned(); 1870 xen_mark_init_mm_pinned();
1820} 1871}
1821 1872
1822
1823const struct pv_mmu_ops xen_mmu_ops __initdata = { 1873const struct pv_mmu_ops xen_mmu_ops __initdata = {
1824 .pagetable_setup_start = xen_pagetable_setup_start, 1874 .pagetable_setup_start = xen_pagetable_setup_start,
1825 .pagetable_setup_done = xen_pagetable_setup_done, 1875 .pagetable_setup_done = xen_pagetable_setup_done,
@@ -1843,9 +1893,9 @@ const struct pv_mmu_ops xen_mmu_ops __initdata = {
1843 1893
1844 .alloc_pte = xen_alloc_pte_init, 1894 .alloc_pte = xen_alloc_pte_init,
1845 .release_pte = xen_release_pte_init, 1895 .release_pte = xen_release_pte_init,
1846 .alloc_pmd = xen_alloc_pte_init, 1896 .alloc_pmd = xen_alloc_pmd_init,
1847 .alloc_pmd_clone = paravirt_nop, 1897 .alloc_pmd_clone = paravirt_nop,
1848 .release_pmd = xen_release_pte_init, 1898 .release_pmd = xen_release_pmd_init,
1849 1899
1850#ifdef CONFIG_HIGHPTE 1900#ifdef CONFIG_HIGHPTE
1851 .kmap_atomic_pte = xen_kmap_atomic_pte, 1901 .kmap_atomic_pte = xen_kmap_atomic_pte,
@@ -1883,8 +1933,8 @@ const struct pv_mmu_ops xen_mmu_ops __initdata = {
1883 .make_pud = PV_CALLEE_SAVE(xen_make_pud), 1933 .make_pud = PV_CALLEE_SAVE(xen_make_pud),
1884 .set_pgd = xen_set_pgd_hyper, 1934 .set_pgd = xen_set_pgd_hyper,
1885 1935
1886 .alloc_pud = xen_alloc_pte_init, 1936 .alloc_pud = xen_alloc_pmd_init,
1887 .release_pud = xen_release_pte_init, 1937 .release_pud = xen_release_pmd_init,
1888#endif /* PAGETABLE_LEVELS == 4 */ 1938#endif /* PAGETABLE_LEVELS == 4 */
1889 1939
1890 .activate_mm = xen_activate_mm, 1940 .activate_mm = xen_activate_mm,
diff --git a/arch/x86/xen/mmu.h b/arch/x86/xen/mmu.h
index 24d1b44a337d..da7302624897 100644
--- a/arch/x86/xen/mmu.h
+++ b/arch/x86/xen/mmu.h
@@ -11,6 +11,9 @@ enum pt_level {
11}; 11};
12 12
13 13
14bool __set_phys_to_machine(unsigned long pfn, unsigned long mfn);
15bool install_p2mtop_page(unsigned long pfn, unsigned long *p);
16
14void set_pte_mfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags); 17void set_pte_mfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags);
15 18
16 19
diff --git a/arch/x86/xen/smp.c b/arch/x86/xen/smp.c
index 585a6e330837..429834ec1687 100644
--- a/arch/x86/xen/smp.c
+++ b/arch/x86/xen/smp.c
@@ -317,7 +317,7 @@ static int __cpuinit xen_cpu_up(unsigned int cpu)
317 BUG_ON(rc); 317 BUG_ON(rc);
318 318
319 while(per_cpu(cpu_state, cpu) != CPU_ONLINE) { 319 while(per_cpu(cpu_state, cpu) != CPU_ONLINE) {
320 HYPERVISOR_sched_op(SCHEDOP_yield, 0); 320 HYPERVISOR_sched_op(SCHEDOP_yield, NULL);
321 barrier(); 321 barrier();
322 } 322 }
323 323
@@ -422,7 +422,7 @@ static void xen_smp_send_call_function_ipi(const struct cpumask *mask)
422 /* Make sure other vcpus get a chance to run if they need to. */ 422 /* Make sure other vcpus get a chance to run if they need to. */
423 for_each_cpu(cpu, mask) { 423 for_each_cpu(cpu, mask) {
424 if (xen_vcpu_stolen(cpu)) { 424 if (xen_vcpu_stolen(cpu)) {
425 HYPERVISOR_sched_op(SCHEDOP_yield, 0); 425 HYPERVISOR_sched_op(SCHEDOP_yield, NULL);
426 break; 426 break;
427 } 427 }
428 } 428 }
diff --git a/arch/x86/xen/xen-ops.h b/arch/x86/xen/xen-ops.h
index 2f5ef2632ea2..20139464943c 100644
--- a/arch/x86/xen/xen-ops.h
+++ b/arch/x86/xen/xen-ops.h
@@ -57,8 +57,6 @@ irqreturn_t xen_debug_interrupt(int irq, void *dev_id);
57 57
58bool xen_vcpu_stolen(int vcpu); 58bool xen_vcpu_stolen(int vcpu);
59 59
60void xen_mark_init_mm_pinned(void);
61
62void xen_setup_vcpu_info_placement(void); 60void xen_setup_vcpu_info_placement(void);
63 61
64#ifdef CONFIG_SMP 62#ifdef CONFIG_SMP
diff --git a/block/as-iosched.c b/block/as-iosched.c
index 631f6f44460a..c48fa670d221 100644
--- a/block/as-iosched.c
+++ b/block/as-iosched.c
@@ -17,9 +17,6 @@
17#include <linux/rbtree.h> 17#include <linux/rbtree.h>
18#include <linux/interrupt.h> 18#include <linux/interrupt.h>
19 19
20#define REQ_SYNC 1
21#define REQ_ASYNC 0
22
23/* 20/*
24 * See Documentation/block/as-iosched.txt 21 * See Documentation/block/as-iosched.txt
25 */ 22 */
@@ -93,7 +90,7 @@ struct as_data {
93 struct list_head fifo_list[2]; 90 struct list_head fifo_list[2];
94 91
95 struct request *next_rq[2]; /* next in sort order */ 92 struct request *next_rq[2]; /* next in sort order */
96 sector_t last_sector[2]; /* last REQ_SYNC & REQ_ASYNC sectors */ 93 sector_t last_sector[2]; /* last SYNC & ASYNC sectors */
97 94
98 unsigned long exit_prob; /* probability a task will exit while 95 unsigned long exit_prob; /* probability a task will exit while
99 being waited on */ 96 being waited on */
@@ -109,7 +106,7 @@ struct as_data {
109 unsigned long last_check_fifo[2]; 106 unsigned long last_check_fifo[2];
110 int changed_batch; /* 1: waiting for old batch to end */ 107 int changed_batch; /* 1: waiting for old batch to end */
111 int new_batch; /* 1: waiting on first read complete */ 108 int new_batch; /* 1: waiting on first read complete */
112 int batch_data_dir; /* current batch REQ_SYNC / REQ_ASYNC */ 109 int batch_data_dir; /* current batch SYNC / ASYNC */
113 int write_batch_count; /* max # of reqs in a write batch */ 110 int write_batch_count; /* max # of reqs in a write batch */
114 int current_write_count; /* how many requests left this batch */ 111 int current_write_count; /* how many requests left this batch */
115 int write_batch_idled; /* has the write batch gone idle? */ 112 int write_batch_idled; /* has the write batch gone idle? */
@@ -554,7 +551,7 @@ static void as_update_iohist(struct as_data *ad, struct as_io_context *aic,
554 if (aic == NULL) 551 if (aic == NULL)
555 return; 552 return;
556 553
557 if (data_dir == REQ_SYNC) { 554 if (data_dir == BLK_RW_SYNC) {
558 unsigned long in_flight = atomic_read(&aic->nr_queued) 555 unsigned long in_flight = atomic_read(&aic->nr_queued)
559 + atomic_read(&aic->nr_dispatched); 556 + atomic_read(&aic->nr_dispatched);
560 spin_lock(&aic->lock); 557 spin_lock(&aic->lock);
@@ -811,7 +808,7 @@ static void as_update_rq(struct as_data *ad, struct request *rq)
811 */ 808 */
812static void update_write_batch(struct as_data *ad) 809static void update_write_batch(struct as_data *ad)
813{ 810{
814 unsigned long batch = ad->batch_expire[REQ_ASYNC]; 811 unsigned long batch = ad->batch_expire[BLK_RW_ASYNC];
815 long write_time; 812 long write_time;
816 813
817 write_time = (jiffies - ad->current_batch_expires) + batch; 814 write_time = (jiffies - ad->current_batch_expires) + batch;
@@ -855,7 +852,7 @@ static void as_completed_request(struct request_queue *q, struct request *rq)
855 kblockd_schedule_work(q, &ad->antic_work); 852 kblockd_schedule_work(q, &ad->antic_work);
856 ad->changed_batch = 0; 853 ad->changed_batch = 0;
857 854
858 if (ad->batch_data_dir == REQ_SYNC) 855 if (ad->batch_data_dir == BLK_RW_SYNC)
859 ad->new_batch = 1; 856 ad->new_batch = 1;
860 } 857 }
861 WARN_ON(ad->nr_dispatched == 0); 858 WARN_ON(ad->nr_dispatched == 0);
@@ -869,7 +866,7 @@ static void as_completed_request(struct request_queue *q, struct request *rq)
869 if (ad->new_batch && ad->batch_data_dir == rq_is_sync(rq)) { 866 if (ad->new_batch && ad->batch_data_dir == rq_is_sync(rq)) {
870 update_write_batch(ad); 867 update_write_batch(ad);
871 ad->current_batch_expires = jiffies + 868 ad->current_batch_expires = jiffies +
872 ad->batch_expire[REQ_SYNC]; 869 ad->batch_expire[BLK_RW_SYNC];
873 ad->new_batch = 0; 870 ad->new_batch = 0;
874 } 871 }
875 872
@@ -960,7 +957,7 @@ static inline int as_batch_expired(struct as_data *ad)
960 if (ad->changed_batch || ad->new_batch) 957 if (ad->changed_batch || ad->new_batch)
961 return 0; 958 return 0;
962 959
963 if (ad->batch_data_dir == REQ_SYNC) 960 if (ad->batch_data_dir == BLK_RW_SYNC)
964 /* TODO! add a check so a complete fifo gets written? */ 961 /* TODO! add a check so a complete fifo gets written? */
965 return time_after(jiffies, ad->current_batch_expires); 962 return time_after(jiffies, ad->current_batch_expires);
966 963
@@ -986,7 +983,7 @@ static void as_move_to_dispatch(struct as_data *ad, struct request *rq)
986 */ 983 */
987 ad->last_sector[data_dir] = rq->sector + rq->nr_sectors; 984 ad->last_sector[data_dir] = rq->sector + rq->nr_sectors;
988 985
989 if (data_dir == REQ_SYNC) { 986 if (data_dir == BLK_RW_SYNC) {
990 struct io_context *ioc = RQ_IOC(rq); 987 struct io_context *ioc = RQ_IOC(rq);
991 /* In case we have to anticipate after this */ 988 /* In case we have to anticipate after this */
992 copy_io_context(&ad->io_context, &ioc); 989 copy_io_context(&ad->io_context, &ioc);
@@ -1025,41 +1022,41 @@ static void as_move_to_dispatch(struct as_data *ad, struct request *rq)
1025static int as_dispatch_request(struct request_queue *q, int force) 1022static int as_dispatch_request(struct request_queue *q, int force)
1026{ 1023{
1027 struct as_data *ad = q->elevator->elevator_data; 1024 struct as_data *ad = q->elevator->elevator_data;
1028 const int reads = !list_empty(&ad->fifo_list[REQ_SYNC]); 1025 const int reads = !list_empty(&ad->fifo_list[BLK_RW_SYNC]);
1029 const int writes = !list_empty(&ad->fifo_list[REQ_ASYNC]); 1026 const int writes = !list_empty(&ad->fifo_list[BLK_RW_ASYNC]);
1030 struct request *rq; 1027 struct request *rq;
1031 1028
1032 if (unlikely(force)) { 1029 if (unlikely(force)) {
1033 /* 1030 /*
1034 * Forced dispatch, accounting is useless. Reset 1031 * Forced dispatch, accounting is useless. Reset
1035 * accounting states and dump fifo_lists. Note that 1032 * accounting states and dump fifo_lists. Note that
1036 * batch_data_dir is reset to REQ_SYNC to avoid 1033 * batch_data_dir is reset to BLK_RW_SYNC to avoid
1037 * screwing write batch accounting as write batch 1034 * screwing write batch accounting as write batch
1038 * accounting occurs on W->R transition. 1035 * accounting occurs on W->R transition.
1039 */ 1036 */
1040 int dispatched = 0; 1037 int dispatched = 0;
1041 1038
1042 ad->batch_data_dir = REQ_SYNC; 1039 ad->batch_data_dir = BLK_RW_SYNC;
1043 ad->changed_batch = 0; 1040 ad->changed_batch = 0;
1044 ad->new_batch = 0; 1041 ad->new_batch = 0;
1045 1042
1046 while (ad->next_rq[REQ_SYNC]) { 1043 while (ad->next_rq[BLK_RW_SYNC]) {
1047 as_move_to_dispatch(ad, ad->next_rq[REQ_SYNC]); 1044 as_move_to_dispatch(ad, ad->next_rq[BLK_RW_SYNC]);
1048 dispatched++; 1045 dispatched++;
1049 } 1046 }
1050 ad->last_check_fifo[REQ_SYNC] = jiffies; 1047 ad->last_check_fifo[BLK_RW_SYNC] = jiffies;
1051 1048
1052 while (ad->next_rq[REQ_ASYNC]) { 1049 while (ad->next_rq[BLK_RW_ASYNC]) {
1053 as_move_to_dispatch(ad, ad->next_rq[REQ_ASYNC]); 1050 as_move_to_dispatch(ad, ad->next_rq[BLK_RW_ASYNC]);
1054 dispatched++; 1051 dispatched++;
1055 } 1052 }
1056 ad->last_check_fifo[REQ_ASYNC] = jiffies; 1053 ad->last_check_fifo[BLK_RW_ASYNC] = jiffies;
1057 1054
1058 return dispatched; 1055 return dispatched;
1059 } 1056 }
1060 1057
1061 /* Signal that the write batch was uncontended, so we can't time it */ 1058 /* Signal that the write batch was uncontended, so we can't time it */
1062 if (ad->batch_data_dir == REQ_ASYNC && !reads) { 1059 if (ad->batch_data_dir == BLK_RW_ASYNC && !reads) {
1063 if (ad->current_write_count == 0 || !writes) 1060 if (ad->current_write_count == 0 || !writes)
1064 ad->write_batch_idled = 1; 1061 ad->write_batch_idled = 1;
1065 } 1062 }
@@ -1076,8 +1073,8 @@ static int as_dispatch_request(struct request_queue *q, int force)
1076 */ 1073 */
1077 rq = ad->next_rq[ad->batch_data_dir]; 1074 rq = ad->next_rq[ad->batch_data_dir];
1078 1075
1079 if (ad->batch_data_dir == REQ_SYNC && ad->antic_expire) { 1076 if (ad->batch_data_dir == BLK_RW_SYNC && ad->antic_expire) {
1080 if (as_fifo_expired(ad, REQ_SYNC)) 1077 if (as_fifo_expired(ad, BLK_RW_SYNC))
1081 goto fifo_expired; 1078 goto fifo_expired;
1082 1079
1083 if (as_can_anticipate(ad, rq)) { 1080 if (as_can_anticipate(ad, rq)) {
@@ -1090,7 +1087,7 @@ static int as_dispatch_request(struct request_queue *q, int force)
1090 /* we have a "next request" */ 1087 /* we have a "next request" */
1091 if (reads && !writes) 1088 if (reads && !writes)
1092 ad->current_batch_expires = 1089 ad->current_batch_expires =
1093 jiffies + ad->batch_expire[REQ_SYNC]; 1090 jiffies + ad->batch_expire[BLK_RW_SYNC];
1094 goto dispatch_request; 1091 goto dispatch_request;
1095 } 1092 }
1096 } 1093 }
@@ -1101,20 +1098,20 @@ static int as_dispatch_request(struct request_queue *q, int force)
1101 */ 1098 */
1102 1099
1103 if (reads) { 1100 if (reads) {
1104 BUG_ON(RB_EMPTY_ROOT(&ad->sort_list[REQ_SYNC])); 1101 BUG_ON(RB_EMPTY_ROOT(&ad->sort_list[BLK_RW_SYNC]));
1105 1102
1106 if (writes && ad->batch_data_dir == REQ_SYNC) 1103 if (writes && ad->batch_data_dir == BLK_RW_SYNC)
1107 /* 1104 /*
1108 * Last batch was a read, switch to writes 1105 * Last batch was a read, switch to writes
1109 */ 1106 */
1110 goto dispatch_writes; 1107 goto dispatch_writes;
1111 1108
1112 if (ad->batch_data_dir == REQ_ASYNC) { 1109 if (ad->batch_data_dir == BLK_RW_ASYNC) {
1113 WARN_ON(ad->new_batch); 1110 WARN_ON(ad->new_batch);
1114 ad->changed_batch = 1; 1111 ad->changed_batch = 1;
1115 } 1112 }
1116 ad->batch_data_dir = REQ_SYNC; 1113 ad->batch_data_dir = BLK_RW_SYNC;
1117 rq = rq_entry_fifo(ad->fifo_list[REQ_SYNC].next); 1114 rq = rq_entry_fifo(ad->fifo_list[BLK_RW_SYNC].next);
1118 ad->last_check_fifo[ad->batch_data_dir] = jiffies; 1115 ad->last_check_fifo[ad->batch_data_dir] = jiffies;
1119 goto dispatch_request; 1116 goto dispatch_request;
1120 } 1117 }
@@ -1125,9 +1122,9 @@ static int as_dispatch_request(struct request_queue *q, int force)
1125 1122
1126 if (writes) { 1123 if (writes) {
1127dispatch_writes: 1124dispatch_writes:
1128 BUG_ON(RB_EMPTY_ROOT(&ad->sort_list[REQ_ASYNC])); 1125 BUG_ON(RB_EMPTY_ROOT(&ad->sort_list[BLK_RW_ASYNC]));
1129 1126
1130 if (ad->batch_data_dir == REQ_SYNC) { 1127 if (ad->batch_data_dir == BLK_RW_SYNC) {
1131 ad->changed_batch = 1; 1128 ad->changed_batch = 1;
1132 1129
1133 /* 1130 /*
@@ -1137,11 +1134,11 @@ dispatch_writes:
1137 */ 1134 */
1138 ad->new_batch = 0; 1135 ad->new_batch = 0;
1139 } 1136 }
1140 ad->batch_data_dir = REQ_ASYNC; 1137 ad->batch_data_dir = BLK_RW_ASYNC;
1141 ad->current_write_count = ad->write_batch_count; 1138 ad->current_write_count = ad->write_batch_count;
1142 ad->write_batch_idled = 0; 1139 ad->write_batch_idled = 0;
1143 rq = rq_entry_fifo(ad->fifo_list[REQ_ASYNC].next); 1140 rq = rq_entry_fifo(ad->fifo_list[BLK_RW_ASYNC].next);
1144 ad->last_check_fifo[REQ_ASYNC] = jiffies; 1141 ad->last_check_fifo[BLK_RW_ASYNC] = jiffies;
1145 goto dispatch_request; 1142 goto dispatch_request;
1146 } 1143 }
1147 1144
@@ -1164,9 +1161,9 @@ fifo_expired:
1164 if (ad->nr_dispatched) 1161 if (ad->nr_dispatched)
1165 return 0; 1162 return 0;
1166 1163
1167 if (ad->batch_data_dir == REQ_ASYNC) 1164 if (ad->batch_data_dir == BLK_RW_ASYNC)
1168 ad->current_batch_expires = jiffies + 1165 ad->current_batch_expires = jiffies +
1169 ad->batch_expire[REQ_ASYNC]; 1166 ad->batch_expire[BLK_RW_ASYNC];
1170 else 1167 else
1171 ad->new_batch = 1; 1168 ad->new_batch = 1;
1172 1169
@@ -1238,8 +1235,8 @@ static int as_queue_empty(struct request_queue *q)
1238{ 1235{
1239 struct as_data *ad = q->elevator->elevator_data; 1236 struct as_data *ad = q->elevator->elevator_data;
1240 1237
1241 return list_empty(&ad->fifo_list[REQ_ASYNC]) 1238 return list_empty(&ad->fifo_list[BLK_RW_ASYNC])
1242 && list_empty(&ad->fifo_list[REQ_SYNC]); 1239 && list_empty(&ad->fifo_list[BLK_RW_SYNC]);
1243} 1240}
1244 1241
1245static int 1242static int
@@ -1346,8 +1343,8 @@ static void as_exit_queue(struct elevator_queue *e)
1346 del_timer_sync(&ad->antic_timer); 1343 del_timer_sync(&ad->antic_timer);
1347 cancel_work_sync(&ad->antic_work); 1344 cancel_work_sync(&ad->antic_work);
1348 1345
1349 BUG_ON(!list_empty(&ad->fifo_list[REQ_SYNC])); 1346 BUG_ON(!list_empty(&ad->fifo_list[BLK_RW_SYNC]));
1350 BUG_ON(!list_empty(&ad->fifo_list[REQ_ASYNC])); 1347 BUG_ON(!list_empty(&ad->fifo_list[BLK_RW_ASYNC]));
1351 1348
1352 put_io_context(ad->io_context); 1349 put_io_context(ad->io_context);
1353 kfree(ad); 1350 kfree(ad);
@@ -1372,18 +1369,18 @@ static void *as_init_queue(struct request_queue *q)
1372 init_timer(&ad->antic_timer); 1369 init_timer(&ad->antic_timer);
1373 INIT_WORK(&ad->antic_work, as_work_handler); 1370 INIT_WORK(&ad->antic_work, as_work_handler);
1374 1371
1375 INIT_LIST_HEAD(&ad->fifo_list[REQ_SYNC]); 1372 INIT_LIST_HEAD(&ad->fifo_list[BLK_RW_SYNC]);
1376 INIT_LIST_HEAD(&ad->fifo_list[REQ_ASYNC]); 1373 INIT_LIST_HEAD(&ad->fifo_list[BLK_RW_ASYNC]);
1377 ad->sort_list[REQ_SYNC] = RB_ROOT; 1374 ad->sort_list[BLK_RW_SYNC] = RB_ROOT;
1378 ad->sort_list[REQ_ASYNC] = RB_ROOT; 1375 ad->sort_list[BLK_RW_ASYNC] = RB_ROOT;
1379 ad->fifo_expire[REQ_SYNC] = default_read_expire; 1376 ad->fifo_expire[BLK_RW_SYNC] = default_read_expire;
1380 ad->fifo_expire[REQ_ASYNC] = default_write_expire; 1377 ad->fifo_expire[BLK_RW_ASYNC] = default_write_expire;
1381 ad->antic_expire = default_antic_expire; 1378 ad->antic_expire = default_antic_expire;
1382 ad->batch_expire[REQ_SYNC] = default_read_batch_expire; 1379 ad->batch_expire[BLK_RW_SYNC] = default_read_batch_expire;
1383 ad->batch_expire[REQ_ASYNC] = default_write_batch_expire; 1380 ad->batch_expire[BLK_RW_ASYNC] = default_write_batch_expire;
1384 1381
1385 ad->current_batch_expires = jiffies + ad->batch_expire[REQ_SYNC]; 1382 ad->current_batch_expires = jiffies + ad->batch_expire[BLK_RW_SYNC];
1386 ad->write_batch_count = ad->batch_expire[REQ_ASYNC] / 10; 1383 ad->write_batch_count = ad->batch_expire[BLK_RW_ASYNC] / 10;
1387 if (ad->write_batch_count < 2) 1384 if (ad->write_batch_count < 2)
1388 ad->write_batch_count = 2; 1385 ad->write_batch_count = 2;
1389 1386
@@ -1432,11 +1429,11 @@ static ssize_t __FUNC(struct elevator_queue *e, char *page) \
1432 struct as_data *ad = e->elevator_data; \ 1429 struct as_data *ad = e->elevator_data; \
1433 return as_var_show(jiffies_to_msecs((__VAR)), (page)); \ 1430 return as_var_show(jiffies_to_msecs((__VAR)), (page)); \
1434} 1431}
1435SHOW_FUNCTION(as_read_expire_show, ad->fifo_expire[REQ_SYNC]); 1432SHOW_FUNCTION(as_read_expire_show, ad->fifo_expire[BLK_RW_SYNC]);
1436SHOW_FUNCTION(as_write_expire_show, ad->fifo_expire[REQ_ASYNC]); 1433SHOW_FUNCTION(as_write_expire_show, ad->fifo_expire[BLK_RW_ASYNC]);
1437SHOW_FUNCTION(as_antic_expire_show, ad->antic_expire); 1434SHOW_FUNCTION(as_antic_expire_show, ad->antic_expire);
1438SHOW_FUNCTION(as_read_batch_expire_show, ad->batch_expire[REQ_SYNC]); 1435SHOW_FUNCTION(as_read_batch_expire_show, ad->batch_expire[BLK_RW_SYNC]);
1439SHOW_FUNCTION(as_write_batch_expire_show, ad->batch_expire[REQ_ASYNC]); 1436SHOW_FUNCTION(as_write_batch_expire_show, ad->batch_expire[BLK_RW_ASYNC]);
1440#undef SHOW_FUNCTION 1437#undef SHOW_FUNCTION
1441 1438
1442#define STORE_FUNCTION(__FUNC, __PTR, MIN, MAX) \ 1439#define STORE_FUNCTION(__FUNC, __PTR, MIN, MAX) \
@@ -1451,13 +1448,14 @@ static ssize_t __FUNC(struct elevator_queue *e, const char *page, size_t count)
1451 *(__PTR) = msecs_to_jiffies(*(__PTR)); \ 1448 *(__PTR) = msecs_to_jiffies(*(__PTR)); \
1452 return ret; \ 1449 return ret; \
1453} 1450}
1454STORE_FUNCTION(as_read_expire_store, &ad->fifo_expire[REQ_SYNC], 0, INT_MAX); 1451STORE_FUNCTION(as_read_expire_store, &ad->fifo_expire[BLK_RW_SYNC], 0, INT_MAX);
1455STORE_FUNCTION(as_write_expire_store, &ad->fifo_expire[REQ_ASYNC], 0, INT_MAX); 1452STORE_FUNCTION(as_write_expire_store,
1453 &ad->fifo_expire[BLK_RW_ASYNC], 0, INT_MAX);
1456STORE_FUNCTION(as_antic_expire_store, &ad->antic_expire, 0, INT_MAX); 1454STORE_FUNCTION(as_antic_expire_store, &ad->antic_expire, 0, INT_MAX);
1457STORE_FUNCTION(as_read_batch_expire_store, 1455STORE_FUNCTION(as_read_batch_expire_store,
1458 &ad->batch_expire[REQ_SYNC], 0, INT_MAX); 1456 &ad->batch_expire[BLK_RW_SYNC], 0, INT_MAX);
1459STORE_FUNCTION(as_write_batch_expire_store, 1457STORE_FUNCTION(as_write_batch_expire_store,
1460 &ad->batch_expire[REQ_ASYNC], 0, INT_MAX); 1458 &ad->batch_expire[BLK_RW_ASYNC], 0, INT_MAX);
1461#undef STORE_FUNCTION 1459#undef STORE_FUNCTION
1462 1460
1463#define AS_ATTR(name) \ 1461#define AS_ATTR(name) \
diff --git a/block/blk-barrier.c b/block/blk-barrier.c
index f7dae57e6cab..20b4111fa050 100644
--- a/block/blk-barrier.c
+++ b/block/blk-barrier.c
@@ -319,9 +319,6 @@ int blkdev_issue_flush(struct block_device *bdev, sector_t *error_sector)
319 return -ENXIO; 319 return -ENXIO;
320 320
321 bio = bio_alloc(GFP_KERNEL, 0); 321 bio = bio_alloc(GFP_KERNEL, 0);
322 if (!bio)
323 return -ENOMEM;
324
325 bio->bi_end_io = bio_end_empty_barrier; 322 bio->bi_end_io = bio_end_empty_barrier;
326 bio->bi_private = &wait; 323 bio->bi_private = &wait;
327 bio->bi_bdev = bdev; 324 bio->bi_bdev = bdev;
diff --git a/block/blk-sysfs.c b/block/blk-sysfs.c
index 73f36beff5cd..cac4e9febe6a 100644
--- a/block/blk-sysfs.c
+++ b/block/blk-sysfs.c
@@ -209,14 +209,14 @@ static ssize_t queue_iostats_store(struct request_queue *q, const char *page,
209 ssize_t ret = queue_var_store(&stats, page, count); 209 ssize_t ret = queue_var_store(&stats, page, count);
210 210
211 spin_lock_irq(q->queue_lock); 211 spin_lock_irq(q->queue_lock);
212 elv_quisce_start(q); 212 elv_quiesce_start(q);
213 213
214 if (stats) 214 if (stats)
215 queue_flag_set(QUEUE_FLAG_IO_STAT, q); 215 queue_flag_set(QUEUE_FLAG_IO_STAT, q);
216 else 216 else
217 queue_flag_clear(QUEUE_FLAG_IO_STAT, q); 217 queue_flag_clear(QUEUE_FLAG_IO_STAT, q);
218 218
219 elv_quisce_end(q); 219 elv_quiesce_end(q);
220 spin_unlock_irq(q->queue_lock); 220 spin_unlock_irq(q->queue_lock);
221 221
222 return ret; 222 return ret;
diff --git a/block/blk.h b/block/blk.h
index 24fcaeeaf620..5dfc41267a08 100644
--- a/block/blk.h
+++ b/block/blk.h
@@ -70,8 +70,8 @@ void blk_queue_congestion_threshold(struct request_queue *q);
70 70
71int blk_dev_init(void); 71int blk_dev_init(void);
72 72
73void elv_quisce_start(struct request_queue *q); 73void elv_quiesce_start(struct request_queue *q);
74void elv_quisce_end(struct request_queue *q); 74void elv_quiesce_end(struct request_queue *q);
75 75
76 76
77/* 77/*
diff --git a/block/cfq-iosched.c b/block/cfq-iosched.c
index a4809de6fea6..0d3b70de3d80 100644
--- a/block/cfq-iosched.c
+++ b/block/cfq-iosched.c
@@ -56,9 +56,6 @@ static DEFINE_SPINLOCK(ioc_gone_lock);
56#define cfq_class_idle(cfqq) ((cfqq)->ioprio_class == IOPRIO_CLASS_IDLE) 56#define cfq_class_idle(cfqq) ((cfqq)->ioprio_class == IOPRIO_CLASS_IDLE)
57#define cfq_class_rt(cfqq) ((cfqq)->ioprio_class == IOPRIO_CLASS_RT) 57#define cfq_class_rt(cfqq) ((cfqq)->ioprio_class == IOPRIO_CLASS_RT)
58 58
59#define ASYNC (0)
60#define SYNC (1)
61
62#define sample_valid(samples) ((samples) > 80) 59#define sample_valid(samples) ((samples) > 80)
63 60
64/* 61/*
@@ -83,6 +80,14 @@ struct cfq_data {
83 * rr list of queues with requests and the count of them 80 * rr list of queues with requests and the count of them
84 */ 81 */
85 struct cfq_rb_root service_tree; 82 struct cfq_rb_root service_tree;
83
84 /*
85 * Each priority tree is sorted by next_request position. These
86 * trees are used when determining if two or more queues are
87 * interleaving requests (see cfq_close_cooperator).
88 */
89 struct rb_root prio_trees[CFQ_PRIO_LISTS];
90
86 unsigned int busy_queues; 91 unsigned int busy_queues;
87 /* 92 /*
88 * Used to track any pending rt requests so we can pre-empt current 93 * Used to track any pending rt requests so we can pre-empt current
@@ -147,6 +152,8 @@ struct cfq_queue {
147 struct rb_node rb_node; 152 struct rb_node rb_node;
148 /* service_tree key */ 153 /* service_tree key */
149 unsigned long rb_key; 154 unsigned long rb_key;
155 /* prio tree member */
156 struct rb_node p_node;
150 /* sorted list of pending requests */ 157 /* sorted list of pending requests */
151 struct rb_root sort_list; 158 struct rb_root sort_list;
152 /* if fifo isn't expired, next request to serve */ 159 /* if fifo isn't expired, next request to serve */
@@ -185,6 +192,7 @@ enum cfqq_state_flags {
185 CFQ_CFQQ_FLAG_prio_changed, /* task priority has changed */ 192 CFQ_CFQQ_FLAG_prio_changed, /* task priority has changed */
186 CFQ_CFQQ_FLAG_slice_new, /* no requests dispatched in slice */ 193 CFQ_CFQQ_FLAG_slice_new, /* no requests dispatched in slice */
187 CFQ_CFQQ_FLAG_sync, /* synchronous queue */ 194 CFQ_CFQQ_FLAG_sync, /* synchronous queue */
195 CFQ_CFQQ_FLAG_coop, /* has done a coop jump of the queue */
188}; 196};
189 197
190#define CFQ_CFQQ_FNS(name) \ 198#define CFQ_CFQQ_FNS(name) \
@@ -211,6 +219,7 @@ CFQ_CFQQ_FNS(idle_window);
211CFQ_CFQQ_FNS(prio_changed); 219CFQ_CFQQ_FNS(prio_changed);
212CFQ_CFQQ_FNS(slice_new); 220CFQ_CFQQ_FNS(slice_new);
213CFQ_CFQQ_FNS(sync); 221CFQ_CFQQ_FNS(sync);
222CFQ_CFQQ_FNS(coop);
214#undef CFQ_CFQQ_FNS 223#undef CFQ_CFQQ_FNS
215 224
216#define cfq_log_cfqq(cfqd, cfqq, fmt, args...) \ 225#define cfq_log_cfqq(cfqd, cfqq, fmt, args...) \
@@ -419,13 +428,17 @@ static struct cfq_queue *cfq_rb_first(struct cfq_rb_root *root)
419 return NULL; 428 return NULL;
420} 429}
421 430
431static void rb_erase_init(struct rb_node *n, struct rb_root *root)
432{
433 rb_erase(n, root);
434 RB_CLEAR_NODE(n);
435}
436
422static void cfq_rb_erase(struct rb_node *n, struct cfq_rb_root *root) 437static void cfq_rb_erase(struct rb_node *n, struct cfq_rb_root *root)
423{ 438{
424 if (root->left == n) 439 if (root->left == n)
425 root->left = NULL; 440 root->left = NULL;
426 441 rb_erase_init(n, &root->rb);
427 rb_erase(n, &root->rb);
428 RB_CLEAR_NODE(n);
429} 442}
430 443
431/* 444/*
@@ -470,8 +483,8 @@ static unsigned long cfq_slice_offset(struct cfq_data *cfqd,
470 * requests waiting to be processed. It is sorted in the order that 483 * requests waiting to be processed. It is sorted in the order that
471 * we will service the queues. 484 * we will service the queues.
472 */ 485 */
473static void cfq_service_tree_add(struct cfq_data *cfqd, 486static void cfq_service_tree_add(struct cfq_data *cfqd, struct cfq_queue *cfqq,
474 struct cfq_queue *cfqq, int add_front) 487 int add_front)
475{ 488{
476 struct rb_node **p, *parent; 489 struct rb_node **p, *parent;
477 struct cfq_queue *__cfqq; 490 struct cfq_queue *__cfqq;
@@ -544,6 +557,63 @@ static void cfq_service_tree_add(struct cfq_data *cfqd,
544 rb_insert_color(&cfqq->rb_node, &cfqd->service_tree.rb); 557 rb_insert_color(&cfqq->rb_node, &cfqd->service_tree.rb);
545} 558}
546 559
560static struct cfq_queue *
561cfq_prio_tree_lookup(struct cfq_data *cfqd, int ioprio, sector_t sector,
562 struct rb_node **ret_parent, struct rb_node ***rb_link)
563{
564 struct rb_root *root = &cfqd->prio_trees[ioprio];
565 struct rb_node **p, *parent;
566 struct cfq_queue *cfqq = NULL;
567
568 parent = NULL;
569 p = &root->rb_node;
570 while (*p) {
571 struct rb_node **n;
572
573 parent = *p;
574 cfqq = rb_entry(parent, struct cfq_queue, p_node);
575
576 /*
577 * Sort strictly based on sector. Smallest to the left,
578 * largest to the right.
579 */
580 if (sector > cfqq->next_rq->sector)
581 n = &(*p)->rb_right;
582 else if (sector < cfqq->next_rq->sector)
583 n = &(*p)->rb_left;
584 else
585 break;
586 p = n;
587 }
588
589 *ret_parent = parent;
590 if (rb_link)
591 *rb_link = p;
592 return NULL;
593}
594
595static void cfq_prio_tree_add(struct cfq_data *cfqd, struct cfq_queue *cfqq)
596{
597 struct rb_root *root = &cfqd->prio_trees[cfqq->ioprio];
598 struct rb_node **p, *parent;
599 struct cfq_queue *__cfqq;
600
601 if (!RB_EMPTY_NODE(&cfqq->p_node))
602 rb_erase_init(&cfqq->p_node, root);
603
604 if (cfq_class_idle(cfqq))
605 return;
606 if (!cfqq->next_rq)
607 return;
608
609 __cfqq = cfq_prio_tree_lookup(cfqd, cfqq->ioprio, cfqq->next_rq->sector,
610 &parent, &p);
611 BUG_ON(__cfqq);
612
613 rb_link_node(&cfqq->p_node, parent, p);
614 rb_insert_color(&cfqq->p_node, root);
615}
616
547/* 617/*
548 * Update cfqq's position in the service tree. 618 * Update cfqq's position in the service tree.
549 */ 619 */
@@ -552,8 +622,10 @@ static void cfq_resort_rr_list(struct cfq_data *cfqd, struct cfq_queue *cfqq)
552 /* 622 /*
553 * Resorting requires the cfqq to be on the RR list already. 623 * Resorting requires the cfqq to be on the RR list already.
554 */ 624 */
555 if (cfq_cfqq_on_rr(cfqq)) 625 if (cfq_cfqq_on_rr(cfqq)) {
556 cfq_service_tree_add(cfqd, cfqq, 0); 626 cfq_service_tree_add(cfqd, cfqq, 0);
627 cfq_prio_tree_add(cfqd, cfqq);
628 }
557} 629}
558 630
559/* 631/*
@@ -584,6 +656,8 @@ static void cfq_del_cfqq_rr(struct cfq_data *cfqd, struct cfq_queue *cfqq)
584 656
585 if (!RB_EMPTY_NODE(&cfqq->rb_node)) 657 if (!RB_EMPTY_NODE(&cfqq->rb_node))
586 cfq_rb_erase(&cfqq->rb_node, &cfqd->service_tree); 658 cfq_rb_erase(&cfqq->rb_node, &cfqd->service_tree);
659 if (!RB_EMPTY_NODE(&cfqq->p_node))
660 rb_erase_init(&cfqq->p_node, &cfqd->prio_trees[cfqq->ioprio]);
587 661
588 BUG_ON(!cfqd->busy_queues); 662 BUG_ON(!cfqd->busy_queues);
589 cfqd->busy_queues--; 663 cfqd->busy_queues--;
@@ -613,7 +687,7 @@ static void cfq_add_rq_rb(struct request *rq)
613{ 687{
614 struct cfq_queue *cfqq = RQ_CFQQ(rq); 688 struct cfq_queue *cfqq = RQ_CFQQ(rq);
615 struct cfq_data *cfqd = cfqq->cfqd; 689 struct cfq_data *cfqd = cfqq->cfqd;
616 struct request *__alias; 690 struct request *__alias, *prev;
617 691
618 cfqq->queued[rq_is_sync(rq)]++; 692 cfqq->queued[rq_is_sync(rq)]++;
619 693
@@ -630,7 +704,15 @@ static void cfq_add_rq_rb(struct request *rq)
630 /* 704 /*
631 * check if this request is a better next-serve candidate 705 * check if this request is a better next-serve candidate
632 */ 706 */
707 prev = cfqq->next_rq;
633 cfqq->next_rq = cfq_choose_req(cfqd, cfqq->next_rq, rq); 708 cfqq->next_rq = cfq_choose_req(cfqd, cfqq->next_rq, rq);
709
710 /*
711 * adjust priority tree position, if ->next_rq changes
712 */
713 if (prev != cfqq->next_rq)
714 cfq_prio_tree_add(cfqd, cfqq);
715
634 BUG_ON(!cfqq->next_rq); 716 BUG_ON(!cfqq->next_rq);
635} 717}
636 718
@@ -843,11 +925,15 @@ static struct cfq_queue *cfq_get_next_queue(struct cfq_data *cfqd)
843/* 925/*
844 * Get and set a new active queue for service. 926 * Get and set a new active queue for service.
845 */ 927 */
846static struct cfq_queue *cfq_set_active_queue(struct cfq_data *cfqd) 928static struct cfq_queue *cfq_set_active_queue(struct cfq_data *cfqd,
929 struct cfq_queue *cfqq)
847{ 930{
848 struct cfq_queue *cfqq; 931 if (!cfqq) {
932 cfqq = cfq_get_next_queue(cfqd);
933 if (cfqq)
934 cfq_clear_cfqq_coop(cfqq);
935 }
849 936
850 cfqq = cfq_get_next_queue(cfqd);
851 __cfq_set_active_queue(cfqd, cfqq); 937 __cfq_set_active_queue(cfqd, cfqq);
852 return cfqq; 938 return cfqq;
853} 939}
@@ -871,17 +957,89 @@ static inline int cfq_rq_close(struct cfq_data *cfqd, struct request *rq)
871 return cfq_dist_from_last(cfqd, rq) <= cic->seek_mean; 957 return cfq_dist_from_last(cfqd, rq) <= cic->seek_mean;
872} 958}
873 959
874static int cfq_close_cooperator(struct cfq_data *cfq_data, 960static struct cfq_queue *cfqq_close(struct cfq_data *cfqd,
875 struct cfq_queue *cfqq) 961 struct cfq_queue *cur_cfqq)
962{
963 struct rb_root *root = &cfqd->prio_trees[cur_cfqq->ioprio];
964 struct rb_node *parent, *node;
965 struct cfq_queue *__cfqq;
966 sector_t sector = cfqd->last_position;
967
968 if (RB_EMPTY_ROOT(root))
969 return NULL;
970
971 /*
972 * First, if we find a request starting at the end of the last
973 * request, choose it.
974 */
975 __cfqq = cfq_prio_tree_lookup(cfqd, cur_cfqq->ioprio,
976 sector, &parent, NULL);
977 if (__cfqq)
978 return __cfqq;
979
980 /*
981 * If the exact sector wasn't found, the parent of the NULL leaf
982 * will contain the closest sector.
983 */
984 __cfqq = rb_entry(parent, struct cfq_queue, p_node);
985 if (cfq_rq_close(cfqd, __cfqq->next_rq))
986 return __cfqq;
987
988 if (__cfqq->next_rq->sector < sector)
989 node = rb_next(&__cfqq->p_node);
990 else
991 node = rb_prev(&__cfqq->p_node);
992 if (!node)
993 return NULL;
994
995 __cfqq = rb_entry(node, struct cfq_queue, p_node);
996 if (cfq_rq_close(cfqd, __cfqq->next_rq))
997 return __cfqq;
998
999 return NULL;
1000}
1001
1002/*
1003 * cfqd - obvious
1004 * cur_cfqq - passed in so that we don't decide that the current queue is
1005 * closely cooperating with itself.
1006 *
1007 * So, basically we're assuming that that cur_cfqq has dispatched at least
1008 * one request, and that cfqd->last_position reflects a position on the disk
1009 * associated with the I/O issued by cur_cfqq. I'm not sure this is a valid
1010 * assumption.
1011 */
1012static struct cfq_queue *cfq_close_cooperator(struct cfq_data *cfqd,
1013 struct cfq_queue *cur_cfqq,
1014 int probe)
876{ 1015{
1016 struct cfq_queue *cfqq;
1017
1018 /*
1019 * A valid cfq_io_context is necessary to compare requests against
1020 * the seek_mean of the current cfqq.
1021 */
1022 if (!cfqd->active_cic)
1023 return NULL;
1024
877 /* 1025 /*
878 * We should notice if some of the queues are cooperating, eg 1026 * We should notice if some of the queues are cooperating, eg
879 * working closely on the same area of the disk. In that case, 1027 * working closely on the same area of the disk. In that case,
880 * we can group them together and don't waste time idling. 1028 * we can group them together and don't waste time idling.
881 */ 1029 */
882 return 0; 1030 cfqq = cfqq_close(cfqd, cur_cfqq);
1031 if (!cfqq)
1032 return NULL;
1033
1034 if (cfq_cfqq_coop(cfqq))
1035 return NULL;
1036
1037 if (!probe)
1038 cfq_mark_cfqq_coop(cfqq);
1039 return cfqq;
883} 1040}
884 1041
1042
885#define CIC_SEEKY(cic) ((cic)->seek_mean > (8 * 1024)) 1043#define CIC_SEEKY(cic) ((cic)->seek_mean > (8 * 1024))
886 1044
887static void cfq_arm_slice_timer(struct cfq_data *cfqd) 1045static void cfq_arm_slice_timer(struct cfq_data *cfqd)
@@ -920,13 +1078,6 @@ static void cfq_arm_slice_timer(struct cfq_data *cfqd)
920 if (!cic || !atomic_read(&cic->ioc->nr_tasks)) 1078 if (!cic || !atomic_read(&cic->ioc->nr_tasks))
921 return; 1079 return;
922 1080
923 /*
924 * See if this prio level has a good candidate
925 */
926 if (cfq_close_cooperator(cfqd, cfqq) &&
927 (sample_valid(cic->ttime_samples) && cic->ttime_mean > 2))
928 return;
929
930 cfq_mark_cfqq_wait_request(cfqq); 1081 cfq_mark_cfqq_wait_request(cfqq);
931 1082
932 /* 1083 /*
@@ -939,7 +1090,7 @@ static void cfq_arm_slice_timer(struct cfq_data *cfqd)
939 sl = min(sl, msecs_to_jiffies(CFQ_MIN_TT)); 1090 sl = min(sl, msecs_to_jiffies(CFQ_MIN_TT));
940 1091
941 mod_timer(&cfqd->idle_slice_timer, jiffies + sl); 1092 mod_timer(&cfqd->idle_slice_timer, jiffies + sl);
942 cfq_log(cfqd, "arm_idle: %lu", sl); 1093 cfq_log_cfqq(cfqd, cfqq, "arm_idle: %lu", sl);
943} 1094}
944 1095
945/* 1096/*
@@ -1003,7 +1154,7 @@ cfq_prio_to_maxrq(struct cfq_data *cfqd, struct cfq_queue *cfqq)
1003 */ 1154 */
1004static struct cfq_queue *cfq_select_queue(struct cfq_data *cfqd) 1155static struct cfq_queue *cfq_select_queue(struct cfq_data *cfqd)
1005{ 1156{
1006 struct cfq_queue *cfqq; 1157 struct cfq_queue *cfqq, *new_cfqq = NULL;
1007 1158
1008 cfqq = cfqd->active_queue; 1159 cfqq = cfqd->active_queue;
1009 if (!cfqq) 1160 if (!cfqq)
@@ -1037,6 +1188,16 @@ static struct cfq_queue *cfq_select_queue(struct cfq_data *cfqd)
1037 goto keep_queue; 1188 goto keep_queue;
1038 1189
1039 /* 1190 /*
1191 * If another queue has a request waiting within our mean seek
1192 * distance, let it run. The expire code will check for close
1193 * cooperators and put the close queue at the front of the service
1194 * tree.
1195 */
1196 new_cfqq = cfq_close_cooperator(cfqd, cfqq, 0);
1197 if (new_cfqq)
1198 goto expire;
1199
1200 /*
1040 * No requests pending. If the active queue still has requests in 1201 * No requests pending. If the active queue still has requests in
1041 * flight or is idling for a new request, allow either of these 1202 * flight or is idling for a new request, allow either of these
1042 * conditions to happen (or time out) before selecting a new queue. 1203 * conditions to happen (or time out) before selecting a new queue.
@@ -1050,7 +1211,7 @@ static struct cfq_queue *cfq_select_queue(struct cfq_data *cfqd)
1050expire: 1211expire:
1051 cfq_slice_expired(cfqd, 0); 1212 cfq_slice_expired(cfqd, 0);
1052new_queue: 1213new_queue:
1053 cfqq = cfq_set_active_queue(cfqd); 1214 cfqq = cfq_set_active_queue(cfqd, new_cfqq);
1054keep_queue: 1215keep_queue:
1055 return cfqq; 1216 return cfqq;
1056} 1217}
@@ -1333,14 +1494,14 @@ static void __cfq_exit_single_io_context(struct cfq_data *cfqd,
1333 if (ioc->ioc_data == cic) 1494 if (ioc->ioc_data == cic)
1334 rcu_assign_pointer(ioc->ioc_data, NULL); 1495 rcu_assign_pointer(ioc->ioc_data, NULL);
1335 1496
1336 if (cic->cfqq[ASYNC]) { 1497 if (cic->cfqq[BLK_RW_ASYNC]) {
1337 cfq_exit_cfqq(cfqd, cic->cfqq[ASYNC]); 1498 cfq_exit_cfqq(cfqd, cic->cfqq[BLK_RW_ASYNC]);
1338 cic->cfqq[ASYNC] = NULL; 1499 cic->cfqq[BLK_RW_ASYNC] = NULL;
1339 } 1500 }
1340 1501
1341 if (cic->cfqq[SYNC]) { 1502 if (cic->cfqq[BLK_RW_SYNC]) {
1342 cfq_exit_cfqq(cfqd, cic->cfqq[SYNC]); 1503 cfq_exit_cfqq(cfqd, cic->cfqq[BLK_RW_SYNC]);
1343 cic->cfqq[SYNC] = NULL; 1504 cic->cfqq[BLK_RW_SYNC] = NULL;
1344 } 1505 }
1345} 1506}
1346 1507
@@ -1449,17 +1610,18 @@ static void changed_ioprio(struct io_context *ioc, struct cfq_io_context *cic)
1449 1610
1450 spin_lock_irqsave(cfqd->queue->queue_lock, flags); 1611 spin_lock_irqsave(cfqd->queue->queue_lock, flags);
1451 1612
1452 cfqq = cic->cfqq[ASYNC]; 1613 cfqq = cic->cfqq[BLK_RW_ASYNC];
1453 if (cfqq) { 1614 if (cfqq) {
1454 struct cfq_queue *new_cfqq; 1615 struct cfq_queue *new_cfqq;
1455 new_cfqq = cfq_get_queue(cfqd, ASYNC, cic->ioc, GFP_ATOMIC); 1616 new_cfqq = cfq_get_queue(cfqd, BLK_RW_ASYNC, cic->ioc,
1617 GFP_ATOMIC);
1456 if (new_cfqq) { 1618 if (new_cfqq) {
1457 cic->cfqq[ASYNC] = new_cfqq; 1619 cic->cfqq[BLK_RW_ASYNC] = new_cfqq;
1458 cfq_put_queue(cfqq); 1620 cfq_put_queue(cfqq);
1459 } 1621 }
1460 } 1622 }
1461 1623
1462 cfqq = cic->cfqq[SYNC]; 1624 cfqq = cic->cfqq[BLK_RW_SYNC];
1463 if (cfqq) 1625 if (cfqq)
1464 cfq_mark_cfqq_prio_changed(cfqq); 1626 cfq_mark_cfqq_prio_changed(cfqq);
1465 1627
@@ -1510,6 +1672,7 @@ retry:
1510 } 1672 }
1511 1673
1512 RB_CLEAR_NODE(&cfqq->rb_node); 1674 RB_CLEAR_NODE(&cfqq->rb_node);
1675 RB_CLEAR_NODE(&cfqq->p_node);
1513 INIT_LIST_HEAD(&cfqq->fifo); 1676 INIT_LIST_HEAD(&cfqq->fifo);
1514 1677
1515 atomic_set(&cfqq->ref, 0); 1678 atomic_set(&cfqq->ref, 0);
@@ -1905,10 +2068,20 @@ cfq_rq_enqueued(struct cfq_data *cfqd, struct cfq_queue *cfqq,
1905 * Remember that we saw a request from this process, but 2068 * Remember that we saw a request from this process, but
1906 * don't start queuing just yet. Otherwise we risk seeing lots 2069 * don't start queuing just yet. Otherwise we risk seeing lots
1907 * of tiny requests, because we disrupt the normal plugging 2070 * of tiny requests, because we disrupt the normal plugging
1908 * and merging. 2071 * and merging. If the request is already larger than a single
2072 * page, let it rip immediately. For that case we assume that
2073 * merging is already done. Ditto for a busy system that
2074 * has other work pending, don't risk delaying until the
2075 * idle timer unplug to continue working.
1909 */ 2076 */
1910 if (cfq_cfqq_wait_request(cfqq)) 2077 if (cfq_cfqq_wait_request(cfqq)) {
2078 if (blk_rq_bytes(rq) > PAGE_CACHE_SIZE ||
2079 cfqd->busy_queues > 1) {
2080 del_timer(&cfqd->idle_slice_timer);
2081 blk_start_queueing(cfqd->queue);
2082 }
1911 cfq_mark_cfqq_must_dispatch(cfqq); 2083 cfq_mark_cfqq_must_dispatch(cfqq);
2084 }
1912 } else if (cfq_should_preempt(cfqd, cfqq, rq)) { 2085 } else if (cfq_should_preempt(cfqd, cfqq, rq)) {
1913 /* 2086 /*
1914 * not the active queue - expire current slice if it is 2087 * not the active queue - expire current slice if it is
@@ -1992,16 +2165,24 @@ static void cfq_completed_request(struct request_queue *q, struct request *rq)
1992 * or if we want to idle in case it has no pending requests. 2165 * or if we want to idle in case it has no pending requests.
1993 */ 2166 */
1994 if (cfqd->active_queue == cfqq) { 2167 if (cfqd->active_queue == cfqq) {
2168 const bool cfqq_empty = RB_EMPTY_ROOT(&cfqq->sort_list);
2169
1995 if (cfq_cfqq_slice_new(cfqq)) { 2170 if (cfq_cfqq_slice_new(cfqq)) {
1996 cfq_set_prio_slice(cfqd, cfqq); 2171 cfq_set_prio_slice(cfqd, cfqq);
1997 cfq_clear_cfqq_slice_new(cfqq); 2172 cfq_clear_cfqq_slice_new(cfqq);
1998 } 2173 }
2174 /*
2175 * If there are no requests waiting in this queue, and
2176 * there are other queues ready to issue requests, AND
2177 * those other queues are issuing requests within our
2178 * mean seek distance, give them a chance to run instead
2179 * of idling.
2180 */
1999 if (cfq_slice_used(cfqq) || cfq_class_idle(cfqq)) 2181 if (cfq_slice_used(cfqq) || cfq_class_idle(cfqq))
2000 cfq_slice_expired(cfqd, 1); 2182 cfq_slice_expired(cfqd, 1);
2001 else if (sync && !rq_noidle(rq) && 2183 else if (cfqq_empty && !cfq_close_cooperator(cfqd, cfqq, 1) &&
2002 RB_EMPTY_ROOT(&cfqq->sort_list)) { 2184 sync && !rq_noidle(rq))
2003 cfq_arm_slice_timer(cfqd); 2185 cfq_arm_slice_timer(cfqd);
2004 }
2005 } 2186 }
2006 2187
2007 if (!cfqd->rq_in_driver) 2188 if (!cfqd->rq_in_driver)
@@ -2062,7 +2243,7 @@ static int cfq_may_queue(struct request_queue *q, int rw)
2062 if (!cic) 2243 if (!cic)
2063 return ELV_MQUEUE_MAY; 2244 return ELV_MQUEUE_MAY;
2064 2245
2065 cfqq = cic_to_cfqq(cic, rw & REQ_RW_SYNC); 2246 cfqq = cic_to_cfqq(cic, rw_is_sync(rw));
2066 if (cfqq) { 2247 if (cfqq) {
2067 cfq_init_prio_data(cfqq, cic->ioc); 2248 cfq_init_prio_data(cfqq, cic->ioc);
2068 cfq_prio_boost(cfqq); 2249 cfq_prio_boost(cfqq);
@@ -2152,11 +2333,10 @@ static void cfq_kick_queue(struct work_struct *work)
2152 struct cfq_data *cfqd = 2333 struct cfq_data *cfqd =
2153 container_of(work, struct cfq_data, unplug_work); 2334 container_of(work, struct cfq_data, unplug_work);
2154 struct request_queue *q = cfqd->queue; 2335 struct request_queue *q = cfqd->queue;
2155 unsigned long flags;
2156 2336
2157 spin_lock_irqsave(q->queue_lock, flags); 2337 spin_lock_irq(q->queue_lock);
2158 blk_start_queueing(q); 2338 blk_start_queueing(q);
2159 spin_unlock_irqrestore(q->queue_lock, flags); 2339 spin_unlock_irq(q->queue_lock);
2160} 2340}
2161 2341
2162/* 2342/*
diff --git a/block/elevator.c b/block/elevator.c
index fb81bcc14a8c..7073a9072577 100644
--- a/block/elevator.c
+++ b/block/elevator.c
@@ -590,7 +590,7 @@ void elv_drain_elevator(struct request_queue *q)
590/* 590/*
591 * Call with queue lock held, interrupts disabled 591 * Call with queue lock held, interrupts disabled
592 */ 592 */
593void elv_quisce_start(struct request_queue *q) 593void elv_quiesce_start(struct request_queue *q)
594{ 594{
595 queue_flag_set(QUEUE_FLAG_ELVSWITCH, q); 595 queue_flag_set(QUEUE_FLAG_ELVSWITCH, q);
596 596
@@ -607,7 +607,7 @@ void elv_quisce_start(struct request_queue *q)
607 } 607 }
608} 608}
609 609
610void elv_quisce_end(struct request_queue *q) 610void elv_quiesce_end(struct request_queue *q)
611{ 611{
612 queue_flag_clear(QUEUE_FLAG_ELVSWITCH, q); 612 queue_flag_clear(QUEUE_FLAG_ELVSWITCH, q);
613} 613}
@@ -1126,7 +1126,7 @@ static int elevator_switch(struct request_queue *q, struct elevator_type *new_e)
1126 * Turn on BYPASS and drain all requests w/ elevator private data 1126 * Turn on BYPASS and drain all requests w/ elevator private data
1127 */ 1127 */
1128 spin_lock_irq(q->queue_lock); 1128 spin_lock_irq(q->queue_lock);
1129 elv_quisce_start(q); 1129 elv_quiesce_start(q);
1130 1130
1131 /* 1131 /*
1132 * Remember old elevator. 1132 * Remember old elevator.
@@ -1150,7 +1150,7 @@ static int elevator_switch(struct request_queue *q, struct elevator_type *new_e)
1150 */ 1150 */
1151 elevator_exit(old_elevator); 1151 elevator_exit(old_elevator);
1152 spin_lock_irq(q->queue_lock); 1152 spin_lock_irq(q->queue_lock);
1153 elv_quisce_end(q); 1153 elv_quiesce_end(q);
1154 spin_unlock_irq(q->queue_lock); 1154 spin_unlock_irq(q->queue_lock);
1155 1155
1156 blk_add_trace_msg(q, "elv switch: %s", e->elevator_type->elevator_name); 1156 blk_add_trace_msg(q, "elv switch: %s", e->elevator_type->elevator_name);
diff --git a/block/ioctl.c b/block/ioctl.c
index 0f22e629b13c..ad474d4bbcce 100644
--- a/block/ioctl.c
+++ b/block/ioctl.c
@@ -146,8 +146,6 @@ static int blk_ioctl_discard(struct block_device *bdev, uint64_t start,
146 struct bio *bio; 146 struct bio *bio;
147 147
148 bio = bio_alloc(GFP_KERNEL, 0); 148 bio = bio_alloc(GFP_KERNEL, 0);
149 if (!bio)
150 return -ENOMEM;
151 149
152 bio->bi_end_io = blk_ioc_discard_endio; 150 bio->bi_end_io = blk_ioc_discard_endio;
153 bio->bi_bdev = bdev; 151 bio->bi_bdev = bdev;
diff --git a/block/scsi_ioctl.c b/block/scsi_ioctl.c
index 626ee274c5c4..84b7f8709f41 100644
--- a/block/scsi_ioctl.c
+++ b/block/scsi_ioctl.c
@@ -217,7 +217,7 @@ static int blk_fill_sghdr_rq(struct request_queue *q, struct request *rq,
217static int blk_complete_sghdr_rq(struct request *rq, struct sg_io_hdr *hdr, 217static int blk_complete_sghdr_rq(struct request *rq, struct sg_io_hdr *hdr,
218 struct bio *bio) 218 struct bio *bio)
219{ 219{
220 int ret = 0; 220 int r, ret = 0;
221 221
222 /* 222 /*
223 * fill in all the output members 223 * fill in all the output members
@@ -242,7 +242,9 @@ static int blk_complete_sghdr_rq(struct request *rq, struct sg_io_hdr *hdr,
242 ret = -EFAULT; 242 ret = -EFAULT;
243 } 243 }
244 244
245 blk_rq_unmap_user(bio); 245 r = blk_rq_unmap_user(bio);
246 if (!ret)
247 ret = r;
246 blk_put_request(rq); 248 blk_put_request(rq);
247 249
248 return ret; 250 return ret;
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 065507c46644..17c5d48a75d2 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -1231,6 +1231,9 @@ unsigned int ata_dev_classify(const struct ata_taskfile *tf)
1231 * 1231 *
1232 * We follow the current spec and consider that 0x69/0x96 1232 * We follow the current spec and consider that 0x69/0x96
1233 * identifies a port multiplier and 0x3c/0xc3 a SEMB device. 1233 * identifies a port multiplier and 0x3c/0xc3 a SEMB device.
1234 * Unfortunately, WDC WD1600JS-62MHB5 (a hard drive) reports
1235 * SEMB signature. This is worked around in
1236 * ata_dev_read_id().
1234 */ 1237 */
1235 if ((tf->lbam == 0) && (tf->lbah == 0)) { 1238 if ((tf->lbam == 0) && (tf->lbah == 0)) {
1236 DPRINTK("found ATA device by sig\n"); 1239 DPRINTK("found ATA device by sig\n");
@@ -1248,8 +1251,8 @@ unsigned int ata_dev_classify(const struct ata_taskfile *tf)
1248 } 1251 }
1249 1252
1250 if ((tf->lbam == 0x3c) && (tf->lbah == 0xc3)) { 1253 if ((tf->lbam == 0x3c) && (tf->lbah == 0xc3)) {
1251 printk(KERN_INFO "ata: SEMB device ignored\n"); 1254 DPRINTK("found SEMB device by sig (could be ATA device)\n");
1252 return ATA_DEV_SEMB_UNSUP; /* not yet */ 1255 return ATA_DEV_SEMB;
1253 } 1256 }
1254 1257
1255 DPRINTK("unknown device\n"); 1258 DPRINTK("unknown device\n");
@@ -1653,8 +1656,8 @@ unsigned long ata_id_xfermask(const u16 *id)
1653 /* 1656 /*
1654 * Process compact flash extended modes 1657 * Process compact flash extended modes
1655 */ 1658 */
1656 int pio = id[163] & 0x7; 1659 int pio = (id[ATA_ID_CFA_MODES] >> 0) & 0x7;
1657 int dma = (id[163] >> 3) & 7; 1660 int dma = (id[ATA_ID_CFA_MODES] >> 3) & 0x7;
1658 1661
1659 if (pio) 1662 if (pio)
1660 pio_mask |= (1 << 5); 1663 pio_mask |= (1 << 5);
@@ -2080,6 +2083,7 @@ int ata_dev_read_id(struct ata_device *dev, unsigned int *p_class,
2080 struct ata_taskfile tf; 2083 struct ata_taskfile tf;
2081 unsigned int err_mask = 0; 2084 unsigned int err_mask = 0;
2082 const char *reason; 2085 const char *reason;
2086 bool is_semb = class == ATA_DEV_SEMB;
2083 int may_fallback = 1, tried_spinup = 0; 2087 int may_fallback = 1, tried_spinup = 0;
2084 int rc; 2088 int rc;
2085 2089
@@ -2090,6 +2094,8 @@ retry:
2090 ata_tf_init(dev, &tf); 2094 ata_tf_init(dev, &tf);
2091 2095
2092 switch (class) { 2096 switch (class) {
2097 case ATA_DEV_SEMB:
2098 class = ATA_DEV_ATA; /* some hard drives report SEMB sig */
2093 case ATA_DEV_ATA: 2099 case ATA_DEV_ATA:
2094 tf.command = ATA_CMD_ID_ATA; 2100 tf.command = ATA_CMD_ID_ATA;
2095 break; 2101 break;
@@ -2126,6 +2132,14 @@ retry:
2126 return -ENOENT; 2132 return -ENOENT;
2127 } 2133 }
2128 2134
2135 if (is_semb) {
2136 ata_dev_printk(dev, KERN_INFO, "IDENTIFY failed on "
2137 "device w/ SEMB sig, disabled\n");
2138 /* SEMB is not supported yet */
2139 *p_class = ATA_DEV_SEMB_UNSUP;
2140 return 0;
2141 }
2142
2129 if ((err_mask == AC_ERR_DEV) && (tf.feature & ATA_ABORTED)) { 2143 if ((err_mask == AC_ERR_DEV) && (tf.feature & ATA_ABORTED)) {
2130 /* Device or controller might have reported 2144 /* Device or controller might have reported
2131 * the wrong device class. Give a shot at the 2145 * the wrong device class. Give a shot at the
@@ -2412,7 +2426,8 @@ int ata_dev_configure(struct ata_device *dev)
2412 /* ATA-specific feature tests */ 2426 /* ATA-specific feature tests */
2413 if (dev->class == ATA_DEV_ATA) { 2427 if (dev->class == ATA_DEV_ATA) {
2414 if (ata_id_is_cfa(id)) { 2428 if (ata_id_is_cfa(id)) {
2415 if (id[162] & 1) /* CPRM may make this media unusable */ 2429 /* CPRM may make this media unusable */
2430 if (id[ATA_ID_CFA_KEY_MGMT] & 1)
2416 ata_dev_printk(dev, KERN_WARNING, 2431 ata_dev_printk(dev, KERN_WARNING,
2417 "supports DRM functions and may " 2432 "supports DRM functions and may "
2418 "not be fully accessable.\n"); 2433 "not be fully accessable.\n");
diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
index b9747fa59e54..2733b0c90b75 100644
--- a/drivers/ata/libata-scsi.c
+++ b/drivers/ata/libata-scsi.c
@@ -647,23 +647,45 @@ int ata_task_ioctl(struct scsi_device *scsidev, void __user *arg)
647 return rc; 647 return rc;
648} 648}
649 649
650static int ata_ioc32(struct ata_port *ap)
651{
652 if (ap->flags & ATA_FLAG_PIO_DMA)
653 return 1;
654 if (ap->pflags & ATA_PFLAG_PIO32)
655 return 1;
656 return 0;
657}
658
650int ata_sas_scsi_ioctl(struct ata_port *ap, struct scsi_device *scsidev, 659int ata_sas_scsi_ioctl(struct ata_port *ap, struct scsi_device *scsidev,
651 int cmd, void __user *arg) 660 int cmd, void __user *arg)
652{ 661{
653 int val = -EINVAL, rc = -EINVAL; 662 int val = -EINVAL, rc = -EINVAL;
663 unsigned long flags;
654 664
655 switch (cmd) { 665 switch (cmd) {
656 case ATA_IOC_GET_IO32: 666 case ATA_IOC_GET_IO32:
657 val = 0; 667 spin_lock_irqsave(ap->lock, flags);
668 val = ata_ioc32(ap);
669 spin_unlock_irqrestore(ap->lock, flags);
658 if (copy_to_user(arg, &val, 1)) 670 if (copy_to_user(arg, &val, 1))
659 return -EFAULT; 671 return -EFAULT;
660 return 0; 672 return 0;
661 673
662 case ATA_IOC_SET_IO32: 674 case ATA_IOC_SET_IO32:
663 val = (unsigned long) arg; 675 val = (unsigned long) arg;
664 if (val != 0) 676 rc = 0;
665 return -EINVAL; 677 spin_lock_irqsave(ap->lock, flags);
666 return 0; 678 if (ap->pflags & ATA_PFLAG_PIO32CHANGE) {
679 if (val)
680 ap->pflags |= ATA_PFLAG_PIO32;
681 else
682 ap->pflags &= ~ATA_PFLAG_PIO32;
683 } else {
684 if (val != ata_ioc32(ap))
685 rc = -EINVAL;
686 }
687 spin_unlock_irqrestore(ap->lock, flags);
688 return rc;
667 689
668 case HDIO_GET_IDENTITY: 690 case HDIO_GET_IDENTITY:
669 return ata_get_identity(ap, scsidev, arg); 691 return ata_get_identity(ap, scsidev, arg);
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index 8332e97a9de3..bb18415d3d63 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -87,6 +87,7 @@ const struct ata_port_operations ata_bmdma32_port_ops = {
87 .inherits = &ata_bmdma_port_ops, 87 .inherits = &ata_bmdma_port_ops,
88 88
89 .sff_data_xfer = ata_sff_data_xfer32, 89 .sff_data_xfer = ata_sff_data_xfer32,
90 .port_start = ata_sff_port_start32,
90}; 91};
91EXPORT_SYMBOL_GPL(ata_bmdma32_port_ops); 92EXPORT_SYMBOL_GPL(ata_bmdma32_port_ops);
92 93
@@ -769,6 +770,9 @@ unsigned int ata_sff_data_xfer32(struct ata_device *dev, unsigned char *buf,
769 void __iomem *data_addr = ap->ioaddr.data_addr; 770 void __iomem *data_addr = ap->ioaddr.data_addr;
770 unsigned int words = buflen >> 2; 771 unsigned int words = buflen >> 2;
771 int slop = buflen & 3; 772 int slop = buflen & 3;
773
774 if (!(ap->pflags & ATA_PFLAG_PIO32))
775 return ata_sff_data_xfer(dev, buf, buflen, rw);
772 776
773 /* Transfer multiple of 4 bytes */ 777 /* Transfer multiple of 4 bytes */
774 if (rw == READ) 778 if (rw == READ)
@@ -2402,6 +2406,29 @@ int ata_sff_port_start(struct ata_port *ap)
2402EXPORT_SYMBOL_GPL(ata_sff_port_start); 2406EXPORT_SYMBOL_GPL(ata_sff_port_start);
2403 2407
2404/** 2408/**
2409 * ata_sff_port_start32 - Set port up for dma.
2410 * @ap: Port to initialize
2411 *
2412 * Called just after data structures for each port are
2413 * initialized. Allocates space for PRD table if the device
2414 * is DMA capable SFF.
2415 *
2416 * May be used as the port_start() entry in ata_port_operations for
2417 * devices that are capable of 32bit PIO.
2418 *
2419 * LOCKING:
2420 * Inherited from caller.
2421 */
2422int ata_sff_port_start32(struct ata_port *ap)
2423{
2424 ap->pflags |= ATA_PFLAG_PIO32 | ATA_PFLAG_PIO32CHANGE;
2425 if (ap->ioaddr.bmdma_addr)
2426 return ata_port_start(ap);
2427 return 0;
2428}
2429EXPORT_SYMBOL_GPL(ata_sff_port_start32);
2430
2431/**
2405 * ata_sff_std_ports - initialize ioaddr with standard port offsets. 2432 * ata_sff_std_ports - initialize ioaddr with standard port offsets.
2406 * @ioaddr: IO address structure to be initialized 2433 * @ioaddr: IO address structure to be initialized
2407 * 2434 *
diff --git a/drivers/ata/pata_hpt37x.c b/drivers/ata/pata_hpt37x.c
index 81ab57003aba..122c786449a9 100644
--- a/drivers/ata/pata_hpt37x.c
+++ b/drivers/ata/pata_hpt37x.c
@@ -8,7 +8,7 @@
8 * Copyright (C) 1999-2003 Andre Hedrick <andre@linux-ide.org> 8 * Copyright (C) 1999-2003 Andre Hedrick <andre@linux-ide.org>
9 * Portions Copyright (C) 2001 Sun Microsystems, Inc. 9 * Portions Copyright (C) 2001 Sun Microsystems, Inc.
10 * Portions Copyright (C) 2003 Red Hat Inc 10 * Portions Copyright (C) 2003 Red Hat Inc
11 * Portions Copyright (C) 2005-2007 MontaVista Software, Inc. 11 * Portions Copyright (C) 2005-2009 MontaVista Software, Inc.
12 * 12 *
13 * TODO 13 * TODO
14 * Look into engine reset on timeout errors. Should not be required. 14 * Look into engine reset on timeout errors. Should not be required.
@@ -24,7 +24,7 @@
24#include <linux/libata.h> 24#include <linux/libata.h>
25 25
26#define DRV_NAME "pata_hpt37x" 26#define DRV_NAME "pata_hpt37x"
27#define DRV_VERSION "0.6.11" 27#define DRV_VERSION "0.6.12"
28 28
29struct hpt_clock { 29struct hpt_clock {
30 u8 xfer_speed; 30 u8 xfer_speed;
@@ -445,23 +445,6 @@ static void hpt370_set_dmamode(struct ata_port *ap, struct ata_device *adev)
445} 445}
446 446
447/** 447/**
448 * hpt370_bmdma_start - DMA engine begin
449 * @qc: ATA command
450 *
451 * The 370 and 370A want us to reset the DMA engine each time we
452 * use it. The 372 and later are fine.
453 */
454
455static void hpt370_bmdma_start(struct ata_queued_cmd *qc)
456{
457 struct ata_port *ap = qc->ap;
458 struct pci_dev *pdev = to_pci_dev(ap->host->dev);
459 pci_write_config_byte(pdev, 0x50 + 4 * ap->port_no, 0x37);
460 udelay(10);
461 ata_bmdma_start(qc);
462}
463
464/**
465 * hpt370_bmdma_end - DMA engine stop 448 * hpt370_bmdma_end - DMA engine stop
466 * @qc: ATA command 449 * @qc: ATA command
467 * 450 *
@@ -598,7 +581,6 @@ static struct scsi_host_template hpt37x_sht = {
598static struct ata_port_operations hpt370_port_ops = { 581static struct ata_port_operations hpt370_port_ops = {
599 .inherits = &ata_bmdma_port_ops, 582 .inherits = &ata_bmdma_port_ops,
600 583
601 .bmdma_start = hpt370_bmdma_start,
602 .bmdma_stop = hpt370_bmdma_stop, 584 .bmdma_stop = hpt370_bmdma_stop,
603 585
604 .mode_filter = hpt370_filter, 586 .mode_filter = hpt370_filter,
diff --git a/drivers/ata/pata_legacy.c b/drivers/ata/pata_legacy.c
index 3f830f0fe2cc..6f985bed8cbb 100644
--- a/drivers/ata/pata_legacy.c
+++ b/drivers/ata/pata_legacy.c
@@ -108,6 +108,7 @@ struct legacy_controller {
108 struct ata_port_operations *ops; 108 struct ata_port_operations *ops;
109 unsigned int pio_mask; 109 unsigned int pio_mask;
110 unsigned int flags; 110 unsigned int flags;
111 unsigned int pflags;
111 int (*setup)(struct platform_device *, struct legacy_probe *probe, 112 int (*setup)(struct platform_device *, struct legacy_probe *probe,
112 struct legacy_data *data); 113 struct legacy_data *data);
113}; 114};
@@ -285,7 +286,8 @@ static unsigned int pdc_data_xfer_vlb(struct ata_device *dev,
285{ 286{
286 int slop = buflen & 3; 287 int slop = buflen & 3;
287 /* 32bit I/O capable *and* we need to write a whole number of dwords */ 288 /* 32bit I/O capable *and* we need to write a whole number of dwords */
288 if (ata_id_has_dword_io(dev->id) && (slop == 0 || slop == 3)) { 289 if (ata_id_has_dword_io(dev->id) && (slop == 0 || slop == 3)
290 && (ap->pflags & ATA_PFLAG_PIO32)) {
289 struct ata_port *ap = dev->link->ap; 291 struct ata_port *ap = dev->link->ap;
290 unsigned long flags; 292 unsigned long flags;
291 293
@@ -736,7 +738,8 @@ static unsigned int vlb32_data_xfer(struct ata_device *adev, unsigned char *buf,
736 struct ata_port *ap = adev->link->ap; 738 struct ata_port *ap = adev->link->ap;
737 int slop = buflen & 3; 739 int slop = buflen & 3;
738 740
739 if (ata_id_has_dword_io(adev->id) && (slop == 0 || slop == 3)) { 741 if (ata_id_has_dword_io(adev->id) && (slop == 0 || slop == 3)
742 && (ap->pflags & ATA_PFLAG_PIO32)) {
740 if (rw == WRITE) 743 if (rw == WRITE)
741 iowrite32_rep(ap->ioaddr.data_addr, buf, buflen >> 2); 744 iowrite32_rep(ap->ioaddr.data_addr, buf, buflen >> 2);
742 else 745 else
@@ -858,27 +861,30 @@ static struct ata_port_operations winbond_port_ops = {
858 861
859static struct legacy_controller controllers[] = { 862static struct legacy_controller controllers[] = {
860 {"BIOS", &legacy_port_ops, 0x1F, 863 {"BIOS", &legacy_port_ops, 0x1F,
861 ATA_FLAG_NO_IORDY, NULL }, 864 ATA_FLAG_NO_IORDY, 0, NULL },
862 {"Snooping", &simple_port_ops, 0x1F, 865 {"Snooping", &simple_port_ops, 0x1F,
863 0 , NULL }, 866 0, 0, NULL },
864 {"PDC20230", &pdc20230_port_ops, 0x7, 867 {"PDC20230", &pdc20230_port_ops, 0x7,
865 ATA_FLAG_NO_IORDY, NULL }, 868 ATA_FLAG_NO_IORDY,
869 ATA_PFLAG_PIO32 | ATA_PFLAG_PIO32_CHANGE, NULL },
866 {"HT6560A", &ht6560a_port_ops, 0x07, 870 {"HT6560A", &ht6560a_port_ops, 0x07,
867 ATA_FLAG_NO_IORDY, NULL }, 871 ATA_FLAG_NO_IORDY, 0, NULL },
868 {"HT6560B", &ht6560b_port_ops, 0x1F, 872 {"HT6560B", &ht6560b_port_ops, 0x1F,
869 ATA_FLAG_NO_IORDY, NULL }, 873 ATA_FLAG_NO_IORDY, 0, NULL },
870 {"OPTI82C611A", &opti82c611a_port_ops, 0x0F, 874 {"OPTI82C611A", &opti82c611a_port_ops, 0x0F,
871 0 , NULL }, 875 0, 0, NULL },
872 {"OPTI82C46X", &opti82c46x_port_ops, 0x0F, 876 {"OPTI82C46X", &opti82c46x_port_ops, 0x0F,
873 0 , NULL }, 877 0, 0, NULL },
874 {"QDI6500", &qdi6500_port_ops, 0x07, 878 {"QDI6500", &qdi6500_port_ops, 0x07,
875 ATA_FLAG_NO_IORDY, qdi_port }, 879 ATA_FLAG_NO_IORDY,
880 ATA_PFLAG_PIO32 | ATA_PFLAG_PIO32_CHANGE, qdi_port },
876 {"QDI6580", &qdi6580_port_ops, 0x1F, 881 {"QDI6580", &qdi6580_port_ops, 0x1F,
877 0 , qdi_port }, 882 0, ATA_PFLAG_PIO32 | ATA_PFLAG_PIO32_CHANGE, qdi_port },
878 {"QDI6580DP", &qdi6580dp_port_ops, 0x1F, 883 {"QDI6580DP", &qdi6580dp_port_ops, 0x1F,
879 0 , qdi_port }, 884 0, ATA_PFLAG_PIO32 | ATA_PFLAG_PIO32_CHANGE, qdi_port },
880 {"W83759A", &winbond_port_ops, 0x1F, 885 {"W83759A", &winbond_port_ops, 0x1F,
881 0 , winbond_port } 886 0, ATA_PFLAG_PIO32 | ATA_PFLAG_PIO32_CHANGE,
887 winbond_port }
882}; 888};
883 889
884/** 890/**
@@ -1008,6 +1014,7 @@ static __init int legacy_init_one(struct legacy_probe *probe)
1008 ap->ops = ops; 1014 ap->ops = ops;
1009 ap->pio_mask = pio_modes; 1015 ap->pio_mask = pio_modes;
1010 ap->flags |= ATA_FLAG_SLAVE_POSS | iordy; 1016 ap->flags |= ATA_FLAG_SLAVE_POSS | iordy;
1017 ap->pflags |= controller->pflags;
1011 ap->ioaddr.cmd_addr = io_addr; 1018 ap->ioaddr.cmd_addr = io_addr;
1012 ap->ioaddr.altstatus_addr = ctrl_addr; 1019 ap->ioaddr.altstatus_addr = ctrl_addr;
1013 ap->ioaddr.ctl_addr = ctrl_addr; 1020 ap->ioaddr.ctl_addr = ctrl_addr;
@@ -1032,6 +1039,7 @@ static __init int legacy_init_one(struct legacy_probe *probe)
1032 return 0; 1039 return 0;
1033 } 1040 }
1034 } 1041 }
1042 ata_host_detach(host);
1035fail: 1043fail:
1036 platform_device_unregister(pdev); 1044 platform_device_unregister(pdev);
1037 return ret; 1045 return ret;
diff --git a/drivers/ata/pata_ninja32.c b/drivers/ata/pata_ninja32.c
index 0fb6b1b1e634..dd53a66b19e3 100644
--- a/drivers/ata/pata_ninja32.c
+++ b/drivers/ata/pata_ninja32.c
@@ -44,7 +44,7 @@
44#include <linux/libata.h> 44#include <linux/libata.h>
45 45
46#define DRV_NAME "pata_ninja32" 46#define DRV_NAME "pata_ninja32"
47#define DRV_VERSION "0.1.3" 47#define DRV_VERSION "0.1.5"
48 48
49 49
50/** 50/**
@@ -86,6 +86,7 @@ static struct ata_port_operations ninja32_port_ops = {
86 .sff_dev_select = ninja32_dev_select, 86 .sff_dev_select = ninja32_dev_select,
87 .cable_detect = ata_cable_40wire, 87 .cable_detect = ata_cable_40wire,
88 .set_piomode = ninja32_set_piomode, 88 .set_piomode = ninja32_set_piomode,
89 .sff_data_xfer = ata_sff_data_xfer32
89}; 90};
90 91
91static void ninja32_program(void __iomem *base) 92static void ninja32_program(void __iomem *base)
@@ -144,6 +145,7 @@ static int ninja32_init_one(struct pci_dev *dev, const struct pci_device_id *id)
144 ap->ioaddr.altstatus_addr = base + 0x1E; 145 ap->ioaddr.altstatus_addr = base + 0x1E;
145 ap->ioaddr.bmdma_addr = base; 146 ap->ioaddr.bmdma_addr = base;
146 ata_sff_std_ports(&ap->ioaddr); 147 ata_sff_std_ports(&ap->ioaddr);
148 ap->pflags = ATA_PFLAG_PIO32 | ATA_PFLAG_PIO32CHANGE;
147 149
148 ninja32_program(base); 150 ninja32_program(base);
149 /* FIXME: Should we disable them at remove ? */ 151 /* FIXME: Should we disable them at remove ? */
diff --git a/drivers/atm/solos-pci.c b/drivers/atm/solos-pci.c
index be204308cc1b..9359613addc5 100644
--- a/drivers/atm/solos-pci.c
+++ b/drivers/atm/solos-pci.c
@@ -1059,7 +1059,7 @@ static int fpga_probe(struct pci_dev *dev, const struct pci_device_id *id)
1059 goto out; 1059 goto out;
1060 } 1060 }
1061 1061
1062 err = pci_set_dma_mask(dev, DMA_32BIT_MASK); 1062 err = pci_set_dma_mask(dev, DMA_BIT_MASK(32));
1063 if (err) { 1063 if (err) {
1064 dev_warn(&dev->dev, "Failed to set 32-bit DMA mask\n"); 1064 dev_warn(&dev->dev, "Failed to set 32-bit DMA mask\n");
1065 goto out; 1065 goto out;
diff --git a/drivers/block/brd.c b/drivers/block/brd.c
index bdd4f5f45575..5f7e64ba87e5 100644
--- a/drivers/block/brd.c
+++ b/drivers/block/brd.c
@@ -275,8 +275,10 @@ static int brd_do_bvec(struct brd_device *brd, struct page *page,
275 if (rw == READ) { 275 if (rw == READ) {
276 copy_from_brd(mem + off, brd, sector, len); 276 copy_from_brd(mem + off, brd, sector, len);
277 flush_dcache_page(page); 277 flush_dcache_page(page);
278 } else 278 } else {
279 flush_dcache_page(page);
279 copy_to_brd(brd, mem + off, sector, len); 280 copy_to_brd(brd, mem + off, sector, len);
281 }
280 kunmap_atomic(mem, KM_USER0); 282 kunmap_atomic(mem, KM_USER0);
281 283
282out: 284out:
@@ -436,6 +438,7 @@ static struct brd_device *brd_alloc(int i)
436 if (!brd->brd_queue) 438 if (!brd->brd_queue)
437 goto out_free_dev; 439 goto out_free_dev;
438 blk_queue_make_request(brd->brd_queue, brd_make_request); 440 blk_queue_make_request(brd->brd_queue, brd_make_request);
441 blk_queue_ordered(brd->brd_queue, QUEUE_ORDERED_TAG, NULL);
439 blk_queue_max_sectors(brd->brd_queue, 1024); 442 blk_queue_max_sectors(brd->brd_queue, 1024);
440 blk_queue_bounce_limit(brd->brd_queue, BLK_BOUNCE_ANY); 443 blk_queue_bounce_limit(brd->brd_queue, BLK_BOUNCE_ANY);
441 444
diff --git a/drivers/block/cciss.c b/drivers/block/cciss.c
index 0ef6f08aa6ea..4d4d5e0d3fa6 100644
--- a/drivers/block/cciss.c
+++ b/drivers/block/cciss.c
@@ -3505,7 +3505,7 @@ static __devinit int cciss_message(struct pci_dev *pdev, unsigned char opcode, u
3505 /* The Inbound Post Queue only accepts 32-bit physical addresses for the 3505 /* The Inbound Post Queue only accepts 32-bit physical addresses for the
3506 CCISS commands, so they must be allocated from the lower 4GiB of 3506 CCISS commands, so they must be allocated from the lower 4GiB of
3507 memory. */ 3507 memory. */
3508 err = pci_set_consistent_dma_mask(pdev, DMA_32BIT_MASK); 3508 err = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32));
3509 if (err) { 3509 if (err) {
3510 iounmap(vaddr); 3510 iounmap(vaddr);
3511 return -ENOMEM; 3511 return -ENOMEM;
diff --git a/drivers/char/agp/intel-agp.c b/drivers/char/agp/intel-agp.c
index 9d9490e22e07..3686912427ba 100644
--- a/drivers/char/agp/intel-agp.c
+++ b/drivers/char/agp/intel-agp.c
@@ -2131,6 +2131,8 @@ static const struct intel_driver_description {
2131 { PCI_DEVICE_ID_INTEL_82845G_HB, PCI_DEVICE_ID_INTEL_82845G_IG, 0, "830M", 2131 { PCI_DEVICE_ID_INTEL_82845G_HB, PCI_DEVICE_ID_INTEL_82845G_IG, 0, "830M",
2132 &intel_845_driver, &intel_830_driver }, 2132 &intel_845_driver, &intel_830_driver },
2133 { PCI_DEVICE_ID_INTEL_82850_HB, 0, 0, "i850", &intel_850_driver, NULL }, 2133 { PCI_DEVICE_ID_INTEL_82850_HB, 0, 0, "i850", &intel_850_driver, NULL },
2134 { PCI_DEVICE_ID_INTEL_82854_HB, PCI_DEVICE_ID_INTEL_82854_IG, 0, "854",
2135 &intel_845_driver, &intel_830_driver },
2134 { PCI_DEVICE_ID_INTEL_82855PM_HB, 0, 0, "855PM", &intel_845_driver, NULL }, 2136 { PCI_DEVICE_ID_INTEL_82855PM_HB, 0, 0, "855PM", &intel_845_driver, NULL },
2135 { PCI_DEVICE_ID_INTEL_82855GM_HB, PCI_DEVICE_ID_INTEL_82855GM_IG, 0, "855GM", 2137 { PCI_DEVICE_ID_INTEL_82855GM_HB, PCI_DEVICE_ID_INTEL_82855GM_IG, 0, "855GM",
2136 &intel_845_driver, &intel_830_driver }, 2138 &intel_845_driver, &intel_830_driver },
@@ -2355,6 +2357,7 @@ static struct pci_device_id agp_intel_pci_table[] = {
2355 ID(PCI_DEVICE_ID_INTEL_82845_HB), 2357 ID(PCI_DEVICE_ID_INTEL_82845_HB),
2356 ID(PCI_DEVICE_ID_INTEL_82845G_HB), 2358 ID(PCI_DEVICE_ID_INTEL_82845G_HB),
2357 ID(PCI_DEVICE_ID_INTEL_82850_HB), 2359 ID(PCI_DEVICE_ID_INTEL_82850_HB),
2360 ID(PCI_DEVICE_ID_INTEL_82854_HB),
2358 ID(PCI_DEVICE_ID_INTEL_82855PM_HB), 2361 ID(PCI_DEVICE_ID_INTEL_82855PM_HB),
2359 ID(PCI_DEVICE_ID_INTEL_82855GM_HB), 2362 ID(PCI_DEVICE_ID_INTEL_82855GM_HB),
2360 ID(PCI_DEVICE_ID_INTEL_82860_HB), 2363 ID(PCI_DEVICE_ID_INTEL_82860_HB),
diff --git a/drivers/char/sysrq.c b/drivers/char/sysrq.c
index 6de020d078e1..b0a6a3e51924 100644
--- a/drivers/char/sysrq.c
+++ b/drivers/char/sysrq.c
@@ -35,7 +35,6 @@
35#include <linux/vt_kern.h> 35#include <linux/vt_kern.h>
36#include <linux/workqueue.h> 36#include <linux/workqueue.h>
37#include <linux/kexec.h> 37#include <linux/kexec.h>
38#include <linux/interrupt.h>
39#include <linux/hrtimer.h> 38#include <linux/hrtimer.h>
40#include <linux/oom.h> 39#include <linux/oom.h>
41 40
diff --git a/drivers/edac/edac_core.h b/drivers/edac/edac_core.h
index 28f2c3f959b5..6ad95c8d6363 100644
--- a/drivers/edac/edac_core.h
+++ b/drivers/edac/edac_core.h
@@ -767,11 +767,19 @@ static inline void pci_write_bits16(struct pci_dev *pdev, int offset,
767 pci_write_config_word(pdev, offset, value); 767 pci_write_config_word(pdev, offset, value);
768} 768}
769 769
770/* write all or some bits in a dword-register*/ 770/*
771 * pci_write_bits32
772 *
773 * edac local routine to do pci_write_config_dword, but adds
774 * a mask parameter. If mask is all ones, ignore the mask.
775 * Otherwise utilize the mask to isolate specified bits
776 *
777 * write all or some bits in a dword-register
778 */
771static inline void pci_write_bits32(struct pci_dev *pdev, int offset, 779static inline void pci_write_bits32(struct pci_dev *pdev, int offset,
772 u32 value, u32 mask) 780 u32 value, u32 mask)
773{ 781{
774 if (mask != 0xffff) { 782 if (mask != 0xffffffff) {
775 u32 buf; 783 u32 buf;
776 784
777 pci_read_config_dword(pdev, offset, &buf); 785 pci_read_config_dword(pdev, offset, &buf);
diff --git a/drivers/edac/edac_device.c b/drivers/edac/edac_device.c
index ca9113e1c106..a7d2c717d033 100644
--- a/drivers/edac/edac_device.c
+++ b/drivers/edac/edac_device.c
@@ -389,7 +389,7 @@ static void del_edac_device_from_global_list(struct edac_device_ctl_info
389 */ 389 */
390static void edac_device_workq_function(struct work_struct *work_req) 390static void edac_device_workq_function(struct work_struct *work_req)
391{ 391{
392 struct delayed_work *d_work = (struct delayed_work *)work_req; 392 struct delayed_work *d_work = to_delayed_work(work_req);
393 struct edac_device_ctl_info *edac_dev = to_edac_device_ctl_work(d_work); 393 struct edac_device_ctl_info *edac_dev = to_edac_device_ctl_work(d_work);
394 394
395 mutex_lock(&device_ctls_mutex); 395 mutex_lock(&device_ctls_mutex);
diff --git a/drivers/edac/edac_mc.c b/drivers/edac/edac_mc.c
index 25d66940b4fa..335b7ebdb11c 100644
--- a/drivers/edac/edac_mc.c
+++ b/drivers/edac/edac_mc.c
@@ -260,7 +260,7 @@ static int edac_mc_assert_error_check_and_clear(void)
260 */ 260 */
261static void edac_mc_workq_function(struct work_struct *work_req) 261static void edac_mc_workq_function(struct work_struct *work_req)
262{ 262{
263 struct delayed_work *d_work = (struct delayed_work *)work_req; 263 struct delayed_work *d_work = to_delayed_work(work_req);
264 struct mem_ctl_info *mci = to_edac_mem_ctl_work(d_work); 264 struct mem_ctl_info *mci = to_edac_mem_ctl_work(d_work);
265 265
266 mutex_lock(&mem_ctls_mutex); 266 mutex_lock(&mem_ctls_mutex);
diff --git a/drivers/edac/edac_pci.c b/drivers/edac/edac_pci.c
index 5b150aea703a..30b585b1d60b 100644
--- a/drivers/edac/edac_pci.c
+++ b/drivers/edac/edac_pci.c
@@ -233,7 +233,7 @@ EXPORT_SYMBOL_GPL(edac_pci_find);
233 */ 233 */
234static void edac_pci_workq_function(struct work_struct *work_req) 234static void edac_pci_workq_function(struct work_struct *work_req)
235{ 235{
236 struct delayed_work *d_work = (struct delayed_work *)work_req; 236 struct delayed_work *d_work = to_delayed_work(work_req);
237 struct edac_pci_ctl_info *pci = to_edac_pci_ctl_work(d_work); 237 struct edac_pci_ctl_info *pci = to_edac_pci_ctl_work(d_work);
238 int msec; 238 int msec;
239 unsigned long delay; 239 unsigned long delay;
diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h
index 3750d8003048..473a8f7fbdb5 100644
--- a/drivers/gpu/drm/i915/i915_drv.h
+++ b/drivers/gpu/drm/i915/i915_drv.h
@@ -446,6 +446,9 @@ struct drm_i915_gem_object {
446 uint32_t tiling_mode; 446 uint32_t tiling_mode;
447 uint32_t stride; 447 uint32_t stride;
448 448
449 /** Record of address bit 17 of each page at last unbind. */
450 long *bit_17;
451
449 /** AGP mapping type (AGP_USER_MEMORY or AGP_USER_CACHED_MEMORY */ 452 /** AGP mapping type (AGP_USER_MEMORY or AGP_USER_CACHED_MEMORY */
450 uint32_t agp_type; 453 uint32_t agp_type;
451 454
@@ -635,9 +638,13 @@ int i915_gem_attach_phys_object(struct drm_device *dev,
635void i915_gem_detach_phys_object(struct drm_device *dev, 638void i915_gem_detach_phys_object(struct drm_device *dev,
636 struct drm_gem_object *obj); 639 struct drm_gem_object *obj);
637void i915_gem_free_all_phys_object(struct drm_device *dev); 640void i915_gem_free_all_phys_object(struct drm_device *dev);
641int i915_gem_object_get_pages(struct drm_gem_object *obj);
642void i915_gem_object_put_pages(struct drm_gem_object *obj);
638 643
639/* i915_gem_tiling.c */ 644/* i915_gem_tiling.c */
640void i915_gem_detect_bit_6_swizzle(struct drm_device *dev); 645void i915_gem_detect_bit_6_swizzle(struct drm_device *dev);
646void i915_gem_object_do_bit_17_swizzle(struct drm_gem_object *obj);
647void i915_gem_object_save_bit_17_swizzle(struct drm_gem_object *obj);
641 648
642/* i915_gem_debug.c */ 649/* i915_gem_debug.c */
643void i915_gem_dump_object(struct drm_gem_object *obj, int len, 650void i915_gem_dump_object(struct drm_gem_object *obj, int len,
diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c
index 1449b452cc63..4642115902d6 100644
--- a/drivers/gpu/drm/i915/i915_gem.c
+++ b/drivers/gpu/drm/i915/i915_gem.c
@@ -43,8 +43,6 @@ static int i915_gem_object_set_cpu_read_domain_range(struct drm_gem_object *obj,
43 uint64_t offset, 43 uint64_t offset,
44 uint64_t size); 44 uint64_t size);
45static void i915_gem_object_set_to_full_cpu_read_domain(struct drm_gem_object *obj); 45static void i915_gem_object_set_to_full_cpu_read_domain(struct drm_gem_object *obj);
46static int i915_gem_object_get_pages(struct drm_gem_object *obj);
47static void i915_gem_object_put_pages(struct drm_gem_object *obj);
48static int i915_gem_object_wait_rendering(struct drm_gem_object *obj); 46static int i915_gem_object_wait_rendering(struct drm_gem_object *obj);
49static int i915_gem_object_bind_to_gtt(struct drm_gem_object *obj, 47static int i915_gem_object_bind_to_gtt(struct drm_gem_object *obj,
50 unsigned alignment); 48 unsigned alignment);
@@ -143,15 +141,27 @@ fast_shmem_read(struct page **pages,
143 int length) 141 int length)
144{ 142{
145 char __iomem *vaddr; 143 char __iomem *vaddr;
146 int ret; 144 int unwritten;
147 145
148 vaddr = kmap_atomic(pages[page_base >> PAGE_SHIFT], KM_USER0); 146 vaddr = kmap_atomic(pages[page_base >> PAGE_SHIFT], KM_USER0);
149 if (vaddr == NULL) 147 if (vaddr == NULL)
150 return -ENOMEM; 148 return -ENOMEM;
151 ret = __copy_to_user_inatomic(data, vaddr + page_offset, length); 149 unwritten = __copy_to_user_inatomic(data, vaddr + page_offset, length);
152 kunmap_atomic(vaddr, KM_USER0); 150 kunmap_atomic(vaddr, KM_USER0);
153 151
154 return ret; 152 if (unwritten)
153 return -EFAULT;
154
155 return 0;
156}
157
158static int i915_gem_object_needs_bit17_swizzle(struct drm_gem_object *obj)
159{
160 drm_i915_private_t *dev_priv = obj->dev->dev_private;
161 struct drm_i915_gem_object *obj_priv = obj->driver_private;
162
163 return dev_priv->mm.bit_6_swizzle_x == I915_BIT_6_SWIZZLE_9_10_17 &&
164 obj_priv->tiling_mode != I915_TILING_NONE;
155} 165}
156 166
157static inline int 167static inline int
@@ -181,6 +191,64 @@ slow_shmem_copy(struct page *dst_page,
181 return 0; 191 return 0;
182} 192}
183 193
194static inline int
195slow_shmem_bit17_copy(struct page *gpu_page,
196 int gpu_offset,
197 struct page *cpu_page,
198 int cpu_offset,
199 int length,
200 int is_read)
201{
202 char *gpu_vaddr, *cpu_vaddr;
203
204 /* Use the unswizzled path if this page isn't affected. */
205 if ((page_to_phys(gpu_page) & (1 << 17)) == 0) {
206 if (is_read)
207 return slow_shmem_copy(cpu_page, cpu_offset,
208 gpu_page, gpu_offset, length);
209 else
210 return slow_shmem_copy(gpu_page, gpu_offset,
211 cpu_page, cpu_offset, length);
212 }
213
214 gpu_vaddr = kmap_atomic(gpu_page, KM_USER0);
215 if (gpu_vaddr == NULL)
216 return -ENOMEM;
217
218 cpu_vaddr = kmap_atomic(cpu_page, KM_USER1);
219 if (cpu_vaddr == NULL) {
220 kunmap_atomic(gpu_vaddr, KM_USER0);
221 return -ENOMEM;
222 }
223
224 /* Copy the data, XORing A6 with A17 (1). The user already knows he's
225 * XORing with the other bits (A9 for Y, A9 and A10 for X)
226 */
227 while (length > 0) {
228 int cacheline_end = ALIGN(gpu_offset + 1, 64);
229 int this_length = min(cacheline_end - gpu_offset, length);
230 int swizzled_gpu_offset = gpu_offset ^ 64;
231
232 if (is_read) {
233 memcpy(cpu_vaddr + cpu_offset,
234 gpu_vaddr + swizzled_gpu_offset,
235 this_length);
236 } else {
237 memcpy(gpu_vaddr + swizzled_gpu_offset,
238 cpu_vaddr + cpu_offset,
239 this_length);
240 }
241 cpu_offset += this_length;
242 gpu_offset += this_length;
243 length -= this_length;
244 }
245
246 kunmap_atomic(cpu_vaddr, KM_USER1);
247 kunmap_atomic(gpu_vaddr, KM_USER0);
248
249 return 0;
250}
251
184/** 252/**
185 * This is the fast shmem pread path, which attempts to copy_from_user directly 253 * This is the fast shmem pread path, which attempts to copy_from_user directly
186 * from the backing pages of the object to the user's address space. On a 254 * from the backing pages of the object to the user's address space. On a
@@ -269,6 +337,7 @@ i915_gem_shmem_pread_slow(struct drm_device *dev, struct drm_gem_object *obj,
269 int page_length; 337 int page_length;
270 int ret; 338 int ret;
271 uint64_t data_ptr = args->data_ptr; 339 uint64_t data_ptr = args->data_ptr;
340 int do_bit17_swizzling;
272 341
273 remain = args->size; 342 remain = args->size;
274 343
@@ -286,13 +355,15 @@ i915_gem_shmem_pread_slow(struct drm_device *dev, struct drm_gem_object *obj,
286 355
287 down_read(&mm->mmap_sem); 356 down_read(&mm->mmap_sem);
288 pinned_pages = get_user_pages(current, mm, (uintptr_t)args->data_ptr, 357 pinned_pages = get_user_pages(current, mm, (uintptr_t)args->data_ptr,
289 num_pages, 0, 0, user_pages, NULL); 358 num_pages, 1, 0, user_pages, NULL);
290 up_read(&mm->mmap_sem); 359 up_read(&mm->mmap_sem);
291 if (pinned_pages < num_pages) { 360 if (pinned_pages < num_pages) {
292 ret = -EFAULT; 361 ret = -EFAULT;
293 goto fail_put_user_pages; 362 goto fail_put_user_pages;
294 } 363 }
295 364
365 do_bit17_swizzling = i915_gem_object_needs_bit17_swizzle(obj);
366
296 mutex_lock(&dev->struct_mutex); 367 mutex_lock(&dev->struct_mutex);
297 368
298 ret = i915_gem_object_get_pages(obj); 369 ret = i915_gem_object_get_pages(obj);
@@ -327,11 +398,20 @@ i915_gem_shmem_pread_slow(struct drm_device *dev, struct drm_gem_object *obj,
327 if ((data_page_offset + page_length) > PAGE_SIZE) 398 if ((data_page_offset + page_length) > PAGE_SIZE)
328 page_length = PAGE_SIZE - data_page_offset; 399 page_length = PAGE_SIZE - data_page_offset;
329 400
330 ret = slow_shmem_copy(user_pages[data_page_index], 401 if (do_bit17_swizzling) {
331 data_page_offset, 402 ret = slow_shmem_bit17_copy(obj_priv->pages[shmem_page_index],
332 obj_priv->pages[shmem_page_index], 403 shmem_page_offset,
333 shmem_page_offset, 404 user_pages[data_page_index],
334 page_length); 405 data_page_offset,
406 page_length,
407 1);
408 } else {
409 ret = slow_shmem_copy(user_pages[data_page_index],
410 data_page_offset,
411 obj_priv->pages[shmem_page_index],
412 shmem_page_offset,
413 page_length);
414 }
335 if (ret) 415 if (ret)
336 goto fail_put_pages; 416 goto fail_put_pages;
337 417
@@ -383,9 +463,14 @@ i915_gem_pread_ioctl(struct drm_device *dev, void *data,
383 return -EINVAL; 463 return -EINVAL;
384 } 464 }
385 465
386 ret = i915_gem_shmem_pread_fast(dev, obj, args, file_priv); 466 if (i915_gem_object_needs_bit17_swizzle(obj)) {
387 if (ret != 0)
388 ret = i915_gem_shmem_pread_slow(dev, obj, args, file_priv); 467 ret = i915_gem_shmem_pread_slow(dev, obj, args, file_priv);
468 } else {
469 ret = i915_gem_shmem_pread_fast(dev, obj, args, file_priv);
470 if (ret != 0)
471 ret = i915_gem_shmem_pread_slow(dev, obj, args,
472 file_priv);
473 }
389 474
390 drm_gem_object_unreference(obj); 475 drm_gem_object_unreference(obj);
391 476
@@ -727,6 +812,7 @@ i915_gem_shmem_pwrite_slow(struct drm_device *dev, struct drm_gem_object *obj,
727 int page_length; 812 int page_length;
728 int ret; 813 int ret;
729 uint64_t data_ptr = args->data_ptr; 814 uint64_t data_ptr = args->data_ptr;
815 int do_bit17_swizzling;
730 816
731 remain = args->size; 817 remain = args->size;
732 818
@@ -751,6 +837,8 @@ i915_gem_shmem_pwrite_slow(struct drm_device *dev, struct drm_gem_object *obj,
751 goto fail_put_user_pages; 837 goto fail_put_user_pages;
752 } 838 }
753 839
840 do_bit17_swizzling = i915_gem_object_needs_bit17_swizzle(obj);
841
754 mutex_lock(&dev->struct_mutex); 842 mutex_lock(&dev->struct_mutex);
755 843
756 ret = i915_gem_object_get_pages(obj); 844 ret = i915_gem_object_get_pages(obj);
@@ -785,11 +873,20 @@ i915_gem_shmem_pwrite_slow(struct drm_device *dev, struct drm_gem_object *obj,
785 if ((data_page_offset + page_length) > PAGE_SIZE) 873 if ((data_page_offset + page_length) > PAGE_SIZE)
786 page_length = PAGE_SIZE - data_page_offset; 874 page_length = PAGE_SIZE - data_page_offset;
787 875
788 ret = slow_shmem_copy(obj_priv->pages[shmem_page_index], 876 if (do_bit17_swizzling) {
789 shmem_page_offset, 877 ret = slow_shmem_bit17_copy(obj_priv->pages[shmem_page_index],
790 user_pages[data_page_index], 878 shmem_page_offset,
791 data_page_offset, 879 user_pages[data_page_index],
792 page_length); 880 data_page_offset,
881 page_length,
882 0);
883 } else {
884 ret = slow_shmem_copy(obj_priv->pages[shmem_page_index],
885 shmem_page_offset,
886 user_pages[data_page_index],
887 data_page_offset,
888 page_length);
889 }
793 if (ret) 890 if (ret)
794 goto fail_put_pages; 891 goto fail_put_pages;
795 892
@@ -854,6 +951,8 @@ i915_gem_pwrite_ioctl(struct drm_device *dev, void *data,
854 ret = i915_gem_gtt_pwrite_slow(dev, obj, args, 951 ret = i915_gem_gtt_pwrite_slow(dev, obj, args,
855 file_priv); 952 file_priv);
856 } 953 }
954 } else if (i915_gem_object_needs_bit17_swizzle(obj)) {
955 ret = i915_gem_shmem_pwrite_slow(dev, obj, args, file_priv);
857 } else { 956 } else {
858 ret = i915_gem_shmem_pwrite_fast(dev, obj, args, file_priv); 957 ret = i915_gem_shmem_pwrite_fast(dev, obj, args, file_priv);
859 if (ret == -EFAULT) { 958 if (ret == -EFAULT) {
@@ -1285,7 +1384,7 @@ i915_gem_mmap_gtt_ioctl(struct drm_device *dev, void *data,
1285 return 0; 1384 return 0;
1286} 1385}
1287 1386
1288static void 1387void
1289i915_gem_object_put_pages(struct drm_gem_object *obj) 1388i915_gem_object_put_pages(struct drm_gem_object *obj)
1290{ 1389{
1291 struct drm_i915_gem_object *obj_priv = obj->driver_private; 1390 struct drm_i915_gem_object *obj_priv = obj->driver_private;
@@ -1297,6 +1396,9 @@ i915_gem_object_put_pages(struct drm_gem_object *obj)
1297 if (--obj_priv->pages_refcount != 0) 1396 if (--obj_priv->pages_refcount != 0)
1298 return; 1397 return;
1299 1398
1399 if (obj_priv->tiling_mode != I915_TILING_NONE)
1400 i915_gem_object_save_bit_17_swizzle(obj);
1401
1300 for (i = 0; i < page_count; i++) 1402 for (i = 0; i < page_count; i++)
1301 if (obj_priv->pages[i] != NULL) { 1403 if (obj_priv->pages[i] != NULL) {
1302 if (obj_priv->dirty) 1404 if (obj_priv->dirty)
@@ -1494,8 +1596,19 @@ i915_gem_retire_request(struct drm_device *dev,
1494 1596
1495 if (obj->write_domain != 0) 1597 if (obj->write_domain != 0)
1496 i915_gem_object_move_to_flushing(obj); 1598 i915_gem_object_move_to_flushing(obj);
1497 else 1599 else {
1600 /* Take a reference on the object so it won't be
1601 * freed while the spinlock is held. The list
1602 * protection for this spinlock is safe when breaking
1603 * the lock like this since the next thing we do
1604 * is just get the head of the list again.
1605 */
1606 drm_gem_object_reference(obj);
1498 i915_gem_object_move_to_inactive(obj); 1607 i915_gem_object_move_to_inactive(obj);
1608 spin_unlock(&dev_priv->mm.active_list_lock);
1609 drm_gem_object_unreference(obj);
1610 spin_lock(&dev_priv->mm.active_list_lock);
1611 }
1499 } 1612 }
1500out: 1613out:
1501 spin_unlock(&dev_priv->mm.active_list_lock); 1614 spin_unlock(&dev_priv->mm.active_list_lock);
@@ -1884,7 +1997,7 @@ i915_gem_evict_everything(struct drm_device *dev)
1884 return ret; 1997 return ret;
1885} 1998}
1886 1999
1887static int 2000int
1888i915_gem_object_get_pages(struct drm_gem_object *obj) 2001i915_gem_object_get_pages(struct drm_gem_object *obj)
1889{ 2002{
1890 struct drm_i915_gem_object *obj_priv = obj->driver_private; 2003 struct drm_i915_gem_object *obj_priv = obj->driver_private;
@@ -1922,6 +2035,10 @@ i915_gem_object_get_pages(struct drm_gem_object *obj)
1922 } 2035 }
1923 obj_priv->pages[i] = page; 2036 obj_priv->pages[i] = page;
1924 } 2037 }
2038
2039 if (obj_priv->tiling_mode != I915_TILING_NONE)
2040 i915_gem_object_do_bit_17_swizzle(obj);
2041
1925 return 0; 2042 return 0;
1926} 2043}
1927 2044
@@ -3002,13 +3119,13 @@ i915_gem_get_relocs_from_user(struct drm_i915_gem_exec_object *exec_list,
3002 drm_free(*relocs, reloc_count * sizeof(**relocs), 3119 drm_free(*relocs, reloc_count * sizeof(**relocs),
3003 DRM_MEM_DRIVER); 3120 DRM_MEM_DRIVER);
3004 *relocs = NULL; 3121 *relocs = NULL;
3005 return ret; 3122 return -EFAULT;
3006 } 3123 }
3007 3124
3008 reloc_index += exec_list[i].relocation_count; 3125 reloc_index += exec_list[i].relocation_count;
3009 } 3126 }
3010 3127
3011 return ret; 3128 return 0;
3012} 3129}
3013 3130
3014static int 3131static int
@@ -3017,23 +3134,28 @@ i915_gem_put_relocs_to_user(struct drm_i915_gem_exec_object *exec_list,
3017 struct drm_i915_gem_relocation_entry *relocs) 3134 struct drm_i915_gem_relocation_entry *relocs)
3018{ 3135{
3019 uint32_t reloc_count = 0, i; 3136 uint32_t reloc_count = 0, i;
3020 int ret; 3137 int ret = 0;
3021 3138
3022 for (i = 0; i < buffer_count; i++) { 3139 for (i = 0; i < buffer_count; i++) {
3023 struct drm_i915_gem_relocation_entry __user *user_relocs; 3140 struct drm_i915_gem_relocation_entry __user *user_relocs;
3141 int unwritten;
3024 3142
3025 user_relocs = (void __user *)(uintptr_t)exec_list[i].relocs_ptr; 3143 user_relocs = (void __user *)(uintptr_t)exec_list[i].relocs_ptr;
3026 3144
3027 if (ret == 0) { 3145 unwritten = copy_to_user(user_relocs,
3028 ret = copy_to_user(user_relocs, 3146 &relocs[reloc_count],
3029 &relocs[reloc_count], 3147 exec_list[i].relocation_count *
3030 exec_list[i].relocation_count * 3148 sizeof(*relocs));
3031 sizeof(*relocs)); 3149
3150 if (unwritten) {
3151 ret = -EFAULT;
3152 goto err;
3032 } 3153 }
3033 3154
3034 reloc_count += exec_list[i].relocation_count; 3155 reloc_count += exec_list[i].relocation_count;
3035 } 3156 }
3036 3157
3158err:
3037 drm_free(relocs, reloc_count * sizeof(*relocs), DRM_MEM_DRIVER); 3159 drm_free(relocs, reloc_count * sizeof(*relocs), DRM_MEM_DRIVER);
3038 3160
3039 return ret; 3161 return ret;
@@ -3243,7 +3365,7 @@ i915_gem_execbuffer(struct drm_device *dev, void *data,
3243 exec_offset = exec_list[args->buffer_count - 1].offset; 3365 exec_offset = exec_list[args->buffer_count - 1].offset;
3244 3366
3245#if WATCH_EXEC 3367#if WATCH_EXEC
3246 i915_gem_dump_object(object_list[args->buffer_count - 1], 3368 i915_gem_dump_object(batch_obj,
3247 args->batch_len, 3369 args->batch_len,
3248 __func__, 3370 __func__,
3249 ~0); 3371 ~0);
@@ -3308,10 +3430,12 @@ err:
3308 (uintptr_t) args->buffers_ptr, 3430 (uintptr_t) args->buffers_ptr,
3309 exec_list, 3431 exec_list,
3310 sizeof(*exec_list) * args->buffer_count); 3432 sizeof(*exec_list) * args->buffer_count);
3311 if (ret) 3433 if (ret) {
3434 ret = -EFAULT;
3312 DRM_ERROR("failed to copy %d exec entries " 3435 DRM_ERROR("failed to copy %d exec entries "
3313 "back to user (%d)\n", 3436 "back to user (%d)\n",
3314 args->buffer_count, ret); 3437 args->buffer_count, ret);
3438 }
3315 } 3439 }
3316 3440
3317 /* Copy the updated relocations out regardless of current error 3441 /* Copy the updated relocations out regardless of current error
@@ -3593,6 +3717,7 @@ void i915_gem_free_object(struct drm_gem_object *obj)
3593 i915_gem_free_mmap_offset(obj); 3717 i915_gem_free_mmap_offset(obj);
3594 3718
3595 drm_free(obj_priv->page_cpu_valid, 1, DRM_MEM_DRIVER); 3719 drm_free(obj_priv->page_cpu_valid, 1, DRM_MEM_DRIVER);
3720 kfree(obj_priv->bit_17);
3596 drm_free(obj->driver_private, 1, DRM_MEM_DRIVER); 3721 drm_free(obj->driver_private, 1, DRM_MEM_DRIVER);
3597} 3722}
3598 3723
diff --git a/drivers/gpu/drm/i915/i915_gem_debugfs.c b/drivers/gpu/drm/i915/i915_gem_debugfs.c
index a1ac0c5e7307..986f1082c596 100644
--- a/drivers/gpu/drm/i915/i915_gem_debugfs.c
+++ b/drivers/gpu/drm/i915/i915_gem_debugfs.c
@@ -234,6 +234,96 @@ static int i915_hws_info(struct seq_file *m, void *data)
234 return 0; 234 return 0;
235} 235}
236 236
237static void i915_dump_pages(struct seq_file *m, struct page **pages, int page_count)
238{
239 int page, i;
240 uint32_t *mem;
241
242 for (page = 0; page < page_count; page++) {
243 mem = kmap(pages[page]);
244 for (i = 0; i < PAGE_SIZE; i += 4)
245 seq_printf(m, "%08x : %08x\n", i, mem[i / 4]);
246 kunmap(pages[page]);
247 }
248}
249
250static int i915_batchbuffer_info(struct seq_file *m, void *data)
251{
252 struct drm_info_node *node = (struct drm_info_node *) m->private;
253 struct drm_device *dev = node->minor->dev;
254 drm_i915_private_t *dev_priv = dev->dev_private;
255 struct drm_gem_object *obj;
256 struct drm_i915_gem_object *obj_priv;
257 int ret;
258
259 spin_lock(&dev_priv->mm.active_list_lock);
260
261 list_for_each_entry(obj_priv, &dev_priv->mm.active_list, list) {
262 obj = obj_priv->obj;
263 if (obj->read_domains & I915_GEM_DOMAIN_COMMAND) {
264 ret = i915_gem_object_get_pages(obj);
265 if (ret) {
266 DRM_ERROR("Failed to get pages: %d\n", ret);
267 spin_unlock(&dev_priv->mm.active_list_lock);
268 return ret;
269 }
270
271 seq_printf(m, "--- gtt_offset = 0x%08x\n", obj_priv->gtt_offset);
272 i915_dump_pages(m, obj_priv->pages, obj->size / PAGE_SIZE);
273
274 i915_gem_object_put_pages(obj);
275 }
276 }
277
278 spin_unlock(&dev_priv->mm.active_list_lock);
279
280 return 0;
281}
282
283static int i915_ringbuffer_data(struct seq_file *m, void *data)
284{
285 struct drm_info_node *node = (struct drm_info_node *) m->private;
286 struct drm_device *dev = node->minor->dev;
287 drm_i915_private_t *dev_priv = dev->dev_private;
288 u8 *virt;
289 uint32_t *ptr, off;
290
291 if (!dev_priv->ring.ring_obj) {
292 seq_printf(m, "No ringbuffer setup\n");
293 return 0;
294 }
295
296 virt = dev_priv->ring.virtual_start;
297
298 for (off = 0; off < dev_priv->ring.Size; off += 4) {
299 ptr = (uint32_t *)(virt + off);
300 seq_printf(m, "%08x : %08x\n", off, *ptr);
301 }
302
303 return 0;
304}
305
306static int i915_ringbuffer_info(struct seq_file *m, void *data)
307{
308 struct drm_info_node *node = (struct drm_info_node *) m->private;
309 struct drm_device *dev = node->minor->dev;
310 drm_i915_private_t *dev_priv = dev->dev_private;
311 unsigned int head, tail, mask;
312
313 head = I915_READ(PRB0_HEAD) & HEAD_ADDR;
314 tail = I915_READ(PRB0_TAIL) & TAIL_ADDR;
315 mask = dev_priv->ring.tail_mask;
316
317 seq_printf(m, "RingHead : %08x\n", head);
318 seq_printf(m, "RingTail : %08x\n", tail);
319 seq_printf(m, "RingMask : %08x\n", mask);
320 seq_printf(m, "RingSize : %08lx\n", dev_priv->ring.Size);
321 seq_printf(m, "Acthd : %08x\n", I915_READ(IS_I965G(dev) ? ACTHD_I965 : ACTHD));
322
323 return 0;
324}
325
326
237static struct drm_info_list i915_gem_debugfs_list[] = { 327static struct drm_info_list i915_gem_debugfs_list[] = {
238 {"i915_gem_active", i915_gem_object_list_info, 0, (void *) ACTIVE_LIST}, 328 {"i915_gem_active", i915_gem_object_list_info, 0, (void *) ACTIVE_LIST},
239 {"i915_gem_flushing", i915_gem_object_list_info, 0, (void *) FLUSHING_LIST}, 329 {"i915_gem_flushing", i915_gem_object_list_info, 0, (void *) FLUSHING_LIST},
@@ -243,6 +333,9 @@ static struct drm_info_list i915_gem_debugfs_list[] = {
243 {"i915_gem_fence_regs", i915_gem_fence_regs_info, 0}, 333 {"i915_gem_fence_regs", i915_gem_fence_regs_info, 0},
244 {"i915_gem_interrupt", i915_interrupt_info, 0}, 334 {"i915_gem_interrupt", i915_interrupt_info, 0},
245 {"i915_gem_hws", i915_hws_info, 0}, 335 {"i915_gem_hws", i915_hws_info, 0},
336 {"i915_ringbuffer_data", i915_ringbuffer_data, 0},
337 {"i915_ringbuffer_info", i915_ringbuffer_info, 0},
338 {"i915_batchbuffers", i915_batchbuffer_info, 0},
246}; 339};
247#define I915_GEM_DEBUGFS_ENTRIES ARRAY_SIZE(i915_gem_debugfs_list) 340#define I915_GEM_DEBUGFS_ENTRIES ARRAY_SIZE(i915_gem_debugfs_list)
248 341
diff --git a/drivers/gpu/drm/i915/i915_gem_tiling.c b/drivers/gpu/drm/i915/i915_gem_tiling.c
index 6be3f927c86a..f27e523c764f 100644
--- a/drivers/gpu/drm/i915/i915_gem_tiling.c
+++ b/drivers/gpu/drm/i915/i915_gem_tiling.c
@@ -25,6 +25,8 @@
25 * 25 *
26 */ 26 */
27 27
28#include "linux/string.h"
29#include "linux/bitops.h"
28#include "drmP.h" 30#include "drmP.h"
29#include "drm.h" 31#include "drm.h"
30#include "i915_drm.h" 32#include "i915_drm.h"
@@ -127,8 +129,8 @@ i915_gem_detect_bit_6_swizzle(struct drm_device *dev)
127 swizzle_y = I915_BIT_6_SWIZZLE_9_11; 129 swizzle_y = I915_BIT_6_SWIZZLE_9_11;
128 } else { 130 } else {
129 /* Bit 17 swizzling by the CPU in addition. */ 131 /* Bit 17 swizzling by the CPU in addition. */
130 swizzle_x = I915_BIT_6_SWIZZLE_UNKNOWN; 132 swizzle_x = I915_BIT_6_SWIZZLE_9_10_17;
131 swizzle_y = I915_BIT_6_SWIZZLE_UNKNOWN; 133 swizzle_y = I915_BIT_6_SWIZZLE_9_17;
132 } 134 }
133 break; 135 break;
134 } 136 }
@@ -288,6 +290,19 @@ i915_gem_set_tiling(struct drm_device *dev, void *data,
288 args->swizzle_mode = dev_priv->mm.bit_6_swizzle_x; 290 args->swizzle_mode = dev_priv->mm.bit_6_swizzle_x;
289 else 291 else
290 args->swizzle_mode = dev_priv->mm.bit_6_swizzle_y; 292 args->swizzle_mode = dev_priv->mm.bit_6_swizzle_y;
293
294 /* Hide bit 17 swizzling from the user. This prevents old Mesa
295 * from aborting the application on sw fallbacks to bit 17,
296 * and we use the pread/pwrite bit17 paths to swizzle for it.
297 * If there was a user that was relying on the swizzle
298 * information for drm_intel_bo_map()ed reads/writes this would
299 * break it, but we don't have any of those.
300 */
301 if (args->swizzle_mode == I915_BIT_6_SWIZZLE_9_17)
302 args->swizzle_mode = I915_BIT_6_SWIZZLE_9;
303 if (args->swizzle_mode == I915_BIT_6_SWIZZLE_9_10_17)
304 args->swizzle_mode = I915_BIT_6_SWIZZLE_9_10;
305
291 /* If we can't handle the swizzling, make it untiled. */ 306 /* If we can't handle the swizzling, make it untiled. */
292 if (args->swizzle_mode == I915_BIT_6_SWIZZLE_UNKNOWN) { 307 if (args->swizzle_mode == I915_BIT_6_SWIZZLE_UNKNOWN) {
293 args->tiling_mode = I915_TILING_NONE; 308 args->tiling_mode = I915_TILING_NONE;
@@ -354,8 +369,100 @@ i915_gem_get_tiling(struct drm_device *dev, void *data,
354 DRM_ERROR("unknown tiling mode\n"); 369 DRM_ERROR("unknown tiling mode\n");
355 } 370 }
356 371
372 /* Hide bit 17 from the user -- see comment in i915_gem_set_tiling */
373 if (args->swizzle_mode == I915_BIT_6_SWIZZLE_9_17)
374 args->swizzle_mode = I915_BIT_6_SWIZZLE_9;
375 if (args->swizzle_mode == I915_BIT_6_SWIZZLE_9_10_17)
376 args->swizzle_mode = I915_BIT_6_SWIZZLE_9_10;
377
357 drm_gem_object_unreference(obj); 378 drm_gem_object_unreference(obj);
358 mutex_unlock(&dev->struct_mutex); 379 mutex_unlock(&dev->struct_mutex);
359 380
360 return 0; 381 return 0;
361} 382}
383
384/**
385 * Swap every 64 bytes of this page around, to account for it having a new
386 * bit 17 of its physical address and therefore being interpreted differently
387 * by the GPU.
388 */
389static int
390i915_gem_swizzle_page(struct page *page)
391{
392 char *vaddr;
393 int i;
394 char temp[64];
395
396 vaddr = kmap(page);
397 if (vaddr == NULL)
398 return -ENOMEM;
399
400 for (i = 0; i < PAGE_SIZE; i += 128) {
401 memcpy(temp, &vaddr[i], 64);
402 memcpy(&vaddr[i], &vaddr[i + 64], 64);
403 memcpy(&vaddr[i + 64], temp, 64);
404 }
405
406 kunmap(page);
407
408 return 0;
409}
410
411void
412i915_gem_object_do_bit_17_swizzle(struct drm_gem_object *obj)
413{
414 struct drm_device *dev = obj->dev;
415 drm_i915_private_t *dev_priv = dev->dev_private;
416 struct drm_i915_gem_object *obj_priv = obj->driver_private;
417 int page_count = obj->size >> PAGE_SHIFT;
418 int i;
419
420 if (dev_priv->mm.bit_6_swizzle_x != I915_BIT_6_SWIZZLE_9_10_17)
421 return;
422
423 if (obj_priv->bit_17 == NULL)
424 return;
425
426 for (i = 0; i < page_count; i++) {
427 char new_bit_17 = page_to_phys(obj_priv->pages[i]) >> 17;
428 if ((new_bit_17 & 0x1) !=
429 (test_bit(i, obj_priv->bit_17) != 0)) {
430 int ret = i915_gem_swizzle_page(obj_priv->pages[i]);
431 if (ret != 0) {
432 DRM_ERROR("Failed to swizzle page\n");
433 return;
434 }
435 set_page_dirty(obj_priv->pages[i]);
436 }
437 }
438}
439
440void
441i915_gem_object_save_bit_17_swizzle(struct drm_gem_object *obj)
442{
443 struct drm_device *dev = obj->dev;
444 drm_i915_private_t *dev_priv = dev->dev_private;
445 struct drm_i915_gem_object *obj_priv = obj->driver_private;
446 int page_count = obj->size >> PAGE_SHIFT;
447 int i;
448
449 if (dev_priv->mm.bit_6_swizzle_x != I915_BIT_6_SWIZZLE_9_10_17)
450 return;
451
452 if (obj_priv->bit_17 == NULL) {
453 obj_priv->bit_17 = kmalloc(BITS_TO_LONGS(page_count) *
454 sizeof(long), GFP_KERNEL);
455 if (obj_priv->bit_17 == NULL) {
456 DRM_ERROR("Failed to allocate memory for bit 17 "
457 "record\n");
458 return;
459 }
460 }
461
462 for (i = 0; i < page_count; i++) {
463 if (page_to_phys(obj_priv->pages[i]) & (1 << 17))
464 __set_bit(i, obj_priv->bit_17);
465 else
466 __clear_bit(i, obj_priv->bit_17);
467 }
468}
diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c
index 64773ce52964..c2c8e95ff14d 100644
--- a/drivers/gpu/drm/i915/intel_display.c
+++ b/drivers/gpu/drm/i915/intel_display.c
@@ -367,6 +367,7 @@ static const intel_limit_t intel_limits[] = {
367 .p1 = { .min = I9XX_P1_MIN, .max = I9XX_P1_MAX }, 367 .p1 = { .min = I9XX_P1_MIN, .max = I9XX_P1_MAX },
368 .p2 = { .dot_limit = I9XX_P2_SDVO_DAC_SLOW_LIMIT, 368 .p2 = { .dot_limit = I9XX_P2_SDVO_DAC_SLOW_LIMIT,
369 .p2_slow = I9XX_P2_SDVO_DAC_SLOW, .p2_fast = I9XX_P2_SDVO_DAC_FAST }, 369 .p2_slow = I9XX_P2_SDVO_DAC_SLOW, .p2_fast = I9XX_P2_SDVO_DAC_FAST },
370 .find_pll = intel_find_best_PLL,
370 }, 371 },
371 { /* INTEL_LIMIT_IGD_LVDS */ 372 { /* INTEL_LIMIT_IGD_LVDS */
372 .dot = { .min = I9XX_DOT_MIN, .max = I9XX_DOT_MAX }, 373 .dot = { .min = I9XX_DOT_MIN, .max = I9XX_DOT_MAX },
@@ -380,6 +381,7 @@ static const intel_limit_t intel_limits[] = {
380 /* IGD only supports single-channel mode. */ 381 /* IGD only supports single-channel mode. */
381 .p2 = { .dot_limit = I9XX_P2_LVDS_SLOW_LIMIT, 382 .p2 = { .dot_limit = I9XX_P2_LVDS_SLOW_LIMIT,
382 .p2_slow = I9XX_P2_LVDS_SLOW, .p2_fast = I9XX_P2_LVDS_SLOW }, 383 .p2_slow = I9XX_P2_LVDS_SLOW, .p2_fast = I9XX_P2_LVDS_SLOW },
384 .find_pll = intel_find_best_PLL,
383 }, 385 },
384 386
385}; 387};
diff --git a/drivers/gpu/drm/i915/intel_fb.c b/drivers/gpu/drm/i915/intel_fb.c
index b7f0ebe9f810..3e094beecb99 100644
--- a/drivers/gpu/drm/i915/intel_fb.c
+++ b/drivers/gpu/drm/i915/intel_fb.c
@@ -864,8 +864,8 @@ static void intelfb_sysrq(int dummy1, struct tty_struct *dummy3)
864 864
865static struct sysrq_key_op sysrq_intelfb_restore_op = { 865static struct sysrq_key_op sysrq_intelfb_restore_op = {
866 .handler = intelfb_sysrq, 866 .handler = intelfb_sysrq,
867 .help_msg = "force fb", 867 .help_msg = "force-fb(G)",
868 .action_msg = "force restore of fb console", 868 .action_msg = "Restore framebuffer console",
869}; 869};
870 870
871int intelfb_probe(struct drm_device *dev) 871int intelfb_probe(struct drm_device *dev)
diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c
index b06a4a3ff08d..550374225388 100644
--- a/drivers/gpu/drm/i915/intel_hdmi.c
+++ b/drivers/gpu/drm/i915/intel_hdmi.c
@@ -38,7 +38,7 @@
38struct intel_hdmi_priv { 38struct intel_hdmi_priv {
39 u32 sdvox_reg; 39 u32 sdvox_reg;
40 u32 save_SDVOX; 40 u32 save_SDVOX;
41 int has_hdmi_sink; 41 bool has_hdmi_sink;
42}; 42};
43 43
44static void intel_hdmi_mode_set(struct drm_encoder *encoder, 44static void intel_hdmi_mode_set(struct drm_encoder *encoder,
@@ -128,6 +128,22 @@ static bool intel_hdmi_mode_fixup(struct drm_encoder *encoder,
128 return true; 128 return true;
129} 129}
130 130
131static void
132intel_hdmi_sink_detect(struct drm_connector *connector)
133{
134 struct intel_output *intel_output = to_intel_output(connector);
135 struct intel_hdmi_priv *hdmi_priv = intel_output->dev_priv;
136 struct edid *edid = NULL;
137
138 edid = drm_get_edid(&intel_output->base,
139 &intel_output->ddc_bus->adapter);
140 if (edid != NULL) {
141 hdmi_priv->has_hdmi_sink = drm_detect_hdmi_monitor(edid);
142 kfree(edid);
143 intel_output->base.display_info.raw_edid = NULL;
144 }
145}
146
131static enum drm_connector_status 147static enum drm_connector_status
132intel_hdmi_detect(struct drm_connector *connector) 148intel_hdmi_detect(struct drm_connector *connector)
133{ 149{
@@ -158,9 +174,10 @@ intel_hdmi_detect(struct drm_connector *connector)
158 return connector_status_unknown; 174 return connector_status_unknown;
159 } 175 }
160 176
161 if ((I915_READ(PORT_HOTPLUG_STAT) & bit) != 0) 177 if ((I915_READ(PORT_HOTPLUG_STAT) & bit) != 0) {
178 intel_hdmi_sink_detect(connector);
162 return connector_status_connected; 179 return connector_status_connected;
163 else 180 } else
164 return connector_status_disconnected; 181 return connector_status_disconnected;
165} 182}
166 183
diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c
index 7b31f55f55c8..9913651c1e17 100644
--- a/drivers/gpu/drm/i915/intel_sdvo.c
+++ b/drivers/gpu/drm/i915/intel_sdvo.c
@@ -1357,6 +1357,23 @@ void intel_sdvo_set_hotplug(struct drm_connector *connector, int on)
1357 intel_sdvo_read_response(intel_output, &response, 2); 1357 intel_sdvo_read_response(intel_output, &response, 2);
1358} 1358}
1359 1359
1360static void
1361intel_sdvo_hdmi_sink_detect(struct drm_connector *connector)
1362{
1363 struct intel_output *intel_output = to_intel_output(connector);
1364 struct intel_sdvo_priv *sdvo_priv = intel_output->dev_priv;
1365 struct edid *edid = NULL;
1366
1367 intel_sdvo_set_control_bus_switch(intel_output, sdvo_priv->ddc_bus);
1368 edid = drm_get_edid(&intel_output->base,
1369 &intel_output->ddc_bus->adapter);
1370 if (edid != NULL) {
1371 sdvo_priv->is_hdmi = drm_detect_hdmi_monitor(edid);
1372 kfree(edid);
1373 intel_output->base.display_info.raw_edid = NULL;
1374 }
1375}
1376
1360static enum drm_connector_status intel_sdvo_detect(struct drm_connector *connector) 1377static enum drm_connector_status intel_sdvo_detect(struct drm_connector *connector)
1361{ 1378{
1362 u8 response[2]; 1379 u8 response[2];
@@ -1371,9 +1388,10 @@ static enum drm_connector_status intel_sdvo_detect(struct drm_connector *connect
1371 if (status != SDVO_CMD_STATUS_SUCCESS) 1388 if (status != SDVO_CMD_STATUS_SUCCESS)
1372 return connector_status_unknown; 1389 return connector_status_unknown;
1373 1390
1374 if ((response[0] != 0) || (response[1] != 0)) 1391 if ((response[0] != 0) || (response[1] != 0)) {
1392 intel_sdvo_hdmi_sink_detect(connector);
1375 return connector_status_connected; 1393 return connector_status_connected;
1376 else 1394 } else
1377 return connector_status_disconnected; 1395 return connector_status_disconnected;
1378} 1396}
1379 1397
diff --git a/drivers/hwmon/Kconfig b/drivers/hwmon/Kconfig
index 0e8a9185f676..d73f5f473e38 100644
--- a/drivers/hwmon/Kconfig
+++ b/drivers/hwmon/Kconfig
@@ -692,6 +692,16 @@ config SENSORS_PCF8591
692 These devices are hard to detect and rarely found on mainstream 692 These devices are hard to detect and rarely found on mainstream
693 hardware. If unsure, say N. 693 hardware. If unsure, say N.
694 694
695config SENSORS_SHT15
696 tristate "Sensiron humidity and temperature sensors. SHT15 and compat."
697 depends on GENERIC_GPIO
698 help
699 If you say yes here you get support for the Sensiron SHT10, SHT11,
700 SHT15, SHT71, SHT75 humidity and temperature sensors.
701
702 This driver can also be built as a module. If so, the module
703 will be called sht15.
704
695config SENSORS_SIS5595 705config SENSORS_SIS5595
696 tristate "Silicon Integrated Systems Corp. SiS5595" 706 tristate "Silicon Integrated Systems Corp. SiS5595"
697 depends on PCI 707 depends on PCI
diff --git a/drivers/hwmon/Makefile b/drivers/hwmon/Makefile
index 1d3757837b4f..0ae26984ba45 100644
--- a/drivers/hwmon/Makefile
+++ b/drivers/hwmon/Makefile
@@ -76,6 +76,7 @@ obj-$(CONFIG_SENSORS_MAX6650) += max6650.o
76obj-$(CONFIG_SENSORS_PC87360) += pc87360.o 76obj-$(CONFIG_SENSORS_PC87360) += pc87360.o
77obj-$(CONFIG_SENSORS_PC87427) += pc87427.o 77obj-$(CONFIG_SENSORS_PC87427) += pc87427.o
78obj-$(CONFIG_SENSORS_PCF8591) += pcf8591.o 78obj-$(CONFIG_SENSORS_PCF8591) += pcf8591.o
79obj-$(CONFIG_SENSORS_SHT15) += sht15.o
79obj-$(CONFIG_SENSORS_SIS5595) += sis5595.o 80obj-$(CONFIG_SENSORS_SIS5595) += sis5595.o
80obj-$(CONFIG_SENSORS_SMSC47B397)+= smsc47b397.o 81obj-$(CONFIG_SENSORS_SMSC47B397)+= smsc47b397.o
81obj-$(CONFIG_SENSORS_SMSC47M1) += smsc47m1.o 82obj-$(CONFIG_SENSORS_SMSC47M1) += smsc47m1.o
diff --git a/drivers/hwmon/hp_accel.c b/drivers/hwmon/hp_accel.c
index 55d3dc565be6..abca7e9f953b 100644
--- a/drivers/hwmon/hp_accel.c
+++ b/drivers/hwmon/hp_accel.c
@@ -34,7 +34,6 @@
34#include <linux/wait.h> 34#include <linux/wait.h>
35#include <linux/poll.h> 35#include <linux/poll.h>
36#include <linux/freezer.h> 36#include <linux/freezer.h>
37#include <linux/version.h>
38#include <linux/uaccess.h> 37#include <linux/uaccess.h>
39#include <linux/leds.h> 38#include <linux/leds.h>
40#include <acpi/acpi_drivers.h> 39#include <acpi/acpi_drivers.h>
diff --git a/drivers/hwmon/sht15.c b/drivers/hwmon/sht15.c
new file mode 100644
index 000000000000..6cbdc2fea734
--- /dev/null
+++ b/drivers/hwmon/sht15.c
@@ -0,0 +1,692 @@
1/*
2 * sht15.c - support for the SHT15 Temperature and Humidity Sensor
3 *
4 * Copyright (c) 2009 Jonathan Cameron
5 *
6 * Copyright (c) 2007 Wouter Horre
7 *
8 * This program is free software; you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License version 2 as
10 * published by the Free Software Foundation.
11 *
12 * Currently ignoring checksum on readings.
13 * Default resolution only (14bit temp, 12bit humidity)
14 * Ignoring battery status.
15 * Heater not enabled.
16 * Timings are all conservative.
17 *
18 * Data sheet available (1/2009) at
19 * http://www.sensirion.ch/en/pdf/product_information/Datasheet-humidity-sensor-SHT1x.pdf
20 *
21 * Regulator supply name = vcc
22 */
23
24#include <linux/interrupt.h>
25#include <linux/irq.h>
26#include <linux/gpio.h>
27#include <linux/module.h>
28#include <linux/init.h>
29#include <linux/hwmon.h>
30#include <linux/hwmon-sysfs.h>
31#include <linux/mutex.h>
32#include <linux/platform_device.h>
33#include <linux/delay.h>
34#include <linux/jiffies.h>
35#include <linux/err.h>
36#include <linux/sht15.h>
37#include <linux/regulator/consumer.h>
38#include <asm/atomic.h>
39
40#define SHT15_MEASURE_TEMP 3
41#define SHT15_MEASURE_RH 5
42
43#define SHT15_READING_NOTHING 0
44#define SHT15_READING_TEMP 1
45#define SHT15_READING_HUMID 2
46
47/* Min timings in nsecs */
48#define SHT15_TSCKL 100 /* clock low */
49#define SHT15_TSCKH 100 /* clock high */
50#define SHT15_TSU 150 /* data setup time */
51
52/**
53 * struct sht15_temppair - elements of voltage dependant temp calc
54 * @vdd: supply voltage in microvolts
55 * @d1: see data sheet
56 */
57struct sht15_temppair {
58 int vdd; /* microvolts */
59 int d1;
60};
61
62/* Table 9 from data sheet - relates temperature calculation
63 * to supply voltage.
64 */
65static const struct sht15_temppair temppoints[] = {
66 { 2500000, -39400 },
67 { 3000000, -39600 },
68 { 3500000, -39700 },
69 { 4000000, -39800 },
70 { 5000000, -40100 },
71};
72
73/**
74 * struct sht15_data - device instance specific data
75 * @pdata: platform data (gpio's etc)
76 * @read_work: bh of interrupt handler
77 * @wait_queue: wait queue for getting values from device
78 * @val_temp: last temperature value read from device
79 * @val_humid: last humidity value read from device
80 * @flag: status flag used to identify what the last request was
81 * @valid: are the current stored values valid (start condition)
82 * @last_updat: time of last update
83 * @read_lock: mutex to ensure only one read in progress
84 * at a time.
85 * @dev: associate device structure
86 * @hwmon_dev: device associated with hwmon subsystem
87 * @reg: associated regulator (if specified)
88 * @nb: notifier block to handle notifications of voltage changes
89 * @supply_uV: local copy of supply voltage used to allow
90 * use of regulator consumer if available
91 * @supply_uV_valid: indicates that an updated value has not yet
92 * been obtained from the regulator and so any calculations
93 * based upon it will be invalid.
94 * @update_supply_work: work struct that is used to update the supply_uV
95 * @interrupt_handled: flag used to indicate a hander has been scheduled
96 */
97struct sht15_data {
98 struct sht15_platform_data *pdata;
99 struct work_struct read_work;
100 wait_queue_head_t wait_queue;
101 uint16_t val_temp;
102 uint16_t val_humid;
103 u8 flag;
104 u8 valid;
105 unsigned long last_updat;
106 struct mutex read_lock;
107 struct device *dev;
108 struct device *hwmon_dev;
109 struct regulator *reg;
110 struct notifier_block nb;
111 int supply_uV;
112 int supply_uV_valid;
113 struct work_struct update_supply_work;
114 atomic_t interrupt_handled;
115};
116
117/**
118 * sht15_connection_reset() - reset the comms interface
119 * @data: sht15 specific data
120 *
121 * This implements section 3.4 of the data sheet
122 */
123static void sht15_connection_reset(struct sht15_data *data)
124{
125 int i;
126 gpio_direction_output(data->pdata->gpio_data, 1);
127 ndelay(SHT15_TSCKL);
128 gpio_set_value(data->pdata->gpio_sck, 0);
129 ndelay(SHT15_TSCKL);
130 for (i = 0; i < 9; ++i) {
131 gpio_set_value(data->pdata->gpio_sck, 1);
132 ndelay(SHT15_TSCKH);
133 gpio_set_value(data->pdata->gpio_sck, 0);
134 ndelay(SHT15_TSCKL);
135 }
136}
137/**
138 * sht15_send_bit() - send an individual bit to the device
139 * @data: device state data
140 * @val: value of bit to be sent
141 **/
142static inline void sht15_send_bit(struct sht15_data *data, int val)
143{
144
145 gpio_set_value(data->pdata->gpio_data, val);
146 ndelay(SHT15_TSU);
147 gpio_set_value(data->pdata->gpio_sck, 1);
148 ndelay(SHT15_TSCKH);
149 gpio_set_value(data->pdata->gpio_sck, 0);
150 ndelay(SHT15_TSCKL); /* clock low time */
151}
152
153/**
154 * sht15_transmission_start() - specific sequence for new transmission
155 *
156 * @data: device state data
157 * Timings for this are not documented on the data sheet, so very
158 * conservative ones used in implementation. This implements
159 * figure 12 on the data sheet.
160 **/
161static void sht15_transmission_start(struct sht15_data *data)
162{
163 /* ensure data is high and output */
164 gpio_direction_output(data->pdata->gpio_data, 1);
165 ndelay(SHT15_TSU);
166 gpio_set_value(data->pdata->gpio_sck, 0);
167 ndelay(SHT15_TSCKL);
168 gpio_set_value(data->pdata->gpio_sck, 1);
169 ndelay(SHT15_TSCKH);
170 gpio_set_value(data->pdata->gpio_data, 0);
171 ndelay(SHT15_TSU);
172 gpio_set_value(data->pdata->gpio_sck, 0);
173 ndelay(SHT15_TSCKL);
174 gpio_set_value(data->pdata->gpio_sck, 1);
175 ndelay(SHT15_TSCKH);
176 gpio_set_value(data->pdata->gpio_data, 1);
177 ndelay(SHT15_TSU);
178 gpio_set_value(data->pdata->gpio_sck, 0);
179 ndelay(SHT15_TSCKL);
180}
181/**
182 * sht15_send_byte() - send a single byte to the device
183 * @data: device state
184 * @byte: value to be sent
185 **/
186static void sht15_send_byte(struct sht15_data *data, u8 byte)
187{
188 int i;
189 for (i = 0; i < 8; i++) {
190 sht15_send_bit(data, !!(byte & 0x80));
191 byte <<= 1;
192 }
193}
194/**
195 * sht15_wait_for_response() - checks for ack from device
196 * @data: device state
197 **/
198static int sht15_wait_for_response(struct sht15_data *data)
199{
200 gpio_direction_input(data->pdata->gpio_data);
201 gpio_set_value(data->pdata->gpio_sck, 1);
202 ndelay(SHT15_TSCKH);
203 if (gpio_get_value(data->pdata->gpio_data)) {
204 gpio_set_value(data->pdata->gpio_sck, 0);
205 dev_err(data->dev, "Command not acknowledged\n");
206 sht15_connection_reset(data);
207 return -EIO;
208 }
209 gpio_set_value(data->pdata->gpio_sck, 0);
210 ndelay(SHT15_TSCKL);
211 return 0;
212}
213
214/**
215 * sht15_send_cmd() - Sends a command to the device.
216 * @data: device state
217 * @cmd: command byte to be sent
218 *
219 * On entry, sck is output low, data is output pull high
220 * and the interrupt disabled.
221 **/
222static int sht15_send_cmd(struct sht15_data *data, u8 cmd)
223{
224 int ret = 0;
225 sht15_transmission_start(data);
226 sht15_send_byte(data, cmd);
227 ret = sht15_wait_for_response(data);
228 return ret;
229}
230/**
231 * sht15_update_single_val() - get a new value from device
232 * @data: device instance specific data
233 * @command: command sent to request value
234 * @timeout_msecs: timeout after which comms are assumed
235 * to have failed are reset.
236 **/
237static inline int sht15_update_single_val(struct sht15_data *data,
238 int command,
239 int timeout_msecs)
240{
241 int ret;
242 ret = sht15_send_cmd(data, command);
243 if (ret)
244 return ret;
245
246 gpio_direction_input(data->pdata->gpio_data);
247 atomic_set(&data->interrupt_handled, 0);
248
249 enable_irq(gpio_to_irq(data->pdata->gpio_data));
250 if (gpio_get_value(data->pdata->gpio_data) == 0) {
251 disable_irq_nosync(gpio_to_irq(data->pdata->gpio_data));
252 /* Only relevant if the interrupt hasn't occured. */
253 if (!atomic_read(&data->interrupt_handled))
254 schedule_work(&data->read_work);
255 }
256 ret = wait_event_timeout(data->wait_queue,
257 (data->flag == SHT15_READING_NOTHING),
258 msecs_to_jiffies(timeout_msecs));
259 if (ret == 0) {/* timeout occurred */
260 disable_irq_nosync(gpio_to_irq(data->pdata->gpio_data));;
261 sht15_connection_reset(data);
262 return -ETIME;
263 }
264 return 0;
265}
266
267/**
268 * sht15_update_vals() - get updated readings from device if too old
269 * @data: device state
270 **/
271static int sht15_update_vals(struct sht15_data *data)
272{
273 int ret = 0;
274 int timeout = HZ;
275
276 mutex_lock(&data->read_lock);
277 if (time_after(jiffies, data->last_updat + timeout)
278 || !data->valid) {
279 data->flag = SHT15_READING_HUMID;
280 ret = sht15_update_single_val(data, SHT15_MEASURE_RH, 160);
281 if (ret)
282 goto error_ret;
283 data->flag = SHT15_READING_TEMP;
284 ret = sht15_update_single_val(data, SHT15_MEASURE_TEMP, 400);
285 if (ret)
286 goto error_ret;
287 data->valid = 1;
288 data->last_updat = jiffies;
289 }
290error_ret:
291 mutex_unlock(&data->read_lock);
292
293 return ret;
294}
295
296/**
297 * sht15_calc_temp() - convert the raw reading to a temperature
298 * @data: device state
299 *
300 * As per section 4.3 of the data sheet.
301 **/
302static inline int sht15_calc_temp(struct sht15_data *data)
303{
304 int d1 = 0;
305 int i;
306
307 for (i = 1; i < ARRAY_SIZE(temppoints) - 1; i++)
308 /* Find pointer to interpolate */
309 if (data->supply_uV > temppoints[i - 1].vdd) {
310 d1 = (data->supply_uV/1000 - temppoints[i - 1].vdd)
311 * (temppoints[i].d1 - temppoints[i - 1].d1)
312 / (temppoints[i].vdd - temppoints[i - 1].vdd)
313 + temppoints[i - 1].d1;
314 break;
315 }
316
317 return data->val_temp*10 + d1;
318}
319
320/**
321 * sht15_calc_humid() - using last temperature convert raw to humid
322 * @data: device state
323 *
324 * This is the temperature compensated version as per section 4.2 of
325 * the data sheet.
326 **/
327static inline int sht15_calc_humid(struct sht15_data *data)
328{
329 int RHlinear; /* milli percent */
330 int temp = sht15_calc_temp(data);
331
332 const int c1 = -4;
333 const int c2 = 40500; /* x 10 ^ -6 */
334 const int c3 = 2800; /* x10 ^ -9 */
335
336 RHlinear = c1*1000
337 + c2 * data->val_humid/1000
338 + (data->val_humid * data->val_humid * c3)/1000000;
339 return (temp - 25000) * (10000 + 800 * data->val_humid)
340 / 1000000 + RHlinear;
341}
342
343static ssize_t sht15_show_temp(struct device *dev,
344 struct device_attribute *attr,
345 char *buf)
346{
347 int ret;
348 struct sht15_data *data = dev_get_drvdata(dev);
349
350 /* Technically no need to read humidity as well */
351 ret = sht15_update_vals(data);
352
353 return ret ? ret : sprintf(buf, "%d\n",
354 sht15_calc_temp(data));
355}
356
357static ssize_t sht15_show_humidity(struct device *dev,
358 struct device_attribute *attr,
359 char *buf)
360{
361 int ret;
362 struct sht15_data *data = dev_get_drvdata(dev);
363
364 ret = sht15_update_vals(data);
365
366 return ret ? ret : sprintf(buf, "%d\n", sht15_calc_humid(data));
367
368};
369static ssize_t show_name(struct device *dev,
370 struct device_attribute *attr,
371 char *buf)
372{
373 struct platform_device *pdev = to_platform_device(dev);
374 return sprintf(buf, "%s\n", pdev->name);
375}
376
377static SENSOR_DEVICE_ATTR(temp1_input,
378 S_IRUGO, sht15_show_temp,
379 NULL, 0);
380static SENSOR_DEVICE_ATTR(humidity1_input,
381 S_IRUGO, sht15_show_humidity,
382 NULL, 0);
383static DEVICE_ATTR(name, S_IRUGO, show_name, NULL);
384static struct attribute *sht15_attrs[] = {
385 &sensor_dev_attr_temp1_input.dev_attr.attr,
386 &sensor_dev_attr_humidity1_input.dev_attr.attr,
387 &dev_attr_name.attr,
388 NULL,
389};
390
391static const struct attribute_group sht15_attr_group = {
392 .attrs = sht15_attrs,
393};
394
395static irqreturn_t sht15_interrupt_fired(int irq, void *d)
396{
397 struct sht15_data *data = d;
398 /* First disable the interrupt */
399 disable_irq_nosync(irq);
400 atomic_inc(&data->interrupt_handled);
401 /* Then schedule a reading work struct */
402 if (data->flag != SHT15_READING_NOTHING)
403 schedule_work(&data->read_work);
404 return IRQ_HANDLED;
405}
406
407/* Each byte of data is acknowledged by pulling the data line
408 * low for one clock pulse.
409 */
410static void sht15_ack(struct sht15_data *data)
411{
412 gpio_direction_output(data->pdata->gpio_data, 0);
413 ndelay(SHT15_TSU);
414 gpio_set_value(data->pdata->gpio_sck, 1);
415 ndelay(SHT15_TSU);
416 gpio_set_value(data->pdata->gpio_sck, 0);
417 ndelay(SHT15_TSU);
418 gpio_set_value(data->pdata->gpio_data, 1);
419
420 gpio_direction_input(data->pdata->gpio_data);
421}
422/**
423 * sht15_end_transmission() - notify device of end of transmission
424 * @data: device state
425 *
426 * This is basically a NAK. (single clock pulse, data high)
427 **/
428static void sht15_end_transmission(struct sht15_data *data)
429{
430 gpio_direction_output(data->pdata->gpio_data, 1);
431 ndelay(SHT15_TSU);
432 gpio_set_value(data->pdata->gpio_sck, 1);
433 ndelay(SHT15_TSCKH);
434 gpio_set_value(data->pdata->gpio_sck, 0);
435 ndelay(SHT15_TSCKL);
436}
437
438static void sht15_bh_read_data(struct work_struct *work_s)
439{
440 int i;
441 uint16_t val = 0;
442 struct sht15_data *data
443 = container_of(work_s, struct sht15_data,
444 read_work);
445 /* Firstly, verify the line is low */
446 if (gpio_get_value(data->pdata->gpio_data)) {
447 /* If not, then start the interrupt again - care
448 here as could have gone low in meantime so verify
449 it hasn't!
450 */
451 atomic_set(&data->interrupt_handled, 0);
452 enable_irq(gpio_to_irq(data->pdata->gpio_data));
453 /* If still not occured or another handler has been scheduled */
454 if (gpio_get_value(data->pdata->gpio_data)
455 || atomic_read(&data->interrupt_handled))
456 return;
457 }
458 /* Read the data back from the device */
459 for (i = 0; i < 16; ++i) {
460 val <<= 1;
461 gpio_set_value(data->pdata->gpio_sck, 1);
462 ndelay(SHT15_TSCKH);
463 val |= !!gpio_get_value(data->pdata->gpio_data);
464 gpio_set_value(data->pdata->gpio_sck, 0);
465 ndelay(SHT15_TSCKL);
466 if (i == 7)
467 sht15_ack(data);
468 }
469 /* Tell the device we are done */
470 sht15_end_transmission(data);
471
472 switch (data->flag) {
473 case SHT15_READING_TEMP:
474 data->val_temp = val;
475 break;
476 case SHT15_READING_HUMID:
477 data->val_humid = val;
478 break;
479 }
480
481 data->flag = SHT15_READING_NOTHING;
482 wake_up(&data->wait_queue);
483}
484
485static void sht15_update_voltage(struct work_struct *work_s)
486{
487 struct sht15_data *data
488 = container_of(work_s, struct sht15_data,
489 update_supply_work);
490 data->supply_uV = regulator_get_voltage(data->reg);
491}
492
493/**
494 * sht15_invalidate_voltage() - mark supply voltage invalid when notified by reg
495 * @nb: associated notification structure
496 * @event: voltage regulator state change event code
497 * @ignored: function parameter - ignored here
498 *
499 * Note that as the notification code holds the regulator lock, we have
500 * to schedule an update of the supply voltage rather than getting it directly.
501 **/
502static int sht15_invalidate_voltage(struct notifier_block *nb,
503 unsigned long event,
504 void *ignored)
505{
506 struct sht15_data *data = container_of(nb, struct sht15_data, nb);
507
508 if (event == REGULATOR_EVENT_VOLTAGE_CHANGE)
509 data->supply_uV_valid = false;
510 schedule_work(&data->update_supply_work);
511
512 return NOTIFY_OK;
513}
514
515static int __devinit sht15_probe(struct platform_device *pdev)
516{
517 int ret = 0;
518 struct sht15_data *data = kzalloc(sizeof(*data), GFP_KERNEL);
519
520 if (!data) {
521 ret = -ENOMEM;
522 dev_err(&pdev->dev, "kzalloc failed");
523 goto error_ret;
524 }
525
526 INIT_WORK(&data->read_work, sht15_bh_read_data);
527 INIT_WORK(&data->update_supply_work, sht15_update_voltage);
528 platform_set_drvdata(pdev, data);
529 mutex_init(&data->read_lock);
530 data->dev = &pdev->dev;
531 init_waitqueue_head(&data->wait_queue);
532
533 if (pdev->dev.platform_data == NULL) {
534 dev_err(&pdev->dev, "no platform data supplied");
535 goto err_free_data;
536 }
537 data->pdata = pdev->dev.platform_data;
538 data->supply_uV = data->pdata->supply_mv*1000;
539
540/* If a regulator is available, query what the supply voltage actually is!*/
541 data->reg = regulator_get(data->dev, "vcc");
542 if (!IS_ERR(data->reg)) {
543 data->supply_uV = regulator_get_voltage(data->reg);
544 regulator_enable(data->reg);
545 /* setup a notifier block to update this if another device
546 * causes the voltage to change */
547 data->nb.notifier_call = &sht15_invalidate_voltage;
548 ret = regulator_register_notifier(data->reg, &data->nb);
549 }
550/* Try requesting the GPIOs */
551 ret = gpio_request(data->pdata->gpio_sck, "SHT15 sck");
552 if (ret) {
553 dev_err(&pdev->dev, "gpio request failed");
554 goto err_free_data;
555 }
556 gpio_direction_output(data->pdata->gpio_sck, 0);
557 ret = gpio_request(data->pdata->gpio_data, "SHT15 data");
558 if (ret) {
559 dev_err(&pdev->dev, "gpio request failed");
560 goto err_release_gpio_sck;
561 }
562 ret = sysfs_create_group(&pdev->dev.kobj, &sht15_attr_group);
563 if (ret) {
564 dev_err(&pdev->dev, "sysfs create failed");
565 goto err_free_data;
566 }
567
568 ret = request_irq(gpio_to_irq(data->pdata->gpio_data),
569 sht15_interrupt_fired,
570 IRQF_TRIGGER_FALLING,
571 "sht15 data",
572 data);
573 if (ret) {
574 dev_err(&pdev->dev, "failed to get irq for data line");
575 goto err_release_gpio_data;
576 }
577 disable_irq_nosync(gpio_to_irq(data->pdata->gpio_data));
578 sht15_connection_reset(data);
579 sht15_send_cmd(data, 0x1E);
580
581 data->hwmon_dev = hwmon_device_register(data->dev);
582 if (IS_ERR(data->hwmon_dev)) {
583 ret = PTR_ERR(data->hwmon_dev);
584 goto err_release_gpio_data;
585 }
586 return 0;
587
588err_release_gpio_data:
589 gpio_free(data->pdata->gpio_data);
590err_release_gpio_sck:
591 gpio_free(data->pdata->gpio_sck);
592err_free_data:
593 kfree(data);
594error_ret:
595
596 return ret;
597}
598
599static int __devexit sht15_remove(struct platform_device *pdev)
600{
601 struct sht15_data *data = platform_get_drvdata(pdev);
602
603 /* Make sure any reads from the device are done and
604 * prevent new ones beginnning */
605 mutex_lock(&data->read_lock);
606 hwmon_device_unregister(data->hwmon_dev);
607 sysfs_remove_group(&pdev->dev.kobj, &sht15_attr_group);
608 if (!IS_ERR(data->reg)) {
609 regulator_unregister_notifier(data->reg, &data->nb);
610 regulator_disable(data->reg);
611 regulator_put(data->reg);
612 }
613
614 free_irq(gpio_to_irq(data->pdata->gpio_data), data);
615 gpio_free(data->pdata->gpio_data);
616 gpio_free(data->pdata->gpio_sck);
617 mutex_unlock(&data->read_lock);
618 kfree(data);
619 return 0;
620}
621
622
623static struct platform_driver sht_drivers[] = {
624 {
625 .driver = {
626 .name = "sht10",
627 .owner = THIS_MODULE,
628 },
629 .probe = sht15_probe,
630 .remove = sht15_remove,
631 }, {
632 .driver = {
633 .name = "sht11",
634 .owner = THIS_MODULE,
635 },
636 .probe = sht15_probe,
637 .remove = sht15_remove,
638 }, {
639 .driver = {
640 .name = "sht15",
641 .owner = THIS_MODULE,
642 },
643 .probe = sht15_probe,
644 .remove = sht15_remove,
645 }, {
646 .driver = {
647 .name = "sht71",
648 .owner = THIS_MODULE,
649 },
650 .probe = sht15_probe,
651 .remove = sht15_remove,
652 }, {
653 .driver = {
654 .name = "sht75",
655 .owner = THIS_MODULE,
656 },
657 .probe = sht15_probe,
658 .remove = sht15_remove,
659 },
660};
661
662
663static int __init sht15_init(void)
664{
665 int ret;
666 int i;
667
668 for (i = 0; i < ARRAY_SIZE(sht_drivers); i++) {
669 ret = platform_driver_register(&sht_drivers[i]);
670 if (ret)
671 goto error_unreg;
672 }
673
674 return 0;
675
676error_unreg:
677 while (--i >= 0)
678 platform_driver_unregister(&sht_drivers[i]);
679
680 return ret;
681}
682module_init(sht15_init);
683
684static void __exit sht15_exit(void)
685{
686 int i;
687 for (i = ARRAY_SIZE(sht_drivers) - 1; i >= 0; i--)
688 platform_driver_unregister(&sht_drivers[i]);
689}
690module_exit(sht15_exit);
691
692MODULE_LICENSE("GPL");
diff --git a/drivers/md/dm-bio-list.h b/drivers/md/dm-bio-list.h
deleted file mode 100644
index 345098b4ca77..000000000000
--- a/drivers/md/dm-bio-list.h
+++ /dev/null
@@ -1,117 +0,0 @@
1/*
2 * Copyright (C) 2004 Red Hat UK Ltd.
3 *
4 * This file is released under the GPL.
5 */
6
7#ifndef DM_BIO_LIST_H
8#define DM_BIO_LIST_H
9
10#include <linux/bio.h>
11
12#ifdef CONFIG_BLOCK
13
14struct bio_list {
15 struct bio *head;
16 struct bio *tail;
17};
18
19static inline int bio_list_empty(const struct bio_list *bl)
20{
21 return bl->head == NULL;
22}
23
24static inline void bio_list_init(struct bio_list *bl)
25{
26 bl->head = bl->tail = NULL;
27}
28
29#define bio_list_for_each(bio, bl) \
30 for (bio = (bl)->head; bio; bio = bio->bi_next)
31
32static inline unsigned bio_list_size(const struct bio_list *bl)
33{
34 unsigned sz = 0;
35 struct bio *bio;
36
37 bio_list_for_each(bio, bl)
38 sz++;
39
40 return sz;
41}
42
43static inline void bio_list_add(struct bio_list *bl, struct bio *bio)
44{
45 bio->bi_next = NULL;
46
47 if (bl->tail)
48 bl->tail->bi_next = bio;
49 else
50 bl->head = bio;
51
52 bl->tail = bio;
53}
54
55static inline void bio_list_add_head(struct bio_list *bl, struct bio *bio)
56{
57 bio->bi_next = bl->head;
58
59 bl->head = bio;
60
61 if (!bl->tail)
62 bl->tail = bio;
63}
64
65static inline void bio_list_merge(struct bio_list *bl, struct bio_list *bl2)
66{
67 if (!bl2->head)
68 return;
69
70 if (bl->tail)
71 bl->tail->bi_next = bl2->head;
72 else
73 bl->head = bl2->head;
74
75 bl->tail = bl2->tail;
76}
77
78static inline void bio_list_merge_head(struct bio_list *bl,
79 struct bio_list *bl2)
80{
81 if (!bl2->head)
82 return;
83
84 if (bl->head)
85 bl2->tail->bi_next = bl->head;
86 else
87 bl->tail = bl2->tail;
88
89 bl->head = bl2->head;
90}
91
92static inline struct bio *bio_list_pop(struct bio_list *bl)
93{
94 struct bio *bio = bl->head;
95
96 if (bio) {
97 bl->head = bl->head->bi_next;
98 if (!bl->head)
99 bl->tail = NULL;
100
101 bio->bi_next = NULL;
102 }
103
104 return bio;
105}
106
107static inline struct bio *bio_list_get(struct bio_list *bl)
108{
109 struct bio *bio = bl->head;
110
111 bl->head = bl->tail = NULL;
112
113 return bio;
114}
115
116#endif /* CONFIG_BLOCK */
117#endif
diff --git a/drivers/md/dm-delay.c b/drivers/md/dm-delay.c
index 59ee1b015d2d..559dbb52bc85 100644
--- a/drivers/md/dm-delay.c
+++ b/drivers/md/dm-delay.c
@@ -15,8 +15,6 @@
15 15
16#include <linux/device-mapper.h> 16#include <linux/device-mapper.h>
17 17
18#include "dm-bio-list.h"
19
20#define DM_MSG_PREFIX "delay" 18#define DM_MSG_PREFIX "delay"
21 19
22struct delay_c { 20struct delay_c {
diff --git a/drivers/md/dm-mpath.c b/drivers/md/dm-mpath.c
index 095f77bf9681..6a386ab4f7eb 100644
--- a/drivers/md/dm-mpath.c
+++ b/drivers/md/dm-mpath.c
@@ -8,7 +8,6 @@
8#include <linux/device-mapper.h> 8#include <linux/device-mapper.h>
9 9
10#include "dm-path-selector.h" 10#include "dm-path-selector.h"
11#include "dm-bio-list.h"
12#include "dm-bio-record.h" 11#include "dm-bio-record.h"
13#include "dm-uevent.h" 12#include "dm-uevent.h"
14 13
diff --git a/drivers/md/dm-raid1.c b/drivers/md/dm-raid1.c
index 536ef0bef154..076fbb4e967a 100644
--- a/drivers/md/dm-raid1.c
+++ b/drivers/md/dm-raid1.c
@@ -5,7 +5,6 @@
5 * This file is released under the GPL. 5 * This file is released under the GPL.
6 */ 6 */
7 7
8#include "dm-bio-list.h"
9#include "dm-bio-record.h" 8#include "dm-bio-record.h"
10 9
11#include <linux/init.h> 10#include <linux/init.h>
diff --git a/drivers/md/dm-region-hash.c b/drivers/md/dm-region-hash.c
index 59f8d9df9e1a..7b899be0b087 100644
--- a/drivers/md/dm-region-hash.c
+++ b/drivers/md/dm-region-hash.c
@@ -14,7 +14,6 @@
14#include <linux/vmalloc.h> 14#include <linux/vmalloc.h>
15 15
16#include "dm.h" 16#include "dm.h"
17#include "dm-bio-list.h"
18 17
19#define DM_MSG_PREFIX "region hash" 18#define DM_MSG_PREFIX "region hash"
20 19
diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c
index 981a0413068f..d73f17fc7778 100644
--- a/drivers/md/dm-snap.c
+++ b/drivers/md/dm-snap.c
@@ -22,7 +22,6 @@
22#include <linux/workqueue.h> 22#include <linux/workqueue.h>
23 23
24#include "dm-exception-store.h" 24#include "dm-exception-store.h"
25#include "dm-bio-list.h"
26 25
27#define DM_MSG_PREFIX "snapshots" 26#define DM_MSG_PREFIX "snapshots"
28 27
diff --git a/drivers/md/dm.c b/drivers/md/dm.c
index 8a994be035ba..424f7b048c30 100644
--- a/drivers/md/dm.c
+++ b/drivers/md/dm.c
@@ -6,7 +6,6 @@
6 */ 6 */
7 7
8#include "dm.h" 8#include "dm.h"
9#include "dm-bio-list.h"
10#include "dm-uevent.h" 9#include "dm-uevent.h"
11 10
12#include <linux/init.h> 11#include <linux/init.h>
diff --git a/drivers/md/raid1.c b/drivers/md/raid1.c
index 274b491a11c1..36df9109cde1 100644
--- a/drivers/md/raid1.c
+++ b/drivers/md/raid1.c
@@ -35,7 +35,6 @@
35#include <linux/blkdev.h> 35#include <linux/blkdev.h>
36#include <linux/seq_file.h> 36#include <linux/seq_file.h>
37#include "md.h" 37#include "md.h"
38#include "dm-bio-list.h"
39#include "raid1.h" 38#include "raid1.h"
40#include "bitmap.h" 39#include "bitmap.h"
41 40
diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index e293d92641ac..81a54f17417e 100644
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -22,7 +22,6 @@
22#include <linux/blkdev.h> 22#include <linux/blkdev.h>
23#include <linux/seq_file.h> 23#include <linux/seq_file.h>
24#include "md.h" 24#include "md.h"
25#include "dm-bio-list.h"
26#include "raid10.h" 25#include "raid10.h"
27#include "bitmap.h" 26#include "bitmap.h"
28 27
diff --git a/drivers/misc/eeprom/at24.c b/drivers/misc/eeprom/at24.c
index d184dfab9631..db39f4a52f53 100644
--- a/drivers/misc/eeprom/at24.c
+++ b/drivers/misc/eeprom/at24.c
@@ -278,7 +278,7 @@ static ssize_t at24_bin_read(struct kobject *kobj, struct bin_attribute *attr,
278 * We only use page mode writes; the alternative is sloooow. This routine 278 * We only use page mode writes; the alternative is sloooow. This routine
279 * writes at most one page. 279 * writes at most one page.
280 */ 280 */
281static ssize_t at24_eeprom_write(struct at24_data *at24, char *buf, 281static ssize_t at24_eeprom_write(struct at24_data *at24, const char *buf,
282 unsigned offset, size_t count) 282 unsigned offset, size_t count)
283{ 283{
284 struct i2c_client *client; 284 struct i2c_client *client;
@@ -347,8 +347,8 @@ static ssize_t at24_eeprom_write(struct at24_data *at24, char *buf,
347 return -ETIMEDOUT; 347 return -ETIMEDOUT;
348} 348}
349 349
350static ssize_t at24_write(struct at24_data *at24, 350static ssize_t at24_write(struct at24_data *at24, const char *buf, loff_t off,
351 char *buf, loff_t off, size_t count) 351 size_t count)
352{ 352{
353 ssize_t retval = 0; 353 ssize_t retval = 0;
354 354
@@ -406,7 +406,7 @@ static ssize_t at24_macc_read(struct memory_accessor *macc, char *buf,
406 return at24_read(at24, buf, offset, count); 406 return at24_read(at24, buf, offset, count);
407} 407}
408 408
409static ssize_t at24_macc_write(struct memory_accessor *macc, char *buf, 409static ssize_t at24_macc_write(struct memory_accessor *macc, const char *buf,
410 off_t offset, size_t count) 410 off_t offset, size_t count)
411{ 411{
412 struct at24_data *at24 = container_of(macc, struct at24_data, macc); 412 struct at24_data *at24 = container_of(macc, struct at24_data, macc);
diff --git a/drivers/misc/eeprom/at25.c b/drivers/misc/eeprom/at25.c
index 6bc0dac5c1e8..b34cb5f79eea 100644
--- a/drivers/misc/eeprom/at25.c
+++ b/drivers/misc/eeprom/at25.c
@@ -140,7 +140,8 @@ at25_bin_read(struct kobject *kobj, struct bin_attribute *bin_attr,
140 140
141 141
142static ssize_t 142static ssize_t
143at25_ee_write(struct at25_data *at25, char *buf, loff_t off, size_t count) 143at25_ee_write(struct at25_data *at25, const char *buf, loff_t off,
144 size_t count)
144{ 145{
145 ssize_t status = 0; 146 ssize_t status = 0;
146 unsigned written = 0; 147 unsigned written = 0;
@@ -276,7 +277,7 @@ static ssize_t at25_mem_read(struct memory_accessor *mem, char *buf,
276 return at25_ee_read(at25, buf, offset, count); 277 return at25_ee_read(at25, buf, offset, count);
277} 278}
278 279
279static ssize_t at25_mem_write(struct memory_accessor *mem, char *buf, 280static ssize_t at25_mem_write(struct memory_accessor *mem, const char *buf,
280 off_t offset, size_t count) 281 off_t offset, size_t count)
281{ 282{
282 struct at25_data *at25 = container_of(mem, struct at25_data, mem); 283 struct at25_data *at25 = container_of(mem, struct at25_data, mem);
diff --git a/drivers/misc/sgi-xp/xpc.h b/drivers/misc/sgi-xp/xpc.h
index 114444cfd496..b94d5f767703 100644
--- a/drivers/misc/sgi-xp/xpc.h
+++ b/drivers/misc/sgi-xp/xpc.h
@@ -90,18 +90,21 @@ struct xpc_rsvd_page {
90 short max_npartitions; /* value of XPC_MAX_PARTITIONS */ 90 short max_npartitions; /* value of XPC_MAX_PARTITIONS */
91 u8 version; 91 u8 version;
92 u8 pad1[3]; /* align to next u64 in 1st 64-byte cacheline */ 92 u8 pad1[3]; /* align to next u64 in 1st 64-byte cacheline */
93 unsigned long ts_jiffies; /* timestamp when rsvd pg was setup by XPC */
93 union { 94 union {
94 unsigned long vars_pa; /* phys address of struct xpc_vars */ 95 struct {
95 unsigned long activate_gru_mq_desc_gpa; /* phys addr of */ 96 unsigned long vars_pa; /* phys addr */
96 /* activate mq's */ 97 } sn2;
97 /* gru mq descriptor */ 98 struct {
99 unsigned long heartbeat_gpa; /* phys addr */
100 unsigned long activate_gru_mq_desc_gpa; /* phys addr */
101 } uv;
98 } sn; 102 } sn;
99 unsigned long ts_jiffies; /* timestamp when rsvd pg was setup by XPC */ 103 u64 pad2[9]; /* align to last u64 in 2nd 64-byte cacheline */
100 u64 pad2[10]; /* align to last u64 in 2nd 64-byte cacheline */
101 u64 SAL_nasids_size; /* SAL: size of each nasid mask in bytes */ 104 u64 SAL_nasids_size; /* SAL: size of each nasid mask in bytes */
102}; 105};
103 106
104#define XPC_RP_VERSION _XPC_VERSION(2, 0) /* version 2.0 of the reserved page */ 107#define XPC_RP_VERSION _XPC_VERSION(3, 0) /* version 3.0 of the reserved page */
105 108
106/* 109/*
107 * Define the structures by which XPC variables can be exported to other 110 * Define the structures by which XPC variables can be exported to other
@@ -182,6 +185,17 @@ struct xpc_vars_part_sn2 {
182 (XPC_RP_MACH_NASIDS(_rp) + \ 185 (XPC_RP_MACH_NASIDS(_rp) + \
183 xpc_nasid_mask_nlongs)) 186 xpc_nasid_mask_nlongs))
184 187
188
189/*
190 * The following structure describes the partition's heartbeat info which
191 * will be periodically read by other partitions to determine whether this
192 * XPC is still 'alive'.
193 */
194struct xpc_heartbeat_uv {
195 unsigned long value;
196 unsigned long offline; /* if 0, heartbeat should be changing */
197};
198
185/* 199/*
186 * Info pertinent to a GRU message queue using a watch list for irq generation. 200 * Info pertinent to a GRU message queue using a watch list for irq generation.
187 */ 201 */
@@ -198,7 +212,7 @@ struct xpc_gru_mq_uv {
198 212
199/* 213/*
200 * The activate_mq is used to send/receive GRU messages that affect XPC's 214 * The activate_mq is used to send/receive GRU messages that affect XPC's
201 * heartbeat, partition active state, and channel state. This is UV only. 215 * partition active state and channel state. This is uv only.
202 */ 216 */
203struct xpc_activate_mq_msghdr_uv { 217struct xpc_activate_mq_msghdr_uv {
204 unsigned int gru_msg_hdr; /* FOR GRU INTERNAL USE ONLY */ 218 unsigned int gru_msg_hdr; /* FOR GRU INTERNAL USE ONLY */
@@ -210,33 +224,27 @@ struct xpc_activate_mq_msghdr_uv {
210 224
211/* activate_mq defined message types */ 225/* activate_mq defined message types */
212#define XPC_ACTIVATE_MQ_MSG_SYNC_ACT_STATE_UV 0 226#define XPC_ACTIVATE_MQ_MSG_SYNC_ACT_STATE_UV 0
213#define XPC_ACTIVATE_MQ_MSG_INC_HEARTBEAT_UV 1
214#define XPC_ACTIVATE_MQ_MSG_OFFLINE_HEARTBEAT_UV 2
215#define XPC_ACTIVATE_MQ_MSG_ONLINE_HEARTBEAT_UV 3
216 227
217#define XPC_ACTIVATE_MQ_MSG_ACTIVATE_REQ_UV 4 228#define XPC_ACTIVATE_MQ_MSG_ACTIVATE_REQ_UV 1
218#define XPC_ACTIVATE_MQ_MSG_DEACTIVATE_REQ_UV 5 229#define XPC_ACTIVATE_MQ_MSG_DEACTIVATE_REQ_UV 2
219 230
220#define XPC_ACTIVATE_MQ_MSG_CHCTL_CLOSEREQUEST_UV 6 231#define XPC_ACTIVATE_MQ_MSG_CHCTL_CLOSEREQUEST_UV 3
221#define XPC_ACTIVATE_MQ_MSG_CHCTL_CLOSEREPLY_UV 7 232#define XPC_ACTIVATE_MQ_MSG_CHCTL_CLOSEREPLY_UV 4
222#define XPC_ACTIVATE_MQ_MSG_CHCTL_OPENREQUEST_UV 8 233#define XPC_ACTIVATE_MQ_MSG_CHCTL_OPENREQUEST_UV 5
223#define XPC_ACTIVATE_MQ_MSG_CHCTL_OPENREPLY_UV 9 234#define XPC_ACTIVATE_MQ_MSG_CHCTL_OPENREPLY_UV 6
235#define XPC_ACTIVATE_MQ_MSG_CHCTL_OPENCOMPLETE_UV 7
224 236
225#define XPC_ACTIVATE_MQ_MSG_MARK_ENGAGED_UV 10 237#define XPC_ACTIVATE_MQ_MSG_MARK_ENGAGED_UV 8
226#define XPC_ACTIVATE_MQ_MSG_MARK_DISENGAGED_UV 11 238#define XPC_ACTIVATE_MQ_MSG_MARK_DISENGAGED_UV 9
227 239
228struct xpc_activate_mq_msg_uv { 240struct xpc_activate_mq_msg_uv {
229 struct xpc_activate_mq_msghdr_uv hdr; 241 struct xpc_activate_mq_msghdr_uv hdr;
230}; 242};
231 243
232struct xpc_activate_mq_msg_heartbeat_req_uv {
233 struct xpc_activate_mq_msghdr_uv hdr;
234 u64 heartbeat;
235};
236
237struct xpc_activate_mq_msg_activate_req_uv { 244struct xpc_activate_mq_msg_activate_req_uv {
238 struct xpc_activate_mq_msghdr_uv hdr; 245 struct xpc_activate_mq_msghdr_uv hdr;
239 unsigned long rp_gpa; 246 unsigned long rp_gpa;
247 unsigned long heartbeat_gpa;
240 unsigned long activate_gru_mq_desc_gpa; 248 unsigned long activate_gru_mq_desc_gpa;
241}; 249};
242 250
@@ -271,6 +279,11 @@ struct xpc_activate_mq_msg_chctl_openreply_uv {
271 unsigned long notify_gru_mq_desc_gpa; 279 unsigned long notify_gru_mq_desc_gpa;
272}; 280};
273 281
282struct xpc_activate_mq_msg_chctl_opencomplete_uv {
283 struct xpc_activate_mq_msghdr_uv hdr;
284 short ch_number;
285};
286
274/* 287/*
275 * Functions registered by add_timer() or called by kernel_thread() only 288 * Functions registered by add_timer() or called by kernel_thread() only
276 * allow for a single 64-bit argument. The following macros can be used to 289 * allow for a single 64-bit argument. The following macros can be used to
@@ -576,30 +589,32 @@ struct xpc_channel {
576 589
577#define XPC_C_WASCONNECTED 0x00000001 /* channel was connected */ 590#define XPC_C_WASCONNECTED 0x00000001 /* channel was connected */
578 591
579#define XPC_C_ROPENREPLY 0x00000002 /* remote open channel reply */ 592#define XPC_C_ROPENCOMPLETE 0x00000002 /* remote open channel complete */
580#define XPC_C_OPENREPLY 0x00000004 /* local open channel reply */ 593#define XPC_C_OPENCOMPLETE 0x00000004 /* local open channel complete */
581#define XPC_C_ROPENREQUEST 0x00000008 /* remote open channel request */ 594#define XPC_C_ROPENREPLY 0x00000008 /* remote open channel reply */
582#define XPC_C_OPENREQUEST 0x00000010 /* local open channel request */ 595#define XPC_C_OPENREPLY 0x00000010 /* local open channel reply */
596#define XPC_C_ROPENREQUEST 0x00000020 /* remote open channel request */
597#define XPC_C_OPENREQUEST 0x00000040 /* local open channel request */
583 598
584#define XPC_C_SETUP 0x00000020 /* channel's msgqueues are alloc'd */ 599#define XPC_C_SETUP 0x00000080 /* channel's msgqueues are alloc'd */
585#define XPC_C_CONNECTEDCALLOUT 0x00000040 /* connected callout initiated */ 600#define XPC_C_CONNECTEDCALLOUT 0x00000100 /* connected callout initiated */
586#define XPC_C_CONNECTEDCALLOUT_MADE \ 601#define XPC_C_CONNECTEDCALLOUT_MADE \
587 0x00000080 /* connected callout completed */ 602 0x00000200 /* connected callout completed */
588#define XPC_C_CONNECTED 0x00000100 /* local channel is connected */ 603#define XPC_C_CONNECTED 0x00000400 /* local channel is connected */
589#define XPC_C_CONNECTING 0x00000200 /* channel is being connected */ 604#define XPC_C_CONNECTING 0x00000800 /* channel is being connected */
590 605
591#define XPC_C_RCLOSEREPLY 0x00000400 /* remote close channel reply */ 606#define XPC_C_RCLOSEREPLY 0x00001000 /* remote close channel reply */
592#define XPC_C_CLOSEREPLY 0x00000800 /* local close channel reply */ 607#define XPC_C_CLOSEREPLY 0x00002000 /* local close channel reply */
593#define XPC_C_RCLOSEREQUEST 0x00001000 /* remote close channel request */ 608#define XPC_C_RCLOSEREQUEST 0x00004000 /* remote close channel request */
594#define XPC_C_CLOSEREQUEST 0x00002000 /* local close channel request */ 609#define XPC_C_CLOSEREQUEST 0x00008000 /* local close channel request */
595 610
596#define XPC_C_DISCONNECTED 0x00004000 /* channel is disconnected */ 611#define XPC_C_DISCONNECTED 0x00010000 /* channel is disconnected */
597#define XPC_C_DISCONNECTING 0x00008000 /* channel is being disconnected */ 612#define XPC_C_DISCONNECTING 0x00020000 /* channel is being disconnected */
598#define XPC_C_DISCONNECTINGCALLOUT \ 613#define XPC_C_DISCONNECTINGCALLOUT \
599 0x00010000 /* disconnecting callout initiated */ 614 0x00040000 /* disconnecting callout initiated */
600#define XPC_C_DISCONNECTINGCALLOUT_MADE \ 615#define XPC_C_DISCONNECTINGCALLOUT_MADE \
601 0x00020000 /* disconnecting callout completed */ 616 0x00080000 /* disconnecting callout completed */
602#define XPC_C_WDISCONNECT 0x00040000 /* waiting for channel disconnect */ 617#define XPC_C_WDISCONNECT 0x00100000 /* waiting for channel disconnect */
603 618
604/* 619/*
605 * The channel control flags (chctl) union consists of a 64-bit variable which 620 * The channel control flags (chctl) union consists of a 64-bit variable which
@@ -618,11 +633,13 @@ union xpc_channel_ctl_flags {
618#define XPC_CHCTL_CLOSEREPLY 0x02 633#define XPC_CHCTL_CLOSEREPLY 0x02
619#define XPC_CHCTL_OPENREQUEST 0x04 634#define XPC_CHCTL_OPENREQUEST 0x04
620#define XPC_CHCTL_OPENREPLY 0x08 635#define XPC_CHCTL_OPENREPLY 0x08
621#define XPC_CHCTL_MSGREQUEST 0x10 636#define XPC_CHCTL_OPENCOMPLETE 0x10
637#define XPC_CHCTL_MSGREQUEST 0x20
622 638
623#define XPC_OPENCLOSE_CHCTL_FLAGS \ 639#define XPC_OPENCLOSE_CHCTL_FLAGS \
624 (XPC_CHCTL_CLOSEREQUEST | XPC_CHCTL_CLOSEREPLY | \ 640 (XPC_CHCTL_CLOSEREQUEST | XPC_CHCTL_CLOSEREPLY | \
625 XPC_CHCTL_OPENREQUEST | XPC_CHCTL_OPENREPLY) 641 XPC_CHCTL_OPENREQUEST | XPC_CHCTL_OPENREPLY | \
642 XPC_CHCTL_OPENCOMPLETE)
626#define XPC_MSG_CHCTL_FLAGS XPC_CHCTL_MSGREQUEST 643#define XPC_MSG_CHCTL_FLAGS XPC_CHCTL_MSGREQUEST
627 644
628static inline int 645static inline int
@@ -687,6 +704,9 @@ struct xpc_partition_sn2 {
687}; 704};
688 705
689struct xpc_partition_uv { 706struct xpc_partition_uv {
707 unsigned long heartbeat_gpa; /* phys addr of partition's heartbeat */
708 struct xpc_heartbeat_uv cached_heartbeat; /* cached copy of */
709 /* partition's heartbeat */
690 unsigned long activate_gru_mq_desc_gpa; /* phys addr of parititon's */ 710 unsigned long activate_gru_mq_desc_gpa; /* phys addr of parititon's */
691 /* activate mq's gru mq */ 711 /* activate mq's gru mq */
692 /* descriptor */ 712 /* descriptor */
@@ -698,14 +718,12 @@ struct xpc_partition_uv {
698 u8 remote_act_state; /* remote partition's act_state */ 718 u8 remote_act_state; /* remote partition's act_state */
699 u8 act_state_req; /* act_state request from remote partition */ 719 u8 act_state_req; /* act_state request from remote partition */
700 enum xp_retval reason; /* reason for deactivate act_state request */ 720 enum xp_retval reason; /* reason for deactivate act_state request */
701 u64 heartbeat; /* incremented by remote partition */
702}; 721};
703 722
704/* struct xpc_partition_uv flags */ 723/* struct xpc_partition_uv flags */
705 724
706#define XPC_P_HEARTBEAT_OFFLINE_UV 0x00000001 725#define XPC_P_CACHED_ACTIVATE_GRU_MQ_DESC_UV 0x00000001
707#define XPC_P_ENGAGED_UV 0x00000002 726#define XPC_P_ENGAGED_UV 0x00000002
708#define XPC_P_CACHED_ACTIVATE_GRU_MQ_DESC_UV 0x00000004
709 727
710/* struct xpc_partition_uv act_state change requests */ 728/* struct xpc_partition_uv act_state change requests */
711 729
@@ -762,6 +780,62 @@ struct xpc_partition {
762 780
763} ____cacheline_aligned; 781} ____cacheline_aligned;
764 782
783struct xpc_arch_operations {
784 int (*setup_partitions) (void);
785 void (*teardown_partitions) (void);
786 void (*process_activate_IRQ_rcvd) (void);
787 enum xp_retval (*get_partition_rsvd_page_pa)
788 (void *, u64 *, unsigned long *, size_t *);
789 int (*setup_rsvd_page) (struct xpc_rsvd_page *);
790
791 void (*allow_hb) (short);
792 void (*disallow_hb) (short);
793 void (*disallow_all_hbs) (void);
794 void (*increment_heartbeat) (void);
795 void (*offline_heartbeat) (void);
796 void (*online_heartbeat) (void);
797 void (*heartbeat_init) (void);
798 void (*heartbeat_exit) (void);
799 enum xp_retval (*get_remote_heartbeat) (struct xpc_partition *);
800
801 void (*request_partition_activation) (struct xpc_rsvd_page *,
802 unsigned long, int);
803 void (*request_partition_reactivation) (struct xpc_partition *);
804 void (*request_partition_deactivation) (struct xpc_partition *);
805 void (*cancel_partition_deactivation_request) (struct xpc_partition *);
806 enum xp_retval (*setup_ch_structures) (struct xpc_partition *);
807 void (*teardown_ch_structures) (struct xpc_partition *);
808
809 enum xp_retval (*make_first_contact) (struct xpc_partition *);
810
811 u64 (*get_chctl_all_flags) (struct xpc_partition *);
812 void (*send_chctl_closerequest) (struct xpc_channel *, unsigned long *);
813 void (*send_chctl_closereply) (struct xpc_channel *, unsigned long *);
814 void (*send_chctl_openrequest) (struct xpc_channel *, unsigned long *);
815 void (*send_chctl_openreply) (struct xpc_channel *, unsigned long *);
816 void (*send_chctl_opencomplete) (struct xpc_channel *, unsigned long *);
817 void (*process_msg_chctl_flags) (struct xpc_partition *, int);
818
819 enum xp_retval (*save_remote_msgqueue_pa) (struct xpc_channel *,
820 unsigned long);
821
822 enum xp_retval (*setup_msg_structures) (struct xpc_channel *);
823 void (*teardown_msg_structures) (struct xpc_channel *);
824
825 void (*indicate_partition_engaged) (struct xpc_partition *);
826 void (*indicate_partition_disengaged) (struct xpc_partition *);
827 void (*assume_partition_disengaged) (short);
828 int (*partition_engaged) (short);
829 int (*any_partition_engaged) (void);
830
831 int (*n_of_deliverable_payloads) (struct xpc_channel *);
832 enum xp_retval (*send_payload) (struct xpc_channel *, u32, void *,
833 u16, u8, xpc_notify_func, void *);
834 void *(*get_deliverable_payload) (struct xpc_channel *);
835 void (*received_payload) (struct xpc_channel *, void *);
836 void (*notify_senders_of_disconnect) (struct xpc_channel *);
837};
838
765/* struct xpc_partition act_state values (for XPC HB) */ 839/* struct xpc_partition act_state values (for XPC HB) */
766 840
767#define XPC_P_AS_INACTIVE 0x00 /* partition is not active */ 841#define XPC_P_AS_INACTIVE 0x00 /* partition is not active */
@@ -802,67 +876,17 @@ extern struct xpc_registration xpc_registrations[];
802/* found in xpc_main.c */ 876/* found in xpc_main.c */
803extern struct device *xpc_part; 877extern struct device *xpc_part;
804extern struct device *xpc_chan; 878extern struct device *xpc_chan;
879extern struct xpc_arch_operations xpc_arch_ops;
805extern int xpc_disengage_timelimit; 880extern int xpc_disengage_timelimit;
806extern int xpc_disengage_timedout; 881extern int xpc_disengage_timedout;
807extern int xpc_activate_IRQ_rcvd; 882extern int xpc_activate_IRQ_rcvd;
808extern spinlock_t xpc_activate_IRQ_rcvd_lock; 883extern spinlock_t xpc_activate_IRQ_rcvd_lock;
809extern wait_queue_head_t xpc_activate_IRQ_wq; 884extern wait_queue_head_t xpc_activate_IRQ_wq;
810extern void *xpc_heartbeating_to_mask;
811extern void *xpc_kzalloc_cacheline_aligned(size_t, gfp_t, void **); 885extern void *xpc_kzalloc_cacheline_aligned(size_t, gfp_t, void **);
812extern void xpc_activate_partition(struct xpc_partition *); 886extern void xpc_activate_partition(struct xpc_partition *);
813extern void xpc_activate_kthreads(struct xpc_channel *, int); 887extern void xpc_activate_kthreads(struct xpc_channel *, int);
814extern void xpc_create_kthreads(struct xpc_channel *, int, int); 888extern void xpc_create_kthreads(struct xpc_channel *, int, int);
815extern void xpc_disconnect_wait(int); 889extern void xpc_disconnect_wait(int);
816extern int (*xpc_setup_partitions_sn) (void);
817extern void (*xpc_teardown_partitions_sn) (void);
818extern enum xp_retval (*xpc_get_partition_rsvd_page_pa) (void *, u64 *,
819 unsigned long *,
820 size_t *);
821extern int (*xpc_setup_rsvd_page_sn) (struct xpc_rsvd_page *);
822extern void (*xpc_heartbeat_init) (void);
823extern void (*xpc_heartbeat_exit) (void);
824extern void (*xpc_increment_heartbeat) (void);
825extern void (*xpc_offline_heartbeat) (void);
826extern void (*xpc_online_heartbeat) (void);
827extern enum xp_retval (*xpc_get_remote_heartbeat) (struct xpc_partition *);
828extern enum xp_retval (*xpc_make_first_contact) (struct xpc_partition *);
829extern u64 (*xpc_get_chctl_all_flags) (struct xpc_partition *);
830extern enum xp_retval (*xpc_setup_msg_structures) (struct xpc_channel *);
831extern void (*xpc_teardown_msg_structures) (struct xpc_channel *);
832extern void (*xpc_notify_senders_of_disconnect) (struct xpc_channel *);
833extern void (*xpc_process_msg_chctl_flags) (struct xpc_partition *, int);
834extern int (*xpc_n_of_deliverable_payloads) (struct xpc_channel *);
835extern void *(*xpc_get_deliverable_payload) (struct xpc_channel *);
836extern void (*xpc_request_partition_activation) (struct xpc_rsvd_page *,
837 unsigned long, int);
838extern void (*xpc_request_partition_reactivation) (struct xpc_partition *);
839extern void (*xpc_request_partition_deactivation) (struct xpc_partition *);
840extern void (*xpc_cancel_partition_deactivation_request) (
841 struct xpc_partition *);
842extern void (*xpc_process_activate_IRQ_rcvd) (void);
843extern enum xp_retval (*xpc_setup_ch_structures_sn) (struct xpc_partition *);
844extern void (*xpc_teardown_ch_structures_sn) (struct xpc_partition *);
845
846extern void (*xpc_indicate_partition_engaged) (struct xpc_partition *);
847extern int (*xpc_partition_engaged) (short);
848extern int (*xpc_any_partition_engaged) (void);
849extern void (*xpc_indicate_partition_disengaged) (struct xpc_partition *);
850extern void (*xpc_assume_partition_disengaged) (short);
851
852extern void (*xpc_send_chctl_closerequest) (struct xpc_channel *,
853 unsigned long *);
854extern void (*xpc_send_chctl_closereply) (struct xpc_channel *,
855 unsigned long *);
856extern void (*xpc_send_chctl_openrequest) (struct xpc_channel *,
857 unsigned long *);
858extern void (*xpc_send_chctl_openreply) (struct xpc_channel *, unsigned long *);
859
860extern enum xp_retval (*xpc_save_remote_msgqueue_pa) (struct xpc_channel *,
861 unsigned long);
862
863extern enum xp_retval (*xpc_send_payload) (struct xpc_channel *, u32, void *,
864 u16, u8, xpc_notify_func, void *);
865extern void (*xpc_received_payload) (struct xpc_channel *, void *);
866 890
867/* found in xpc_sn2.c */ 891/* found in xpc_sn2.c */
868extern int xpc_init_sn2(void); 892extern int xpc_init_sn2(void);
@@ -909,40 +933,6 @@ extern void xpc_disconnect_channel(const int, struct xpc_channel *,
909extern void xpc_disconnect_callout(struct xpc_channel *, enum xp_retval); 933extern void xpc_disconnect_callout(struct xpc_channel *, enum xp_retval);
910extern void xpc_partition_going_down(struct xpc_partition *, enum xp_retval); 934extern void xpc_partition_going_down(struct xpc_partition *, enum xp_retval);
911 935
912static inline int
913xpc_hb_allowed(short partid, void *heartbeating_to_mask)
914{
915 return test_bit(partid, heartbeating_to_mask);
916}
917
918static inline int
919xpc_any_hbs_allowed(void)
920{
921 DBUG_ON(xpc_heartbeating_to_mask == NULL);
922 return !bitmap_empty(xpc_heartbeating_to_mask, xp_max_npartitions);
923}
924
925static inline void
926xpc_allow_hb(short partid)
927{
928 DBUG_ON(xpc_heartbeating_to_mask == NULL);
929 set_bit(partid, xpc_heartbeating_to_mask);
930}
931
932static inline void
933xpc_disallow_hb(short partid)
934{
935 DBUG_ON(xpc_heartbeating_to_mask == NULL);
936 clear_bit(partid, xpc_heartbeating_to_mask);
937}
938
939static inline void
940xpc_disallow_all_hbs(void)
941{
942 DBUG_ON(xpc_heartbeating_to_mask == NULL);
943 bitmap_zero(xpc_heartbeating_to_mask, xp_max_npartitions);
944}
945
946static inline void 936static inline void
947xpc_wakeup_channel_mgr(struct xpc_partition *part) 937xpc_wakeup_channel_mgr(struct xpc_partition *part)
948{ 938{
diff --git a/drivers/misc/sgi-xp/xpc_channel.c b/drivers/misc/sgi-xp/xpc_channel.c
index 99a2534c38a1..652593fc486d 100644
--- a/drivers/misc/sgi-xp/xpc_channel.c
+++ b/drivers/misc/sgi-xp/xpc_channel.c
@@ -3,7 +3,7 @@
3 * License. See the file "COPYING" in the main directory of this archive 3 * License. See the file "COPYING" in the main directory of this archive
4 * for more details. 4 * for more details.
5 * 5 *
6 * Copyright (c) 2004-2008 Silicon Graphics, Inc. All Rights Reserved. 6 * Copyright (c) 2004-2009 Silicon Graphics, Inc. All Rights Reserved.
7 */ 7 */
8 8
9/* 9/*
@@ -39,34 +39,38 @@ xpc_process_connect(struct xpc_channel *ch, unsigned long *irq_flags)
39 39
40 if (!(ch->flags & XPC_C_SETUP)) { 40 if (!(ch->flags & XPC_C_SETUP)) {
41 spin_unlock_irqrestore(&ch->lock, *irq_flags); 41 spin_unlock_irqrestore(&ch->lock, *irq_flags);
42 ret = xpc_setup_msg_structures(ch); 42 ret = xpc_arch_ops.setup_msg_structures(ch);
43 spin_lock_irqsave(&ch->lock, *irq_flags); 43 spin_lock_irqsave(&ch->lock, *irq_flags);
44 44
45 if (ret != xpSuccess) 45 if (ret != xpSuccess)
46 XPC_DISCONNECT_CHANNEL(ch, ret, irq_flags); 46 XPC_DISCONNECT_CHANNEL(ch, ret, irq_flags);
47 else
48 ch->flags |= XPC_C_SETUP;
47 49
48 ch->flags |= XPC_C_SETUP; 50 if (ch->flags & XPC_C_DISCONNECTING)
49
50 if (ch->flags & (XPC_C_CONNECTED | XPC_C_DISCONNECTING))
51 return; 51 return;
52 } 52 }
53 53
54 if (!(ch->flags & XPC_C_OPENREPLY)) { 54 if (!(ch->flags & XPC_C_OPENREPLY)) {
55 ch->flags |= XPC_C_OPENREPLY; 55 ch->flags |= XPC_C_OPENREPLY;
56 xpc_send_chctl_openreply(ch, irq_flags); 56 xpc_arch_ops.send_chctl_openreply(ch, irq_flags);
57 } 57 }
58 58
59 if (!(ch->flags & XPC_C_ROPENREPLY)) 59 if (!(ch->flags & XPC_C_ROPENREPLY))
60 return; 60 return;
61 61
62 ch->flags = (XPC_C_CONNECTED | XPC_C_SETUP); /* clear all else */ 62 if (!(ch->flags & XPC_C_OPENCOMPLETE)) {
63 ch->flags |= (XPC_C_OPENCOMPLETE | XPC_C_CONNECTED);
64 xpc_arch_ops.send_chctl_opencomplete(ch, irq_flags);
65 }
66
67 if (!(ch->flags & XPC_C_ROPENCOMPLETE))
68 return;
63 69
64 dev_info(xpc_chan, "channel %d to partition %d connected\n", 70 dev_info(xpc_chan, "channel %d to partition %d connected\n",
65 ch->number, ch->partid); 71 ch->number, ch->partid);
66 72
67 spin_unlock_irqrestore(&ch->lock, *irq_flags); 73 ch->flags = (XPC_C_CONNECTED | XPC_C_SETUP); /* clear all else */
68 xpc_create_kthreads(ch, 1, 0);
69 spin_lock_irqsave(&ch->lock, *irq_flags);
70} 74}
71 75
72/* 76/*
@@ -96,7 +100,7 @@ xpc_process_disconnect(struct xpc_channel *ch, unsigned long *irq_flags)
96 100
97 if (part->act_state == XPC_P_AS_DEACTIVATING) { 101 if (part->act_state == XPC_P_AS_DEACTIVATING) {
98 /* can't proceed until the other side disengages from us */ 102 /* can't proceed until the other side disengages from us */
99 if (xpc_partition_engaged(ch->partid)) 103 if (xpc_arch_ops.partition_engaged(ch->partid))
100 return; 104 return;
101 105
102 } else { 106 } else {
@@ -108,7 +112,7 @@ xpc_process_disconnect(struct xpc_channel *ch, unsigned long *irq_flags)
108 112
109 if (!(ch->flags & XPC_C_CLOSEREPLY)) { 113 if (!(ch->flags & XPC_C_CLOSEREPLY)) {
110 ch->flags |= XPC_C_CLOSEREPLY; 114 ch->flags |= XPC_C_CLOSEREPLY;
111 xpc_send_chctl_closereply(ch, irq_flags); 115 xpc_arch_ops.send_chctl_closereply(ch, irq_flags);
112 } 116 }
113 117
114 if (!(ch->flags & XPC_C_RCLOSEREPLY)) 118 if (!(ch->flags & XPC_C_RCLOSEREPLY))
@@ -118,7 +122,7 @@ xpc_process_disconnect(struct xpc_channel *ch, unsigned long *irq_flags)
118 /* wake those waiting for notify completion */ 122 /* wake those waiting for notify completion */
119 if (atomic_read(&ch->n_to_notify) > 0) { 123 if (atomic_read(&ch->n_to_notify) > 0) {
120 /* we do callout while holding ch->lock, callout can't block */ 124 /* we do callout while holding ch->lock, callout can't block */
121 xpc_notify_senders_of_disconnect(ch); 125 xpc_arch_ops.notify_senders_of_disconnect(ch);
122 } 126 }
123 127
124 /* both sides are disconnected now */ 128 /* both sides are disconnected now */
@@ -132,7 +136,7 @@ xpc_process_disconnect(struct xpc_channel *ch, unsigned long *irq_flags)
132 DBUG_ON(atomic_read(&ch->n_to_notify) != 0); 136 DBUG_ON(atomic_read(&ch->n_to_notify) != 0);
133 137
134 /* it's now safe to free the channel's message queues */ 138 /* it's now safe to free the channel's message queues */
135 xpc_teardown_msg_structures(ch); 139 xpc_arch_ops.teardown_msg_structures(ch);
136 140
137 ch->func = NULL; 141 ch->func = NULL;
138 ch->key = NULL; 142 ch->key = NULL;
@@ -144,8 +148,9 @@ xpc_process_disconnect(struct xpc_channel *ch, unsigned long *irq_flags)
144 148
145 /* 149 /*
146 * Mark the channel disconnected and clear all other flags, including 150 * Mark the channel disconnected and clear all other flags, including
147 * XPC_C_SETUP (because of call to xpc_teardown_msg_structures()) but 151 * XPC_C_SETUP (because of call to
148 * not including XPC_C_WDISCONNECT (if it was set). 152 * xpc_arch_ops.teardown_msg_structures()) but not including
153 * XPC_C_WDISCONNECT (if it was set).
149 */ 154 */
150 ch->flags = (XPC_C_DISCONNECTED | (ch->flags & XPC_C_WDISCONNECT)); 155 ch->flags = (XPC_C_DISCONNECTED | (ch->flags & XPC_C_WDISCONNECT));
151 156
@@ -184,6 +189,7 @@ xpc_process_openclose_chctl_flags(struct xpc_partition *part, int ch_number,
184 struct xpc_channel *ch = &part->channels[ch_number]; 189 struct xpc_channel *ch = &part->channels[ch_number];
185 enum xp_retval reason; 190 enum xp_retval reason;
186 enum xp_retval ret; 191 enum xp_retval ret;
192 int create_kthread = 0;
187 193
188 spin_lock_irqsave(&ch->lock, irq_flags); 194 spin_lock_irqsave(&ch->lock, irq_flags);
189 195
@@ -196,8 +202,7 @@ again:
196 * has had a chance to see that the channel is disconnected. 202 * has had a chance to see that the channel is disconnected.
197 */ 203 */
198 ch->delayed_chctl_flags |= chctl_flags; 204 ch->delayed_chctl_flags |= chctl_flags;
199 spin_unlock_irqrestore(&ch->lock, irq_flags); 205 goto out;
200 return;
201 } 206 }
202 207
203 if (chctl_flags & XPC_CHCTL_CLOSEREQUEST) { 208 if (chctl_flags & XPC_CHCTL_CLOSEREQUEST) {
@@ -239,8 +244,7 @@ again:
239 XPC_CHCTL_CLOSEREQUEST; 244 XPC_CHCTL_CLOSEREQUEST;
240 spin_unlock(&part->chctl_lock); 245 spin_unlock(&part->chctl_lock);
241 } 246 }
242 spin_unlock_irqrestore(&ch->lock, irq_flags); 247 goto out;
243 return;
244 } 248 }
245 249
246 XPC_SET_REASON(ch, 0, 0); 250 XPC_SET_REASON(ch, 0, 0);
@@ -250,7 +254,8 @@ again:
250 ch->flags |= (XPC_C_CONNECTING | XPC_C_ROPENREQUEST); 254 ch->flags |= (XPC_C_CONNECTING | XPC_C_ROPENREQUEST);
251 } 255 }
252 256
253 chctl_flags &= ~(XPC_CHCTL_OPENREQUEST | XPC_CHCTL_OPENREPLY); 257 chctl_flags &= ~(XPC_CHCTL_OPENREQUEST | XPC_CHCTL_OPENREPLY |
258 XPC_CHCTL_OPENCOMPLETE);
254 259
255 /* 260 /*
256 * The meaningful CLOSEREQUEST connection state fields are: 261 * The meaningful CLOSEREQUEST connection state fields are:
@@ -269,8 +274,7 @@ again:
269 XPC_DISCONNECT_CHANNEL(ch, reason, &irq_flags); 274 XPC_DISCONNECT_CHANNEL(ch, reason, &irq_flags);
270 275
271 DBUG_ON(chctl_flags & XPC_CHCTL_CLOSEREPLY); 276 DBUG_ON(chctl_flags & XPC_CHCTL_CLOSEREPLY);
272 spin_unlock_irqrestore(&ch->lock, irq_flags); 277 goto out;
273 return;
274 } 278 }
275 279
276 xpc_process_disconnect(ch, &irq_flags); 280 xpc_process_disconnect(ch, &irq_flags);
@@ -283,8 +287,7 @@ again:
283 287
284 if (ch->flags & XPC_C_DISCONNECTED) { 288 if (ch->flags & XPC_C_DISCONNECTED) {
285 DBUG_ON(part->act_state != XPC_P_AS_DEACTIVATING); 289 DBUG_ON(part->act_state != XPC_P_AS_DEACTIVATING);
286 spin_unlock_irqrestore(&ch->lock, irq_flags); 290 goto out;
287 return;
288 } 291 }
289 292
290 DBUG_ON(!(ch->flags & XPC_C_CLOSEREQUEST)); 293 DBUG_ON(!(ch->flags & XPC_C_CLOSEREQUEST));
@@ -299,8 +302,7 @@ again:
299 XPC_CHCTL_CLOSEREPLY; 302 XPC_CHCTL_CLOSEREPLY;
300 spin_unlock(&part->chctl_lock); 303 spin_unlock(&part->chctl_lock);
301 } 304 }
302 spin_unlock_irqrestore(&ch->lock, irq_flags); 305 goto out;
303 return;
304 } 306 }
305 307
306 ch->flags |= XPC_C_RCLOSEREPLY; 308 ch->flags |= XPC_C_RCLOSEREPLY;
@@ -320,14 +322,12 @@ again:
320 322
321 if (part->act_state == XPC_P_AS_DEACTIVATING || 323 if (part->act_state == XPC_P_AS_DEACTIVATING ||
322 (ch->flags & XPC_C_ROPENREQUEST)) { 324 (ch->flags & XPC_C_ROPENREQUEST)) {
323 spin_unlock_irqrestore(&ch->lock, irq_flags); 325 goto out;
324 return;
325 } 326 }
326 327
327 if (ch->flags & (XPC_C_DISCONNECTING | XPC_C_WDISCONNECT)) { 328 if (ch->flags & (XPC_C_DISCONNECTING | XPC_C_WDISCONNECT)) {
328 ch->delayed_chctl_flags |= XPC_CHCTL_OPENREQUEST; 329 ch->delayed_chctl_flags |= XPC_CHCTL_OPENREQUEST;
329 spin_unlock_irqrestore(&ch->lock, irq_flags); 330 goto out;
330 return;
331 } 331 }
332 DBUG_ON(!(ch->flags & (XPC_C_DISCONNECTED | 332 DBUG_ON(!(ch->flags & (XPC_C_DISCONNECTED |
333 XPC_C_OPENREQUEST))); 333 XPC_C_OPENREQUEST)));
@@ -341,8 +341,7 @@ again:
341 */ 341 */
342 if (args->entry_size == 0 || args->local_nentries == 0) { 342 if (args->entry_size == 0 || args->local_nentries == 0) {
343 /* assume OPENREQUEST was delayed by mistake */ 343 /* assume OPENREQUEST was delayed by mistake */
344 spin_unlock_irqrestore(&ch->lock, irq_flags); 344 goto out;
345 return;
346 } 345 }
347 346
348 ch->flags |= (XPC_C_ROPENREQUEST | XPC_C_CONNECTING); 347 ch->flags |= (XPC_C_ROPENREQUEST | XPC_C_CONNECTING);
@@ -352,8 +351,7 @@ again:
352 if (args->entry_size != ch->entry_size) { 351 if (args->entry_size != ch->entry_size) {
353 XPC_DISCONNECT_CHANNEL(ch, xpUnequalMsgSizes, 352 XPC_DISCONNECT_CHANNEL(ch, xpUnequalMsgSizes,
354 &irq_flags); 353 &irq_flags);
355 spin_unlock_irqrestore(&ch->lock, irq_flags); 354 goto out;
356 return;
357 } 355 }
358 } else { 356 } else {
359 ch->entry_size = args->entry_size; 357 ch->entry_size = args->entry_size;
@@ -375,15 +373,13 @@ again:
375 args->local_msgqueue_pa, args->local_nentries, 373 args->local_msgqueue_pa, args->local_nentries,
376 args->remote_nentries, ch->partid, ch->number); 374 args->remote_nentries, ch->partid, ch->number);
377 375
378 if (ch->flags & (XPC_C_DISCONNECTING | XPC_C_DISCONNECTED)) { 376 if (ch->flags & (XPC_C_DISCONNECTING | XPC_C_DISCONNECTED))
379 spin_unlock_irqrestore(&ch->lock, irq_flags); 377 goto out;
380 return; 378
381 }
382 if (!(ch->flags & XPC_C_OPENREQUEST)) { 379 if (!(ch->flags & XPC_C_OPENREQUEST)) {
383 XPC_DISCONNECT_CHANNEL(ch, xpOpenCloseError, 380 XPC_DISCONNECT_CHANNEL(ch, xpOpenCloseError,
384 &irq_flags); 381 &irq_flags);
385 spin_unlock_irqrestore(&ch->lock, irq_flags); 382 goto out;
386 return;
387 } 383 }
388 384
389 DBUG_ON(!(ch->flags & XPC_C_ROPENREQUEST)); 385 DBUG_ON(!(ch->flags & XPC_C_ROPENREQUEST));
@@ -400,11 +396,11 @@ again:
400 DBUG_ON(args->local_nentries == 0); 396 DBUG_ON(args->local_nentries == 0);
401 DBUG_ON(args->remote_nentries == 0); 397 DBUG_ON(args->remote_nentries == 0);
402 398
403 ret = xpc_save_remote_msgqueue_pa(ch, args->local_msgqueue_pa); 399 ret = xpc_arch_ops.save_remote_msgqueue_pa(ch,
400 args->local_msgqueue_pa);
404 if (ret != xpSuccess) { 401 if (ret != xpSuccess) {
405 XPC_DISCONNECT_CHANNEL(ch, ret, &irq_flags); 402 XPC_DISCONNECT_CHANNEL(ch, ret, &irq_flags);
406 spin_unlock_irqrestore(&ch->lock, irq_flags); 403 goto out;
407 return;
408 } 404 }
409 ch->flags |= XPC_C_ROPENREPLY; 405 ch->flags |= XPC_C_ROPENREPLY;
410 406
@@ -430,7 +426,36 @@ again:
430 xpc_process_connect(ch, &irq_flags); 426 xpc_process_connect(ch, &irq_flags);
431 } 427 }
432 428
429 if (chctl_flags & XPC_CHCTL_OPENCOMPLETE) {
430
431 dev_dbg(xpc_chan, "XPC_CHCTL_OPENCOMPLETE received from "
432 "partid=%d, channel=%d\n", ch->partid, ch->number);
433
434 if (ch->flags & (XPC_C_DISCONNECTING | XPC_C_DISCONNECTED))
435 goto out;
436
437 if (!(ch->flags & XPC_C_OPENREQUEST) ||
438 !(ch->flags & XPC_C_OPENREPLY)) {
439 XPC_DISCONNECT_CHANNEL(ch, xpOpenCloseError,
440 &irq_flags);
441 goto out;
442 }
443
444 DBUG_ON(!(ch->flags & XPC_C_ROPENREQUEST));
445 DBUG_ON(!(ch->flags & XPC_C_ROPENREPLY));
446 DBUG_ON(!(ch->flags & XPC_C_CONNECTED));
447
448 ch->flags |= XPC_C_ROPENCOMPLETE;
449
450 xpc_process_connect(ch, &irq_flags);
451 create_kthread = 1;
452 }
453
454out:
433 spin_unlock_irqrestore(&ch->lock, irq_flags); 455 spin_unlock_irqrestore(&ch->lock, irq_flags);
456
457 if (create_kthread)
458 xpc_create_kthreads(ch, 1, 0);
434} 459}
435 460
436/* 461/*
@@ -508,7 +533,7 @@ xpc_connect_channel(struct xpc_channel *ch)
508 /* initiate the connection */ 533 /* initiate the connection */
509 534
510 ch->flags |= (XPC_C_OPENREQUEST | XPC_C_CONNECTING); 535 ch->flags |= (XPC_C_OPENREQUEST | XPC_C_CONNECTING);
511 xpc_send_chctl_openrequest(ch, &irq_flags); 536 xpc_arch_ops.send_chctl_openrequest(ch, &irq_flags);
512 537
513 xpc_process_connect(ch, &irq_flags); 538 xpc_process_connect(ch, &irq_flags);
514 539
@@ -526,7 +551,7 @@ xpc_process_sent_chctl_flags(struct xpc_partition *part)
526 int ch_number; 551 int ch_number;
527 u32 ch_flags; 552 u32 ch_flags;
528 553
529 chctl.all_flags = xpc_get_chctl_all_flags(part); 554 chctl.all_flags = xpc_arch_ops.get_chctl_all_flags(part);
530 555
531 /* 556 /*
532 * Initiate channel connections for registered channels. 557 * Initiate channel connections for registered channels.
@@ -564,10 +589,6 @@ xpc_process_sent_chctl_flags(struct xpc_partition *part)
564 if (!(ch_flags & XPC_C_OPENREQUEST)) { 589 if (!(ch_flags & XPC_C_OPENREQUEST)) {
565 DBUG_ON(ch_flags & XPC_C_SETUP); 590 DBUG_ON(ch_flags & XPC_C_SETUP);
566 (void)xpc_connect_channel(ch); 591 (void)xpc_connect_channel(ch);
567 } else {
568 spin_lock_irqsave(&ch->lock, irq_flags);
569 xpc_process_connect(ch, &irq_flags);
570 spin_unlock_irqrestore(&ch->lock, irq_flags);
571 } 592 }
572 continue; 593 continue;
573 } 594 }
@@ -579,7 +600,7 @@ xpc_process_sent_chctl_flags(struct xpc_partition *part)
579 */ 600 */
580 601
581 if (chctl.flags[ch_number] & XPC_MSG_CHCTL_FLAGS) 602 if (chctl.flags[ch_number] & XPC_MSG_CHCTL_FLAGS)
582 xpc_process_msg_chctl_flags(part, ch_number); 603 xpc_arch_ops.process_msg_chctl_flags(part, ch_number);
583 } 604 }
584} 605}
585 606
@@ -755,7 +776,7 @@ xpc_disconnect_channel(const int line, struct xpc_channel *ch,
755 XPC_C_ROPENREQUEST | XPC_C_ROPENREPLY | 776 XPC_C_ROPENREQUEST | XPC_C_ROPENREPLY |
756 XPC_C_CONNECTING | XPC_C_CONNECTED); 777 XPC_C_CONNECTING | XPC_C_CONNECTED);
757 778
758 xpc_send_chctl_closerequest(ch, irq_flags); 779 xpc_arch_ops.send_chctl_closerequest(ch, irq_flags);
759 780
760 if (channel_was_connected) 781 if (channel_was_connected)
761 ch->flags |= XPC_C_WASCONNECTED; 782 ch->flags |= XPC_C_WASCONNECTED;
@@ -862,8 +883,8 @@ xpc_initiate_send(short partid, int ch_number, u32 flags, void *payload,
862 DBUG_ON(payload == NULL); 883 DBUG_ON(payload == NULL);
863 884
864 if (xpc_part_ref(part)) { 885 if (xpc_part_ref(part)) {
865 ret = xpc_send_payload(&part->channels[ch_number], flags, 886 ret = xpc_arch_ops.send_payload(&part->channels[ch_number],
866 payload, payload_size, 0, NULL, NULL); 887 flags, payload, payload_size, 0, NULL, NULL);
867 xpc_part_deref(part); 888 xpc_part_deref(part);
868 } 889 }
869 890
@@ -914,9 +935,8 @@ xpc_initiate_send_notify(short partid, int ch_number, u32 flags, void *payload,
914 DBUG_ON(func == NULL); 935 DBUG_ON(func == NULL);
915 936
916 if (xpc_part_ref(part)) { 937 if (xpc_part_ref(part)) {
917 ret = xpc_send_payload(&part->channels[ch_number], flags, 938 ret = xpc_arch_ops.send_payload(&part->channels[ch_number],
918 payload, payload_size, XPC_N_CALL, func, 939 flags, payload, payload_size, XPC_N_CALL, func, key);
919 key);
920 xpc_part_deref(part); 940 xpc_part_deref(part);
921 } 941 }
922 return ret; 942 return ret;
@@ -930,7 +950,7 @@ xpc_deliver_payload(struct xpc_channel *ch)
930{ 950{
931 void *payload; 951 void *payload;
932 952
933 payload = xpc_get_deliverable_payload(ch); 953 payload = xpc_arch_ops.get_deliverable_payload(ch);
934 if (payload != NULL) { 954 if (payload != NULL) {
935 955
936 /* 956 /*
@@ -984,7 +1004,7 @@ xpc_initiate_received(short partid, int ch_number, void *payload)
984 DBUG_ON(ch_number < 0 || ch_number >= part->nchannels); 1004 DBUG_ON(ch_number < 0 || ch_number >= part->nchannels);
985 1005
986 ch = &part->channels[ch_number]; 1006 ch = &part->channels[ch_number];
987 xpc_received_payload(ch, payload); 1007 xpc_arch_ops.received_payload(ch, payload);
988 1008
989 /* the call to xpc_msgqueue_ref() was done by xpc_deliver_payload() */ 1009 /* the call to xpc_msgqueue_ref() was done by xpc_deliver_payload() */
990 xpc_msgqueue_deref(ch); 1010 xpc_msgqueue_deref(ch);
diff --git a/drivers/misc/sgi-xp/xpc_main.c b/drivers/misc/sgi-xp/xpc_main.c
index 1ab9fda87fab..fd3688a3e23f 100644
--- a/drivers/misc/sgi-xp/xpc_main.c
+++ b/drivers/misc/sgi-xp/xpc_main.c
@@ -3,7 +3,7 @@
3 * License. See the file "COPYING" in the main directory of this archive 3 * License. See the file "COPYING" in the main directory of this archive
4 * for more details. 4 * for more details.
5 * 5 *
6 * Copyright (c) 2004-2008 Silicon Graphics, Inc. All Rights Reserved. 6 * Copyright (c) 2004-2009 Silicon Graphics, Inc. All Rights Reserved.
7 */ 7 */
8 8
9/* 9/*
@@ -150,7 +150,6 @@ DECLARE_WAIT_QUEUE_HEAD(xpc_activate_IRQ_wq);
150 150
151static unsigned long xpc_hb_check_timeout; 151static unsigned long xpc_hb_check_timeout;
152static struct timer_list xpc_hb_timer; 152static struct timer_list xpc_hb_timer;
153void *xpc_heartbeating_to_mask;
154 153
155/* notification that the xpc_hb_checker thread has exited */ 154/* notification that the xpc_hb_checker thread has exited */
156static DECLARE_COMPLETION(xpc_hb_checker_exited); 155static DECLARE_COMPLETION(xpc_hb_checker_exited);
@@ -170,62 +169,7 @@ static struct notifier_block xpc_die_notifier = {
170 .notifier_call = xpc_system_die, 169 .notifier_call = xpc_system_die,
171}; 170};
172 171
173int (*xpc_setup_partitions_sn) (void); 172struct xpc_arch_operations xpc_arch_ops;
174void (*xpc_teardown_partitions_sn) (void);
175enum xp_retval (*xpc_get_partition_rsvd_page_pa) (void *buf, u64 *cookie,
176 unsigned long *rp_pa,
177 size_t *len);
178int (*xpc_setup_rsvd_page_sn) (struct xpc_rsvd_page *rp);
179void (*xpc_heartbeat_init) (void);
180void (*xpc_heartbeat_exit) (void);
181void (*xpc_increment_heartbeat) (void);
182void (*xpc_offline_heartbeat) (void);
183void (*xpc_online_heartbeat) (void);
184enum xp_retval (*xpc_get_remote_heartbeat) (struct xpc_partition *part);
185
186enum xp_retval (*xpc_make_first_contact) (struct xpc_partition *part);
187void (*xpc_notify_senders_of_disconnect) (struct xpc_channel *ch);
188u64 (*xpc_get_chctl_all_flags) (struct xpc_partition *part);
189enum xp_retval (*xpc_setup_msg_structures) (struct xpc_channel *ch);
190void (*xpc_teardown_msg_structures) (struct xpc_channel *ch);
191void (*xpc_process_msg_chctl_flags) (struct xpc_partition *part, int ch_number);
192int (*xpc_n_of_deliverable_payloads) (struct xpc_channel *ch);
193void *(*xpc_get_deliverable_payload) (struct xpc_channel *ch);
194
195void (*xpc_request_partition_activation) (struct xpc_rsvd_page *remote_rp,
196 unsigned long remote_rp_pa,
197 int nasid);
198void (*xpc_request_partition_reactivation) (struct xpc_partition *part);
199void (*xpc_request_partition_deactivation) (struct xpc_partition *part);
200void (*xpc_cancel_partition_deactivation_request) (struct xpc_partition *part);
201
202void (*xpc_process_activate_IRQ_rcvd) (void);
203enum xp_retval (*xpc_setup_ch_structures_sn) (struct xpc_partition *part);
204void (*xpc_teardown_ch_structures_sn) (struct xpc_partition *part);
205
206void (*xpc_indicate_partition_engaged) (struct xpc_partition *part);
207int (*xpc_partition_engaged) (short partid);
208int (*xpc_any_partition_engaged) (void);
209void (*xpc_indicate_partition_disengaged) (struct xpc_partition *part);
210void (*xpc_assume_partition_disengaged) (short partid);
211
212void (*xpc_send_chctl_closerequest) (struct xpc_channel *ch,
213 unsigned long *irq_flags);
214void (*xpc_send_chctl_closereply) (struct xpc_channel *ch,
215 unsigned long *irq_flags);
216void (*xpc_send_chctl_openrequest) (struct xpc_channel *ch,
217 unsigned long *irq_flags);
218void (*xpc_send_chctl_openreply) (struct xpc_channel *ch,
219 unsigned long *irq_flags);
220
221enum xp_retval (*xpc_save_remote_msgqueue_pa) (struct xpc_channel *ch,
222 unsigned long msgqueue_pa);
223
224enum xp_retval (*xpc_send_payload) (struct xpc_channel *ch, u32 flags,
225 void *payload, u16 payload_size,
226 u8 notify_type, xpc_notify_func func,
227 void *key);
228void (*xpc_received_payload) (struct xpc_channel *ch, void *payload);
229 173
230/* 174/*
231 * Timer function to enforce the timelimit on the partition disengage. 175 * Timer function to enforce the timelimit on the partition disengage.
@@ -240,7 +184,7 @@ xpc_timeout_partition_disengage(unsigned long data)
240 (void)xpc_partition_disengaged(part); 184 (void)xpc_partition_disengaged(part);
241 185
242 DBUG_ON(part->disengage_timeout != 0); 186 DBUG_ON(part->disengage_timeout != 0);
243 DBUG_ON(xpc_partition_engaged(XPC_PARTID(part))); 187 DBUG_ON(xpc_arch_ops.partition_engaged(XPC_PARTID(part)));
244} 188}
245 189
246/* 190/*
@@ -251,7 +195,7 @@ xpc_timeout_partition_disengage(unsigned long data)
251static void 195static void
252xpc_hb_beater(unsigned long dummy) 196xpc_hb_beater(unsigned long dummy)
253{ 197{
254 xpc_increment_heartbeat(); 198 xpc_arch_ops.increment_heartbeat();
255 199
256 if (time_is_before_eq_jiffies(xpc_hb_check_timeout)) 200 if (time_is_before_eq_jiffies(xpc_hb_check_timeout))
257 wake_up_interruptible(&xpc_activate_IRQ_wq); 201 wake_up_interruptible(&xpc_activate_IRQ_wq);
@@ -263,7 +207,7 @@ xpc_hb_beater(unsigned long dummy)
263static void 207static void
264xpc_start_hb_beater(void) 208xpc_start_hb_beater(void)
265{ 209{
266 xpc_heartbeat_init(); 210 xpc_arch_ops.heartbeat_init();
267 init_timer(&xpc_hb_timer); 211 init_timer(&xpc_hb_timer);
268 xpc_hb_timer.function = xpc_hb_beater; 212 xpc_hb_timer.function = xpc_hb_beater;
269 xpc_hb_beater(0); 213 xpc_hb_beater(0);
@@ -273,7 +217,7 @@ static void
273xpc_stop_hb_beater(void) 217xpc_stop_hb_beater(void)
274{ 218{
275 del_timer_sync(&xpc_hb_timer); 219 del_timer_sync(&xpc_hb_timer);
276 xpc_heartbeat_exit(); 220 xpc_arch_ops.heartbeat_exit();
277} 221}
278 222
279/* 223/*
@@ -302,7 +246,7 @@ xpc_check_remote_hb(void)
302 continue; 246 continue;
303 } 247 }
304 248
305 ret = xpc_get_remote_heartbeat(part); 249 ret = xpc_arch_ops.get_remote_heartbeat(part);
306 if (ret != xpSuccess) 250 if (ret != xpSuccess)
307 XPC_DEACTIVATE_PARTITION(part, ret); 251 XPC_DEACTIVATE_PARTITION(part, ret);
308 } 252 }
@@ -353,7 +297,7 @@ xpc_hb_checker(void *ignore)
353 force_IRQ = 0; 297 force_IRQ = 0;
354 dev_dbg(xpc_part, "processing activate IRQs " 298 dev_dbg(xpc_part, "processing activate IRQs "
355 "received\n"); 299 "received\n");
356 xpc_process_activate_IRQ_rcvd(); 300 xpc_arch_ops.process_activate_IRQ_rcvd();
357 } 301 }
358 302
359 /* wait for IRQ or timeout */ 303 /* wait for IRQ or timeout */
@@ -528,7 +472,7 @@ xpc_setup_ch_structures(struct xpc_partition *part)
528 init_waitqueue_head(&ch->idle_wq); 472 init_waitqueue_head(&ch->idle_wq);
529 } 473 }
530 474
531 ret = xpc_setup_ch_structures_sn(part); 475 ret = xpc_arch_ops.setup_ch_structures(part);
532 if (ret != xpSuccess) 476 if (ret != xpSuccess)
533 goto out_2; 477 goto out_2;
534 478
@@ -572,7 +516,7 @@ xpc_teardown_ch_structures(struct xpc_partition *part)
572 516
573 /* now we can begin tearing down the infrastructure */ 517 /* now we can begin tearing down the infrastructure */
574 518
575 xpc_teardown_ch_structures_sn(part); 519 xpc_arch_ops.teardown_ch_structures(part);
576 520
577 kfree(part->remote_openclose_args_base); 521 kfree(part->remote_openclose_args_base);
578 part->remote_openclose_args = NULL; 522 part->remote_openclose_args = NULL;
@@ -620,12 +564,12 @@ xpc_activating(void *__partid)
620 564
621 dev_dbg(xpc_part, "activating partition %d\n", partid); 565 dev_dbg(xpc_part, "activating partition %d\n", partid);
622 566
623 xpc_allow_hb(partid); 567 xpc_arch_ops.allow_hb(partid);
624 568
625 if (xpc_setup_ch_structures(part) == xpSuccess) { 569 if (xpc_setup_ch_structures(part) == xpSuccess) {
626 (void)xpc_part_ref(part); /* this will always succeed */ 570 (void)xpc_part_ref(part); /* this will always succeed */
627 571
628 if (xpc_make_first_contact(part) == xpSuccess) { 572 if (xpc_arch_ops.make_first_contact(part) == xpSuccess) {
629 xpc_mark_partition_active(part); 573 xpc_mark_partition_active(part);
630 xpc_channel_mgr(part); 574 xpc_channel_mgr(part);
631 /* won't return until partition is deactivating */ 575 /* won't return until partition is deactivating */
@@ -635,12 +579,12 @@ xpc_activating(void *__partid)
635 xpc_teardown_ch_structures(part); 579 xpc_teardown_ch_structures(part);
636 } 580 }
637 581
638 xpc_disallow_hb(partid); 582 xpc_arch_ops.disallow_hb(partid);
639 xpc_mark_partition_inactive(part); 583 xpc_mark_partition_inactive(part);
640 584
641 if (part->reason == xpReactivating) { 585 if (part->reason == xpReactivating) {
642 /* interrupting ourselves results in activating partition */ 586 /* interrupting ourselves results in activating partition */
643 xpc_request_partition_reactivation(part); 587 xpc_arch_ops.request_partition_reactivation(part);
644 } 588 }
645 589
646 return 0; 590 return 0;
@@ -713,10 +657,13 @@ xpc_activate_kthreads(struct xpc_channel *ch, int needed)
713static void 657static void
714xpc_kthread_waitmsgs(struct xpc_partition *part, struct xpc_channel *ch) 658xpc_kthread_waitmsgs(struct xpc_partition *part, struct xpc_channel *ch)
715{ 659{
660 int (*n_of_deliverable_payloads) (struct xpc_channel *) =
661 xpc_arch_ops.n_of_deliverable_payloads;
662
716 do { 663 do {
717 /* deliver messages to their intended recipients */ 664 /* deliver messages to their intended recipients */
718 665
719 while (xpc_n_of_deliverable_payloads(ch) > 0 && 666 while (n_of_deliverable_payloads(ch) > 0 &&
720 !(ch->flags & XPC_C_DISCONNECTING)) { 667 !(ch->flags & XPC_C_DISCONNECTING)) {
721 xpc_deliver_payload(ch); 668 xpc_deliver_payload(ch);
722 } 669 }
@@ -732,7 +679,7 @@ xpc_kthread_waitmsgs(struct xpc_partition *part, struct xpc_channel *ch)
732 "wait_event_interruptible_exclusive()\n"); 679 "wait_event_interruptible_exclusive()\n");
733 680
734 (void)wait_event_interruptible_exclusive(ch->idle_wq, 681 (void)wait_event_interruptible_exclusive(ch->idle_wq,
735 (xpc_n_of_deliverable_payloads(ch) > 0 || 682 (n_of_deliverable_payloads(ch) > 0 ||
736 (ch->flags & XPC_C_DISCONNECTING))); 683 (ch->flags & XPC_C_DISCONNECTING)));
737 684
738 atomic_dec(&ch->kthreads_idle); 685 atomic_dec(&ch->kthreads_idle);
@@ -749,6 +696,8 @@ xpc_kthread_start(void *args)
749 struct xpc_channel *ch; 696 struct xpc_channel *ch;
750 int n_needed; 697 int n_needed;
751 unsigned long irq_flags; 698 unsigned long irq_flags;
699 int (*n_of_deliverable_payloads) (struct xpc_channel *) =
700 xpc_arch_ops.n_of_deliverable_payloads;
752 701
753 dev_dbg(xpc_chan, "kthread starting, partid=%d, channel=%d\n", 702 dev_dbg(xpc_chan, "kthread starting, partid=%d, channel=%d\n",
754 partid, ch_number); 703 partid, ch_number);
@@ -777,7 +726,7 @@ xpc_kthread_start(void *args)
777 * additional kthreads to help deliver them. We only 726 * additional kthreads to help deliver them. We only
778 * need one less than total #of messages to deliver. 727 * need one less than total #of messages to deliver.
779 */ 728 */
780 n_needed = xpc_n_of_deliverable_payloads(ch) - 1; 729 n_needed = n_of_deliverable_payloads(ch) - 1;
781 if (n_needed > 0 && !(ch->flags & XPC_C_DISCONNECTING)) 730 if (n_needed > 0 && !(ch->flags & XPC_C_DISCONNECTING))
782 xpc_activate_kthreads(ch, n_needed); 731 xpc_activate_kthreads(ch, n_needed);
783 732
@@ -805,7 +754,7 @@ xpc_kthread_start(void *args)
805 754
806 if (atomic_dec_return(&ch->kthreads_assigned) == 0 && 755 if (atomic_dec_return(&ch->kthreads_assigned) == 0 &&
807 atomic_dec_return(&part->nchannels_engaged) == 0) { 756 atomic_dec_return(&part->nchannels_engaged) == 0) {
808 xpc_indicate_partition_disengaged(part); 757 xpc_arch_ops.indicate_partition_disengaged(part);
809 } 758 }
810 759
811 xpc_msgqueue_deref(ch); 760 xpc_msgqueue_deref(ch);
@@ -837,6 +786,8 @@ xpc_create_kthreads(struct xpc_channel *ch, int needed,
837 u64 args = XPC_PACK_ARGS(ch->partid, ch->number); 786 u64 args = XPC_PACK_ARGS(ch->partid, ch->number);
838 struct xpc_partition *part = &xpc_partitions[ch->partid]; 787 struct xpc_partition *part = &xpc_partitions[ch->partid];
839 struct task_struct *kthread; 788 struct task_struct *kthread;
789 void (*indicate_partition_disengaged) (struct xpc_partition *) =
790 xpc_arch_ops.indicate_partition_disengaged;
840 791
841 while (needed-- > 0) { 792 while (needed-- > 0) {
842 793
@@ -858,7 +809,7 @@ xpc_create_kthreads(struct xpc_channel *ch, int needed,
858 809
859 } else if (atomic_inc_return(&ch->kthreads_assigned) == 1 && 810 } else if (atomic_inc_return(&ch->kthreads_assigned) == 1 &&
860 atomic_inc_return(&part->nchannels_engaged) == 1) { 811 atomic_inc_return(&part->nchannels_engaged) == 1) {
861 xpc_indicate_partition_engaged(part); 812 xpc_arch_ops.indicate_partition_engaged(part);
862 } 813 }
863 (void)xpc_part_ref(part); 814 (void)xpc_part_ref(part);
864 xpc_msgqueue_ref(ch); 815 xpc_msgqueue_ref(ch);
@@ -880,7 +831,7 @@ xpc_create_kthreads(struct xpc_channel *ch, int needed,
880 831
881 if (atomic_dec_return(&ch->kthreads_assigned) == 0 && 832 if (atomic_dec_return(&ch->kthreads_assigned) == 0 &&
882 atomic_dec_return(&part->nchannels_engaged) == 0) { 833 atomic_dec_return(&part->nchannels_engaged) == 0) {
883 xpc_indicate_partition_disengaged(part); 834 indicate_partition_disengaged(part);
884 } 835 }
885 xpc_msgqueue_deref(ch); 836 xpc_msgqueue_deref(ch);
886 xpc_part_deref(part); 837 xpc_part_deref(part);
@@ -993,13 +944,13 @@ xpc_setup_partitions(void)
993 atomic_set(&part->references, 0); 944 atomic_set(&part->references, 0);
994 } 945 }
995 946
996 return xpc_setup_partitions_sn(); 947 return xpc_arch_ops.setup_partitions();
997} 948}
998 949
999static void 950static void
1000xpc_teardown_partitions(void) 951xpc_teardown_partitions(void)
1001{ 952{
1002 xpc_teardown_partitions_sn(); 953 xpc_arch_ops.teardown_partitions();
1003 kfree(xpc_partitions); 954 kfree(xpc_partitions);
1004} 955}
1005 956
@@ -1055,7 +1006,7 @@ xpc_do_exit(enum xp_retval reason)
1055 disengage_timeout = part->disengage_timeout; 1006 disengage_timeout = part->disengage_timeout;
1056 } 1007 }
1057 1008
1058 if (xpc_any_partition_engaged()) { 1009 if (xpc_arch_ops.any_partition_engaged()) {
1059 if (time_is_before_jiffies(printmsg_time)) { 1010 if (time_is_before_jiffies(printmsg_time)) {
1060 dev_info(xpc_part, "waiting for remote " 1011 dev_info(xpc_part, "waiting for remote "
1061 "partitions to deactivate, timeout in " 1012 "partitions to deactivate, timeout in "
@@ -1086,8 +1037,7 @@ xpc_do_exit(enum xp_retval reason)
1086 1037
1087 } while (1); 1038 } while (1);
1088 1039
1089 DBUG_ON(xpc_any_partition_engaged()); 1040 DBUG_ON(xpc_arch_ops.any_partition_engaged());
1090 DBUG_ON(xpc_any_hbs_allowed() != 0);
1091 1041
1092 xpc_teardown_rsvd_page(); 1042 xpc_teardown_rsvd_page();
1093 1043
@@ -1152,15 +1102,15 @@ xpc_die_deactivate(void)
1152 /* keep xpc_hb_checker thread from doing anything (just in case) */ 1102 /* keep xpc_hb_checker thread from doing anything (just in case) */
1153 xpc_exiting = 1; 1103 xpc_exiting = 1;
1154 1104
1155 xpc_disallow_all_hbs(); /*indicate we're deactivated */ 1105 xpc_arch_ops.disallow_all_hbs(); /*indicate we're deactivated */
1156 1106
1157 for (partid = 0; partid < xp_max_npartitions; partid++) { 1107 for (partid = 0; partid < xp_max_npartitions; partid++) {
1158 part = &xpc_partitions[partid]; 1108 part = &xpc_partitions[partid];
1159 1109
1160 if (xpc_partition_engaged(partid) || 1110 if (xpc_arch_ops.partition_engaged(partid) ||
1161 part->act_state != XPC_P_AS_INACTIVE) { 1111 part->act_state != XPC_P_AS_INACTIVE) {
1162 xpc_request_partition_deactivation(part); 1112 xpc_arch_ops.request_partition_deactivation(part);
1163 xpc_indicate_partition_disengaged(part); 1113 xpc_arch_ops.indicate_partition_disengaged(part);
1164 } 1114 }
1165 } 1115 }
1166 1116
@@ -1177,7 +1127,7 @@ xpc_die_deactivate(void)
1177 wait_to_print = XPC_DEACTIVATE_PRINTMSG_INTERVAL * 1000 * 5; 1127 wait_to_print = XPC_DEACTIVATE_PRINTMSG_INTERVAL * 1000 * 5;
1178 1128
1179 while (1) { 1129 while (1) {
1180 any_engaged = xpc_any_partition_engaged(); 1130 any_engaged = xpc_arch_ops.any_partition_engaged();
1181 if (!any_engaged) { 1131 if (!any_engaged) {
1182 dev_info(xpc_part, "all partitions have deactivated\n"); 1132 dev_info(xpc_part, "all partitions have deactivated\n");
1183 break; 1133 break;
@@ -1186,7 +1136,7 @@ xpc_die_deactivate(void)
1186 if (!keep_waiting--) { 1136 if (!keep_waiting--) {
1187 for (partid = 0; partid < xp_max_npartitions; 1137 for (partid = 0; partid < xp_max_npartitions;
1188 partid++) { 1138 partid++) {
1189 if (xpc_partition_engaged(partid)) { 1139 if (xpc_arch_ops.partition_engaged(partid)) {
1190 dev_info(xpc_part, "deactivate from " 1140 dev_info(xpc_part, "deactivate from "
1191 "remote partition %d timed " 1141 "remote partition %d timed "
1192 "out\n", partid); 1142 "out\n", partid);
@@ -1233,7 +1183,7 @@ xpc_system_die(struct notifier_block *nb, unsigned long event, void *unused)
1233 /* fall through */ 1183 /* fall through */
1234 case DIE_MCA_MONARCH_ENTER: 1184 case DIE_MCA_MONARCH_ENTER:
1235 case DIE_INIT_MONARCH_ENTER: 1185 case DIE_INIT_MONARCH_ENTER:
1236 xpc_offline_heartbeat(); 1186 xpc_arch_ops.offline_heartbeat();
1237 break; 1187 break;
1238 1188
1239 case DIE_KDEBUG_LEAVE: 1189 case DIE_KDEBUG_LEAVE:
@@ -1244,7 +1194,7 @@ xpc_system_die(struct notifier_block *nb, unsigned long event, void *unused)
1244 /* fall through */ 1194 /* fall through */
1245 case DIE_MCA_MONARCH_LEAVE: 1195 case DIE_MCA_MONARCH_LEAVE:
1246 case DIE_INIT_MONARCH_LEAVE: 1196 case DIE_INIT_MONARCH_LEAVE:
1247 xpc_online_heartbeat(); 1197 xpc_arch_ops.online_heartbeat();
1248 break; 1198 break;
1249 } 1199 }
1250#else 1200#else
diff --git a/drivers/misc/sgi-xp/xpc_partition.c b/drivers/misc/sgi-xp/xpc_partition.c
index 6722f6fe4dc7..65877bc5edaa 100644
--- a/drivers/misc/sgi-xp/xpc_partition.c
+++ b/drivers/misc/sgi-xp/xpc_partition.c
@@ -70,6 +70,9 @@ xpc_get_rsvd_page_pa(int nasid)
70 size_t buf_len = 0; 70 size_t buf_len = 0;
71 void *buf = buf; 71 void *buf = buf;
72 void *buf_base = NULL; 72 void *buf_base = NULL;
73 enum xp_retval (*get_partition_rsvd_page_pa)
74 (void *, u64 *, unsigned long *, size_t *) =
75 xpc_arch_ops.get_partition_rsvd_page_pa;
73 76
74 while (1) { 77 while (1) {
75 78
@@ -79,8 +82,7 @@ xpc_get_rsvd_page_pa(int nasid)
79 * ??? function or have two versions? Rename rp_pa for UV to 82 * ??? function or have two versions? Rename rp_pa for UV to
80 * ??? rp_gpa? 83 * ??? rp_gpa?
81 */ 84 */
82 ret = xpc_get_partition_rsvd_page_pa(buf, &cookie, &rp_pa, 85 ret = get_partition_rsvd_page_pa(buf, &cookie, &rp_pa, &len);
83 &len);
84 86
85 dev_dbg(xpc_part, "SAL returned with ret=%d, cookie=0x%016lx, " 87 dev_dbg(xpc_part, "SAL returned with ret=%d, cookie=0x%016lx, "
86 "address=0x%016lx, len=0x%016lx\n", ret, 88 "address=0x%016lx, len=0x%016lx\n", ret,
@@ -172,7 +174,7 @@ xpc_setup_rsvd_page(void)
172 xpc_part_nasids = XPC_RP_PART_NASIDS(rp); 174 xpc_part_nasids = XPC_RP_PART_NASIDS(rp);
173 xpc_mach_nasids = XPC_RP_MACH_NASIDS(rp); 175 xpc_mach_nasids = XPC_RP_MACH_NASIDS(rp);
174 176
175 ret = xpc_setup_rsvd_page_sn(rp); 177 ret = xpc_arch_ops.setup_rsvd_page(rp);
176 if (ret != 0) 178 if (ret != 0)
177 return ret; 179 return ret;
178 180
@@ -264,7 +266,7 @@ xpc_partition_disengaged(struct xpc_partition *part)
264 short partid = XPC_PARTID(part); 266 short partid = XPC_PARTID(part);
265 int disengaged; 267 int disengaged;
266 268
267 disengaged = !xpc_partition_engaged(partid); 269 disengaged = !xpc_arch_ops.partition_engaged(partid);
268 if (part->disengage_timeout) { 270 if (part->disengage_timeout) {
269 if (!disengaged) { 271 if (!disengaged) {
270 if (time_is_after_jiffies(part->disengage_timeout)) { 272 if (time_is_after_jiffies(part->disengage_timeout)) {
@@ -280,7 +282,7 @@ xpc_partition_disengaged(struct xpc_partition *part)
280 dev_info(xpc_part, "deactivate request to remote " 282 dev_info(xpc_part, "deactivate request to remote "
281 "partition %d timed out\n", partid); 283 "partition %d timed out\n", partid);
282 xpc_disengage_timedout = 1; 284 xpc_disengage_timedout = 1;
283 xpc_assume_partition_disengaged(partid); 285 xpc_arch_ops.assume_partition_disengaged(partid);
284 disengaged = 1; 286 disengaged = 1;
285 } 287 }
286 part->disengage_timeout = 0; 288 part->disengage_timeout = 0;
@@ -294,7 +296,7 @@ xpc_partition_disengaged(struct xpc_partition *part)
294 if (part->act_state != XPC_P_AS_INACTIVE) 296 if (part->act_state != XPC_P_AS_INACTIVE)
295 xpc_wakeup_channel_mgr(part); 297 xpc_wakeup_channel_mgr(part);
296 298
297 xpc_cancel_partition_deactivation_request(part); 299 xpc_arch_ops.cancel_partition_deactivation_request(part);
298 } 300 }
299 return disengaged; 301 return disengaged;
300} 302}
@@ -339,7 +341,7 @@ xpc_deactivate_partition(const int line, struct xpc_partition *part,
339 spin_unlock_irqrestore(&part->act_lock, irq_flags); 341 spin_unlock_irqrestore(&part->act_lock, irq_flags);
340 if (reason == xpReactivating) { 342 if (reason == xpReactivating) {
341 /* we interrupt ourselves to reactivate partition */ 343 /* we interrupt ourselves to reactivate partition */
342 xpc_request_partition_reactivation(part); 344 xpc_arch_ops.request_partition_reactivation(part);
343 } 345 }
344 return; 346 return;
345 } 347 }
@@ -358,7 +360,7 @@ xpc_deactivate_partition(const int line, struct xpc_partition *part,
358 spin_unlock_irqrestore(&part->act_lock, irq_flags); 360 spin_unlock_irqrestore(&part->act_lock, irq_flags);
359 361
360 /* ask remote partition to deactivate with regard to us */ 362 /* ask remote partition to deactivate with regard to us */
361 xpc_request_partition_deactivation(part); 363 xpc_arch_ops.request_partition_deactivation(part);
362 364
363 /* set a timelimit on the disengage phase of the deactivation request */ 365 /* set a timelimit on the disengage phase of the deactivation request */
364 part->disengage_timeout = jiffies + (xpc_disengage_timelimit * HZ); 366 part->disengage_timeout = jiffies + (xpc_disengage_timelimit * HZ);
@@ -496,7 +498,7 @@ xpc_discovery(void)
496 continue; 498 continue;
497 } 499 }
498 500
499 xpc_request_partition_activation(remote_rp, 501 xpc_arch_ops.request_partition_activation(remote_rp,
500 remote_rp_pa, nasid); 502 remote_rp_pa, nasid);
501 } 503 }
502 } 504 }
diff --git a/drivers/misc/sgi-xp/xpc_sn2.c b/drivers/misc/sgi-xp/xpc_sn2.c
index eaaa964942de..915a3b495da5 100644
--- a/drivers/misc/sgi-xp/xpc_sn2.c
+++ b/drivers/misc/sgi-xp/xpc_sn2.c
@@ -3,7 +3,7 @@
3 * License. See the file "COPYING" in the main directory of this archive 3 * License. See the file "COPYING" in the main directory of this archive
4 * for more details. 4 * for more details.
5 * 5 *
6 * Copyright (c) 2008 Silicon Graphics, Inc. All Rights Reserved. 6 * Copyright (c) 2008-2009 Silicon Graphics, Inc. All Rights Reserved.
7 */ 7 */
8 8
9/* 9/*
@@ -60,14 +60,14 @@ static struct xpc_vars_sn2 *xpc_vars_sn2;
60static struct xpc_vars_part_sn2 *xpc_vars_part_sn2; 60static struct xpc_vars_part_sn2 *xpc_vars_part_sn2;
61 61
62static int 62static int
63xpc_setup_partitions_sn_sn2(void) 63xpc_setup_partitions_sn2(void)
64{ 64{
65 /* nothing needs to be done */ 65 /* nothing needs to be done */
66 return 0; 66 return 0;
67} 67}
68 68
69static void 69static void
70xpc_teardown_partitions_sn_sn2(void) 70xpc_teardown_partitions_sn2(void)
71{ 71{
72 /* nothing needs to be done */ 72 /* nothing needs to be done */
73} 73}
@@ -431,6 +431,13 @@ xpc_send_chctl_openreply_sn2(struct xpc_channel *ch, unsigned long *irq_flags)
431} 431}
432 432
433static void 433static void
434xpc_send_chctl_opencomplete_sn2(struct xpc_channel *ch,
435 unsigned long *irq_flags)
436{
437 XPC_SEND_NOTIFY_IRQ_SN2(ch, XPC_CHCTL_OPENCOMPLETE, irq_flags);
438}
439
440static void
434xpc_send_chctl_msgrequest_sn2(struct xpc_channel *ch) 441xpc_send_chctl_msgrequest_sn2(struct xpc_channel *ch)
435{ 442{
436 XPC_SEND_NOTIFY_IRQ_SN2(ch, XPC_CHCTL_MSGREQUEST, NULL); 443 XPC_SEND_NOTIFY_IRQ_SN2(ch, XPC_CHCTL_MSGREQUEST, NULL);
@@ -621,7 +628,7 @@ xpc_get_partition_rsvd_page_pa_sn2(void *buf, u64 *cookie, unsigned long *rp_pa,
621 628
622 629
623static int 630static int
624xpc_setup_rsvd_page_sn_sn2(struct xpc_rsvd_page *rp) 631xpc_setup_rsvd_page_sn2(struct xpc_rsvd_page *rp)
625{ 632{
626 struct amo *amos_page; 633 struct amo *amos_page;
627 int i; 634 int i;
@@ -629,7 +636,7 @@ xpc_setup_rsvd_page_sn_sn2(struct xpc_rsvd_page *rp)
629 636
630 xpc_vars_sn2 = XPC_RP_VARS(rp); 637 xpc_vars_sn2 = XPC_RP_VARS(rp);
631 638
632 rp->sn.vars_pa = xp_pa(xpc_vars_sn2); 639 rp->sn.sn2.vars_pa = xp_pa(xpc_vars_sn2);
633 640
634 /* vars_part array follows immediately after vars */ 641 /* vars_part array follows immediately after vars */
635 xpc_vars_part_sn2 = (struct xpc_vars_part_sn2 *)((u8 *)XPC_RP_VARS(rp) + 642 xpc_vars_part_sn2 = (struct xpc_vars_part_sn2 *)((u8 *)XPC_RP_VARS(rp) +
@@ -693,6 +700,33 @@ xpc_setup_rsvd_page_sn_sn2(struct xpc_rsvd_page *rp)
693 return 0; 700 return 0;
694} 701}
695 702
703static int
704xpc_hb_allowed_sn2(short partid, void *heartbeating_to_mask)
705{
706 return test_bit(partid, heartbeating_to_mask);
707}
708
709static void
710xpc_allow_hb_sn2(short partid)
711{
712 DBUG_ON(xpc_vars_sn2 == NULL);
713 set_bit(partid, xpc_vars_sn2->heartbeating_to_mask);
714}
715
716static void
717xpc_disallow_hb_sn2(short partid)
718{
719 DBUG_ON(xpc_vars_sn2 == NULL);
720 clear_bit(partid, xpc_vars_sn2->heartbeating_to_mask);
721}
722
723static void
724xpc_disallow_all_hbs_sn2(void)
725{
726 DBUG_ON(xpc_vars_sn2 == NULL);
727 bitmap_zero(xpc_vars_sn2->heartbeating_to_mask, xp_max_npartitions);
728}
729
696static void 730static void
697xpc_increment_heartbeat_sn2(void) 731xpc_increment_heartbeat_sn2(void)
698{ 732{
@@ -719,7 +753,6 @@ xpc_heartbeat_init_sn2(void)
719 DBUG_ON(xpc_vars_sn2 == NULL); 753 DBUG_ON(xpc_vars_sn2 == NULL);
720 754
721 bitmap_zero(xpc_vars_sn2->heartbeating_to_mask, XP_MAX_NPARTITIONS_SN2); 755 bitmap_zero(xpc_vars_sn2->heartbeating_to_mask, XP_MAX_NPARTITIONS_SN2);
722 xpc_heartbeating_to_mask = &xpc_vars_sn2->heartbeating_to_mask[0];
723 xpc_online_heartbeat_sn2(); 756 xpc_online_heartbeat_sn2();
724} 757}
725 758
@@ -751,9 +784,9 @@ xpc_get_remote_heartbeat_sn2(struct xpc_partition *part)
751 remote_vars->heartbeating_to_mask[0]); 784 remote_vars->heartbeating_to_mask[0]);
752 785
753 if ((remote_vars->heartbeat == part->last_heartbeat && 786 if ((remote_vars->heartbeat == part->last_heartbeat &&
754 remote_vars->heartbeat_offline == 0) || 787 !remote_vars->heartbeat_offline) ||
755 !xpc_hb_allowed(sn_partition_id, 788 !xpc_hb_allowed_sn2(sn_partition_id,
756 &remote_vars->heartbeating_to_mask)) { 789 remote_vars->heartbeating_to_mask)) {
757 ret = xpNoHeartbeat; 790 ret = xpNoHeartbeat;
758 } else { 791 } else {
759 part->last_heartbeat = remote_vars->heartbeat; 792 part->last_heartbeat = remote_vars->heartbeat;
@@ -972,7 +1005,7 @@ xpc_identify_activate_IRQ_req_sn2(int nasid)
972 return; 1005 return;
973 } 1006 }
974 1007
975 remote_vars_pa = remote_rp->sn.vars_pa; 1008 remote_vars_pa = remote_rp->sn.sn2.vars_pa;
976 remote_rp_version = remote_rp->version; 1009 remote_rp_version = remote_rp->version;
977 remote_rp_ts_jiffies = remote_rp->ts_jiffies; 1010 remote_rp_ts_jiffies = remote_rp->ts_jiffies;
978 1011
@@ -1129,7 +1162,7 @@ xpc_process_activate_IRQ_rcvd_sn2(void)
1129 * Setup the channel structures that are sn2 specific. 1162 * Setup the channel structures that are sn2 specific.
1130 */ 1163 */
1131static enum xp_retval 1164static enum xp_retval
1132xpc_setup_ch_structures_sn_sn2(struct xpc_partition *part) 1165xpc_setup_ch_structures_sn2(struct xpc_partition *part)
1133{ 1166{
1134 struct xpc_partition_sn2 *part_sn2 = &part->sn.sn2; 1167 struct xpc_partition_sn2 *part_sn2 = &part->sn.sn2;
1135 struct xpc_channel_sn2 *ch_sn2; 1168 struct xpc_channel_sn2 *ch_sn2;
@@ -1251,7 +1284,7 @@ out_1:
1251 * Teardown the channel structures that are sn2 specific. 1284 * Teardown the channel structures that are sn2 specific.
1252 */ 1285 */
1253static void 1286static void
1254xpc_teardown_ch_structures_sn_sn2(struct xpc_partition *part) 1287xpc_teardown_ch_structures_sn2(struct xpc_partition *part)
1255{ 1288{
1256 struct xpc_partition_sn2 *part_sn2 = &part->sn.sn2; 1289 struct xpc_partition_sn2 *part_sn2 = &part->sn.sn2;
1257 short partid = XPC_PARTID(part); 1290 short partid = XPC_PARTID(part);
@@ -2315,61 +2348,70 @@ xpc_received_payload_sn2(struct xpc_channel *ch, void *payload)
2315 xpc_acknowledge_msgs_sn2(ch, get, msg->flags); 2348 xpc_acknowledge_msgs_sn2(ch, get, msg->flags);
2316} 2349}
2317 2350
2351static struct xpc_arch_operations xpc_arch_ops_sn2 = {
2352 .setup_partitions = xpc_setup_partitions_sn2,
2353 .teardown_partitions = xpc_teardown_partitions_sn2,
2354 .process_activate_IRQ_rcvd = xpc_process_activate_IRQ_rcvd_sn2,
2355 .get_partition_rsvd_page_pa = xpc_get_partition_rsvd_page_pa_sn2,
2356 .setup_rsvd_page = xpc_setup_rsvd_page_sn2,
2357
2358 .allow_hb = xpc_allow_hb_sn2,
2359 .disallow_hb = xpc_disallow_hb_sn2,
2360 .disallow_all_hbs = xpc_disallow_all_hbs_sn2,
2361 .increment_heartbeat = xpc_increment_heartbeat_sn2,
2362 .offline_heartbeat = xpc_offline_heartbeat_sn2,
2363 .online_heartbeat = xpc_online_heartbeat_sn2,
2364 .heartbeat_init = xpc_heartbeat_init_sn2,
2365 .heartbeat_exit = xpc_heartbeat_exit_sn2,
2366 .get_remote_heartbeat = xpc_get_remote_heartbeat_sn2,
2367
2368 .request_partition_activation =
2369 xpc_request_partition_activation_sn2,
2370 .request_partition_reactivation =
2371 xpc_request_partition_reactivation_sn2,
2372 .request_partition_deactivation =
2373 xpc_request_partition_deactivation_sn2,
2374 .cancel_partition_deactivation_request =
2375 xpc_cancel_partition_deactivation_request_sn2,
2376
2377 .setup_ch_structures = xpc_setup_ch_structures_sn2,
2378 .teardown_ch_structures = xpc_teardown_ch_structures_sn2,
2379
2380 .make_first_contact = xpc_make_first_contact_sn2,
2381
2382 .get_chctl_all_flags = xpc_get_chctl_all_flags_sn2,
2383 .send_chctl_closerequest = xpc_send_chctl_closerequest_sn2,
2384 .send_chctl_closereply = xpc_send_chctl_closereply_sn2,
2385 .send_chctl_openrequest = xpc_send_chctl_openrequest_sn2,
2386 .send_chctl_openreply = xpc_send_chctl_openreply_sn2,
2387 .send_chctl_opencomplete = xpc_send_chctl_opencomplete_sn2,
2388 .process_msg_chctl_flags = xpc_process_msg_chctl_flags_sn2,
2389
2390 .save_remote_msgqueue_pa = xpc_save_remote_msgqueue_pa_sn2,
2391
2392 .setup_msg_structures = xpc_setup_msg_structures_sn2,
2393 .teardown_msg_structures = xpc_teardown_msg_structures_sn2,
2394
2395 .indicate_partition_engaged = xpc_indicate_partition_engaged_sn2,
2396 .indicate_partition_disengaged = xpc_indicate_partition_disengaged_sn2,
2397 .partition_engaged = xpc_partition_engaged_sn2,
2398 .any_partition_engaged = xpc_any_partition_engaged_sn2,
2399 .assume_partition_disengaged = xpc_assume_partition_disengaged_sn2,
2400
2401 .n_of_deliverable_payloads = xpc_n_of_deliverable_payloads_sn2,
2402 .send_payload = xpc_send_payload_sn2,
2403 .get_deliverable_payload = xpc_get_deliverable_payload_sn2,
2404 .received_payload = xpc_received_payload_sn2,
2405 .notify_senders_of_disconnect = xpc_notify_senders_of_disconnect_sn2,
2406};
2407
2318int 2408int
2319xpc_init_sn2(void) 2409xpc_init_sn2(void)
2320{ 2410{
2321 int ret; 2411 int ret;
2322 size_t buf_size; 2412 size_t buf_size;
2323 2413
2324 xpc_setup_partitions_sn = xpc_setup_partitions_sn_sn2; 2414 xpc_arch_ops = xpc_arch_ops_sn2;
2325 xpc_teardown_partitions_sn = xpc_teardown_partitions_sn_sn2;
2326 xpc_get_partition_rsvd_page_pa = xpc_get_partition_rsvd_page_pa_sn2;
2327 xpc_setup_rsvd_page_sn = xpc_setup_rsvd_page_sn_sn2;
2328 xpc_increment_heartbeat = xpc_increment_heartbeat_sn2;
2329 xpc_offline_heartbeat = xpc_offline_heartbeat_sn2;
2330 xpc_online_heartbeat = xpc_online_heartbeat_sn2;
2331 xpc_heartbeat_init = xpc_heartbeat_init_sn2;
2332 xpc_heartbeat_exit = xpc_heartbeat_exit_sn2;
2333 xpc_get_remote_heartbeat = xpc_get_remote_heartbeat_sn2;
2334
2335 xpc_request_partition_activation = xpc_request_partition_activation_sn2;
2336 xpc_request_partition_reactivation =
2337 xpc_request_partition_reactivation_sn2;
2338 xpc_request_partition_deactivation =
2339 xpc_request_partition_deactivation_sn2;
2340 xpc_cancel_partition_deactivation_request =
2341 xpc_cancel_partition_deactivation_request_sn2;
2342
2343 xpc_process_activate_IRQ_rcvd = xpc_process_activate_IRQ_rcvd_sn2;
2344 xpc_setup_ch_structures_sn = xpc_setup_ch_structures_sn_sn2;
2345 xpc_teardown_ch_structures_sn = xpc_teardown_ch_structures_sn_sn2;
2346 xpc_make_first_contact = xpc_make_first_contact_sn2;
2347
2348 xpc_get_chctl_all_flags = xpc_get_chctl_all_flags_sn2;
2349 xpc_send_chctl_closerequest = xpc_send_chctl_closerequest_sn2;
2350 xpc_send_chctl_closereply = xpc_send_chctl_closereply_sn2;
2351 xpc_send_chctl_openrequest = xpc_send_chctl_openrequest_sn2;
2352 xpc_send_chctl_openreply = xpc_send_chctl_openreply_sn2;
2353
2354 xpc_save_remote_msgqueue_pa = xpc_save_remote_msgqueue_pa_sn2;
2355
2356 xpc_setup_msg_structures = xpc_setup_msg_structures_sn2;
2357 xpc_teardown_msg_structures = xpc_teardown_msg_structures_sn2;
2358
2359 xpc_notify_senders_of_disconnect = xpc_notify_senders_of_disconnect_sn2;
2360 xpc_process_msg_chctl_flags = xpc_process_msg_chctl_flags_sn2;
2361 xpc_n_of_deliverable_payloads = xpc_n_of_deliverable_payloads_sn2;
2362 xpc_get_deliverable_payload = xpc_get_deliverable_payload_sn2;
2363
2364 xpc_indicate_partition_engaged = xpc_indicate_partition_engaged_sn2;
2365 xpc_indicate_partition_disengaged =
2366 xpc_indicate_partition_disengaged_sn2;
2367 xpc_partition_engaged = xpc_partition_engaged_sn2;
2368 xpc_any_partition_engaged = xpc_any_partition_engaged_sn2;
2369 xpc_assume_partition_disengaged = xpc_assume_partition_disengaged_sn2;
2370
2371 xpc_send_payload = xpc_send_payload_sn2;
2372 xpc_received_payload = xpc_received_payload_sn2;
2373 2415
2374 if (offsetof(struct xpc_msg_sn2, payload) > XPC_MSG_HDR_MAX_SIZE) { 2416 if (offsetof(struct xpc_msg_sn2, payload) > XPC_MSG_HDR_MAX_SIZE) {
2375 dev_err(xpc_part, "header portion of struct xpc_msg_sn2 is " 2417 dev_err(xpc_part, "header portion of struct xpc_msg_sn2 is "
diff --git a/drivers/misc/sgi-xp/xpc_uv.c b/drivers/misc/sgi-xp/xpc_uv.c
index f7fff4727edb..9172fcdee4e2 100644
--- a/drivers/misc/sgi-xp/xpc_uv.c
+++ b/drivers/misc/sgi-xp/xpc_uv.c
@@ -46,8 +46,7 @@ struct uv_IO_APIC_route_entry {
46}; 46};
47#endif 47#endif
48 48
49static atomic64_t xpc_heartbeat_uv; 49static struct xpc_heartbeat_uv *xpc_heartbeat_uv;
50static DECLARE_BITMAP(xpc_heartbeating_to_mask_uv, XP_MAX_NPARTITIONS_UV);
51 50
52#define XPC_ACTIVATE_MSG_SIZE_UV (1 * GRU_CACHE_LINE_BYTES) 51#define XPC_ACTIVATE_MSG_SIZE_UV (1 * GRU_CACHE_LINE_BYTES)
53#define XPC_ACTIVATE_MQ_SIZE_UV (4 * XP_MAX_NPARTITIONS_UV * \ 52#define XPC_ACTIVATE_MQ_SIZE_UV (4 * XP_MAX_NPARTITIONS_UV * \
@@ -63,7 +62,7 @@ static struct xpc_gru_mq_uv *xpc_activate_mq_uv;
63static struct xpc_gru_mq_uv *xpc_notify_mq_uv; 62static struct xpc_gru_mq_uv *xpc_notify_mq_uv;
64 63
65static int 64static int
66xpc_setup_partitions_sn_uv(void) 65xpc_setup_partitions_uv(void)
67{ 66{
68 short partid; 67 short partid;
69 struct xpc_partition_uv *part_uv; 68 struct xpc_partition_uv *part_uv;
@@ -79,7 +78,7 @@ xpc_setup_partitions_sn_uv(void)
79} 78}
80 79
81static void 80static void
82xpc_teardown_partitions_sn_uv(void) 81xpc_teardown_partitions_uv(void)
83{ 82{
84 short partid; 83 short partid;
85 struct xpc_partition_uv *part_uv; 84 struct xpc_partition_uv *part_uv;
@@ -423,41 +422,6 @@ xpc_handle_activate_mq_msg_uv(struct xpc_partition *part,
423 /* syncing of remote_act_state was just done above */ 422 /* syncing of remote_act_state was just done above */
424 break; 423 break;
425 424
426 case XPC_ACTIVATE_MQ_MSG_INC_HEARTBEAT_UV: {
427 struct xpc_activate_mq_msg_heartbeat_req_uv *msg;
428
429 msg = container_of(msg_hdr,
430 struct xpc_activate_mq_msg_heartbeat_req_uv,
431 hdr);
432 part_uv->heartbeat = msg->heartbeat;
433 break;
434 }
435 case XPC_ACTIVATE_MQ_MSG_OFFLINE_HEARTBEAT_UV: {
436 struct xpc_activate_mq_msg_heartbeat_req_uv *msg;
437
438 msg = container_of(msg_hdr,
439 struct xpc_activate_mq_msg_heartbeat_req_uv,
440 hdr);
441 part_uv->heartbeat = msg->heartbeat;
442
443 spin_lock_irqsave(&part_uv->flags_lock, irq_flags);
444 part_uv->flags |= XPC_P_HEARTBEAT_OFFLINE_UV;
445 spin_unlock_irqrestore(&part_uv->flags_lock, irq_flags);
446 break;
447 }
448 case XPC_ACTIVATE_MQ_MSG_ONLINE_HEARTBEAT_UV: {
449 struct xpc_activate_mq_msg_heartbeat_req_uv *msg;
450
451 msg = container_of(msg_hdr,
452 struct xpc_activate_mq_msg_heartbeat_req_uv,
453 hdr);
454 part_uv->heartbeat = msg->heartbeat;
455
456 spin_lock_irqsave(&part_uv->flags_lock, irq_flags);
457 part_uv->flags &= ~XPC_P_HEARTBEAT_OFFLINE_UV;
458 spin_unlock_irqrestore(&part_uv->flags_lock, irq_flags);
459 break;
460 }
461 case XPC_ACTIVATE_MQ_MSG_ACTIVATE_REQ_UV: { 425 case XPC_ACTIVATE_MQ_MSG_ACTIVATE_REQ_UV: {
462 struct xpc_activate_mq_msg_activate_req_uv *msg; 426 struct xpc_activate_mq_msg_activate_req_uv *msg;
463 427
@@ -475,6 +439,7 @@ xpc_handle_activate_mq_msg_uv(struct xpc_partition *part,
475 part_uv->act_state_req = XPC_P_ASR_ACTIVATE_UV; 439 part_uv->act_state_req = XPC_P_ASR_ACTIVATE_UV;
476 part->remote_rp_pa = msg->rp_gpa; /* !!! _pa is _gpa */ 440 part->remote_rp_pa = msg->rp_gpa; /* !!! _pa is _gpa */
477 part->remote_rp_ts_jiffies = msg_hdr->rp_ts_jiffies; 441 part->remote_rp_ts_jiffies = msg_hdr->rp_ts_jiffies;
442 part_uv->heartbeat_gpa = msg->heartbeat_gpa;
478 443
479 if (msg->activate_gru_mq_desc_gpa != 444 if (msg->activate_gru_mq_desc_gpa !=
480 part_uv->activate_gru_mq_desc_gpa) { 445 part_uv->activate_gru_mq_desc_gpa) {
@@ -569,6 +534,17 @@ xpc_handle_activate_mq_msg_uv(struct xpc_partition *part,
569 xpc_wakeup_channel_mgr(part); 534 xpc_wakeup_channel_mgr(part);
570 break; 535 break;
571 } 536 }
537 case XPC_ACTIVATE_MQ_MSG_CHCTL_OPENCOMPLETE_UV: {
538 struct xpc_activate_mq_msg_chctl_opencomplete_uv *msg;
539
540 msg = container_of(msg_hdr, struct
541 xpc_activate_mq_msg_chctl_opencomplete_uv, hdr);
542 spin_lock_irqsave(&part->chctl_lock, irq_flags);
543 part->chctl.flags[msg->ch_number] |= XPC_CHCTL_OPENCOMPLETE;
544 spin_unlock_irqrestore(&part->chctl_lock, irq_flags);
545
546 xpc_wakeup_channel_mgr(part);
547 }
572 case XPC_ACTIVATE_MQ_MSG_MARK_ENGAGED_UV: 548 case XPC_ACTIVATE_MQ_MSG_MARK_ENGAGED_UV:
573 spin_lock_irqsave(&part_uv->flags_lock, irq_flags); 549 spin_lock_irqsave(&part_uv->flags_lock, irq_flags);
574 part_uv->flags |= XPC_P_ENGAGED_UV; 550 part_uv->flags |= XPC_P_ENGAGED_UV;
@@ -759,7 +735,7 @@ xpc_send_local_activate_IRQ_uv(struct xpc_partition *part, int act_state_req)
759 735
760 /* 736 /*
761 * !!! Make our side think that the remote partition sent an activate 737 * !!! Make our side think that the remote partition sent an activate
762 * !!! message our way by doing what the activate IRQ handler would 738 * !!! mq message our way by doing what the activate IRQ handler would
763 * !!! do had one really been sent. 739 * !!! do had one really been sent.
764 */ 740 */
765 741
@@ -806,90 +782,82 @@ xpc_get_partition_rsvd_page_pa_uv(void *buf, u64 *cookie, unsigned long *rp_pa,
806} 782}
807 783
808static int 784static int
809xpc_setup_rsvd_page_sn_uv(struct xpc_rsvd_page *rp) 785xpc_setup_rsvd_page_uv(struct xpc_rsvd_page *rp)
810{ 786{
811 rp->sn.activate_gru_mq_desc_gpa = 787 xpc_heartbeat_uv =
788 &xpc_partitions[sn_partition_id].sn.uv.cached_heartbeat;
789 rp->sn.uv.heartbeat_gpa = uv_gpa(xpc_heartbeat_uv);
790 rp->sn.uv.activate_gru_mq_desc_gpa =
812 uv_gpa(xpc_activate_mq_uv->gru_mq_desc); 791 uv_gpa(xpc_activate_mq_uv->gru_mq_desc);
813 return 0; 792 return 0;
814} 793}
815 794
816static void 795static void
817xpc_send_heartbeat_uv(int msg_type) 796xpc_allow_hb_uv(short partid)
818{ 797{
819 short partid; 798}
820 struct xpc_partition *part;
821 struct xpc_activate_mq_msg_heartbeat_req_uv msg;
822
823 /*
824 * !!! On uv we're broadcasting a heartbeat message every 5 seconds.
825 * !!! Whereas on sn2 we're bte_copy'ng the heartbeat info every 20
826 * !!! seconds. This is an increase in numalink traffic.
827 * ??? Is this good?
828 */
829
830 msg.heartbeat = atomic64_inc_return(&xpc_heartbeat_uv);
831
832 partid = find_first_bit(xpc_heartbeating_to_mask_uv,
833 XP_MAX_NPARTITIONS_UV);
834
835 while (partid < XP_MAX_NPARTITIONS_UV) {
836 part = &xpc_partitions[partid];
837 799
838 xpc_send_activate_IRQ_part_uv(part, &msg, sizeof(msg), 800static void
839 msg_type); 801xpc_disallow_hb_uv(short partid)
802{
803}
840 804
841 partid = find_next_bit(xpc_heartbeating_to_mask_uv, 805static void
842 XP_MAX_NPARTITIONS_UV, partid + 1); 806xpc_disallow_all_hbs_uv(void)
843 } 807{
844} 808}
845 809
846static void 810static void
847xpc_increment_heartbeat_uv(void) 811xpc_increment_heartbeat_uv(void)
848{ 812{
849 xpc_send_heartbeat_uv(XPC_ACTIVATE_MQ_MSG_INC_HEARTBEAT_UV); 813 xpc_heartbeat_uv->value++;
850} 814}
851 815
852static void 816static void
853xpc_offline_heartbeat_uv(void) 817xpc_offline_heartbeat_uv(void)
854{ 818{
855 xpc_send_heartbeat_uv(XPC_ACTIVATE_MQ_MSG_OFFLINE_HEARTBEAT_UV); 819 xpc_increment_heartbeat_uv();
820 xpc_heartbeat_uv->offline = 1;
856} 821}
857 822
858static void 823static void
859xpc_online_heartbeat_uv(void) 824xpc_online_heartbeat_uv(void)
860{ 825{
861 xpc_send_heartbeat_uv(XPC_ACTIVATE_MQ_MSG_ONLINE_HEARTBEAT_UV); 826 xpc_increment_heartbeat_uv();
827 xpc_heartbeat_uv->offline = 0;
862} 828}
863 829
864static void 830static void
865xpc_heartbeat_init_uv(void) 831xpc_heartbeat_init_uv(void)
866{ 832{
867 atomic64_set(&xpc_heartbeat_uv, 0); 833 xpc_heartbeat_uv->value = 1;
868 bitmap_zero(xpc_heartbeating_to_mask_uv, XP_MAX_NPARTITIONS_UV); 834 xpc_heartbeat_uv->offline = 0;
869 xpc_heartbeating_to_mask = &xpc_heartbeating_to_mask_uv[0];
870} 835}
871 836
872static void 837static void
873xpc_heartbeat_exit_uv(void) 838xpc_heartbeat_exit_uv(void)
874{ 839{
875 xpc_send_heartbeat_uv(XPC_ACTIVATE_MQ_MSG_OFFLINE_HEARTBEAT_UV); 840 xpc_offline_heartbeat_uv();
876} 841}
877 842
878static enum xp_retval 843static enum xp_retval
879xpc_get_remote_heartbeat_uv(struct xpc_partition *part) 844xpc_get_remote_heartbeat_uv(struct xpc_partition *part)
880{ 845{
881 struct xpc_partition_uv *part_uv = &part->sn.uv; 846 struct xpc_partition_uv *part_uv = &part->sn.uv;
882 enum xp_retval ret = xpNoHeartbeat; 847 enum xp_retval ret;
883 848
884 if (part_uv->remote_act_state != XPC_P_AS_INACTIVE && 849 ret = xp_remote_memcpy(uv_gpa(&part_uv->cached_heartbeat),
885 part_uv->remote_act_state != XPC_P_AS_DEACTIVATING) { 850 part_uv->heartbeat_gpa,
851 sizeof(struct xpc_heartbeat_uv));
852 if (ret != xpSuccess)
853 return ret;
886 854
887 if (part_uv->heartbeat != part->last_heartbeat || 855 if (part_uv->cached_heartbeat.value == part->last_heartbeat &&
888 (part_uv->flags & XPC_P_HEARTBEAT_OFFLINE_UV)) { 856 !part_uv->cached_heartbeat.offline) {
889 857
890 part->last_heartbeat = part_uv->heartbeat; 858 ret = xpNoHeartbeat;
891 ret = xpSuccess; 859 } else {
892 } 860 part->last_heartbeat = part_uv->cached_heartbeat.value;
893 } 861 }
894 return ret; 862 return ret;
895} 863}
@@ -904,8 +872,9 @@ xpc_request_partition_activation_uv(struct xpc_rsvd_page *remote_rp,
904 872
905 part->remote_rp_pa = remote_rp_gpa; /* !!! _pa here is really _gpa */ 873 part->remote_rp_pa = remote_rp_gpa; /* !!! _pa here is really _gpa */
906 part->remote_rp_ts_jiffies = remote_rp->ts_jiffies; 874 part->remote_rp_ts_jiffies = remote_rp->ts_jiffies;
875 part->sn.uv.heartbeat_gpa = remote_rp->sn.uv.heartbeat_gpa;
907 part->sn.uv.activate_gru_mq_desc_gpa = 876 part->sn.uv.activate_gru_mq_desc_gpa =
908 remote_rp->sn.activate_gru_mq_desc_gpa; 877 remote_rp->sn.uv.activate_gru_mq_desc_gpa;
909 878
910 /* 879 /*
911 * ??? Is it a good idea to make this conditional on what is 880 * ??? Is it a good idea to make this conditional on what is
@@ -913,8 +882,9 @@ xpc_request_partition_activation_uv(struct xpc_rsvd_page *remote_rp,
913 */ 882 */
914 if (part->sn.uv.remote_act_state == XPC_P_AS_INACTIVE) { 883 if (part->sn.uv.remote_act_state == XPC_P_AS_INACTIVE) {
915 msg.rp_gpa = uv_gpa(xpc_rsvd_page); 884 msg.rp_gpa = uv_gpa(xpc_rsvd_page);
885 msg.heartbeat_gpa = xpc_rsvd_page->sn.uv.heartbeat_gpa;
916 msg.activate_gru_mq_desc_gpa = 886 msg.activate_gru_mq_desc_gpa =
917 xpc_rsvd_page->sn.activate_gru_mq_desc_gpa; 887 xpc_rsvd_page->sn.uv.activate_gru_mq_desc_gpa;
918 xpc_send_activate_IRQ_part_uv(part, &msg, sizeof(msg), 888 xpc_send_activate_IRQ_part_uv(part, &msg, sizeof(msg),
919 XPC_ACTIVATE_MQ_MSG_ACTIVATE_REQ_UV); 889 XPC_ACTIVATE_MQ_MSG_ACTIVATE_REQ_UV);
920 } 890 }
@@ -1010,7 +980,7 @@ xpc_n_of_fifo_entries_uv(struct xpc_fifo_head_uv *head)
1010 * Setup the channel structures that are uv specific. 980 * Setup the channel structures that are uv specific.
1011 */ 981 */
1012static enum xp_retval 982static enum xp_retval
1013xpc_setup_ch_structures_sn_uv(struct xpc_partition *part) 983xpc_setup_ch_structures_uv(struct xpc_partition *part)
1014{ 984{
1015 struct xpc_channel_uv *ch_uv; 985 struct xpc_channel_uv *ch_uv;
1016 int ch_number; 986 int ch_number;
@@ -1029,7 +999,7 @@ xpc_setup_ch_structures_sn_uv(struct xpc_partition *part)
1029 * Teardown the channel structures that are uv specific. 999 * Teardown the channel structures that are uv specific.
1030 */ 1000 */
1031static void 1001static void
1032xpc_teardown_ch_structures_sn_uv(struct xpc_partition *part) 1002xpc_teardown_ch_structures_uv(struct xpc_partition *part)
1033{ 1003{
1034 /* nothing needs to be done */ 1004 /* nothing needs to be done */
1035 return; 1005 return;
@@ -1243,6 +1213,16 @@ xpc_send_chctl_openreply_uv(struct xpc_channel *ch, unsigned long *irq_flags)
1243} 1213}
1244 1214
1245static void 1215static void
1216xpc_send_chctl_opencomplete_uv(struct xpc_channel *ch, unsigned long *irq_flags)
1217{
1218 struct xpc_activate_mq_msg_chctl_opencomplete_uv msg;
1219
1220 msg.ch_number = ch->number;
1221 xpc_send_activate_IRQ_ch_uv(ch, irq_flags, &msg, sizeof(msg),
1222 XPC_ACTIVATE_MQ_MSG_CHCTL_OPENCOMPLETE_UV);
1223}
1224
1225static void
1246xpc_send_chctl_local_msgrequest_uv(struct xpc_partition *part, int ch_number) 1226xpc_send_chctl_local_msgrequest_uv(struct xpc_partition *part, int ch_number)
1247{ 1227{
1248 unsigned long irq_flags; 1228 unsigned long irq_flags;
@@ -1669,58 +1649,67 @@ xpc_received_payload_uv(struct xpc_channel *ch, void *payload)
1669 msg->hdr.msg_slot_number += ch->remote_nentries; 1649 msg->hdr.msg_slot_number += ch->remote_nentries;
1670} 1650}
1671 1651
1652static struct xpc_arch_operations xpc_arch_ops_uv = {
1653 .setup_partitions = xpc_setup_partitions_uv,
1654 .teardown_partitions = xpc_teardown_partitions_uv,
1655 .process_activate_IRQ_rcvd = xpc_process_activate_IRQ_rcvd_uv,
1656 .get_partition_rsvd_page_pa = xpc_get_partition_rsvd_page_pa_uv,
1657 .setup_rsvd_page = xpc_setup_rsvd_page_uv,
1658
1659 .allow_hb = xpc_allow_hb_uv,
1660 .disallow_hb = xpc_disallow_hb_uv,
1661 .disallow_all_hbs = xpc_disallow_all_hbs_uv,
1662 .increment_heartbeat = xpc_increment_heartbeat_uv,
1663 .offline_heartbeat = xpc_offline_heartbeat_uv,
1664 .online_heartbeat = xpc_online_heartbeat_uv,
1665 .heartbeat_init = xpc_heartbeat_init_uv,
1666 .heartbeat_exit = xpc_heartbeat_exit_uv,
1667 .get_remote_heartbeat = xpc_get_remote_heartbeat_uv,
1668
1669 .request_partition_activation =
1670 xpc_request_partition_activation_uv,
1671 .request_partition_reactivation =
1672 xpc_request_partition_reactivation_uv,
1673 .request_partition_deactivation =
1674 xpc_request_partition_deactivation_uv,
1675 .cancel_partition_deactivation_request =
1676 xpc_cancel_partition_deactivation_request_uv,
1677
1678 .setup_ch_structures = xpc_setup_ch_structures_uv,
1679 .teardown_ch_structures = xpc_teardown_ch_structures_uv,
1680
1681 .make_first_contact = xpc_make_first_contact_uv,
1682
1683 .get_chctl_all_flags = xpc_get_chctl_all_flags_uv,
1684 .send_chctl_closerequest = xpc_send_chctl_closerequest_uv,
1685 .send_chctl_closereply = xpc_send_chctl_closereply_uv,
1686 .send_chctl_openrequest = xpc_send_chctl_openrequest_uv,
1687 .send_chctl_openreply = xpc_send_chctl_openreply_uv,
1688 .send_chctl_opencomplete = xpc_send_chctl_opencomplete_uv,
1689 .process_msg_chctl_flags = xpc_process_msg_chctl_flags_uv,
1690
1691 .save_remote_msgqueue_pa = xpc_save_remote_msgqueue_pa_uv,
1692
1693 .setup_msg_structures = xpc_setup_msg_structures_uv,
1694 .teardown_msg_structures = xpc_teardown_msg_structures_uv,
1695
1696 .indicate_partition_engaged = xpc_indicate_partition_engaged_uv,
1697 .indicate_partition_disengaged = xpc_indicate_partition_disengaged_uv,
1698 .assume_partition_disengaged = xpc_assume_partition_disengaged_uv,
1699 .partition_engaged = xpc_partition_engaged_uv,
1700 .any_partition_engaged = xpc_any_partition_engaged_uv,
1701
1702 .n_of_deliverable_payloads = xpc_n_of_deliverable_payloads_uv,
1703 .send_payload = xpc_send_payload_uv,
1704 .get_deliverable_payload = xpc_get_deliverable_payload_uv,
1705 .received_payload = xpc_received_payload_uv,
1706 .notify_senders_of_disconnect = xpc_notify_senders_of_disconnect_uv,
1707};
1708
1672int 1709int
1673xpc_init_uv(void) 1710xpc_init_uv(void)
1674{ 1711{
1675 xpc_setup_partitions_sn = xpc_setup_partitions_sn_uv; 1712 xpc_arch_ops = xpc_arch_ops_uv;
1676 xpc_teardown_partitions_sn = xpc_teardown_partitions_sn_uv;
1677 xpc_process_activate_IRQ_rcvd = xpc_process_activate_IRQ_rcvd_uv;
1678 xpc_get_partition_rsvd_page_pa = xpc_get_partition_rsvd_page_pa_uv;
1679 xpc_setup_rsvd_page_sn = xpc_setup_rsvd_page_sn_uv;
1680 xpc_increment_heartbeat = xpc_increment_heartbeat_uv;
1681 xpc_offline_heartbeat = xpc_offline_heartbeat_uv;
1682 xpc_online_heartbeat = xpc_online_heartbeat_uv;
1683 xpc_heartbeat_init = xpc_heartbeat_init_uv;
1684 xpc_heartbeat_exit = xpc_heartbeat_exit_uv;
1685 xpc_get_remote_heartbeat = xpc_get_remote_heartbeat_uv;
1686
1687 xpc_request_partition_activation = xpc_request_partition_activation_uv;
1688 xpc_request_partition_reactivation =
1689 xpc_request_partition_reactivation_uv;
1690 xpc_request_partition_deactivation =
1691 xpc_request_partition_deactivation_uv;
1692 xpc_cancel_partition_deactivation_request =
1693 xpc_cancel_partition_deactivation_request_uv;
1694
1695 xpc_setup_ch_structures_sn = xpc_setup_ch_structures_sn_uv;
1696 xpc_teardown_ch_structures_sn = xpc_teardown_ch_structures_sn_uv;
1697
1698 xpc_make_first_contact = xpc_make_first_contact_uv;
1699
1700 xpc_get_chctl_all_flags = xpc_get_chctl_all_flags_uv;
1701 xpc_send_chctl_closerequest = xpc_send_chctl_closerequest_uv;
1702 xpc_send_chctl_closereply = xpc_send_chctl_closereply_uv;
1703 xpc_send_chctl_openrequest = xpc_send_chctl_openrequest_uv;
1704 xpc_send_chctl_openreply = xpc_send_chctl_openreply_uv;
1705
1706 xpc_save_remote_msgqueue_pa = xpc_save_remote_msgqueue_pa_uv;
1707
1708 xpc_setup_msg_structures = xpc_setup_msg_structures_uv;
1709 xpc_teardown_msg_structures = xpc_teardown_msg_structures_uv;
1710
1711 xpc_indicate_partition_engaged = xpc_indicate_partition_engaged_uv;
1712 xpc_indicate_partition_disengaged =
1713 xpc_indicate_partition_disengaged_uv;
1714 xpc_assume_partition_disengaged = xpc_assume_partition_disengaged_uv;
1715 xpc_partition_engaged = xpc_partition_engaged_uv;
1716 xpc_any_partition_engaged = xpc_any_partition_engaged_uv;
1717
1718 xpc_n_of_deliverable_payloads = xpc_n_of_deliverable_payloads_uv;
1719 xpc_process_msg_chctl_flags = xpc_process_msg_chctl_flags_uv;
1720 xpc_send_payload = xpc_send_payload_uv;
1721 xpc_notify_senders_of_disconnect = xpc_notify_senders_of_disconnect_uv;
1722 xpc_get_deliverable_payload = xpc_get_deliverable_payload_uv;
1723 xpc_received_payload = xpc_received_payload_uv;
1724 1713
1725 if (sizeof(struct xpc_notify_mq_msghdr_uv) > XPC_MSG_HDR_MAX_SIZE) { 1714 if (sizeof(struct xpc_notify_mq_msghdr_uv) > XPC_MSG_HDR_MAX_SIZE) {
1726 dev_err(xpc_part, "xpc_notify_mq_msghdr_uv is larger than %d\n", 1715 dev_err(xpc_part, "xpc_notify_mq_msghdr_uv is larger than %d\n",
diff --git a/drivers/net/a2065.c b/drivers/net/a2065.c
index d0d0c2fee054..02f64d578641 100644
--- a/drivers/net/a2065.c
+++ b/drivers/net/a2065.c
@@ -692,6 +692,17 @@ static struct zorro_driver a2065_driver = {
692 .remove = __devexit_p(a2065_remove_one), 692 .remove = __devexit_p(a2065_remove_one),
693}; 693};
694 694
695static const struct net_device_ops lance_netdev_ops = {
696 .ndo_open = lance_open,
697 .ndo_stop = lance_close,
698 .ndo_start_xmit = lance_start_xmit,
699 .ndo_tx_timeout = lance_tx_timeout,
700 .ndo_set_multicast_list = lance_set_multicast,
701 .ndo_validate_addr = eth_validate_addr,
702 .ndo_change_mtu = eth_change_mtu,
703 .ndo_set_mac_address = eth_mac_addr,
704};
705
695static int __devinit a2065_init_one(struct zorro_dev *z, 706static int __devinit a2065_init_one(struct zorro_dev *z,
696 const struct zorro_device_id *ent) 707 const struct zorro_device_id *ent)
697{ 708{
@@ -753,12 +764,8 @@ static int __devinit a2065_init_one(struct zorro_dev *z,
753 priv->rx_ring_mod_mask = RX_RING_MOD_MASK; 764 priv->rx_ring_mod_mask = RX_RING_MOD_MASK;
754 priv->tx_ring_mod_mask = TX_RING_MOD_MASK; 765 priv->tx_ring_mod_mask = TX_RING_MOD_MASK;
755 766
756 dev->open = &lance_open; 767 dev->netdev_ops = &lance_netdev_ops;
757 dev->stop = &lance_close;
758 dev->hard_start_xmit = &lance_start_xmit;
759 dev->tx_timeout = &lance_tx_timeout;
760 dev->watchdog_timeo = 5*HZ; 768 dev->watchdog_timeo = 5*HZ;
761 dev->set_multicast_list = &lance_set_multicast;
762 dev->dma = 0; 769 dev->dma = 0;
763 770
764 init_timer(&priv->multicast_timer); 771 init_timer(&priv->multicast_timer);
diff --git a/drivers/net/ariadne.c b/drivers/net/ariadne.c
index e1d72e06f3e1..58e8d522e5bc 100644
--- a/drivers/net/ariadne.c
+++ b/drivers/net/ariadne.c
@@ -155,6 +155,18 @@ static struct zorro_driver ariadne_driver = {
155 .remove = __devexit_p(ariadne_remove_one), 155 .remove = __devexit_p(ariadne_remove_one),
156}; 156};
157 157
158static const struct net_device_ops ariadne_netdev_ops = {
159 .ndo_open = ariadne_open,
160 .ndo_stop = ariadne_close,
161 .ndo_start_xmit = ariadne_start_xmit,
162 .ndo_tx_timeout = ariadne_tx_timeout,
163 .ndo_get_stats = ariadne_get_stats,
164 .ndo_set_multicast_list = set_multicast_list,
165 .ndo_validate_addr = eth_validate_addr,
166 .ndo_change_mtu = eth_change_mtu,
167 .ndo_set_mac_address = eth_mac_addr,
168};
169
158static int __devinit ariadne_init_one(struct zorro_dev *z, 170static int __devinit ariadne_init_one(struct zorro_dev *z,
159 const struct zorro_device_id *ent) 171 const struct zorro_device_id *ent)
160{ 172{
@@ -197,13 +209,8 @@ static int __devinit ariadne_init_one(struct zorro_dev *z,
197 dev->mem_start = ZTWO_VADDR(mem_start); 209 dev->mem_start = ZTWO_VADDR(mem_start);
198 dev->mem_end = dev->mem_start+ARIADNE_RAM_SIZE; 210 dev->mem_end = dev->mem_start+ARIADNE_RAM_SIZE;
199 211
200 dev->open = &ariadne_open; 212 dev->netdev_ops = &ariadne_netdev_ops;
201 dev->stop = &ariadne_close;
202 dev->hard_start_xmit = &ariadne_start_xmit;
203 dev->tx_timeout = &ariadne_tx_timeout;
204 dev->watchdog_timeo = 5*HZ; 213 dev->watchdog_timeo = 5*HZ;
205 dev->get_stats = &ariadne_get_stats;
206 dev->set_multicast_list = &set_multicast_list;
207 214
208 err = register_netdev(dev); 215 err = register_netdev(dev);
209 if (err) { 216 if (err) {
diff --git a/drivers/net/arm/am79c961a.c b/drivers/net/arm/am79c961a.c
index 4bc6901b3819..627bc75da17d 100644
--- a/drivers/net/arm/am79c961a.c
+++ b/drivers/net/arm/am79c961a.c
@@ -665,6 +665,20 @@ static void __init am79c961_banner(void)
665 if (net_debug && version_printed++ == 0) 665 if (net_debug && version_printed++ == 0)
666 printk(KERN_INFO "%s", version); 666 printk(KERN_INFO "%s", version);
667} 667}
668static const struct net_device_ops am79c961_netdev_ops = {
669 .ndo_open = am79c961_open,
670 .ndo_stop = am79c961_close,
671 .ndo_start_xmit = am79c961_sendpacket,
672 .ndo_get_stats = am79c961_getstats,
673 .ndo_set_multicast_list = am79c961_setmulticastlist,
674 .ndo_tx_timeout = am79c961_timeout,
675 .ndo_validate_addr = eth_validate_addr,
676 .ndo_change_mtu = eth_change_mtu,
677 .ndo_set_mac_address = eth_mac_addr,
678#ifdef CONFIG_NET_POLL_CONTROLLER
679 .ndo_poll_controller = am79c961_poll_controller,
680#endif
681};
668 682
669static int __init am79c961_probe(struct platform_device *pdev) 683static int __init am79c961_probe(struct platform_device *pdev)
670{ 684{
@@ -732,15 +746,7 @@ static int __init am79c961_probe(struct platform_device *pdev)
732 if (am79c961_hw_init(dev)) 746 if (am79c961_hw_init(dev))
733 goto release; 747 goto release;
734 748
735 dev->open = am79c961_open; 749 dev->netdev_ops = &am79c961_netdev_ops;
736 dev->stop = am79c961_close;
737 dev->hard_start_xmit = am79c961_sendpacket;
738 dev->get_stats = am79c961_getstats;
739 dev->set_multicast_list = am79c961_setmulticastlist;
740 dev->tx_timeout = am79c961_timeout;
741#ifdef CONFIG_NET_POLL_CONTROLLER
742 dev->poll_controller = am79c961_poll_controller;
743#endif
744 750
745 ret = register_netdev(dev); 751 ret = register_netdev(dev);
746 if (ret == 0) { 752 if (ret == 0) {
diff --git a/drivers/net/arm/at91_ether.c b/drivers/net/arm/at91_ether.c
index 442938d50380..7f4bc8ae5462 100644
--- a/drivers/net/arm/at91_ether.c
+++ b/drivers/net/arm/at91_ether.c
@@ -577,7 +577,7 @@ static void at91ether_sethashtable(struct net_device *dev)
577/* 577/*
578 * Enable/Disable promiscuous and multicast modes. 578 * Enable/Disable promiscuous and multicast modes.
579 */ 579 */
580static void at91ether_set_rx_mode(struct net_device *dev) 580static void at91ether_set_multicast_list(struct net_device *dev)
581{ 581{
582 unsigned long cfg; 582 unsigned long cfg;
583 583
@@ -808,7 +808,7 @@ static int at91ether_close(struct net_device *dev)
808/* 808/*
809 * Transmit packet. 809 * Transmit packet.
810 */ 810 */
811static int at91ether_tx(struct sk_buff *skb, struct net_device *dev) 811static int at91ether_start_xmit(struct sk_buff *skb, struct net_device *dev)
812{ 812{
813 struct at91_private *lp = netdev_priv(dev); 813 struct at91_private *lp = netdev_priv(dev);
814 814
@@ -828,7 +828,7 @@ static int at91ether_tx(struct sk_buff *skb, struct net_device *dev)
828 828
829 dev->trans_start = jiffies; 829 dev->trans_start = jiffies;
830 } else { 830 } else {
831 printk(KERN_ERR "at91_ether.c: at91ether_tx() called, but device is busy!\n"); 831 printk(KERN_ERR "at91_ether.c: at91ether_start_xmit() called, but device is busy!\n");
832 return 1; /* if we return anything but zero, dev.c:1055 calls kfree_skb(skb) 832 return 1; /* if we return anything but zero, dev.c:1055 calls kfree_skb(skb)
833 on this skb, he also reports -ENETDOWN and printk's, so either 833 on this skb, he also reports -ENETDOWN and printk's, so either
834 we free and return(0) or don't free and return 1 */ 834 we free and return(0) or don't free and return 1 */
@@ -965,6 +965,21 @@ static void at91ether_poll_controller(struct net_device *dev)
965} 965}
966#endif 966#endif
967 967
968static const struct net_device_ops at91ether_netdev_ops = {
969 .ndo_open = at91ether_open,
970 .ndo_stop = at91ether_close,
971 .ndo_start_xmit = at91ether_start_xmit,
972 .ndo_get_stats = at91ether_stats,
973 .ndo_set_multicast_list = at91ether_set_multicast_list,
974 .ndo_set_mac_address = set_mac_address,
975 .ndo_do_ioctl = at91ether_ioctl,
976 .ndo_validate_addr = eth_validate_addr,
977 .ndo_change_mtu = eth_change_mtu,
978#ifdef CONFIG_NET_POLL_CONTROLLER
979 .ndo_poll_controller = at91ether_poll_controller,
980#endif
981};
982
968/* 983/*
969 * Initialize the ethernet interface 984 * Initialize the ethernet interface
970 */ 985 */
@@ -1005,17 +1020,8 @@ static int __init at91ether_setup(unsigned long phy_type, unsigned short phy_add
1005 spin_lock_init(&lp->lock); 1020 spin_lock_init(&lp->lock);
1006 1021
1007 ether_setup(dev); 1022 ether_setup(dev);
1008 dev->open = at91ether_open; 1023 dev->netdev_ops = &at91ether_netdev_ops;
1009 dev->stop = at91ether_close;
1010 dev->hard_start_xmit = at91ether_tx;
1011 dev->get_stats = at91ether_stats;
1012 dev->set_multicast_list = at91ether_set_rx_mode;
1013 dev->set_mac_address = set_mac_address;
1014 dev->ethtool_ops = &at91ether_ethtool_ops; 1024 dev->ethtool_ops = &at91ether_ethtool_ops;
1015 dev->do_ioctl = at91ether_ioctl;
1016#ifdef CONFIG_NET_POLL_CONTROLLER
1017 dev->poll_controller = at91ether_poll_controller;
1018#endif
1019 1025
1020 SET_NETDEV_DEV(dev, &pdev->dev); 1026 SET_NETDEV_DEV(dev, &pdev->dev);
1021 1027
diff --git a/drivers/net/arm/ep93xx_eth.c b/drivers/net/arm/ep93xx_eth.c
index cc7708775da0..41736772c1dd 100644
--- a/drivers/net/arm/ep93xx_eth.c
+++ b/drivers/net/arm/ep93xx_eth.c
@@ -770,7 +770,18 @@ static struct ethtool_ops ep93xx_ethtool_ops = {
770 .get_link = ep93xx_get_link, 770 .get_link = ep93xx_get_link,
771}; 771};
772 772
773struct net_device *ep93xx_dev_alloc(struct ep93xx_eth_data *data) 773static const struct net_device_ops ep93xx_netdev_ops = {
774 .ndo_open = ep93xx_open,
775 .ndo_stop = ep93xx_close,
776 .ndo_start_xmit = ep93xx_xmit,
777 .ndo_get_stats = ep93xx_get_stats,
778 .ndo_do_ioctl = ep93xx_ioctl,
779 .ndo_validate_addr = eth_validate_addr,
780 .ndo_change_mtu = eth_change_mtu,
781 .ndo_set_mac_address = eth_mac_addr,
782};
783
784static struct net_device *ep93xx_dev_alloc(struct ep93xx_eth_data *data)
774{ 785{
775 struct net_device *dev; 786 struct net_device *dev;
776 787
@@ -780,12 +791,8 @@ struct net_device *ep93xx_dev_alloc(struct ep93xx_eth_data *data)
780 791
781 memcpy(dev->dev_addr, data->dev_addr, ETH_ALEN); 792 memcpy(dev->dev_addr, data->dev_addr, ETH_ALEN);
782 793
783 dev->get_stats = ep93xx_get_stats;
784 dev->ethtool_ops = &ep93xx_ethtool_ops; 794 dev->ethtool_ops = &ep93xx_ethtool_ops;
785 dev->hard_start_xmit = ep93xx_xmit; 795 dev->netdev_ops = &ep93xx_netdev_ops;
786 dev->open = ep93xx_open;
787 dev->stop = ep93xx_close;
788 dev->do_ioctl = ep93xx_ioctl;
789 796
790 dev->features |= NETIF_F_SG | NETIF_F_HW_CSUM; 797 dev->features |= NETIF_F_SG | NETIF_F_HW_CSUM;
791 798
diff --git a/drivers/net/arm/ether1.c b/drivers/net/arm/ether1.c
index e380de454463..edf770f639fa 100644
--- a/drivers/net/arm/ether1.c
+++ b/drivers/net/arm/ether1.c
@@ -991,6 +991,18 @@ static void __devinit ether1_banner(void)
991 printk(KERN_INFO "%s", version); 991 printk(KERN_INFO "%s", version);
992} 992}
993 993
994static const struct net_device_ops ether1_netdev_ops = {
995 .ndo_open = ether1_open,
996 .ndo_stop = ether1_close,
997 .ndo_start_xmit = ether1_sendpacket,
998 .ndo_get_stats = ether1_getstats,
999 .ndo_set_multicast_list = ether1_setmulticastlist,
1000 .ndo_tx_timeout = ether1_timeout,
1001 .ndo_validate_addr = eth_validate_addr,
1002 .ndo_change_mtu = eth_change_mtu,
1003 .ndo_set_mac_address = eth_mac_addr,
1004};
1005
994static int __devinit 1006static int __devinit
995ether1_probe(struct expansion_card *ec, const struct ecard_id *id) 1007ether1_probe(struct expansion_card *ec, const struct ecard_id *id)
996{ 1008{
@@ -1031,12 +1043,7 @@ ether1_probe(struct expansion_card *ec, const struct ecard_id *id)
1031 goto free; 1043 goto free;
1032 } 1044 }
1033 1045
1034 dev->open = ether1_open; 1046 dev->netdev_ops = &ether1_netdev_ops;
1035 dev->stop = ether1_close;
1036 dev->hard_start_xmit = ether1_sendpacket;
1037 dev->get_stats = ether1_getstats;
1038 dev->set_multicast_list = ether1_setmulticastlist;
1039 dev->tx_timeout = ether1_timeout;
1040 dev->watchdog_timeo = 5 * HZ / 100; 1047 dev->watchdog_timeo = 5 * HZ / 100;
1041 1048
1042 ret = register_netdev(dev); 1049 ret = register_netdev(dev);
diff --git a/drivers/net/arm/ether3.c b/drivers/net/arm/ether3.c
index 21a7bef12d3b..ec8a1ae1e887 100644
--- a/drivers/net/arm/ether3.c
+++ b/drivers/net/arm/ether3.c
@@ -770,6 +770,18 @@ static void __devinit ether3_banner(void)
770 printk(KERN_INFO "%s", version); 770 printk(KERN_INFO "%s", version);
771} 771}
772 772
773static const struct net_device_ops ether3_netdev_ops = {
774 .ndo_open = ether3_open,
775 .ndo_stop = ether3_close,
776 .ndo_start_xmit = ether3_sendpacket,
777 .ndo_get_stats = ether3_getstats,
778 .ndo_set_multicast_list = ether3_setmulticastlist,
779 .ndo_tx_timeout = ether3_timeout,
780 .ndo_validate_addr = eth_validate_addr,
781 .ndo_change_mtu = eth_change_mtu,
782 .ndo_set_mac_address = eth_mac_addr,
783};
784
773static int __devinit 785static int __devinit
774ether3_probe(struct expansion_card *ec, const struct ecard_id *id) 786ether3_probe(struct expansion_card *ec, const struct ecard_id *id)
775{ 787{
@@ -846,12 +858,7 @@ ether3_probe(struct expansion_card *ec, const struct ecard_id *id)
846 goto free; 858 goto free;
847 } 859 }
848 860
849 dev->open = ether3_open; 861 dev->netdev_ops = &ether3_netdev_ops;
850 dev->stop = ether3_close;
851 dev->hard_start_xmit = ether3_sendpacket;
852 dev->get_stats = ether3_getstats;
853 dev->set_multicast_list = ether3_setmulticastlist;
854 dev->tx_timeout = ether3_timeout;
855 dev->watchdog_timeo = 5 * HZ / 100; 862 dev->watchdog_timeo = 5 * HZ / 100;
856 863
857 ret = register_netdev(dev); 864 ret = register_netdev(dev);
diff --git a/drivers/net/atarilance.c b/drivers/net/atarilance.c
index 2d81f6afcb58..5425ab0c38c0 100644
--- a/drivers/net/atarilance.c
+++ b/drivers/net/atarilance.c
@@ -453,6 +453,16 @@ static noinline int __init addr_accessible(volatile void *regp, int wordflag,
453 return( ret ); 453 return( ret );
454} 454}
455 455
456static const struct net_device_ops lance_netdev_ops = {
457 .ndo_open = lance_open,
458 .ndo_stop = lance_close,
459 .ndo_start_xmit = lance_start_xmit,
460 .ndo_set_multicast_list = set_multicast_list,
461 .ndo_set_mac_address = lance_set_mac_address,
462 .ndo_tx_timeout = lance_tx_timeout,
463 .ndo_validate_addr = eth_validate_addr,
464 .ndo_change_mtu = eth_change_mtu,
465};
456 466
457static unsigned long __init lance_probe1( struct net_device *dev, 467static unsigned long __init lance_probe1( struct net_device *dev,
458 struct lance_addr *init_rec ) 468 struct lance_addr *init_rec )
@@ -623,15 +633,9 @@ static unsigned long __init lance_probe1( struct net_device *dev,
623 if (did_version++ == 0) 633 if (did_version++ == 0)
624 DPRINTK( 1, ( version )); 634 DPRINTK( 1, ( version ));
625 635
626 /* The LANCE-specific entries in the device structure. */ 636 dev->netdev_ops = &lance_netdev_ops;
627 dev->open = &lance_open;
628 dev->hard_start_xmit = &lance_start_xmit;
629 dev->stop = &lance_close;
630 dev->set_multicast_list = &set_multicast_list;
631 dev->set_mac_address = &lance_set_mac_address;
632 637
633 /* XXX MSch */ 638 /* XXX MSch */
634 dev->tx_timeout = lance_tx_timeout;
635 dev->watchdog_timeo = TX_TIMEOUT; 639 dev->watchdog_timeo = TX_TIMEOUT;
636 640
637 return( 1 ); 641 return( 1 );
diff --git a/drivers/net/atl1c/atl1c_main.c b/drivers/net/atl1c/atl1c_main.c
index deb7b53167ee..83a12125b94e 100644
--- a/drivers/net/atl1c/atl1c_main.c
+++ b/drivers/net/atl1c/atl1c_main.c
@@ -2532,8 +2532,8 @@ static int __devinit atl1c_probe(struct pci_dev *pdev,
2532 * various kernel subsystems to support the mechanics required by a 2532 * various kernel subsystems to support the mechanics required by a
2533 * fixed-high-32-bit system. 2533 * fixed-high-32-bit system.
2534 */ 2534 */
2535 if ((pci_set_dma_mask(pdev, DMA_32BIT_MASK) != 0) || 2535 if ((pci_set_dma_mask(pdev, DMA_BIT_MASK(32)) != 0) ||
2536 (pci_set_consistent_dma_mask(pdev, DMA_32BIT_MASK) != 0)) { 2536 (pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32)) != 0)) {
2537 dev_err(&pdev->dev, "No usable DMA configuration,aborting\n"); 2537 dev_err(&pdev->dev, "No usable DMA configuration,aborting\n");
2538 goto err_dma; 2538 goto err_dma;
2539 } 2539 }
diff --git a/drivers/net/au1000_eth.c b/drivers/net/au1000_eth.c
index 4274e4ac963b..d58c105fc779 100644
--- a/drivers/net/au1000_eth.c
+++ b/drivers/net/au1000_eth.c
@@ -1004,12 +1004,12 @@ static void au1000_tx_timeout(struct net_device *dev)
1004 netif_wake_queue(dev); 1004 netif_wake_queue(dev);
1005} 1005}
1006 1006
1007static void set_rx_mode(struct net_device *dev) 1007static void au1000_multicast_list(struct net_device *dev)
1008{ 1008{
1009 struct au1000_private *aup = netdev_priv(dev); 1009 struct au1000_private *aup = netdev_priv(dev);
1010 1010
1011 if (au1000_debug > 4) 1011 if (au1000_debug > 4)
1012 printk("%s: set_rx_mode: flags=%x\n", dev->name, dev->flags); 1012 printk("%s: au1000_multicast_list: flags=%x\n", dev->name, dev->flags);
1013 1013
1014 if (dev->flags & IFF_PROMISC) { /* Set promiscuous. */ 1014 if (dev->flags & IFF_PROMISC) { /* Set promiscuous. */
1015 aup->mac->control |= MAC_PROMISCUOUS; 1015 aup->mac->control |= MAC_PROMISCUOUS;
@@ -1047,6 +1047,18 @@ static int au1000_ioctl(struct net_device *dev, struct ifreq *rq, int cmd)
1047 return phy_mii_ioctl(aup->phy_dev, if_mii(rq), cmd); 1047 return phy_mii_ioctl(aup->phy_dev, if_mii(rq), cmd);
1048} 1048}
1049 1049
1050static const struct net_device_ops au1000_netdev_ops = {
1051 .ndo_open = au1000_open,
1052 .ndo_stop = au1000_close,
1053 .ndo_start_xmit = au1000_tx,
1054 .ndo_set_multicast_list = au1000_multicast_list,
1055 .ndo_do_ioctl = au1000_ioctl,
1056 .ndo_tx_timeout = au1000_tx_timeout,
1057 .ndo_set_mac_address = eth_mac_addr,
1058 .ndo_validate_addr = eth_validate_addr,
1059 .ndo_change_mtu = eth_change_mtu,
1060};
1061
1050static struct net_device * au1000_probe(int port_num) 1062static struct net_device * au1000_probe(int port_num)
1051{ 1063{
1052 static unsigned version_printed = 0; 1064 static unsigned version_printed = 0;
@@ -1197,13 +1209,8 @@ static struct net_device * au1000_probe(int port_num)
1197 1209
1198 dev->base_addr = base; 1210 dev->base_addr = base;
1199 dev->irq = irq; 1211 dev->irq = irq;
1200 dev->open = au1000_open; 1212 dev->netdev_ops = &au1000_netdev_ops;
1201 dev->hard_start_xmit = au1000_tx;
1202 dev->stop = au1000_close;
1203 dev->set_multicast_list = &set_rx_mode;
1204 dev->do_ioctl = &au1000_ioctl;
1205 SET_ETHTOOL_OPS(dev, &au1000_ethtool_ops); 1213 SET_ETHTOOL_OPS(dev, &au1000_ethtool_ops);
1206 dev->tx_timeout = au1000_tx_timeout;
1207 dev->watchdog_timeo = ETH_TX_TIMEOUT; 1214 dev->watchdog_timeo = ETH_TX_TIMEOUT;
1208 1215
1209 /* 1216 /*
diff --git a/drivers/net/benet/be_ethtool.c b/drivers/net/benet/be_ethtool.c
index 04f4b73fa8d8..9592f22e4c8c 100644
--- a/drivers/net/benet/be_ethtool.c
+++ b/drivers/net/benet/be_ethtool.c
@@ -319,7 +319,7 @@ be_get_pauseparam(struct net_device *netdev, struct ethtool_pauseparam *ecmd)
319 319
320 be_cmd_get_flow_control(&adapter->ctrl, &ecmd->tx_pause, 320 be_cmd_get_flow_control(&adapter->ctrl, &ecmd->tx_pause,
321 &ecmd->rx_pause); 321 &ecmd->rx_pause);
322 ecmd->autoneg = AUTONEG_ENABLE; 322 ecmd->autoneg = 0;
323} 323}
324 324
325static int 325static int
@@ -328,7 +328,7 @@ be_set_pauseparam(struct net_device *netdev, struct ethtool_pauseparam *ecmd)
328 struct be_adapter *adapter = netdev_priv(netdev); 328 struct be_adapter *adapter = netdev_priv(netdev);
329 int status; 329 int status;
330 330
331 if (ecmd->autoneg != AUTONEG_ENABLE) 331 if (ecmd->autoneg != 0)
332 return -EINVAL; 332 return -EINVAL;
333 333
334 status = be_cmd_set_flow_control(&adapter->ctrl, ecmd->tx_pause, 334 status = be_cmd_set_flow_control(&adapter->ctrl, ecmd->tx_pause,
diff --git a/drivers/net/benet/be_main.c b/drivers/net/benet/be_main.c
index 9b75aa630062..30d0c81c989e 100644
--- a/drivers/net/benet/be_main.c
+++ b/drivers/net/benet/be_main.c
@@ -1821,11 +1821,11 @@ static int __devinit be_probe(struct pci_dev *pdev,
1821 1821
1822 be_msix_enable(adapter); 1822 be_msix_enable(adapter);
1823 1823
1824 status = pci_set_dma_mask(pdev, DMA_64BIT_MASK); 1824 status = pci_set_dma_mask(pdev, DMA_BIT_MASK(64));
1825 if (!status) { 1825 if (!status) {
1826 netdev->features |= NETIF_F_HIGHDMA; 1826 netdev->features |= NETIF_F_HIGHDMA;
1827 } else { 1827 } else {
1828 status = pci_set_dma_mask(pdev, DMA_32BIT_MASK); 1828 status = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
1829 if (status) { 1829 if (status) {
1830 dev_err(&pdev->dev, "Could not set PCI DMA Mask\n"); 1830 dev_err(&pdev->dev, "Could not set PCI DMA Mask\n");
1831 goto free_netdev; 1831 goto free_netdev;
diff --git a/drivers/net/bfin_mac.c b/drivers/net/bfin_mac.c
index 9afe8092dfc4..9f971ed6b58d 100644
--- a/drivers/net/bfin_mac.c
+++ b/drivers/net/bfin_mac.c
@@ -979,6 +979,20 @@ static int bfin_mac_open(struct net_device *dev)
979 return 0; 979 return 0;
980} 980}
981 981
982static const struct net_device_ops bfin_mac_netdev_ops = {
983 .ndo_open = bfin_mac_open,
984 .ndo_stop = bfin_mac_close,
985 .ndo_start_xmit = bfin_mac_hard_start_xmit,
986 .ndo_set_mac_address = bfin_mac_set_mac_address,
987 .ndo_tx_timeout = bfin_mac_timeout,
988 .ndo_set_multicast_list = bfin_mac_set_multicast_list,
989 .ndo_validate_addr = eth_validate_addr,
990 .ndo_change_mtu = eth_change_mtu,
991#ifdef CONFIG_NET_POLL_CONTROLLER
992 .ndo_poll_controller = bfin_mac_poll,
993#endif
994};
995
982/* 996/*
983 * 997 *
984 * this makes the board clean up everything that it can 998 * this makes the board clean up everything that it can
@@ -1086,15 +1100,7 @@ static int __devinit bfin_mac_probe(struct platform_device *pdev)
1086 /* Fill in the fields of the device structure with ethernet values. */ 1100 /* Fill in the fields of the device structure with ethernet values. */
1087 ether_setup(ndev); 1101 ether_setup(ndev);
1088 1102
1089 ndev->open = bfin_mac_open; 1103 ndev->netdev_ops = &bfin_mac_netdev_ops;
1090 ndev->stop = bfin_mac_close;
1091 ndev->hard_start_xmit = bfin_mac_hard_start_xmit;
1092 ndev->set_mac_address = bfin_mac_set_mac_address;
1093 ndev->tx_timeout = bfin_mac_timeout;
1094 ndev->set_multicast_list = bfin_mac_set_multicast_list;
1095#ifdef CONFIG_NET_POLL_CONTROLLER
1096 ndev->poll_controller = bfin_mac_poll;
1097#endif
1098 ndev->ethtool_ops = &bfin_mac_ethtool_ops; 1104 ndev->ethtool_ops = &bfin_mac_ethtool_ops;
1099 1105
1100 spin_lock_init(&lp->lock); 1106 spin_lock_init(&lp->lock);
diff --git a/drivers/net/bonding/bond_alb.c b/drivers/net/bonding/bond_alb.c
index 8dc6fbb9a41e..553a89919778 100644
--- a/drivers/net/bonding/bond_alb.c
+++ b/drivers/net/bonding/bond_alb.c
@@ -370,8 +370,6 @@ static int rlb_arp_recv(struct sk_buff *skb, struct net_device *bond_dev, struct
370 370
371 if (arp->op_code == htons(ARPOP_REPLY)) { 371 if (arp->op_code == htons(ARPOP_REPLY)) {
372 /* update rx hash table for this ARP */ 372 /* update rx hash table for this ARP */
373 printk("rar: update orig %s bond_dev %s\n", orig_dev->name,
374 bond_dev->name);
375 bond = netdev_priv(bond_dev); 373 bond = netdev_priv(bond_dev);
376 rlb_update_entry_from_arp(bond, arp); 374 rlb_update_entry_from_arp(bond, arp);
377 pr_debug("Server received an ARP Reply from client\n"); 375 pr_debug("Server received an ARP Reply from client\n");
diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c
index 99610f358c40..63369b6b14d4 100644
--- a/drivers/net/bonding/bond_main.c
+++ b/drivers/net/bonding/bond_main.c
@@ -2570,7 +2570,7 @@ static void bond_arp_send_all(struct bonding *bond, struct slave *slave)
2570 2570
2571 for (i = 0; (i < BOND_MAX_ARP_TARGETS); i++) { 2571 for (i = 0; (i < BOND_MAX_ARP_TARGETS); i++) {
2572 if (!targets[i]) 2572 if (!targets[i])
2573 continue; 2573 break;
2574 pr_debug("basa: target %x\n", targets[i]); 2574 pr_debug("basa: target %x\n", targets[i]);
2575 if (list_empty(&bond->vlan_list)) { 2575 if (list_empty(&bond->vlan_list)) {
2576 pr_debug("basa: empty vlan: arp_send\n"); 2576 pr_debug("basa: empty vlan: arp_send\n");
@@ -2677,7 +2677,6 @@ static void bond_validate_arp(struct bonding *bond, struct slave *slave, __be32
2677 int i; 2677 int i;
2678 __be32 *targets = bond->params.arp_targets; 2678 __be32 *targets = bond->params.arp_targets;
2679 2679
2680 targets = bond->params.arp_targets;
2681 for (i = 0; (i < BOND_MAX_ARP_TARGETS) && targets[i]; i++) { 2680 for (i = 0; (i < BOND_MAX_ARP_TARGETS) && targets[i]; i++) {
2682 pr_debug("bva: sip %pI4 tip %pI4 t[%d] %pI4 bhti(tip) %d\n", 2681 pr_debug("bva: sip %pI4 tip %pI4 t[%d] %pI4 bhti(tip) %d\n",
2683 &sip, &tip, i, &targets[i], bond_has_this_ip(bond, tip)); 2682 &sip, &tip, i, &targets[i], bond_has_this_ip(bond, tip));
@@ -3303,7 +3302,7 @@ static void bond_info_show_master(struct seq_file *seq)
3303 3302
3304 for(i = 0; (i < BOND_MAX_ARP_TARGETS) ;i++) { 3303 for(i = 0; (i < BOND_MAX_ARP_TARGETS) ;i++) {
3305 if (!bond->params.arp_targets[i]) 3304 if (!bond->params.arp_targets[i])
3306 continue; 3305 break;
3307 if (printed) 3306 if (printed)
3308 seq_printf(seq, ","); 3307 seq_printf(seq, ",");
3309 seq_printf(seq, " %pI4", &bond->params.arp_targets[i]); 3308 seq_printf(seq, " %pI4", &bond->params.arp_targets[i]);
diff --git a/drivers/net/bonding/bond_sysfs.c b/drivers/net/bonding/bond_sysfs.c
index 18cf4787874c..d28731535226 100644
--- a/drivers/net/bonding/bond_sysfs.c
+++ b/drivers/net/bonding/bond_sysfs.c
@@ -684,17 +684,15 @@ static ssize_t bonding_store_arp_targets(struct device *d,
684 goto out; 684 goto out;
685 } 685 }
686 /* look for an empty slot to put the target in, and check for dupes */ 686 /* look for an empty slot to put the target in, and check for dupes */
687 for (i = 0; (i < BOND_MAX_ARP_TARGETS); i++) { 687 for (i = 0; (i < BOND_MAX_ARP_TARGETS) && !done; i++) {
688 if (targets[i] == newtarget) { /* duplicate */ 688 if (targets[i] == newtarget) { /* duplicate */
689 printk(KERN_ERR DRV_NAME 689 printk(KERN_ERR DRV_NAME
690 ": %s: ARP target %pI4 is already present\n", 690 ": %s: ARP target %pI4 is already present\n",
691 bond->dev->name, &newtarget); 691 bond->dev->name, &newtarget);
692 if (done)
693 targets[i] = 0;
694 ret = -EINVAL; 692 ret = -EINVAL;
695 goto out; 693 goto out;
696 } 694 }
697 if (targets[i] == 0 && !done) { 695 if (targets[i] == 0) {
698 printk(KERN_INFO DRV_NAME 696 printk(KERN_INFO DRV_NAME
699 ": %s: adding ARP target %pI4.\n", 697 ": %s: adding ARP target %pI4.\n",
700 bond->dev->name, &newtarget); 698 bond->dev->name, &newtarget);
@@ -720,12 +718,16 @@ static ssize_t bonding_store_arp_targets(struct device *d,
720 goto out; 718 goto out;
721 } 719 }
722 720
723 for (i = 0; (i < BOND_MAX_ARP_TARGETS); i++) { 721 for (i = 0; (i < BOND_MAX_ARP_TARGETS) && !done; i++) {
724 if (targets[i] == newtarget) { 722 if (targets[i] == newtarget) {
723 int j;
725 printk(KERN_INFO DRV_NAME 724 printk(KERN_INFO DRV_NAME
726 ": %s: removing ARP target %pI4.\n", 725 ": %s: removing ARP target %pI4.\n",
727 bond->dev->name, &newtarget); 726 bond->dev->name, &newtarget);
728 targets[i] = 0; 727 for (j = i; (j < (BOND_MAX_ARP_TARGETS-1)) && targets[j+1]; j++)
728 targets[j] = targets[j+1];
729
730 targets[j] = 0;
729 done = 1; 731 done = 1;
730 } 732 }
731 } 733 }
diff --git a/drivers/net/cris/eth_v10.c b/drivers/net/cris/eth_v10.c
index c9806c58b2fd..7a18dc7e5c7f 100644
--- a/drivers/net/cris/eth_v10.c
+++ b/drivers/net/cris/eth_v10.c
@@ -257,6 +257,23 @@ struct transceiver_ops transceivers[] =
257 257
258struct transceiver_ops* transceiver = &transceivers[0]; 258struct transceiver_ops* transceiver = &transceivers[0];
259 259
260static const struct net_device_ops e100_netdev_ops = {
261 .ndo_open = e100_open,
262 .ndo_stop = e100_close,
263 .ndo_start_xmit = e100_send_packet,
264 .ndo_tx_timeout = e100_tx_timeout,
265 .ndo_get_stats = e100_get_stats,
266 .ndo_set_multicast_list = set_multicast_list,
267 .ndo_do_ioctl = e100_ioctl,
268 .ndo_set_mac_address = e100_set_mac_address,
269 .ndo_validate_addr = eth_validate_addr,
270 .ndo_change_mtu = eth_change_mtu,
271 .ndo_set_config = e100_set_config,
272#ifdef CONFIG_NET_POLL_CONTROLLER
273 .ndo_poll_controller = e100_netpoll,
274#endif
275};
276
260#define tx_done(dev) (*R_DMA_CH0_CMD == 0) 277#define tx_done(dev) (*R_DMA_CH0_CMD == 0)
261 278
262/* 279/*
@@ -300,19 +317,8 @@ etrax_ethernet_init(void)
300 317
301 /* fill in our handlers so the network layer can talk to us in the future */ 318 /* fill in our handlers so the network layer can talk to us in the future */
302 319
303 dev->open = e100_open;
304 dev->hard_start_xmit = e100_send_packet;
305 dev->stop = e100_close;
306 dev->get_stats = e100_get_stats;
307 dev->set_multicast_list = set_multicast_list;
308 dev->set_mac_address = e100_set_mac_address;
309 dev->ethtool_ops = &e100_ethtool_ops; 320 dev->ethtool_ops = &e100_ethtool_ops;
310 dev->do_ioctl = e100_ioctl; 321 dev->netdev_ops = &e100_netdev_ops;
311 dev->set_config = e100_set_config;
312 dev->tx_timeout = e100_tx_timeout;
313#ifdef CONFIG_NET_POLL_CONTROLLER
314 dev->poll_controller = e100_netpoll;
315#endif
316 322
317 spin_lock_init(&np->lock); 323 spin_lock_init(&np->lock);
318 spin_lock_init(&np->led_lock); 324 spin_lock_init(&np->led_lock);
diff --git a/drivers/net/declance.c b/drivers/net/declance.c
index 861c867fca87..b62405a69180 100644
--- a/drivers/net/declance.c
+++ b/drivers/net/declance.c
@@ -1010,6 +1010,17 @@ static void lance_set_multicast_retry(unsigned long _opaque)
1010 lance_set_multicast(dev); 1010 lance_set_multicast(dev);
1011} 1011}
1012 1012
1013static const struct net_device_ops lance_netdev_ops = {
1014 .ndo_open = lance_open,
1015 .ndo_stop = lance_close,
1016 .ndo_start_xmit = lance_start_xmit,
1017 .ndo_tx_timeout = lance_tx_timeout,
1018 .ndo_set_multicast_list = lance_set_multicast,
1019 .ndo_change_mtu = eth_change_mtu,
1020 .ndo_validate_addr = eth_validate_addr,
1021 .ndo_set_mac_address = eth_mac_addr,
1022};
1023
1013static int __init dec_lance_probe(struct device *bdev, const int type) 1024static int __init dec_lance_probe(struct device *bdev, const int type)
1014{ 1025{
1015 static unsigned version_printed; 1026 static unsigned version_printed;
@@ -1223,12 +1234,8 @@ static int __init dec_lance_probe(struct device *bdev, const int type)
1223 1234
1224 printk(", addr = %pM, irq = %d\n", dev->dev_addr, dev->irq); 1235 printk(", addr = %pM, irq = %d\n", dev->dev_addr, dev->irq);
1225 1236
1226 dev->open = &lance_open; 1237 dev->netdev_ops = &lance_netdev_ops;
1227 dev->stop = &lance_close;
1228 dev->hard_start_xmit = &lance_start_xmit;
1229 dev->tx_timeout = &lance_tx_timeout;
1230 dev->watchdog_timeo = 5*HZ; 1238 dev->watchdog_timeo = 5*HZ;
1231 dev->set_multicast_list = &lance_set_multicast;
1232 1239
1233 /* lp->ll is the location of the registers for lance card */ 1240 /* lp->ll is the location of the registers for lance card */
1234 lp->ll = ll; 1241 lp->ll = ll;
diff --git a/drivers/net/e1000/e1000_main.c b/drivers/net/e1000/e1000_main.c
index ddc5c533e89c..ef12931d302a 100644
--- a/drivers/net/e1000/e1000_main.c
+++ b/drivers/net/e1000/e1000_main.c
@@ -156,8 +156,8 @@ static void e1000_vlan_rx_add_vid(struct net_device *netdev, u16 vid);
156static void e1000_vlan_rx_kill_vid(struct net_device *netdev, u16 vid); 156static void e1000_vlan_rx_kill_vid(struct net_device *netdev, u16 vid);
157static void e1000_restore_vlan(struct e1000_adapter *adapter); 157static void e1000_restore_vlan(struct e1000_adapter *adapter);
158 158
159static int e1000_suspend(struct pci_dev *pdev, pm_message_t state);
160#ifdef CONFIG_PM 159#ifdef CONFIG_PM
160static int e1000_suspend(struct pci_dev *pdev, pm_message_t state);
161static int e1000_resume(struct pci_dev *pdev); 161static int e1000_resume(struct pci_dev *pdev);
162#endif 162#endif
163static void e1000_shutdown(struct pci_dev *pdev); 163static void e1000_shutdown(struct pci_dev *pdev);
@@ -3834,7 +3834,7 @@ static bool e1000_clean_tx_irq(struct e1000_adapter *adapter,
3834 struct e1000_buffer *buffer_info; 3834 struct e1000_buffer *buffer_info;
3835 unsigned int i, eop; 3835 unsigned int i, eop;
3836 unsigned int count = 0; 3836 unsigned int count = 0;
3837 bool cleaned; 3837 bool cleaned = false;
3838 unsigned int total_tx_bytes=0, total_tx_packets=0; 3838 unsigned int total_tx_bytes=0, total_tx_packets=0;
3839 3839
3840 i = tx_ring->next_to_clean; 3840 i = tx_ring->next_to_clean;
@@ -4601,7 +4601,7 @@ int e1000_set_spd_dplx(struct e1000_adapter *adapter, u16 spddplx)
4601 return 0; 4601 return 0;
4602} 4602}
4603 4603
4604static int e1000_suspend(struct pci_dev *pdev, pm_message_t state) 4604static int __e1000_shutdown(struct pci_dev *pdev, bool *enable_wake)
4605{ 4605{
4606 struct net_device *netdev = pci_get_drvdata(pdev); 4606 struct net_device *netdev = pci_get_drvdata(pdev);
4607 struct e1000_adapter *adapter = netdev_priv(netdev); 4607 struct e1000_adapter *adapter = netdev_priv(netdev);
@@ -4664,22 +4664,18 @@ static int e1000_suspend(struct pci_dev *pdev, pm_message_t state)
4664 4664
4665 ew32(WUC, E1000_WUC_PME_EN); 4665 ew32(WUC, E1000_WUC_PME_EN);
4666 ew32(WUFC, wufc); 4666 ew32(WUFC, wufc);
4667 pci_enable_wake(pdev, PCI_D3hot, 1);
4668 pci_enable_wake(pdev, PCI_D3cold, 1);
4669 } else { 4667 } else {
4670 ew32(WUC, 0); 4668 ew32(WUC, 0);
4671 ew32(WUFC, 0); 4669 ew32(WUFC, 0);
4672 pci_enable_wake(pdev, PCI_D3hot, 0);
4673 pci_enable_wake(pdev, PCI_D3cold, 0);
4674 } 4670 }
4675 4671
4676 e1000_release_manageability(adapter); 4672 e1000_release_manageability(adapter);
4677 4673
4674 *enable_wake = !!wufc;
4675
4678 /* make sure adapter isn't asleep if manageability is enabled */ 4676 /* make sure adapter isn't asleep if manageability is enabled */
4679 if (adapter->en_mng_pt) { 4677 if (adapter->en_mng_pt)
4680 pci_enable_wake(pdev, PCI_D3hot, 1); 4678 *enable_wake = true;
4681 pci_enable_wake(pdev, PCI_D3cold, 1);
4682 }
4683 4679
4684 if (hw->phy_type == e1000_phy_igp_3) 4680 if (hw->phy_type == e1000_phy_igp_3)
4685 e1000_phy_powerdown_workaround(hw); 4681 e1000_phy_powerdown_workaround(hw);
@@ -4693,12 +4689,29 @@ static int e1000_suspend(struct pci_dev *pdev, pm_message_t state)
4693 4689
4694 pci_disable_device(pdev); 4690 pci_disable_device(pdev);
4695 4691
4696 pci_set_power_state(pdev, pci_choose_state(pdev, state));
4697
4698 return 0; 4692 return 0;
4699} 4693}
4700 4694
4701#ifdef CONFIG_PM 4695#ifdef CONFIG_PM
4696static int e1000_suspend(struct pci_dev *pdev, pm_message_t state)
4697{
4698 int retval;
4699 bool wake;
4700
4701 retval = __e1000_shutdown(pdev, &wake);
4702 if (retval)
4703 return retval;
4704
4705 if (wake) {
4706 pci_prepare_to_sleep(pdev);
4707 } else {
4708 pci_wake_from_d3(pdev, false);
4709 pci_set_power_state(pdev, PCI_D3hot);
4710 }
4711
4712 return 0;
4713}
4714
4702static int e1000_resume(struct pci_dev *pdev) 4715static int e1000_resume(struct pci_dev *pdev)
4703{ 4716{
4704 struct net_device *netdev = pci_get_drvdata(pdev); 4717 struct net_device *netdev = pci_get_drvdata(pdev);
@@ -4753,7 +4766,14 @@ static int e1000_resume(struct pci_dev *pdev)
4753 4766
4754static void e1000_shutdown(struct pci_dev *pdev) 4767static void e1000_shutdown(struct pci_dev *pdev)
4755{ 4768{
4756 e1000_suspend(pdev, PMSG_SUSPEND); 4769 bool wake;
4770
4771 __e1000_shutdown(pdev, &wake);
4772
4773 if (system_state == SYSTEM_POWER_OFF) {
4774 pci_wake_from_d3(pdev, wake);
4775 pci_set_power_state(pdev, PCI_D3hot);
4776 }
4757} 4777}
4758 4778
4759#ifdef CONFIG_NET_POLL_CONTROLLER 4779#ifdef CONFIG_NET_POLL_CONTROLLER
diff --git a/drivers/net/e1000e/netdev.c b/drivers/net/e1000e/netdev.c
index 409b58cad0e5..1693ed116b16 100644
--- a/drivers/net/e1000e/netdev.c
+++ b/drivers/net/e1000e/netdev.c
@@ -621,7 +621,7 @@ static bool e1000_clean_tx_irq(struct e1000_adapter *adapter)
621 struct e1000_buffer *buffer_info; 621 struct e1000_buffer *buffer_info;
622 unsigned int i, eop; 622 unsigned int i, eop;
623 unsigned int count = 0; 623 unsigned int count = 0;
624 bool cleaned; 624 bool cleaned = false;
625 unsigned int total_tx_bytes = 0, total_tx_packets = 0; 625 unsigned int total_tx_bytes = 0, total_tx_packets = 0;
626 626
627 i = tx_ring->next_to_clean; 627 i = tx_ring->next_to_clean;
@@ -4346,7 +4346,7 @@ static int e1000_ioctl(struct net_device *netdev, struct ifreq *ifr, int cmd)
4346 } 4346 }
4347} 4347}
4348 4348
4349static int e1000_suspend(struct pci_dev *pdev, pm_message_t state) 4349static int __e1000_shutdown(struct pci_dev *pdev, bool *enable_wake)
4350{ 4350{
4351 struct net_device *netdev = pci_get_drvdata(pdev); 4351 struct net_device *netdev = pci_get_drvdata(pdev);
4352 struct e1000_adapter *adapter = netdev_priv(netdev); 4352 struct e1000_adapter *adapter = netdev_priv(netdev);
@@ -4409,20 +4409,16 @@ static int e1000_suspend(struct pci_dev *pdev, pm_message_t state)
4409 4409
4410 ew32(WUC, E1000_WUC_PME_EN); 4410 ew32(WUC, E1000_WUC_PME_EN);
4411 ew32(WUFC, wufc); 4411 ew32(WUFC, wufc);
4412 pci_enable_wake(pdev, PCI_D3hot, 1);
4413 pci_enable_wake(pdev, PCI_D3cold, 1);
4414 } else { 4412 } else {
4415 ew32(WUC, 0); 4413 ew32(WUC, 0);
4416 ew32(WUFC, 0); 4414 ew32(WUFC, 0);
4417 pci_enable_wake(pdev, PCI_D3hot, 0);
4418 pci_enable_wake(pdev, PCI_D3cold, 0);
4419 } 4415 }
4420 4416
4417 *enable_wake = !!wufc;
4418
4421 /* make sure adapter isn't asleep if manageability is enabled */ 4419 /* make sure adapter isn't asleep if manageability is enabled */
4422 if (adapter->flags & FLAG_MNG_PT_ENABLED) { 4420 if (adapter->flags & FLAG_MNG_PT_ENABLED)
4423 pci_enable_wake(pdev, PCI_D3hot, 1); 4421 *enable_wake = true;
4424 pci_enable_wake(pdev, PCI_D3cold, 1);
4425 }
4426 4422
4427 if (adapter->hw.phy.type == e1000_phy_igp_3) 4423 if (adapter->hw.phy.type == e1000_phy_igp_3)
4428 e1000e_igp3_phy_powerdown_workaround_ich8lan(&adapter->hw); 4424 e1000e_igp3_phy_powerdown_workaround_ich8lan(&adapter->hw);
@@ -4435,6 +4431,26 @@ static int e1000_suspend(struct pci_dev *pdev, pm_message_t state)
4435 4431
4436 pci_disable_device(pdev); 4432 pci_disable_device(pdev);
4437 4433
4434 return 0;
4435}
4436
4437static void e1000_power_off(struct pci_dev *pdev, bool sleep, bool wake)
4438{
4439 if (sleep && wake) {
4440 pci_prepare_to_sleep(pdev);
4441 return;
4442 }
4443
4444 pci_wake_from_d3(pdev, wake);
4445 pci_set_power_state(pdev, PCI_D3hot);
4446}
4447
4448static void e1000_complete_shutdown(struct pci_dev *pdev, bool sleep,
4449 bool wake)
4450{
4451 struct net_device *netdev = pci_get_drvdata(pdev);
4452 struct e1000_adapter *adapter = netdev_priv(netdev);
4453
4438 /* 4454 /*
4439 * The pci-e switch on some quad port adapters will report a 4455 * The pci-e switch on some quad port adapters will report a
4440 * correctable error when the MAC transitions from D0 to D3. To 4456 * correctable error when the MAC transitions from D0 to D3. To
@@ -4450,14 +4466,12 @@ static int e1000_suspend(struct pci_dev *pdev, pm_message_t state)
4450 pci_write_config_word(us_dev, pos + PCI_EXP_DEVCTL, 4466 pci_write_config_word(us_dev, pos + PCI_EXP_DEVCTL,
4451 (devctl & ~PCI_EXP_DEVCTL_CERE)); 4467 (devctl & ~PCI_EXP_DEVCTL_CERE));
4452 4468
4453 pci_set_power_state(pdev, pci_choose_state(pdev, state)); 4469 e1000_power_off(pdev, sleep, wake);
4454 4470
4455 pci_write_config_word(us_dev, pos + PCI_EXP_DEVCTL, devctl); 4471 pci_write_config_word(us_dev, pos + PCI_EXP_DEVCTL, devctl);
4456 } else { 4472 } else {
4457 pci_set_power_state(pdev, pci_choose_state(pdev, state)); 4473 e1000_power_off(pdev, sleep, wake);
4458 } 4474 }
4459
4460 return 0;
4461} 4475}
4462 4476
4463static void e1000e_disable_l1aspm(struct pci_dev *pdev) 4477static void e1000e_disable_l1aspm(struct pci_dev *pdev)
@@ -4486,6 +4500,18 @@ static void e1000e_disable_l1aspm(struct pci_dev *pdev)
4486} 4500}
4487 4501
4488#ifdef CONFIG_PM 4502#ifdef CONFIG_PM
4503static int e1000_suspend(struct pci_dev *pdev, pm_message_t state)
4504{
4505 int retval;
4506 bool wake;
4507
4508 retval = __e1000_shutdown(pdev, &wake);
4509 if (!retval)
4510 e1000_complete_shutdown(pdev, true, wake);
4511
4512 return retval;
4513}
4514
4489static int e1000_resume(struct pci_dev *pdev) 4515static int e1000_resume(struct pci_dev *pdev)
4490{ 4516{
4491 struct net_device *netdev = pci_get_drvdata(pdev); 4517 struct net_device *netdev = pci_get_drvdata(pdev);
@@ -4549,7 +4575,12 @@ static int e1000_resume(struct pci_dev *pdev)
4549 4575
4550static void e1000_shutdown(struct pci_dev *pdev) 4576static void e1000_shutdown(struct pci_dev *pdev)
4551{ 4577{
4552 e1000_suspend(pdev, PMSG_SUSPEND); 4578 bool wake = false;
4579
4580 __e1000_shutdown(pdev, &wake);
4581
4582 if (system_state == SYSTEM_POWER_OFF)
4583 e1000_complete_shutdown(pdev, false, wake);
4553} 4584}
4554 4585
4555#ifdef CONFIG_NET_POLL_CONTROLLER 4586#ifdef CONFIG_NET_POLL_CONTROLLER
diff --git a/drivers/net/ehea/ehea_main.c b/drivers/net/ehea/ehea_main.c
index ac0c5b438e0a..604c844d0769 100644
--- a/drivers/net/ehea/ehea_main.c
+++ b/drivers/net/ehea/ehea_main.c
@@ -3080,7 +3080,8 @@ static const struct net_device_ops ehea_netdev_ops = {
3080 .ndo_change_mtu = ehea_change_mtu, 3080 .ndo_change_mtu = ehea_change_mtu,
3081 .ndo_vlan_rx_register = ehea_vlan_rx_register, 3081 .ndo_vlan_rx_register = ehea_vlan_rx_register,
3082 .ndo_vlan_rx_add_vid = ehea_vlan_rx_add_vid, 3082 .ndo_vlan_rx_add_vid = ehea_vlan_rx_add_vid,
3083 .ndo_vlan_rx_kill_vid = ehea_vlan_rx_kill_vid 3083 .ndo_vlan_rx_kill_vid = ehea_vlan_rx_kill_vid,
3084 .ndo_tx_timeout = ehea_tx_watchdog,
3084}; 3085};
3085 3086
3086struct ehea_port *ehea_setup_single_port(struct ehea_adapter *adapter, 3087struct ehea_port *ehea_setup_single_port(struct ehea_adapter *adapter,
@@ -3142,7 +3143,6 @@ struct ehea_port *ehea_setup_single_port(struct ehea_adapter *adapter,
3142 | NETIF_F_HIGHDMA | NETIF_F_IP_CSUM | NETIF_F_HW_VLAN_TX 3143 | NETIF_F_HIGHDMA | NETIF_F_IP_CSUM | NETIF_F_HW_VLAN_TX
3143 | NETIF_F_HW_VLAN_RX | NETIF_F_HW_VLAN_FILTER 3144 | NETIF_F_HW_VLAN_RX | NETIF_F_HW_VLAN_FILTER
3144 | NETIF_F_LLTX; 3145 | NETIF_F_LLTX;
3145 dev->tx_timeout = &ehea_tx_watchdog;
3146 dev->watchdog_timeo = EHEA_WATCH_DOG_TIMEOUT; 3146 dev->watchdog_timeo = EHEA_WATCH_DOG_TIMEOUT;
3147 3147
3148 INIT_WORK(&port->reset_task, ehea_reset_port); 3148 INIT_WORK(&port->reset_task, ehea_reset_port);
diff --git a/drivers/net/forcedeth.c b/drivers/net/forcedeth.c
index d37465020bcc..11d5db16ed9c 100644
--- a/drivers/net/forcedeth.c
+++ b/drivers/net/forcedeth.c
@@ -3745,14 +3745,14 @@ static int nv_napi_poll(struct napi_struct *napi, int budget)
3745 mod_timer(&np->nic_poll, jiffies + POLL_WAIT); 3745 mod_timer(&np->nic_poll, jiffies + POLL_WAIT);
3746 } 3746 }
3747 spin_unlock_irqrestore(&np->lock, flags); 3747 spin_unlock_irqrestore(&np->lock, flags);
3748 __napi_complete(napi); 3748 napi_complete(napi);
3749 return rx_work; 3749 return rx_work;
3750 } 3750 }
3751 3751
3752 if (rx_work < budget) { 3752 if (rx_work < budget) {
3753 /* re-enable interrupts 3753 /* re-enable interrupts
3754 (msix not enabled in napi) */ 3754 (msix not enabled in napi) */
3755 __napi_complete(napi); 3755 napi_complete(napi);
3756 3756
3757 writel(np->irqmask, base + NvRegIrqMask); 3757 writel(np->irqmask, base + NvRegIrqMask);
3758 } 3758 }
diff --git a/drivers/net/fs_enet/fs_enet-main.c b/drivers/net/fs_enet/fs_enet-main.c
index b037ce9857bf..a9cbc3191a2a 100644
--- a/drivers/net/fs_enet/fs_enet-main.c
+++ b/drivers/net/fs_enet/fs_enet-main.c
@@ -1019,6 +1019,22 @@ out_put_phy:
1019#define IS_FEC(match) 0 1019#define IS_FEC(match) 0
1020#endif 1020#endif
1021 1021
1022static const struct net_device_ops fs_enet_netdev_ops = {
1023 .ndo_open = fs_enet_open,
1024 .ndo_stop = fs_enet_close,
1025 .ndo_get_stats = fs_enet_get_stats,
1026 .ndo_start_xmit = fs_enet_start_xmit,
1027 .ndo_tx_timeout = fs_timeout,
1028 .ndo_set_multicast_list = fs_set_multicast_list,
1029 .ndo_do_ioctl = fs_ioctl,
1030 .ndo_validate_addr = eth_validate_addr,
1031 .ndo_set_mac_address = eth_mac_addr,
1032 .ndo_change_mtu = eth_change_mtu,
1033#ifdef CONFIG_NET_POLL_CONTROLLER
1034 .ndo_poll_controller = fs_enet_netpoll,
1035#endif
1036};
1037
1022static int __devinit fs_enet_probe(struct of_device *ofdev, 1038static int __devinit fs_enet_probe(struct of_device *ofdev,
1023 const struct of_device_id *match) 1039 const struct of_device_id *match)
1024{ 1040{
@@ -1093,22 +1109,13 @@ static int __devinit fs_enet_probe(struct of_device *ofdev,
1093 fep->tx_ring = fpi->tx_ring; 1109 fep->tx_ring = fpi->tx_ring;
1094 fep->rx_ring = fpi->rx_ring; 1110 fep->rx_ring = fpi->rx_ring;
1095 1111
1096 ndev->open = fs_enet_open; 1112 ndev->netdev_ops = &fs_enet_netdev_ops;
1097 ndev->hard_start_xmit = fs_enet_start_xmit;
1098 ndev->tx_timeout = fs_timeout;
1099 ndev->watchdog_timeo = 2 * HZ; 1113 ndev->watchdog_timeo = 2 * HZ;
1100 ndev->stop = fs_enet_close;
1101 ndev->get_stats = fs_enet_get_stats;
1102 ndev->set_multicast_list = fs_set_multicast_list;
1103#ifdef CONFIG_NET_POLL_CONTROLLER
1104 ndev->poll_controller = fs_enet_netpoll;
1105#endif
1106 if (fpi->use_napi) 1114 if (fpi->use_napi)
1107 netif_napi_add(ndev, &fep->napi, fs_enet_rx_napi, 1115 netif_napi_add(ndev, &fep->napi, fs_enet_rx_napi,
1108 fpi->napi_weight); 1116 fpi->napi_weight);
1109 1117
1110 ndev->ethtool_ops = &fs_ethtool_ops; 1118 ndev->ethtool_ops = &fs_ethtool_ops;
1111 ndev->do_ioctl = fs_ioctl;
1112 1119
1113 init_timer(&fep->phy_timer_list); 1120 init_timer(&fep->phy_timer_list);
1114 1121
diff --git a/drivers/net/gianfar.c b/drivers/net/gianfar.c
index 65f55877be95..b2c49679bba7 100644
--- a/drivers/net/gianfar.c
+++ b/drivers/net/gianfar.c
@@ -1583,8 +1583,10 @@ static void gfar_reset_task(struct work_struct *work)
1583 struct net_device *dev = priv->ndev; 1583 struct net_device *dev = priv->ndev;
1584 1584
1585 if (dev->flags & IFF_UP) { 1585 if (dev->flags & IFF_UP) {
1586 netif_stop_queue(dev);
1586 stop_gfar(dev); 1587 stop_gfar(dev);
1587 startup_gfar(dev); 1588 startup_gfar(dev);
1589 netif_start_queue(dev);
1588 } 1590 }
1589 1591
1590 netif_tx_schedule_all(dev); 1592 netif_tx_schedule_all(dev);
diff --git a/drivers/net/ibm_newemac/core.c b/drivers/net/ibm_newemac/core.c
index 77e4b5b52fc8..806533c831c7 100644
--- a/drivers/net/ibm_newemac/core.c
+++ b/drivers/net/ibm_newemac/core.c
@@ -2686,6 +2686,32 @@ static int __devinit emac_init_config(struct emac_instance *dev)
2686 return 0; 2686 return 0;
2687} 2687}
2688 2688
2689static const struct net_device_ops emac_netdev_ops = {
2690 .ndo_open = emac_open,
2691 .ndo_stop = emac_close,
2692 .ndo_get_stats = emac_stats,
2693 .ndo_set_multicast_list = emac_set_multicast_list,
2694 .ndo_do_ioctl = emac_ioctl,
2695 .ndo_tx_timeout = emac_tx_timeout,
2696 .ndo_validate_addr = eth_validate_addr,
2697 .ndo_set_mac_address = eth_mac_addr,
2698 .ndo_start_xmit = emac_start_xmit,
2699 .ndo_change_mtu = eth_change_mtu,
2700};
2701
2702static const struct net_device_ops emac_gige_netdev_ops = {
2703 .ndo_open = emac_open,
2704 .ndo_stop = emac_close,
2705 .ndo_get_stats = emac_stats,
2706 .ndo_set_multicast_list = emac_set_multicast_list,
2707 .ndo_do_ioctl = emac_ioctl,
2708 .ndo_tx_timeout = emac_tx_timeout,
2709 .ndo_validate_addr = eth_validate_addr,
2710 .ndo_set_mac_address = eth_mac_addr,
2711 .ndo_start_xmit = emac_start_xmit_sg,
2712 .ndo_change_mtu = emac_change_mtu,
2713};
2714
2689static int __devinit emac_probe(struct of_device *ofdev, 2715static int __devinit emac_probe(struct of_device *ofdev,
2690 const struct of_device_id *match) 2716 const struct of_device_id *match)
2691{ 2717{
@@ -2827,23 +2853,14 @@ static int __devinit emac_probe(struct of_device *ofdev,
2827 if (err != 0) 2853 if (err != 0)
2828 goto err_detach_tah; 2854 goto err_detach_tah;
2829 2855
2830 /* Fill in the driver function table */
2831 ndev->open = &emac_open;
2832 if (dev->tah_dev) 2856 if (dev->tah_dev)
2833 ndev->features |= NETIF_F_IP_CSUM | NETIF_F_SG; 2857 ndev->features |= NETIF_F_IP_CSUM | NETIF_F_SG;
2834 ndev->tx_timeout = &emac_tx_timeout;
2835 ndev->watchdog_timeo = 5 * HZ; 2858 ndev->watchdog_timeo = 5 * HZ;
2836 ndev->stop = &emac_close;
2837 ndev->get_stats = &emac_stats;
2838 ndev->set_multicast_list = &emac_set_multicast_list;
2839 ndev->do_ioctl = &emac_ioctl;
2840 if (emac_phy_supports_gige(dev->phy_mode)) { 2859 if (emac_phy_supports_gige(dev->phy_mode)) {
2841 ndev->hard_start_xmit = &emac_start_xmit_sg; 2860 ndev->netdev_ops = &emac_gige_netdev_ops;
2842 ndev->change_mtu = &emac_change_mtu;
2843 dev->commac.ops = &emac_commac_sg_ops; 2861 dev->commac.ops = &emac_commac_sg_ops;
2844 } else { 2862 } else
2845 ndev->hard_start_xmit = &emac_start_xmit; 2863 ndev->netdev_ops = &emac_netdev_ops;
2846 }
2847 SET_ETHTOOL_OPS(ndev, &emac_ethtool_ops); 2864 SET_ETHTOOL_OPS(ndev, &emac_ethtool_ops);
2848 2865
2849 netif_carrier_off(ndev); 2866 netif_carrier_off(ndev);
diff --git a/drivers/net/igb/e1000_mac.c b/drivers/net/igb/e1000_mac.c
index f4c315b5a900..472f3f124840 100644
--- a/drivers/net/igb/e1000_mac.c
+++ b/drivers/net/igb/e1000_mac.c
@@ -111,7 +111,7 @@ void igb_clear_vfta(struct e1000_hw *hw)
111 * Writes value at the given offset in the register array which stores 111 * Writes value at the given offset in the register array which stores
112 * the VLAN filter table. 112 * the VLAN filter table.
113 **/ 113 **/
114void igb_write_vfta(struct e1000_hw *hw, u32 offset, u32 value) 114static void igb_write_vfta(struct e1000_hw *hw, u32 offset, u32 value)
115{ 115{
116 array_wr32(E1000_VFTA, offset, value); 116 array_wr32(E1000_VFTA, offset, value);
117 wrfl(); 117 wrfl();
diff --git a/drivers/net/igb/e1000_mac.h b/drivers/net/igb/e1000_mac.h
index a34de5269637..1d690b4c9ae4 100644
--- a/drivers/net/igb/e1000_mac.h
+++ b/drivers/net/igb/e1000_mac.h
@@ -66,7 +66,6 @@ void igb_rar_set(struct e1000_hw *hw, u8 *addr, u32 index);
66s32 igb_check_alt_mac_addr(struct e1000_hw *hw); 66s32 igb_check_alt_mac_addr(struct e1000_hw *hw);
67void igb_reset_adaptive(struct e1000_hw *hw); 67void igb_reset_adaptive(struct e1000_hw *hw);
68void igb_update_adaptive(struct e1000_hw *hw); 68void igb_update_adaptive(struct e1000_hw *hw);
69void igb_write_vfta(struct e1000_hw *hw, u32 offset, u32 value);
70 69
71bool igb_enable_mng_pass_thru(struct e1000_hw *hw); 70bool igb_enable_mng_pass_thru(struct e1000_hw *hw);
72 71
diff --git a/drivers/net/igb/e1000_mbx.c b/drivers/net/igb/e1000_mbx.c
index fe71c7ddaa05..840782fb5736 100644
--- a/drivers/net/igb/e1000_mbx.c
+++ b/drivers/net/igb/e1000_mbx.c
@@ -188,7 +188,7 @@ out:
188 * returns SUCCESS if it successfully received a message notification and 188 * returns SUCCESS if it successfully received a message notification and
189 * copied it into the receive buffer. 189 * copied it into the receive buffer.
190 **/ 190 **/
191s32 igb_read_posted_mbx(struct e1000_hw *hw, u32 *msg, u16 size, u16 mbx_id) 191static s32 igb_read_posted_mbx(struct e1000_hw *hw, u32 *msg, u16 size, u16 mbx_id)
192{ 192{
193 struct e1000_mbx_info *mbx = &hw->mbx; 193 struct e1000_mbx_info *mbx = &hw->mbx;
194 s32 ret_val = -E1000_ERR_MBX; 194 s32 ret_val = -E1000_ERR_MBX;
@@ -214,7 +214,7 @@ out:
214 * returns SUCCESS if it successfully copied message into the buffer and 214 * returns SUCCESS if it successfully copied message into the buffer and
215 * received an ack to that message within delay * timeout period 215 * received an ack to that message within delay * timeout period
216 **/ 216 **/
217s32 igb_write_posted_mbx(struct e1000_hw *hw, u32 *msg, u16 size, u16 mbx_id) 217static s32 igb_write_posted_mbx(struct e1000_hw *hw, u32 *msg, u16 size, u16 mbx_id)
218{ 218{
219 struct e1000_mbx_info *mbx = &hw->mbx; 219 struct e1000_mbx_info *mbx = &hw->mbx;
220 s32 ret_val = 0; 220 s32 ret_val = 0;
@@ -232,19 +232,6 @@ out:
232 return ret_val; 232 return ret_val;
233} 233}
234 234
235/**
236 * e1000_init_mbx_ops_generic - Initialize NVM function pointers
237 * @hw: pointer to the HW structure
238 *
239 * Setups up the function pointers to no-op functions
240 **/
241void e1000_init_mbx_ops_generic(struct e1000_hw *hw)
242{
243 struct e1000_mbx_info *mbx = &hw->mbx;
244 mbx->ops.read_posted = igb_read_posted_mbx;
245 mbx->ops.write_posted = igb_write_posted_mbx;
246}
247
248static s32 igb_check_for_bit_pf(struct e1000_hw *hw, u32 mask) 235static s32 igb_check_for_bit_pf(struct e1000_hw *hw, u32 mask)
249{ 236{
250 u32 mbvficr = rd32(E1000_MBVFICR); 237 u32 mbvficr = rd32(E1000_MBVFICR);
diff --git a/drivers/net/igb/e1000_mbx.h b/drivers/net/igb/e1000_mbx.h
index 6ec9890a8f7a..ebc02ea3f198 100644
--- a/drivers/net/igb/e1000_mbx.h
+++ b/drivers/net/igb/e1000_mbx.h
@@ -67,8 +67,6 @@
67 67
68s32 igb_read_mbx(struct e1000_hw *, u32 *, u16, u16); 68s32 igb_read_mbx(struct e1000_hw *, u32 *, u16, u16);
69s32 igb_write_mbx(struct e1000_hw *, u32 *, u16, u16); 69s32 igb_write_mbx(struct e1000_hw *, u32 *, u16, u16);
70s32 igb_read_posted_mbx(struct e1000_hw *, u32 *, u16, u16);
71s32 igb_write_posted_mbx(struct e1000_hw *, u32 *, u16, u16);
72s32 igb_check_for_msg(struct e1000_hw *, u16); 70s32 igb_check_for_msg(struct e1000_hw *, u16);
73s32 igb_check_for_ack(struct e1000_hw *, u16); 71s32 igb_check_for_ack(struct e1000_hw *, u16);
74s32 igb_check_for_rst(struct e1000_hw *, u16); 72s32 igb_check_for_rst(struct e1000_hw *, u16);
diff --git a/drivers/net/igb/igb_main.c b/drivers/net/igb/igb_main.c
index db7274e62228..08c801490c72 100644
--- a/drivers/net/igb/igb_main.c
+++ b/drivers/net/igb/igb_main.c
@@ -1261,25 +1261,32 @@ static int __devinit igb_probe(struct pci_dev *pdev,
1261 int i; 1261 int i;
1262 unsigned char mac_addr[ETH_ALEN]; 1262 unsigned char mac_addr[ETH_ALEN];
1263 1263
1264 if (num_vfs) 1264 if (num_vfs) {
1265 adapter->vf_data = kcalloc(num_vfs, 1265 adapter->vf_data = kcalloc(num_vfs,
1266 sizeof(struct vf_data_storage), 1266 sizeof(struct vf_data_storage),
1267 GFP_KERNEL); 1267 GFP_KERNEL);
1268 if (!adapter->vf_data) { 1268 if (!adapter->vf_data) {
1269 dev_err(&pdev->dev, "Could not allocate VF private " 1269 dev_err(&pdev->dev,
1270 "data - IOV enable failed\n"); 1270 "Could not allocate VF private data - "
1271 } else { 1271 "IOV enable failed\n");
1272 err = pci_enable_sriov(pdev, num_vfs);
1273 if (!err) {
1274 adapter->vfs_allocated_count = num_vfs;
1275 dev_info(&pdev->dev, "%d vfs allocated\n", num_vfs);
1276 for (i = 0; i < adapter->vfs_allocated_count; i++) {
1277 random_ether_addr(mac_addr);
1278 igb_set_vf_mac(adapter, i, mac_addr);
1279 }
1280 } else { 1272 } else {
1281 kfree(adapter->vf_data); 1273 err = pci_enable_sriov(pdev, num_vfs);
1282 adapter->vf_data = NULL; 1274 if (!err) {
1275 adapter->vfs_allocated_count = num_vfs;
1276 dev_info(&pdev->dev,
1277 "%d vfs allocated\n",
1278 num_vfs);
1279 for (i = 0;
1280 i < adapter->vfs_allocated_count;
1281 i++) {
1282 random_ether_addr(mac_addr);
1283 igb_set_vf_mac(adapter, i,
1284 mac_addr);
1285 }
1286 } else {
1287 kfree(adapter->vf_data);
1288 adapter->vf_data = NULL;
1289 }
1283 } 1290 }
1284 } 1291 }
1285 } 1292 }
diff --git a/drivers/net/igbvf/igbvf.h b/drivers/net/igbvf/igbvf.h
index 936ed2a9435f..4bff35e46871 100644
--- a/drivers/net/igbvf/igbvf.h
+++ b/drivers/net/igbvf/igbvf.h
@@ -321,14 +321,11 @@ extern void igbvf_set_ethtool_ops(struct net_device *);
321extern int igbvf_up(struct igbvf_adapter *); 321extern int igbvf_up(struct igbvf_adapter *);
322extern void igbvf_down(struct igbvf_adapter *); 322extern void igbvf_down(struct igbvf_adapter *);
323extern void igbvf_reinit_locked(struct igbvf_adapter *); 323extern void igbvf_reinit_locked(struct igbvf_adapter *);
324extern void igbvf_reset(struct igbvf_adapter *);
325extern int igbvf_setup_rx_resources(struct igbvf_adapter *, struct igbvf_ring *); 324extern int igbvf_setup_rx_resources(struct igbvf_adapter *, struct igbvf_ring *);
326extern int igbvf_setup_tx_resources(struct igbvf_adapter *, struct igbvf_ring *); 325extern int igbvf_setup_tx_resources(struct igbvf_adapter *, struct igbvf_ring *);
327extern void igbvf_free_rx_resources(struct igbvf_ring *); 326extern void igbvf_free_rx_resources(struct igbvf_ring *);
328extern void igbvf_free_tx_resources(struct igbvf_ring *); 327extern void igbvf_free_tx_resources(struct igbvf_ring *);
329extern void igbvf_update_stats(struct igbvf_adapter *); 328extern void igbvf_update_stats(struct igbvf_adapter *);
330extern void igbvf_set_interrupt_capability(struct igbvf_adapter *);
331extern void igbvf_reset_interrupt_capability(struct igbvf_adapter *);
332 329
333extern unsigned int copybreak; 330extern unsigned int copybreak;
334 331
diff --git a/drivers/net/igbvf/netdev.c b/drivers/net/igbvf/netdev.c
index c5648420dedf..b774666ad3cf 100644
--- a/drivers/net/igbvf/netdev.c
+++ b/drivers/net/igbvf/netdev.c
@@ -52,6 +52,9 @@ static const char igbvf_driver_string[] =
52static const char igbvf_copyright[] = "Copyright (c) 2009 Intel Corporation."; 52static const char igbvf_copyright[] = "Copyright (c) 2009 Intel Corporation.";
53 53
54static int igbvf_poll(struct napi_struct *napi, int budget); 54static int igbvf_poll(struct napi_struct *napi, int budget);
55static void igbvf_reset(struct igbvf_adapter *);
56static void igbvf_set_interrupt_capability(struct igbvf_adapter *);
57static void igbvf_reset_interrupt_capability(struct igbvf_adapter *);
55 58
56static struct igbvf_info igbvf_vf_info = { 59static struct igbvf_info igbvf_vf_info = {
57 .mac = e1000_vfadapt, 60 .mac = e1000_vfadapt,
@@ -990,7 +993,7 @@ static void igbvf_configure_msix(struct igbvf_adapter *adapter)
990 e1e_flush(); 993 e1e_flush();
991} 994}
992 995
993void igbvf_reset_interrupt_capability(struct igbvf_adapter *adapter) 996static void igbvf_reset_interrupt_capability(struct igbvf_adapter *adapter)
994{ 997{
995 if (adapter->msix_entries) { 998 if (adapter->msix_entries) {
996 pci_disable_msix(adapter->pdev); 999 pci_disable_msix(adapter->pdev);
@@ -1005,7 +1008,7 @@ void igbvf_reset_interrupt_capability(struct igbvf_adapter *adapter)
1005 * Attempt to configure interrupts using the best available 1008 * Attempt to configure interrupts using the best available
1006 * capabilities of the hardware and kernel. 1009 * capabilities of the hardware and kernel.
1007 **/ 1010 **/
1008void igbvf_set_interrupt_capability(struct igbvf_adapter *adapter) 1011static void igbvf_set_interrupt_capability(struct igbvf_adapter *adapter)
1009{ 1012{
1010 int err = -ENOMEM; 1013 int err = -ENOMEM;
1011 int i; 1014 int i;
@@ -1447,7 +1450,7 @@ static void igbvf_configure(struct igbvf_adapter *adapter)
1447 * set/changed during runtime. After reset the device needs to be 1450 * set/changed during runtime. After reset the device needs to be
1448 * properly configured for Rx, Tx etc. 1451 * properly configured for Rx, Tx etc.
1449 */ 1452 */
1450void igbvf_reset(struct igbvf_adapter *adapter) 1453static void igbvf_reset(struct igbvf_adapter *adapter)
1451{ 1454{
1452 struct e1000_mac_info *mac = &adapter->hw.mac; 1455 struct e1000_mac_info *mac = &adapter->hw.mac;
1453 struct net_device *netdev = adapter->netdev; 1456 struct net_device *netdev = adapter->netdev;
diff --git a/drivers/net/igbvf/vf.c b/drivers/net/igbvf/vf.c
index aa246c93279d..2a4faf9ade69 100644
--- a/drivers/net/igbvf/vf.c
+++ b/drivers/net/igbvf/vf.c
@@ -44,7 +44,7 @@ static s32 e1000_set_vfta_vf(struct e1000_hw *, u16, bool);
44 * e1000_init_mac_params_vf - Inits MAC params 44 * e1000_init_mac_params_vf - Inits MAC params
45 * @hw: pointer to the HW structure 45 * @hw: pointer to the HW structure
46 **/ 46 **/
47s32 e1000_init_mac_params_vf(struct e1000_hw *hw) 47static s32 e1000_init_mac_params_vf(struct e1000_hw *hw)
48{ 48{
49 struct e1000_mac_info *mac = &hw->mac; 49 struct e1000_mac_info *mac = &hw->mac;
50 50
diff --git a/drivers/net/igbvf/vf.h b/drivers/net/igbvf/vf.h
index ec07228f9478..1e8ce3741a67 100644
--- a/drivers/net/igbvf/vf.h
+++ b/drivers/net/igbvf/vf.h
@@ -259,7 +259,6 @@ struct e1000_hw {
259/* These functions must be implemented by drivers */ 259/* These functions must be implemented by drivers */
260void e1000_rlpml_set_vf(struct e1000_hw *, u16); 260void e1000_rlpml_set_vf(struct e1000_hw *, u16);
261void e1000_init_function_pointers_vf(struct e1000_hw *hw); 261void e1000_init_function_pointers_vf(struct e1000_hw *hw);
262s32 e1000_init_mac_params_vf(struct e1000_hw *hw);
263 262
264 263
265#endif /* _E1000_VF_H_ */ 264#endif /* _E1000_VF_H_ */
diff --git a/drivers/net/ioc3-eth.c b/drivers/net/ioc3-eth.c
index cbc63ff13add..c5593f4665a4 100644
--- a/drivers/net/ioc3-eth.c
+++ b/drivers/net/ioc3-eth.c
@@ -1214,6 +1214,19 @@ static void __devinit ioc3_serial_probe(struct pci_dev *pdev, struct ioc3 *ioc3)
1214} 1214}
1215#endif 1215#endif
1216 1216
1217static const struct net_device_ops ioc3_netdev_ops = {
1218 .ndo_open = ioc3_open,
1219 .ndo_stop = ioc3_close,
1220 .ndo_start_xmit = ioc3_start_xmit,
1221 .ndo_tx_timeout = ioc3_timeout,
1222 .ndo_get_stats = ioc3_get_stats,
1223 .ndo_set_multicast_list = ioc3_set_multicast_list,
1224 .ndo_do_ioctl = ioc3_ioctl,
1225 .ndo_validate_addr = eth_validate_addr,
1226 .ndo_set_mac_address = ioc3_set_mac_address,
1227 .ndo_change_mtu = eth_change_mtu,
1228};
1229
1217static int __devinit ioc3_probe(struct pci_dev *pdev, 1230static int __devinit ioc3_probe(struct pci_dev *pdev,
1218 const struct pci_device_id *ent) 1231 const struct pci_device_id *ent)
1219{ 1232{
@@ -1310,15 +1323,8 @@ static int __devinit ioc3_probe(struct pci_dev *pdev,
1310 ioc3_get_eaddr(ip); 1323 ioc3_get_eaddr(ip);
1311 1324
1312 /* The IOC3-specific entries in the device structure. */ 1325 /* The IOC3-specific entries in the device structure. */
1313 dev->open = ioc3_open;
1314 dev->hard_start_xmit = ioc3_start_xmit;
1315 dev->tx_timeout = ioc3_timeout;
1316 dev->watchdog_timeo = 5 * HZ; 1326 dev->watchdog_timeo = 5 * HZ;
1317 dev->stop = ioc3_close; 1327 dev->netdev_ops = &ioc3_netdev_ops;
1318 dev->get_stats = ioc3_get_stats;
1319 dev->do_ioctl = ioc3_ioctl;
1320 dev->set_multicast_list = ioc3_set_multicast_list;
1321 dev->set_mac_address = ioc3_set_mac_address;
1322 dev->ethtool_ops = &ioc3_ethtool_ops; 1328 dev->ethtool_ops = &ioc3_ethtool_ops;
1323 dev->features = NETIF_F_IP_CSUM; 1329 dev->features = NETIF_F_IP_CSUM;
1324 1330
diff --git a/drivers/net/isa-skeleton.c b/drivers/net/isa-skeleton.c
index 3126678bdd3c..73585fd8f29f 100644
--- a/drivers/net/isa-skeleton.c
+++ b/drivers/net/isa-skeleton.c
@@ -181,6 +181,18 @@ out:
181} 181}
182#endif 182#endif
183 183
184static const struct net_device_ops netcard_netdev_ops = {
185 .ndo_open = net_open,
186 .ndo_stop = net_close,
187 .ndo_start_xmit = net_send_packet,
188 .ndo_get_stats = net_get_stats,
189 .ndo_set_multicast_list = set_multicast_list,
190 .ndo_tx_timeout = net_tx_timeout,
191 .ndo_validate_addr = eth_validate_addr,
192 .ndo_set_mac_address = eth_mac_addr,
193 .ndo_change_mtu = eth_change_mtu,
194};
195
184/* 196/*
185 * This is the real probe routine. Linux has a history of friendly device 197 * This is the real probe routine. Linux has a history of friendly device
186 * probes on the ISA bus. A good device probes avoids doing writes, and 198 * probes on the ISA bus. A good device probes avoids doing writes, and
@@ -303,13 +315,7 @@ static int __init netcard_probe1(struct net_device *dev, int ioaddr)
303 np = netdev_priv(dev); 315 np = netdev_priv(dev);
304 spin_lock_init(&np->lock); 316 spin_lock_init(&np->lock);
305 317
306 dev->open = net_open; 318 dev->netdev_ops = &netcard_netdev_ops;
307 dev->stop = net_close;
308 dev->hard_start_xmit = net_send_packet;
309 dev->get_stats = net_get_stats;
310 dev->set_multicast_list = &set_multicast_list;
311
312 dev->tx_timeout = &net_tx_timeout;
313 dev->watchdog_timeo = MY_TX_TIMEOUT; 319 dev->watchdog_timeo = MY_TX_TIMEOUT;
314 320
315 err = register_netdev(dev); 321 err = register_netdev(dev);
diff --git a/drivers/net/ixgbe/ixgbe_82598.c b/drivers/net/ixgbe/ixgbe_82598.c
index de4db0dc7879..4791238c3f6e 100644
--- a/drivers/net/ixgbe/ixgbe_82598.c
+++ b/drivers/net/ixgbe/ixgbe_82598.c
@@ -885,61 +885,6 @@ static s32 ixgbe_clear_vfta_82598(struct ixgbe_hw *hw)
885} 885}
886 886
887/** 887/**
888 * ixgbe_blink_led_start_82598 - Blink LED based on index.
889 * @hw: pointer to hardware structure
890 * @index: led number to blink
891 **/
892static s32 ixgbe_blink_led_start_82598(struct ixgbe_hw *hw, u32 index)
893{
894 ixgbe_link_speed speed = 0;
895 bool link_up = 0;
896 u32 autoc_reg = IXGBE_READ_REG(hw, IXGBE_AUTOC);
897 u32 led_reg = IXGBE_READ_REG(hw, IXGBE_LEDCTL);
898
899 /*
900 * Link must be up to auto-blink the LEDs on the 82598EB MAC;
901 * force it if link is down.
902 */
903 hw->mac.ops.check_link(hw, &speed, &link_up, false);
904
905 if (!link_up) {
906 autoc_reg |= IXGBE_AUTOC_FLU;
907 IXGBE_WRITE_REG(hw, IXGBE_AUTOC, autoc_reg);
908 msleep(10);
909 }
910
911 led_reg &= ~IXGBE_LED_MODE_MASK(index);
912 led_reg |= IXGBE_LED_BLINK(index);
913 IXGBE_WRITE_REG(hw, IXGBE_LEDCTL, led_reg);
914 IXGBE_WRITE_FLUSH(hw);
915
916 return 0;
917}
918
919/**
920 * ixgbe_blink_led_stop_82598 - Stop blinking LED based on index.
921 * @hw: pointer to hardware structure
922 * @index: led number to stop blinking
923 **/
924static s32 ixgbe_blink_led_stop_82598(struct ixgbe_hw *hw, u32 index)
925{
926 u32 autoc_reg = IXGBE_READ_REG(hw, IXGBE_AUTOC);
927 u32 led_reg = IXGBE_READ_REG(hw, IXGBE_LEDCTL);
928
929 autoc_reg &= ~IXGBE_AUTOC_FLU;
930 autoc_reg |= IXGBE_AUTOC_AN_RESTART;
931 IXGBE_WRITE_REG(hw, IXGBE_AUTOC, autoc_reg);
932
933 led_reg &= ~IXGBE_LED_MODE_MASK(index);
934 led_reg &= ~IXGBE_LED_BLINK(index);
935 led_reg |= IXGBE_LED_LINK_ACTIVE << IXGBE_LED_MODE_SHIFT(index);
936 IXGBE_WRITE_REG(hw, IXGBE_LEDCTL, led_reg);
937 IXGBE_WRITE_FLUSH(hw);
938
939 return 0;
940}
941
942/**
943 * ixgbe_read_analog_reg8_82598 - Reads 8 bit Atlas analog register 888 * ixgbe_read_analog_reg8_82598 - Reads 8 bit Atlas analog register
944 * @hw: pointer to hardware structure 889 * @hw: pointer to hardware structure
945 * @reg: analog register to read 890 * @reg: analog register to read
@@ -1128,8 +1073,8 @@ static struct ixgbe_mac_operations mac_ops_82598 = {
1128 .get_link_capabilities = &ixgbe_get_link_capabilities_82598, 1073 .get_link_capabilities = &ixgbe_get_link_capabilities_82598,
1129 .led_on = &ixgbe_led_on_generic, 1074 .led_on = &ixgbe_led_on_generic,
1130 .led_off = &ixgbe_led_off_generic, 1075 .led_off = &ixgbe_led_off_generic,
1131 .blink_led_start = &ixgbe_blink_led_start_82598, 1076 .blink_led_start = &ixgbe_blink_led_start_generic,
1132 .blink_led_stop = &ixgbe_blink_led_stop_82598, 1077 .blink_led_stop = &ixgbe_blink_led_stop_generic,
1133 .set_rar = &ixgbe_set_rar_generic, 1078 .set_rar = &ixgbe_set_rar_generic,
1134 .clear_rar = &ixgbe_clear_rar_generic, 1079 .clear_rar = &ixgbe_clear_rar_generic,
1135 .set_vmdq = &ixgbe_set_vmdq_82598, 1080 .set_vmdq = &ixgbe_set_vmdq_82598,
diff --git a/drivers/net/ixgbe/ixgbe_82599.c b/drivers/net/ixgbe/ixgbe_82599.c
index beae7e012609..29771fbaa42d 100644
--- a/drivers/net/ixgbe/ixgbe_82599.c
+++ b/drivers/net/ixgbe/ixgbe_82599.c
@@ -68,8 +68,6 @@ s32 ixgbe_clear_vmdq_82599(struct ixgbe_hw *hw, u32 rar, u32 vmdq);
68s32 ixgbe_set_vfta_82599(struct ixgbe_hw *hw, u32 vlan, 68s32 ixgbe_set_vfta_82599(struct ixgbe_hw *hw, u32 vlan,
69 u32 vind, bool vlan_on); 69 u32 vind, bool vlan_on);
70s32 ixgbe_clear_vfta_82599(struct ixgbe_hw *hw); 70s32 ixgbe_clear_vfta_82599(struct ixgbe_hw *hw);
71s32 ixgbe_blink_led_stop_82599(struct ixgbe_hw *hw, u32 index);
72s32 ixgbe_blink_led_start_82599(struct ixgbe_hw *hw, u32 index);
73s32 ixgbe_init_uta_tables_82599(struct ixgbe_hw *hw); 71s32 ixgbe_init_uta_tables_82599(struct ixgbe_hw *hw);
74s32 ixgbe_read_analog_reg8_82599(struct ixgbe_hw *hw, u32 reg, u8 *val); 72s32 ixgbe_read_analog_reg8_82599(struct ixgbe_hw *hw, u32 reg, u8 *val);
75s32 ixgbe_write_analog_reg8_82599(struct ixgbe_hw *hw, u32 reg, u8 val); 73s32 ixgbe_write_analog_reg8_82599(struct ixgbe_hw *hw, u32 reg, u8 val);
@@ -991,40 +989,6 @@ s32 ixgbe_clear_vfta_82599(struct ixgbe_hw *hw)
991} 989}
992 990
993/** 991/**
994 * ixgbe_blink_led_start_82599 - Blink LED based on index.
995 * @hw: pointer to hardware structure
996 * @index: led number to blink
997 **/
998s32 ixgbe_blink_led_start_82599(struct ixgbe_hw *hw, u32 index)
999{
1000 u32 led_reg = IXGBE_READ_REG(hw, IXGBE_LEDCTL);
1001
1002 led_reg &= ~IXGBE_LED_MODE_MASK(index);
1003 led_reg |= IXGBE_LED_BLINK(index);
1004 IXGBE_WRITE_REG(hw, IXGBE_LEDCTL, led_reg);
1005 IXGBE_WRITE_FLUSH(hw);
1006
1007 return 0;
1008}
1009
1010/**
1011 * ixgbe_blink_led_stop_82599 - Stop blinking LED based on index.
1012 * @hw: pointer to hardware structure
1013 * @index: led number to stop blinking
1014 **/
1015s32 ixgbe_blink_led_stop_82599(struct ixgbe_hw *hw, u32 index)
1016{
1017 u32 led_reg = IXGBE_READ_REG(hw, IXGBE_LEDCTL);
1018
1019 led_reg &= ~IXGBE_LED_MODE_MASK(index);
1020 led_reg &= ~IXGBE_LED_BLINK(index);
1021 IXGBE_WRITE_REG(hw, IXGBE_LEDCTL, led_reg);
1022 IXGBE_WRITE_FLUSH(hw);
1023
1024 return 0;
1025}
1026
1027/**
1028 * ixgbe_init_uta_tables_82599 - Initialize the Unicast Table Array 992 * ixgbe_init_uta_tables_82599 - Initialize the Unicast Table Array
1029 * @hw: pointer to hardware structure 993 * @hw: pointer to hardware structure
1030 **/ 994 **/
@@ -1243,8 +1207,8 @@ static struct ixgbe_mac_operations mac_ops_82599 = {
1243 .get_link_capabilities = &ixgbe_get_link_capabilities_82599, 1207 .get_link_capabilities = &ixgbe_get_link_capabilities_82599,
1244 .led_on = &ixgbe_led_on_generic, 1208 .led_on = &ixgbe_led_on_generic,
1245 .led_off = &ixgbe_led_off_generic, 1209 .led_off = &ixgbe_led_off_generic,
1246 .blink_led_start = &ixgbe_blink_led_start_82599, 1210 .blink_led_start = &ixgbe_blink_led_start_generic,
1247 .blink_led_stop = &ixgbe_blink_led_stop_82599, 1211 .blink_led_stop = &ixgbe_blink_led_stop_generic,
1248 .set_rar = &ixgbe_set_rar_generic, 1212 .set_rar = &ixgbe_set_rar_generic,
1249 .clear_rar = &ixgbe_clear_rar_generic, 1213 .clear_rar = &ixgbe_clear_rar_generic,
1250 .set_vmdq = &ixgbe_set_vmdq_82599, 1214 .set_vmdq = &ixgbe_set_vmdq_82599,
diff --git a/drivers/net/ixgbe/ixgbe_common.c b/drivers/net/ixgbe/ixgbe_common.c
index 63ab6671d08e..5567519676d5 100644
--- a/drivers/net/ixgbe/ixgbe_common.c
+++ b/drivers/net/ixgbe/ixgbe_common.c
@@ -2071,3 +2071,58 @@ s32 ixgbe_enable_rx_dma_generic(struct ixgbe_hw *hw, u32 regval)
2071 2071
2072 return 0; 2072 return 0;
2073} 2073}
2074
2075/**
2076 * ixgbe_blink_led_start_generic - Blink LED based on index.
2077 * @hw: pointer to hardware structure
2078 * @index: led number to blink
2079 **/
2080s32 ixgbe_blink_led_start_generic(struct ixgbe_hw *hw, u32 index)
2081{
2082 ixgbe_link_speed speed = 0;
2083 bool link_up = 0;
2084 u32 autoc_reg = IXGBE_READ_REG(hw, IXGBE_AUTOC);
2085 u32 led_reg = IXGBE_READ_REG(hw, IXGBE_LEDCTL);
2086
2087 /*
2088 * Link must be up to auto-blink the LEDs;
2089 * Force it if link is down.
2090 */
2091 hw->mac.ops.check_link(hw, &speed, &link_up, false);
2092
2093 if (!link_up) {
2094 autoc_reg |= IXGBE_AUTOC_FLU;
2095 IXGBE_WRITE_REG(hw, IXGBE_AUTOC, autoc_reg);
2096 msleep(10);
2097 }
2098
2099 led_reg &= ~IXGBE_LED_MODE_MASK(index);
2100 led_reg |= IXGBE_LED_BLINK(index);
2101 IXGBE_WRITE_REG(hw, IXGBE_LEDCTL, led_reg);
2102 IXGBE_WRITE_FLUSH(hw);
2103
2104 return 0;
2105}
2106
2107/**
2108 * ixgbe_blink_led_stop_generic - Stop blinking LED based on index.
2109 * @hw: pointer to hardware structure
2110 * @index: led number to stop blinking
2111 **/
2112s32 ixgbe_blink_led_stop_generic(struct ixgbe_hw *hw, u32 index)
2113{
2114 u32 autoc_reg = IXGBE_READ_REG(hw, IXGBE_AUTOC);
2115 u32 led_reg = IXGBE_READ_REG(hw, IXGBE_LEDCTL);
2116
2117 autoc_reg &= ~IXGBE_AUTOC_FLU;
2118 autoc_reg |= IXGBE_AUTOC_AN_RESTART;
2119 IXGBE_WRITE_REG(hw, IXGBE_AUTOC, autoc_reg);
2120
2121 led_reg &= ~IXGBE_LED_MODE_MASK(index);
2122 led_reg &= ~IXGBE_LED_BLINK(index);
2123 led_reg |= IXGBE_LED_LINK_ACTIVE << IXGBE_LED_MODE_SHIFT(index);
2124 IXGBE_WRITE_REG(hw, IXGBE_LEDCTL, led_reg);
2125 IXGBE_WRITE_FLUSH(hw);
2126
2127 return 0;
2128}
diff --git a/drivers/net/ixgbe/ixgbe_common.h b/drivers/net/ixgbe/ixgbe_common.h
index 24f73e719c3f..dd260890ad0a 100644
--- a/drivers/net/ixgbe/ixgbe_common.h
+++ b/drivers/net/ixgbe/ixgbe_common.h
@@ -76,6 +76,9 @@ s32 ixgbe_disable_pcie_master(struct ixgbe_hw *hw);
76s32 ixgbe_read_analog_reg8_generic(struct ixgbe_hw *hw, u32 reg, u8 *val); 76s32 ixgbe_read_analog_reg8_generic(struct ixgbe_hw *hw, u32 reg, u8 *val);
77s32 ixgbe_write_analog_reg8_generic(struct ixgbe_hw *hw, u32 reg, u8 val); 77s32 ixgbe_write_analog_reg8_generic(struct ixgbe_hw *hw, u32 reg, u8 val);
78 78
79s32 ixgbe_blink_led_start_generic(struct ixgbe_hw *hw, u32 index);
80s32 ixgbe_blink_led_stop_generic(struct ixgbe_hw *hw, u32 index);
81
79#define IXGBE_WRITE_REG(a, reg, value) writel((value), ((a)->hw_addr + (reg))) 82#define IXGBE_WRITE_REG(a, reg, value) writel((value), ((a)->hw_addr + (reg)))
80 83
81#ifndef writeq 84#ifndef writeq
diff --git a/drivers/net/ixgbe/ixgbe_ethtool.c b/drivers/net/ixgbe/ixgbe_ethtool.c
index aafc120f164e..f0a20facc650 100644
--- a/drivers/net/ixgbe/ixgbe_ethtool.c
+++ b/drivers/net/ixgbe/ixgbe_ethtool.c
@@ -943,6 +943,24 @@ static void ixgbe_get_strings(struct net_device *netdev, u32 stringset,
943} 943}
944 944
945 945
946static int ixgbe_wol_exclusion(struct ixgbe_adapter *adapter,
947 struct ethtool_wolinfo *wol)
948{
949 struct ixgbe_hw *hw = &adapter->hw;
950 int retval = 1;
951
952 switch(hw->device_id) {
953 case IXGBE_DEV_ID_82599_KX4:
954 retval = 0;
955 break;
956 default:
957 wol->supported = 0;
958 retval = 0;
959 }
960
961 return retval;
962}
963
946static void ixgbe_get_wol(struct net_device *netdev, 964static void ixgbe_get_wol(struct net_device *netdev,
947 struct ethtool_wolinfo *wol) 965 struct ethtool_wolinfo *wol)
948{ 966{
@@ -952,7 +970,8 @@ static void ixgbe_get_wol(struct net_device *netdev,
952 WAKE_BCAST | WAKE_MAGIC; 970 WAKE_BCAST | WAKE_MAGIC;
953 wol->wolopts = 0; 971 wol->wolopts = 0;
954 972
955 if (!device_can_wakeup(&adapter->pdev->dev)) 973 if (ixgbe_wol_exclusion(adapter, wol) ||
974 !device_can_wakeup(&adapter->pdev->dev))
956 return; 975 return;
957 976
958 if (adapter->wol & IXGBE_WUFC_EX) 977 if (adapter->wol & IXGBE_WUFC_EX)
@@ -974,6 +993,9 @@ static int ixgbe_set_wol(struct net_device *netdev, struct ethtool_wolinfo *wol)
974 if (wol->wolopts & (WAKE_PHY | WAKE_ARP | WAKE_MAGICSECURE)) 993 if (wol->wolopts & (WAKE_PHY | WAKE_ARP | WAKE_MAGICSECURE))
975 return -EOPNOTSUPP; 994 return -EOPNOTSUPP;
976 995
996 if (ixgbe_wol_exclusion(adapter, wol))
997 return wol->wolopts ? -EOPNOTSUPP : 0;
998
977 adapter->wol = 0; 999 adapter->wol = 0;
978 1000
979 if (wol->wolopts & WAKE_UCAST) 1001 if (wol->wolopts & WAKE_UCAST)
diff --git a/drivers/net/ixgbe/ixgbe_main.c b/drivers/net/ixgbe/ixgbe_main.c
index 9ef128ae6458..febde45cf9fa 100644
--- a/drivers/net/ixgbe/ixgbe_main.c
+++ b/drivers/net/ixgbe/ixgbe_main.c
@@ -2723,17 +2723,21 @@ static inline bool ixgbe_set_rss_queues(struct ixgbe_adapter *adapter)
2723 **/ 2723 **/
2724static void ixgbe_set_num_queues(struct ixgbe_adapter *adapter) 2724static void ixgbe_set_num_queues(struct ixgbe_adapter *adapter)
2725{ 2725{
2726 /* Start with base case */
2727 adapter->num_rx_queues = 1;
2728 adapter->num_tx_queues = 1;
2729
2730#ifdef CONFIG_IXGBE_DCB 2726#ifdef CONFIG_IXGBE_DCB
2731 if (ixgbe_set_dcb_queues(adapter)) 2727 if (ixgbe_set_dcb_queues(adapter))
2732 return; 2728 goto done;
2733 2729
2734#endif 2730#endif
2735 if (ixgbe_set_rss_queues(adapter)) 2731 if (ixgbe_set_rss_queues(adapter))
2736 return; 2732 goto done;
2733
2734 /* fallback to base case */
2735 adapter->num_rx_queues = 1;
2736 adapter->num_tx_queues = 1;
2737
2738done:
2739 /* Notify the stack of the (possibly) reduced Tx Queue count. */
2740 adapter->netdev->real_num_tx_queues = adapter->num_tx_queues;
2737} 2741}
2738 2742
2739static void ixgbe_acquire_msix_vectors(struct ixgbe_adapter *adapter, 2743static void ixgbe_acquire_msix_vectors(struct ixgbe_adapter *adapter,
@@ -2992,9 +2996,6 @@ try_msi:
2992 } 2996 }
2993 2997
2994out: 2998out:
2995 /* Notify the stack of the (possibly) reduced Tx Queue count. */
2996 adapter->netdev->real_num_tx_queues = adapter->num_tx_queues;
2997
2998 return err; 2999 return err;
2999} 3000}
3000 3001
@@ -3611,9 +3612,9 @@ static int ixgbe_resume(struct pci_dev *pdev)
3611 3612
3612 return 0; 3613 return 0;
3613} 3614}
3614
3615#endif /* CONFIG_PM */ 3615#endif /* CONFIG_PM */
3616static int ixgbe_suspend(struct pci_dev *pdev, pm_message_t state) 3616
3617static int __ixgbe_shutdown(struct pci_dev *pdev, bool *enable_wake)
3617{ 3618{
3618 struct net_device *netdev = pci_get_drvdata(pdev); 3619 struct net_device *netdev = pci_get_drvdata(pdev);
3619 struct ixgbe_adapter *adapter = netdev_priv(netdev); 3620 struct ixgbe_adapter *adapter = netdev_priv(netdev);
@@ -3672,18 +3673,46 @@ static int ixgbe_suspend(struct pci_dev *pdev, pm_message_t state)
3672 pci_enable_wake(pdev, PCI_D3cold, 0); 3673 pci_enable_wake(pdev, PCI_D3cold, 0);
3673 } 3674 }
3674 3675
3676 *enable_wake = !!wufc;
3677
3675 ixgbe_release_hw_control(adapter); 3678 ixgbe_release_hw_control(adapter);
3676 3679
3677 pci_disable_device(pdev); 3680 pci_disable_device(pdev);
3678 3681
3679 pci_set_power_state(pdev, pci_choose_state(pdev, state)); 3682 return 0;
3683}
3684
3685#ifdef CONFIG_PM
3686static int ixgbe_suspend(struct pci_dev *pdev, pm_message_t state)
3687{
3688 int retval;
3689 bool wake;
3690
3691 retval = __ixgbe_shutdown(pdev, &wake);
3692 if (retval)
3693 return retval;
3694
3695 if (wake) {
3696 pci_prepare_to_sleep(pdev);
3697 } else {
3698 pci_wake_from_d3(pdev, false);
3699 pci_set_power_state(pdev, PCI_D3hot);
3700 }
3680 3701
3681 return 0; 3702 return 0;
3682} 3703}
3704#endif /* CONFIG_PM */
3683 3705
3684static void ixgbe_shutdown(struct pci_dev *pdev) 3706static void ixgbe_shutdown(struct pci_dev *pdev)
3685{ 3707{
3686 ixgbe_suspend(pdev, PMSG_SUSPEND); 3708 bool wake;
3709
3710 __ixgbe_shutdown(pdev, &wake);
3711
3712 if (system_state == SYSTEM_POWER_OFF) {
3713 pci_wake_from_d3(pdev, wake);
3714 pci_set_power_state(pdev, PCI_D3hot);
3715 }
3687} 3716}
3688 3717
3689/** 3718/**
@@ -4342,7 +4371,7 @@ static int ixgbe_xmit_frame(struct sk_buff *skb, struct net_device *netdev)
4342 int count = 0; 4371 int count = 0;
4343 unsigned int f; 4372 unsigned int f;
4344 4373
4345 r_idx = (adapter->num_tx_queues - 1) & skb->queue_mapping; 4374 r_idx = skb->queue_mapping;
4346 tx_ring = &adapter->tx_ring[r_idx]; 4375 tx_ring = &adapter->tx_ring[r_idx];
4347 4376
4348 if (adapter->vlgrp && vlan_tx_tag_present(skb)) { 4377 if (adapter->vlgrp && vlan_tx_tag_present(skb)) {
diff --git a/drivers/net/jme.c b/drivers/net/jme.c
index ece35040288c..621a7c0c46ba 100644
--- a/drivers/net/jme.c
+++ b/drivers/net/jme.c
@@ -2591,13 +2591,13 @@ static int
2591jme_pci_dma64(struct pci_dev *pdev) 2591jme_pci_dma64(struct pci_dev *pdev)
2592{ 2592{
2593 if (pdev->device == PCI_DEVICE_ID_JMICRON_JMC250 && 2593 if (pdev->device == PCI_DEVICE_ID_JMICRON_JMC250 &&
2594 !pci_set_dma_mask(pdev, DMA_64BIT_MASK)) 2594 !pci_set_dma_mask(pdev, DMA_BIT_MASK(64)))
2595 if (!pci_set_consistent_dma_mask(pdev, DMA_64BIT_MASK)) 2595 if (!pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64)))
2596 return 1; 2596 return 1;
2597 2597
2598 if (pdev->device == PCI_DEVICE_ID_JMICRON_JMC250 && 2598 if (pdev->device == PCI_DEVICE_ID_JMICRON_JMC250 &&
2599 !pci_set_dma_mask(pdev, DMA_40BIT_MASK)) 2599 !pci_set_dma_mask(pdev, DMA_BIT_MASK(40)))
2600 if (!pci_set_consistent_dma_mask(pdev, DMA_40BIT_MASK)) 2600 if (!pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(40)))
2601 return 1; 2601 return 1;
2602 2602
2603 if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(32))) 2603 if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(32)))
diff --git a/drivers/net/mac89x0.c b/drivers/net/mac89x0.c
index 380a1a54d530..384e072de2e7 100644
--- a/drivers/net/mac89x0.c
+++ b/drivers/net/mac89x0.c
@@ -168,6 +168,17 @@ writereg(struct net_device *dev, int portno, int value)
168 nubus_writew(swab16(value), dev->mem_start + portno); 168 nubus_writew(swab16(value), dev->mem_start + portno);
169} 169}
170 170
171static const struct net_device_ops mac89x0_netdev_ops = {
172 .ndo_open = net_open,
173 .ndo_stop = net_close,
174 .ndo_start_xmit = net_send_packet,
175 .ndo_get_stats = net_get_stats,
176 .ndo_set_multicast_list = set_multicast_list,
177 .ndo_set_mac_address = set_mac_address,
178 .ndo_validate_addr = eth_validate_addr,
179 .ndo_change_mtu = eth_change_mtu,
180};
181
171/* Probe for the CS8900 card in slot E. We won't bother looking 182/* Probe for the CS8900 card in slot E. We won't bother looking
172 anywhere else until we have a really good reason to do so. */ 183 anywhere else until we have a really good reason to do so. */
173struct net_device * __init mac89x0_probe(int unit) 184struct net_device * __init mac89x0_probe(int unit)
@@ -280,12 +291,7 @@ struct net_device * __init mac89x0_probe(int unit)
280 291
281 printk(" IRQ %d ADDR %pM\n", dev->irq, dev->dev_addr); 292 printk(" IRQ %d ADDR %pM\n", dev->irq, dev->dev_addr);
282 293
283 dev->open = net_open; 294 dev->netdev_ops = &mac89x0_netdev_ops;
284 dev->stop = net_close;
285 dev->hard_start_xmit = net_send_packet;
286 dev->get_stats = net_get_stats;
287 dev->set_multicast_list = &set_multicast_list;
288 dev->set_mac_address = &set_mac_address;
289 295
290 err = register_netdev(dev); 296 err = register_netdev(dev);
291 if (err) 297 if (err)
diff --git a/drivers/net/macb.c b/drivers/net/macb.c
index f50501013b1c..46073de290cf 100644
--- a/drivers/net/macb.c
+++ b/drivers/net/macb.c
@@ -1100,6 +1100,18 @@ static int macb_ioctl(struct net_device *dev, struct ifreq *rq, int cmd)
1100 return phy_mii_ioctl(phydev, if_mii(rq), cmd); 1100 return phy_mii_ioctl(phydev, if_mii(rq), cmd);
1101} 1101}
1102 1102
1103static const struct net_device_ops macb_netdev_ops = {
1104 .ndo_open = macb_open,
1105 .ndo_stop = macb_close,
1106 .ndo_start_xmit = macb_start_xmit,
1107 .ndo_set_multicast_list = macb_set_rx_mode,
1108 .ndo_get_stats = macb_get_stats,
1109 .ndo_do_ioctl = macb_ioctl,
1110 .ndo_validate_addr = eth_validate_addr,
1111 .ndo_change_mtu = eth_change_mtu,
1112 .ndo_set_mac_address = eth_mac_addr,
1113};
1114
1103static int __init macb_probe(struct platform_device *pdev) 1115static int __init macb_probe(struct platform_device *pdev)
1104{ 1116{
1105 struct eth_platform_data *pdata; 1117 struct eth_platform_data *pdata;
@@ -1175,12 +1187,7 @@ static int __init macb_probe(struct platform_device *pdev)
1175 goto err_out_iounmap; 1187 goto err_out_iounmap;
1176 } 1188 }
1177 1189
1178 dev->open = macb_open; 1190 dev->netdev_ops = &macb_netdev_ops;
1179 dev->stop = macb_close;
1180 dev->hard_start_xmit = macb_start_xmit;
1181 dev->get_stats = macb_get_stats;
1182 dev->set_multicast_list = macb_set_rx_mode;
1183 dev->do_ioctl = macb_ioctl;
1184 netif_napi_add(dev, &bp->napi, macb_poll, 64); 1191 netif_napi_add(dev, &bp->napi, macb_poll, 64);
1185 dev->ethtool_ops = &macb_ethtool_ops; 1192 dev->ethtool_ops = &macb_ethtool_ops;
1186 1193
diff --git a/drivers/net/macsonic.c b/drivers/net/macsonic.c
index 527166e35d56..acd143da161d 100644
--- a/drivers/net/macsonic.c
+++ b/drivers/net/macsonic.c
@@ -167,6 +167,18 @@ static int macsonic_close(struct net_device* dev)
167 return err; 167 return err;
168} 168}
169 169
170static const struct net_device_ops macsonic_netdev_ops = {
171 .ndo_open = macsonic_open,
172 .ndo_stop = macsonic_close,
173 .ndo_start_xmit = sonic_send_packet,
174 .ndo_set_multicast_list = sonic_multicast_list,
175 .ndo_tx_timeout = sonic_tx_timeout,
176 .ndo_get_stats = sonic_get_stats,
177 .ndo_validate_addr = eth_validate_addr,
178 .ndo_change_mtu = eth_change_mtu,
179 .ndo_set_mac_address = eth_mac_addr,
180};
181
170static int __init macsonic_init(struct net_device *dev) 182static int __init macsonic_init(struct net_device *dev)
171{ 183{
172 struct sonic_local* lp = netdev_priv(dev); 184 struct sonic_local* lp = netdev_priv(dev);
@@ -198,12 +210,7 @@ static int __init macsonic_init(struct net_device *dev)
198 lp->rra_laddr = lp->rda_laddr + (SIZEOF_SONIC_RD * SONIC_NUM_RDS 210 lp->rra_laddr = lp->rda_laddr + (SIZEOF_SONIC_RD * SONIC_NUM_RDS
199 * SONIC_BUS_SCALE(lp->dma_bitmode)); 211 * SONIC_BUS_SCALE(lp->dma_bitmode));
200 212
201 dev->open = macsonic_open; 213 dev->netdev_ops = &macsonic_netdev_ops;
202 dev->stop = macsonic_close;
203 dev->hard_start_xmit = sonic_send_packet;
204 dev->get_stats = sonic_get_stats;
205 dev->set_multicast_list = &sonic_multicast_list;
206 dev->tx_timeout = sonic_tx_timeout;
207 dev->watchdog_timeo = TX_TIMEOUT; 214 dev->watchdog_timeo = TX_TIMEOUT;
208 215
209 /* 216 /*
diff --git a/drivers/net/myri10ge/myri10ge.c b/drivers/net/myri10ge/myri10ge.c
index 9eed126a82f0..f2c4a665e93f 100644
--- a/drivers/net/myri10ge/myri10ge.c
+++ b/drivers/net/myri10ge/myri10ge.c
@@ -2447,6 +2447,7 @@ static int myri10ge_open(struct net_device *dev)
2447 lro_mgr->lro_arr = ss->rx_done.lro_desc; 2447 lro_mgr->lro_arr = ss->rx_done.lro_desc;
2448 lro_mgr->get_frag_header = myri10ge_get_frag_header; 2448 lro_mgr->get_frag_header = myri10ge_get_frag_header;
2449 lro_mgr->max_aggr = myri10ge_lro_max_pkts; 2449 lro_mgr->max_aggr = myri10ge_lro_max_pkts;
2450 lro_mgr->frag_align_pad = 2;
2450 if (lro_mgr->max_aggr > MAX_SKB_FRAGS) 2451 if (lro_mgr->max_aggr > MAX_SKB_FRAGS)
2451 lro_mgr->max_aggr = MAX_SKB_FRAGS; 2452 lro_mgr->max_aggr = MAX_SKB_FRAGS;
2452 2453
diff --git a/drivers/net/phy/fixed.c b/drivers/net/phy/fixed.c
index cf24cc34debe..e7070515d2e3 100644
--- a/drivers/net/phy/fixed.c
+++ b/drivers/net/phy/fixed.c
@@ -19,6 +19,7 @@
19#include <linux/mii.h> 19#include <linux/mii.h>
20#include <linux/phy.h> 20#include <linux/phy.h>
21#include <linux/phy_fixed.h> 21#include <linux/phy_fixed.h>
22#include <linux/err.h>
22 23
23#define MII_REGS_NUM 29 24#define MII_REGS_NUM 29
24 25
@@ -207,8 +208,8 @@ static int __init fixed_mdio_bus_init(void)
207 int ret; 208 int ret;
208 209
209 pdev = platform_device_register_simple("Fixed MDIO bus", 0, NULL, 0); 210 pdev = platform_device_register_simple("Fixed MDIO bus", 0, NULL, 0);
210 if (!pdev) { 211 if (IS_ERR(pdev)) {
211 ret = -ENOMEM; 212 ret = PTR_ERR(pdev);
212 goto err_pdev; 213 goto err_pdev;
213 } 214 }
214 215
diff --git a/drivers/net/phy/marvell.c b/drivers/net/phy/marvell.c
index eb6411c4694f..7a3ec9d39a9a 100644
--- a/drivers/net/phy/marvell.c
+++ b/drivers/net/phy/marvell.c
@@ -69,6 +69,11 @@
69#define MII_M1111_COPPER 0 69#define MII_M1111_COPPER 0
70#define MII_M1111_FIBER 1 70#define MII_M1111_FIBER 1
71 71
72#define MII_88E1121_PHY_LED_CTRL 16
73#define MII_88E1121_PHY_LED_PAGE 3
74#define MII_88E1121_PHY_LED_DEF 0x0030
75#define MII_88E1121_PHY_PAGE 22
76
72#define MII_M1011_PHY_STATUS 0x11 77#define MII_M1011_PHY_STATUS 0x11
73#define MII_M1011_PHY_STATUS_1000 0x8000 78#define MII_M1011_PHY_STATUS_1000 0x8000
74#define MII_M1011_PHY_STATUS_100 0x4000 79#define MII_M1011_PHY_STATUS_100 0x4000
@@ -154,6 +159,30 @@ static int marvell_config_aneg(struct phy_device *phydev)
154 return err; 159 return err;
155} 160}
156 161
162static int m88e1121_config_aneg(struct phy_device *phydev)
163{
164 int err, temp;
165
166 err = phy_write(phydev, MII_BMCR, BMCR_RESET);
167 if (err < 0)
168 return err;
169
170 err = phy_write(phydev, MII_M1011_PHY_SCR,
171 MII_M1011_PHY_SCR_AUTO_CROSS);
172 if (err < 0)
173 return err;
174
175 temp = phy_read(phydev, MII_88E1121_PHY_PAGE);
176
177 phy_write(phydev, MII_88E1121_PHY_PAGE, MII_88E1121_PHY_LED_PAGE);
178 phy_write(phydev, MII_88E1121_PHY_LED_CTRL, MII_88E1121_PHY_LED_DEF);
179 phy_write(phydev, MII_88E1121_PHY_PAGE, temp);
180
181 err = genphy_config_aneg(phydev);
182
183 return err;
184}
185
157static int m88e1111_config_init(struct phy_device *phydev) 186static int m88e1111_config_init(struct phy_device *phydev)
158{ 187{
159 int err; 188 int err;
@@ -429,6 +458,18 @@ static int marvell_read_status(struct phy_device *phydev)
429 return 0; 458 return 0;
430} 459}
431 460
461static int m88e1121_did_interrupt(struct phy_device *phydev)
462{
463 int imask;
464
465 imask = phy_read(phydev, MII_M1011_IEVENT);
466
467 if (imask & MII_M1011_IMASK_INIT)
468 return 1;
469
470 return 0;
471}
472
432static struct phy_driver marvell_drivers[] = { 473static struct phy_driver marvell_drivers[] = {
433 { 474 {
434 .phy_id = 0x01410c60, 475 .phy_id = 0x01410c60,
@@ -482,6 +523,19 @@ static struct phy_driver marvell_drivers[] = {
482 .driver = {.owner = THIS_MODULE,}, 523 .driver = {.owner = THIS_MODULE,},
483 }, 524 },
484 { 525 {
526 .phy_id = 0x01410cb0,
527 .phy_id_mask = 0xfffffff0,
528 .name = "Marvell 88E1121R",
529 .features = PHY_GBIT_FEATURES,
530 .flags = PHY_HAS_INTERRUPT,
531 .config_aneg = &m88e1121_config_aneg,
532 .read_status = &marvell_read_status,
533 .ack_interrupt = &marvell_ack_interrupt,
534 .config_intr = &marvell_config_intr,
535 .did_interrupt = &m88e1121_did_interrupt,
536 .driver = { .owner = THIS_MODULE },
537 },
538 {
485 .phy_id = 0x01410cd0, 539 .phy_id = 0x01410cd0,
486 .phy_id_mask = 0xfffffff0, 540 .phy_id_mask = 0xfffffff0,
487 .name = "Marvell 88E1145", 541 .name = "Marvell 88E1145",
diff --git a/drivers/net/phy/phy.c b/drivers/net/phy/phy.c
index 3ff1f425f1bb..61755cbd978e 100644
--- a/drivers/net/phy/phy.c
+++ b/drivers/net/phy/phy.c
@@ -434,7 +434,7 @@ void phy_start_machine(struct phy_device *phydev,
434 phydev->adjust_state = handler; 434 phydev->adjust_state = handler;
435 435
436 INIT_DELAYED_WORK(&phydev->state_queue, phy_state_machine); 436 INIT_DELAYED_WORK(&phydev->state_queue, phy_state_machine);
437 schedule_delayed_work(&phydev->state_queue, jiffies + HZ); 437 schedule_delayed_work(&phydev->state_queue, HZ);
438} 438}
439 439
440/** 440/**
@@ -655,6 +655,10 @@ static void phy_change(struct work_struct *work)
655 struct phy_device *phydev = 655 struct phy_device *phydev =
656 container_of(work, struct phy_device, phy_queue); 656 container_of(work, struct phy_device, phy_queue);
657 657
658 if (phydev->drv->did_interrupt &&
659 !phydev->drv->did_interrupt(phydev))
660 goto ignore;
661
658 err = phy_disable_interrupts(phydev); 662 err = phy_disable_interrupts(phydev);
659 663
660 if (err) 664 if (err)
@@ -681,6 +685,11 @@ static void phy_change(struct work_struct *work)
681 685
682 return; 686 return;
683 687
688ignore:
689 atomic_dec(&phydev->irq_disable);
690 enable_irq(phydev->irq);
691 return;
692
684irq_enable_err: 693irq_enable_err:
685 disable_irq(phydev->irq); 694 disable_irq(phydev->irq);
686 atomic_inc(&phydev->irq_disable); 695 atomic_inc(&phydev->irq_disable);
@@ -937,6 +946,5 @@ static void phy_state_machine(struct work_struct *work)
937 if (err < 0) 946 if (err < 0)
938 phy_error(phydev); 947 phy_error(phydev);
939 948
940 schedule_delayed_work(&phydev->state_queue, 949 schedule_delayed_work(&phydev->state_queue, PHY_STATE_TIME * HZ);
941 jiffies + PHY_STATE_TIME * HZ);
942} 950}
diff --git a/drivers/net/sfc/efx.c b/drivers/net/sfc/efx.c
index dee23b159df2..7269a426051c 100644
--- a/drivers/net/sfc/efx.c
+++ b/drivers/net/sfc/efx.c
@@ -448,9 +448,6 @@ static void efx_init_channels(struct efx_nic *efx)
448 448
449 WARN_ON(channel->rx_pkt != NULL); 449 WARN_ON(channel->rx_pkt != NULL);
450 efx_rx_strategy(channel); 450 efx_rx_strategy(channel);
451
452 netif_napi_add(channel->napi_dev, &channel->napi_str,
453 efx_poll, napi_weight);
454 } 451 }
455} 452}
456 453
@@ -1321,6 +1318,8 @@ static int efx_init_napi(struct efx_nic *efx)
1321 1318
1322 efx_for_each_channel(channel, efx) { 1319 efx_for_each_channel(channel, efx) {
1323 channel->napi_dev = efx->net_dev; 1320 channel->napi_dev = efx->net_dev;
1321 netif_napi_add(channel->napi_dev, &channel->napi_str,
1322 efx_poll, napi_weight);
1324 } 1323 }
1325 return 0; 1324 return 0;
1326} 1325}
@@ -1330,6 +1329,8 @@ static void efx_fini_napi(struct efx_nic *efx)
1330 struct efx_channel *channel; 1329 struct efx_channel *channel;
1331 1330
1332 efx_for_each_channel(channel, efx) { 1331 efx_for_each_channel(channel, efx) {
1332 if (channel->napi_dev)
1333 netif_napi_del(&channel->napi_str);
1333 channel->napi_dev = NULL; 1334 channel->napi_dev = NULL;
1334 } 1335 }
1335} 1336}
diff --git a/drivers/net/sfc/falcon.c b/drivers/net/sfc/falcon.c
index d4629ab2c614..466a8abb0053 100644
--- a/drivers/net/sfc/falcon.c
+++ b/drivers/net/sfc/falcon.c
@@ -1176,9 +1176,9 @@ void falcon_sim_phy_event(struct efx_nic *efx)
1176 1176
1177 EFX_POPULATE_QWORD_1(phy_event, EV_CODE, GLOBAL_EV_DECODE); 1177 EFX_POPULATE_QWORD_1(phy_event, EV_CODE, GLOBAL_EV_DECODE);
1178 if (EFX_IS10G(efx)) 1178 if (EFX_IS10G(efx))
1179 EFX_SET_OWORD_FIELD(phy_event, XG_PHY_INTR, 1); 1179 EFX_SET_QWORD_FIELD(phy_event, XG_PHY_INTR, 1);
1180 else 1180 else
1181 EFX_SET_OWORD_FIELD(phy_event, G_PHY0_INTR, 1); 1181 EFX_SET_QWORD_FIELD(phy_event, G_PHY0_INTR, 1);
1182 1182
1183 falcon_generate_event(&efx->channel[0], &phy_event); 1183 falcon_generate_event(&efx->channel[0], &phy_event);
1184} 1184}
diff --git a/drivers/net/sh_eth.c b/drivers/net/sh_eth.c
index 7b1882765a0c..3ab28bb00c12 100644
--- a/drivers/net/sh_eth.c
+++ b/drivers/net/sh_eth.c
@@ -1188,6 +1188,19 @@ out:
1188 return ret; 1188 return ret;
1189} 1189}
1190 1190
1191static const struct net_device_ops sh_eth_netdev_ops = {
1192 .ndo_open = sh_eth_open,
1193 .ndo_stop = sh_eth_close,
1194 .ndo_start_xmit = sh_eth_start_xmit,
1195 .ndo_get_stats = sh_eth_get_stats,
1196 .ndo_set_multicast_list = sh_eth_set_multicast_list,
1197 .ndo_tx_timeout = sh_eth_tx_timeout,
1198 .ndo_do_ioctl = sh_eth_do_ioctl,
1199 .ndo_validate_addr = eth_validate_addr,
1200 .ndo_set_mac_address = eth_mac_addr,
1201 .ndo_change_mtu = eth_change_mtu,
1202};
1203
1191static int sh_eth_drv_probe(struct platform_device *pdev) 1204static int sh_eth_drv_probe(struct platform_device *pdev)
1192{ 1205{
1193 int ret, i, devno = 0; 1206 int ret, i, devno = 0;
@@ -1240,13 +1253,7 @@ static int sh_eth_drv_probe(struct platform_device *pdev)
1240 mdp->edmac_endian = pd->edmac_endian; 1253 mdp->edmac_endian = pd->edmac_endian;
1241 1254
1242 /* set function */ 1255 /* set function */
1243 ndev->open = sh_eth_open; 1256 ndev->netdev_ops = &sh_eth_netdev_ops;
1244 ndev->hard_start_xmit = sh_eth_start_xmit;
1245 ndev->stop = sh_eth_close;
1246 ndev->get_stats = sh_eth_get_stats;
1247 ndev->set_multicast_list = sh_eth_set_multicast_list;
1248 ndev->do_ioctl = sh_eth_do_ioctl;
1249 ndev->tx_timeout = sh_eth_tx_timeout;
1250 ndev->watchdog_timeo = TX_TIMEOUT; 1257 ndev->watchdog_timeo = TX_TIMEOUT;
1251 1258
1252 mdp->post_rx = POST_RX >> (devno << 1); 1259 mdp->post_rx = POST_RX >> (devno << 1);
diff --git a/drivers/net/skge.c b/drivers/net/skge.c
index b8978d4af1b7..c11cdd08ec57 100644
--- a/drivers/net/skge.c
+++ b/drivers/net/skge.c
@@ -2674,7 +2674,7 @@ static int skge_down(struct net_device *dev)
2674 if (netif_msg_ifdown(skge)) 2674 if (netif_msg_ifdown(skge))
2675 printk(KERN_INFO PFX "%s: disabling interface\n", dev->name); 2675 printk(KERN_INFO PFX "%s: disabling interface\n", dev->name);
2676 2676
2677 netif_stop_queue(dev); 2677 netif_tx_disable(dev);
2678 2678
2679 if (hw->chip_id == CHIP_ID_GENESIS && hw->phy_type == SK_PHY_XMAC) 2679 if (hw->chip_id == CHIP_ID_GENESIS && hw->phy_type == SK_PHY_XMAC)
2680 del_timer_sync(&skge->link_timer); 2680 del_timer_sync(&skge->link_timer);
@@ -2881,7 +2881,6 @@ static void skge_tx_clean(struct net_device *dev)
2881 } 2881 }
2882 2882
2883 skge->tx_ring.to_clean = e; 2883 skge->tx_ring.to_clean = e;
2884 netif_wake_queue(dev);
2885} 2884}
2886 2885
2887static void skge_tx_timeout(struct net_device *dev) 2886static void skge_tx_timeout(struct net_device *dev)
@@ -2893,6 +2892,7 @@ static void skge_tx_timeout(struct net_device *dev)
2893 2892
2894 skge_write8(skge->hw, Q_ADDR(txqaddr[skge->port], Q_CSR), CSR_STOP); 2893 skge_write8(skge->hw, Q_ADDR(txqaddr[skge->port], Q_CSR), CSR_STOP);
2895 skge_tx_clean(dev); 2894 skge_tx_clean(dev);
2895 netif_wake_queue(dev);
2896} 2896}
2897 2897
2898static int skge_change_mtu(struct net_device *dev, int new_mtu) 2898static int skge_change_mtu(struct net_device *dev, int new_mtu)
diff --git a/drivers/net/sun3_82586.c b/drivers/net/sun3_82586.c
index e0d84772771c..a39c0b9ba8b6 100644
--- a/drivers/net/sun3_82586.c
+++ b/drivers/net/sun3_82586.c
@@ -331,6 +331,18 @@ out:
331 return ERR_PTR(err); 331 return ERR_PTR(err);
332} 332}
333 333
334static const struct net_device_ops sun3_82586_netdev_ops = {
335 .ndo_open = sun3_82586_open,
336 .ndo_stop = sun3_82586_close,
337 .ndo_start_xmit = sun3_82586_send_packet,
338 .ndo_set_multicast_list = set_multicast_list,
339 .ndo_tx_timeout = sun3_82586_timeout,
340 .ndo_get_stats = sun3_82586_get_stats,
341 .ndo_validate_addr = eth_validate_addr,
342 .ndo_set_mac_address = eth_mac_addr,
343 .ndo_change_mtu = eth_change_mtu,
344};
345
334static int __init sun3_82586_probe1(struct net_device *dev,int ioaddr) 346static int __init sun3_82586_probe1(struct net_device *dev,int ioaddr)
335{ 347{
336 int i, size, retval; 348 int i, size, retval;
@@ -381,13 +393,8 @@ static int __init sun3_82586_probe1(struct net_device *dev,int ioaddr)
381 393
382 printk("Memaddr: 0x%lx, Memsize: %d, IRQ %d\n",dev->mem_start,size, dev->irq); 394 printk("Memaddr: 0x%lx, Memsize: %d, IRQ %d\n",dev->mem_start,size, dev->irq);
383 395
384 dev->open = sun3_82586_open; 396 dev->netdev_ops = &sun3_82586_netdev_ops;
385 dev->stop = sun3_82586_close;
386 dev->get_stats = sun3_82586_get_stats;
387 dev->tx_timeout = sun3_82586_timeout;
388 dev->watchdog_timeo = HZ/20; 397 dev->watchdog_timeo = HZ/20;
389 dev->hard_start_xmit = sun3_82586_send_packet;
390 dev->set_multicast_list = set_multicast_list;
391 398
392 dev->if_port = 0; 399 dev->if_port = 0;
393 return 0; 400 return 0;
diff --git a/drivers/net/tc35815.c b/drivers/net/tc35815.c
index d91e95b237b7..0ce2db6ce2bf 100644
--- a/drivers/net/tc35815.c
+++ b/drivers/net/tc35815.c
@@ -862,6 +862,22 @@ static int __devinit tc35815_init_dev_addr(struct net_device *dev)
862 return 0; 862 return 0;
863} 863}
864 864
865static const struct net_device_ops tc35815_netdev_ops = {
866 .ndo_open = tc35815_open,
867 .ndo_stop = tc35815_close,
868 .ndo_start_xmit = tc35815_send_packet,
869 .ndo_get_stats = tc35815_get_stats,
870 .ndo_set_multicast_list = tc35815_set_multicast_list,
871 .ndo_tx_timeout = tc35815_tx_timeout,
872 .ndo_do_ioctl = tc35815_ioctl,
873 .ndo_validate_addr = eth_validate_addr,
874 .ndo_change_mtu = eth_change_mtu,
875 .ndo_set_mac_address = eth_mac_addr,
876#ifdef CONFIG_NET_POLL_CONTROLLER
877 .ndo_poll_controller = tc35815_poll_controller,
878#endif
879};
880
865static int __devinit tc35815_init_one(struct pci_dev *pdev, 881static int __devinit tc35815_init_one(struct pci_dev *pdev,
866 const struct pci_device_id *ent) 882 const struct pci_device_id *ent)
867{ 883{
@@ -904,21 +920,12 @@ static int __devinit tc35815_init_one(struct pci_dev *pdev,
904 ioaddr = pcim_iomap_table(pdev)[1]; 920 ioaddr = pcim_iomap_table(pdev)[1];
905 921
906 /* Initialize the device structure. */ 922 /* Initialize the device structure. */
907 dev->open = tc35815_open; 923 dev->netdev_ops = &tc35815_netdev_ops;
908 dev->hard_start_xmit = tc35815_send_packet;
909 dev->stop = tc35815_close;
910 dev->get_stats = tc35815_get_stats;
911 dev->set_multicast_list = tc35815_set_multicast_list;
912 dev->do_ioctl = tc35815_ioctl;
913 dev->ethtool_ops = &tc35815_ethtool_ops; 924 dev->ethtool_ops = &tc35815_ethtool_ops;
914 dev->tx_timeout = tc35815_tx_timeout;
915 dev->watchdog_timeo = TC35815_TX_TIMEOUT; 925 dev->watchdog_timeo = TC35815_TX_TIMEOUT;
916#ifdef TC35815_NAPI 926#ifdef TC35815_NAPI
917 netif_napi_add(dev, &lp->napi, tc35815_poll, NAPI_WEIGHT); 927 netif_napi_add(dev, &lp->napi, tc35815_poll, NAPI_WEIGHT);
918#endif 928#endif
919#ifdef CONFIG_NET_POLL_CONTROLLER
920 dev->poll_controller = tc35815_poll_controller;
921#endif
922 929
923 dev->irq = pdev->irq; 930 dev->irq = pdev->irq;
924 dev->base_addr = (unsigned long)ioaddr; 931 dev->base_addr = (unsigned long)ioaddr;
diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c
index 6a736dda3ee2..7a837c465960 100644
--- a/drivers/net/tg3.c
+++ b/drivers/net/tg3.c
@@ -12443,8 +12443,13 @@ static int __devinit tg3_get_device_address(struct tg3 *tp)
12443 /* Next, try NVRAM. */ 12443 /* Next, try NVRAM. */
12444 if (!tg3_nvram_read_be32(tp, mac_offset + 0, &hi) && 12444 if (!tg3_nvram_read_be32(tp, mac_offset + 0, &hi) &&
12445 !tg3_nvram_read_be32(tp, mac_offset + 4, &lo)) { 12445 !tg3_nvram_read_be32(tp, mac_offset + 4, &lo)) {
12446 memcpy(&dev->dev_addr[0], ((char *)&hi) + 2, 2); 12446 dev->dev_addr[0] = ((hi >> 16) & 0xff);
12447 memcpy(&dev->dev_addr[2], (char *)&lo, sizeof(lo)); 12447 dev->dev_addr[1] = ((hi >> 24) & 0xff);
12448 dev->dev_addr[2] = ((lo >> 0) & 0xff);
12449 dev->dev_addr[3] = ((lo >> 8) & 0xff);
12450 dev->dev_addr[4] = ((lo >> 16) & 0xff);
12451 dev->dev_addr[5] = ((lo >> 24) & 0xff);
12452
12448 } 12453 }
12449 /* Finally just fetch it out of the MAC control regs. */ 12454 /* Finally just fetch it out of the MAC control regs. */
12450 else { 12455 else {
diff --git a/drivers/net/tsi108_eth.c b/drivers/net/tsi108_eth.c
index bb43e7fb2a50..0f78f99f9b20 100644
--- a/drivers/net/tsi108_eth.c
+++ b/drivers/net/tsi108_eth.c
@@ -1561,6 +1561,18 @@ static const struct ethtool_ops tsi108_ethtool_ops = {
1561 .set_settings = tsi108_set_settings, 1561 .set_settings = tsi108_set_settings,
1562}; 1562};
1563 1563
1564static const struct net_device_ops tsi108_netdev_ops = {
1565 .ndo_open = tsi108_open,
1566 .ndo_stop = tsi108_close,
1567 .ndo_start_xmit = tsi108_send_packet,
1568 .ndo_set_multicast_list = tsi108_set_rx_mode,
1569 .ndo_get_stats = tsi108_get_stats,
1570 .ndo_do_ioctl = tsi108_do_ioctl,
1571 .ndo_set_mac_address = tsi108_set_mac,
1572 .ndo_validate_addr = eth_validate_addr,
1573 .ndo_change_mtu = eth_change_mtu,
1574};
1575
1564static int 1576static int
1565tsi108_init_one(struct platform_device *pdev) 1577tsi108_init_one(struct platform_device *pdev)
1566{ 1578{
@@ -1616,14 +1628,8 @@ tsi108_init_one(struct platform_device *pdev)
1616 data->phy_type = einfo->phy_type; 1628 data->phy_type = einfo->phy_type;
1617 data->irq_num = einfo->irq_num; 1629 data->irq_num = einfo->irq_num;
1618 data->id = pdev->id; 1630 data->id = pdev->id;
1619 dev->open = tsi108_open;
1620 dev->stop = tsi108_close;
1621 dev->hard_start_xmit = tsi108_send_packet;
1622 dev->set_mac_address = tsi108_set_mac;
1623 dev->set_multicast_list = tsi108_set_rx_mode;
1624 dev->get_stats = tsi108_get_stats;
1625 netif_napi_add(dev, &data->napi, tsi108_poll, 64); 1631 netif_napi_add(dev, &data->napi, tsi108_poll, 64);
1626 dev->do_ioctl = tsi108_do_ioctl; 1632 dev->netdev_ops = &tsi108_netdev_ops;
1627 dev->ethtool_ops = &tsi108_ethtool_ops; 1633 dev->ethtool_ops = &tsi108_ethtool_ops;
1628 1634
1629 /* Apparently, the Linux networking code won't use scatter-gather 1635 /* Apparently, the Linux networking code won't use scatter-gather
diff --git a/drivers/net/tun.c b/drivers/net/tun.c
index a1b0697340ba..16716aef184c 100644
--- a/drivers/net/tun.c
+++ b/drivers/net/tun.c
@@ -518,7 +518,7 @@ static inline struct sk_buff *tun_alloc_skb(struct tun_struct *tun,
518 int err; 518 int err;
519 519
520 /* Under a page? Don't bother with paged skb. */ 520 /* Under a page? Don't bother with paged skb. */
521 if (prepad + len < PAGE_SIZE) 521 if (prepad + len < PAGE_SIZE || !linear)
522 linear = len; 522 linear = len;
523 523
524 skb = sock_alloc_send_pskb(sk, prepad + linear, len - linear, noblock, 524 skb = sock_alloc_send_pskb(sk, prepad + linear, len - linear, noblock,
@@ -565,7 +565,8 @@ static __inline__ ssize_t tun_get_user(struct tun_struct *tun,
565 565
566 if ((tun->flags & TUN_TYPE_MASK) == TUN_TAP_DEV) { 566 if ((tun->flags & TUN_TYPE_MASK) == TUN_TAP_DEV) {
567 align = NET_IP_ALIGN; 567 align = NET_IP_ALIGN;
568 if (unlikely(len < ETH_HLEN)) 568 if (unlikely(len < ETH_HLEN ||
569 (gso.hdr_len && gso.hdr_len < ETH_HLEN)))
569 return -EINVAL; 570 return -EINVAL;
570 } 571 }
571 572
diff --git a/drivers/net/via-velocity.c b/drivers/net/via-velocity.c
index fb53ef872df3..754a4b182c1d 100644
--- a/drivers/net/via-velocity.c
+++ b/drivers/net/via-velocity.c
@@ -377,7 +377,7 @@ static void velocity_print_info(struct velocity_info *vptr);
377static int velocity_open(struct net_device *dev); 377static int velocity_open(struct net_device *dev);
378static int velocity_change_mtu(struct net_device *dev, int mtu); 378static int velocity_change_mtu(struct net_device *dev, int mtu);
379static int velocity_xmit(struct sk_buff *skb, struct net_device *dev); 379static int velocity_xmit(struct sk_buff *skb, struct net_device *dev);
380static int velocity_intr(int irq, void *dev_instance); 380static irqreturn_t velocity_intr(int irq, void *dev_instance);
381static void velocity_set_multi(struct net_device *dev); 381static void velocity_set_multi(struct net_device *dev);
382static struct net_device_stats *velocity_get_stats(struct net_device *dev); 382static struct net_device_stats *velocity_get_stats(struct net_device *dev);
383static int velocity_ioctl(struct net_device *dev, struct ifreq *rq, int cmd); 383static int velocity_ioctl(struct net_device *dev, struct ifreq *rq, int cmd);
@@ -2215,7 +2215,7 @@ out:
2215 * efficiently as possible. 2215 * efficiently as possible.
2216 */ 2216 */
2217 2217
2218static int velocity_intr(int irq, void *dev_instance) 2218static irqreturn_t velocity_intr(int irq, void *dev_instance)
2219{ 2219{
2220 struct net_device *dev = dev_instance; 2220 struct net_device *dev = dev_instance;
2221 struct velocity_info *vptr = netdev_priv(dev); 2221 struct velocity_info *vptr = netdev_priv(dev);
diff --git a/drivers/net/wireless/ath9k/pci.c b/drivers/net/wireless/ath9k/pci.c
index 6dbc58580abb..168411d322a2 100644
--- a/drivers/net/wireless/ath9k/pci.c
+++ b/drivers/net/wireless/ath9k/pci.c
@@ -93,14 +93,14 @@ static int ath_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id)
93 if (pci_enable_device(pdev)) 93 if (pci_enable_device(pdev))
94 return -EIO; 94 return -EIO;
95 95
96 ret = pci_set_dma_mask(pdev, DMA_32BIT_MASK); 96 ret = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
97 97
98 if (ret) { 98 if (ret) {
99 printk(KERN_ERR "ath9k: 32-bit DMA not available\n"); 99 printk(KERN_ERR "ath9k: 32-bit DMA not available\n");
100 goto bad; 100 goto bad;
101 } 101 }
102 102
103 ret = pci_set_consistent_dma_mask(pdev, DMA_32BIT_MASK); 103 ret = pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32));
104 104
105 if (ret) { 105 if (ret) {
106 printk(KERN_ERR "ath9k: 32-bit DMA consistent " 106 printk(KERN_ERR "ath9k: 32-bit DMA consistent "
diff --git a/drivers/net/wireless/p54/p54pci.c b/drivers/net/wireless/p54/p54pci.c
index e3569a0a952d..b1610ea4bb3d 100644
--- a/drivers/net/wireless/p54/p54pci.c
+++ b/drivers/net/wireless/p54/p54pci.c
@@ -492,8 +492,8 @@ static int __devinit p54p_probe(struct pci_dev *pdev,
492 goto err_disable_dev; 492 goto err_disable_dev;
493 } 493 }
494 494
495 if (pci_set_dma_mask(pdev, DMA_32BIT_MASK) || 495 if (pci_set_dma_mask(pdev, DMA_BIT_MASK(32)) ||
496 pci_set_consistent_dma_mask(pdev, DMA_32BIT_MASK)) { 496 pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32))) {
497 dev_err(&pdev->dev, "No suitable DMA available\n"); 497 dev_err(&pdev->dev, "No suitable DMA available\n");
498 goto err_free_reg; 498 goto err_free_reg;
499 } 499 }
diff --git a/drivers/net/xtsonic.c b/drivers/net/xtsonic.c
index a12a7211c982..5a4ad156f63e 100644
--- a/drivers/net/xtsonic.c
+++ b/drivers/net/xtsonic.c
@@ -108,6 +108,18 @@ static int xtsonic_close(struct net_device *dev)
108 return err; 108 return err;
109} 109}
110 110
111static const struct net_device_ops xtsonic_netdev_ops = {
112 .ndo_open = xtsonic_open,
113 .ndo_stop = xtsonic_close,
114 .ndo_start_xmit = sonic_send_packet,
115 .ndo_get_stats = sonic_get_stats,
116 .ndo_set_multicast_list = sonic_multicast_list,
117 .ndo_tx_timeout = sonic_tx_timeout,
118 .ndo_validate_addr = eth_validate_addr,
119 .ndo_change_mtu = eth_change_mtu,
120 .ndo_set_mac_address = eth_mac_addr,
121};
122
111static int __init sonic_probe1(struct net_device *dev) 123static int __init sonic_probe1(struct net_device *dev)
112{ 124{
113 static unsigned version_printed = 0; 125 static unsigned version_printed = 0;
@@ -205,12 +217,7 @@ static int __init sonic_probe1(struct net_device *dev)
205 lp->rra_laddr = lp->rda_laddr + (SIZEOF_SONIC_RD * SONIC_NUM_RDS 217 lp->rra_laddr = lp->rda_laddr + (SIZEOF_SONIC_RD * SONIC_NUM_RDS
206 * SONIC_BUS_SCALE(lp->dma_bitmode)); 218 * SONIC_BUS_SCALE(lp->dma_bitmode));
207 219
208 dev->open = xtsonic_open; 220 dev->netdev_ops = &xtsonic_netdev_ops;
209 dev->stop = xtsonic_close;
210 dev->hard_start_xmit = sonic_send_packet;
211 dev->get_stats = sonic_get_stats;
212 dev->set_multicast_list = &sonic_multicast_list;
213 dev->tx_timeout = sonic_tx_timeout;
214 dev->watchdog_timeo = TX_TIMEOUT; 221 dev->watchdog_timeo = TX_TIMEOUT;
215 222
216 /* 223 /*
diff --git a/drivers/parisc/superio.c b/drivers/parisc/superio.c
index 4fa3bb2ddfe4..33e5ade774ca 100644
--- a/drivers/parisc/superio.c
+++ b/drivers/parisc/superio.c
@@ -434,7 +434,8 @@ static void __init superio_parport_init(void)
434 0 /*base_hi*/, 434 0 /*base_hi*/,
435 PAR_IRQ, 435 PAR_IRQ,
436 PARPORT_DMA_NONE /* dma */, 436 PARPORT_DMA_NONE /* dma */,
437 NULL /*struct pci_dev* */) ) 437 NULL /*struct pci_dev* */),
438 0 /* shared irq flags */ )
438 439
439 printk(KERN_WARNING PFX "Probing parallel port failed.\n"); 440 printk(KERN_WARNING PFX "Probing parallel port failed.\n");
440#endif /* CONFIG_PARPORT_PC */ 441#endif /* CONFIG_PARPORT_PC */
diff --git a/drivers/s390/block/dasd.c b/drivers/s390/block/dasd.c
index 0570794ccf1c..d1815272c435 100644
--- a/drivers/s390/block/dasd.c
+++ b/drivers/s390/block/dasd.c
@@ -20,6 +20,7 @@
20#include <linux/slab.h> 20#include <linux/slab.h>
21#include <linux/buffer_head.h> 21#include <linux/buffer_head.h>
22#include <linux/hdreg.h> 22#include <linux/hdreg.h>
23#include <linux/async.h>
23 24
24#include <asm/ccwdev.h> 25#include <asm/ccwdev.h>
25#include <asm/ebcdic.h> 26#include <asm/ebcdic.h>
@@ -480,8 +481,10 @@ static void dasd_change_state(struct dasd_device *device)
480 if (rc && rc != -EAGAIN) 481 if (rc && rc != -EAGAIN)
481 device->target = device->state; 482 device->target = device->state;
482 483
483 if (device->state == device->target) 484 if (device->state == device->target) {
484 wake_up(&dasd_init_waitq); 485 wake_up(&dasd_init_waitq);
486 dasd_put_device(device);
487 }
485 488
486 /* let user-space know that the device status changed */ 489 /* let user-space know that the device status changed */
487 kobject_uevent(&device->cdev->dev.kobj, KOBJ_CHANGE); 490 kobject_uevent(&device->cdev->dev.kobj, KOBJ_CHANGE);
@@ -513,12 +516,15 @@ void dasd_kick_device(struct dasd_device *device)
513 */ 516 */
514void dasd_set_target_state(struct dasd_device *device, int target) 517void dasd_set_target_state(struct dasd_device *device, int target)
515{ 518{
519 dasd_get_device(device);
516 /* If we are in probeonly mode stop at DASD_STATE_READY. */ 520 /* If we are in probeonly mode stop at DASD_STATE_READY. */
517 if (dasd_probeonly && target > DASD_STATE_READY) 521 if (dasd_probeonly && target > DASD_STATE_READY)
518 target = DASD_STATE_READY; 522 target = DASD_STATE_READY;
519 if (device->target != target) { 523 if (device->target != target) {
520 if (device->state == target) 524 if (device->state == target) {
521 wake_up(&dasd_init_waitq); 525 wake_up(&dasd_init_waitq);
526 dasd_put_device(device);
527 }
522 device->target = target; 528 device->target = target;
523 } 529 }
524 if (device->state != device->target) 530 if (device->state != device->target)
@@ -2148,6 +2154,22 @@ dasd_exit(void)
2148 * SECTION: common functions for ccw_driver use 2154 * SECTION: common functions for ccw_driver use
2149 */ 2155 */
2150 2156
2157static void dasd_generic_auto_online(void *data, async_cookie_t cookie)
2158{
2159 struct ccw_device *cdev = data;
2160 int ret;
2161
2162 ret = ccw_device_set_online(cdev);
2163 if (ret)
2164 pr_warning("%s: Setting the DASD online failed with rc=%d\n",
2165 dev_name(&cdev->dev), ret);
2166 else {
2167 struct dasd_device *device = dasd_device_from_cdev(cdev);
2168 wait_event(dasd_init_waitq, _wait_for_device(device));
2169 dasd_put_device(device);
2170 }
2171}
2172
2151/* 2173/*
2152 * Initial attempt at a probe function. this can be simplified once 2174 * Initial attempt at a probe function. this can be simplified once
2153 * the other detection code is gone. 2175 * the other detection code is gone.
@@ -2180,10 +2202,7 @@ int dasd_generic_probe(struct ccw_device *cdev,
2180 */ 2202 */
2181 if ((dasd_get_feature(cdev, DASD_FEATURE_INITIAL_ONLINE) > 0 ) || 2203 if ((dasd_get_feature(cdev, DASD_FEATURE_INITIAL_ONLINE) > 0 ) ||
2182 (dasd_autodetect && dasd_busid_known(dev_name(&cdev->dev)) != 0)) 2204 (dasd_autodetect && dasd_busid_known(dev_name(&cdev->dev)) != 0))
2183 ret = ccw_device_set_online(cdev); 2205 async_schedule(dasd_generic_auto_online, cdev);
2184 if (ret)
2185 pr_warning("%s: Setting the DASD online failed with rc=%d\n",
2186 dev_name(&cdev->dev), ret);
2187 return 0; 2206 return 0;
2188} 2207}
2189 2208
@@ -2290,13 +2309,7 @@ int dasd_generic_set_online(struct ccw_device *cdev,
2290 } else 2309 } else
2291 pr_debug("dasd_generic device %s found\n", 2310 pr_debug("dasd_generic device %s found\n",
2292 dev_name(&cdev->dev)); 2311 dev_name(&cdev->dev));
2293
2294 /* FIXME: we have to wait for the root device but we don't want
2295 * to wait for each single device but for all at once. */
2296 wait_event(dasd_init_waitq, _wait_for_device(device));
2297
2298 dasd_put_device(device); 2312 dasd_put_device(device);
2299
2300 return rc; 2313 return rc;
2301} 2314}
2302 2315
diff --git a/drivers/s390/block/dasd_eckd.c b/drivers/s390/block/dasd_eckd.c
index 21254793c604..cb52da033f06 100644
--- a/drivers/s390/block/dasd_eckd.c
+++ b/drivers/s390/block/dasd_eckd.c
@@ -2019,15 +2019,23 @@ static struct dasd_ccw_req *dasd_eckd_build_cp_cmd_track(
2019 ccw++; 2019 ccw++;
2020 recid += count; 2020 recid += count;
2021 new_track = 0; 2021 new_track = 0;
2022 /* first idaw for a ccw may start anywhere */
2023 if (!idaw_dst)
2024 idaw_dst = dst;
2022 } 2025 }
2023 /* If we start a new idaw, everything is fine and the 2026 /* If we start a new idaw, we must make sure that it
2024 * start of the new idaw is the start of this segment. 2027 * starts on an IDA_BLOCK_SIZE boundary.
2025 * If we continue an idaw, we must make sure that the 2028 * If we continue an idaw, we must make sure that the
2026 * current segment begins where the so far accumulated 2029 * current segment begins where the so far accumulated
2027 * idaw ends 2030 * idaw ends
2028 */ 2031 */
2029 if (!idaw_dst) 2032 if (!idaw_dst) {
2030 idaw_dst = dst; 2033 if (__pa(dst) & (IDA_BLOCK_SIZE-1)) {
2034 dasd_sfree_request(cqr, startdev);
2035 return ERR_PTR(-ERANGE);
2036 } else
2037 idaw_dst = dst;
2038 }
2031 if ((idaw_dst + idaw_len) != dst) { 2039 if ((idaw_dst + idaw_len) != dst) {
2032 dasd_sfree_request(cqr, startdev); 2040 dasd_sfree_request(cqr, startdev);
2033 return ERR_PTR(-ERANGE); 2041 return ERR_PTR(-ERANGE);
diff --git a/drivers/s390/cio/qdio_main.c b/drivers/s390/cio/qdio_main.c
index 9e8a2914259b..accd957454e7 100644
--- a/drivers/s390/cio/qdio_main.c
+++ b/drivers/s390/cio/qdio_main.c
@@ -881,42 +881,6 @@ no_handler:
881 qdio_set_state(irq_ptr, QDIO_IRQ_STATE_STOPPED); 881 qdio_set_state(irq_ptr, QDIO_IRQ_STATE_STOPPED);
882} 882}
883 883
884static void qdio_call_shutdown(struct work_struct *work)
885{
886 struct ccw_device_private *priv;
887 struct ccw_device *cdev;
888
889 priv = container_of(work, struct ccw_device_private, kick_work);
890 cdev = priv->cdev;
891 qdio_shutdown(cdev, QDIO_FLAG_CLEANUP_USING_CLEAR);
892 put_device(&cdev->dev);
893}
894
895static void qdio_int_error(struct ccw_device *cdev)
896{
897 struct qdio_irq *irq_ptr = cdev->private->qdio_data;
898
899 switch (irq_ptr->state) {
900 case QDIO_IRQ_STATE_INACTIVE:
901 case QDIO_IRQ_STATE_CLEANUP:
902 qdio_set_state(irq_ptr, QDIO_IRQ_STATE_ERR);
903 break;
904 case QDIO_IRQ_STATE_ESTABLISHED:
905 case QDIO_IRQ_STATE_ACTIVE:
906 qdio_set_state(irq_ptr, QDIO_IRQ_STATE_STOPPED);
907 if (get_device(&cdev->dev)) {
908 /* Can't call shutdown from interrupt context. */
909 PREPARE_WORK(&cdev->private->kick_work,
910 qdio_call_shutdown);
911 queue_work(ccw_device_work, &cdev->private->kick_work);
912 }
913 break;
914 default:
915 WARN_ON(1);
916 }
917 wake_up(&cdev->private->wait_q);
918}
919
920static int qdio_establish_check_errors(struct ccw_device *cdev, int cstat, 884static int qdio_establish_check_errors(struct ccw_device *cdev, int cstat,
921 int dstat) 885 int dstat)
922{ 886{
@@ -973,10 +937,8 @@ void qdio_int_handler(struct ccw_device *cdev, unsigned long intparm,
973 switch (PTR_ERR(irb)) { 937 switch (PTR_ERR(irb)) {
974 case -EIO: 938 case -EIO:
975 DBF_ERROR("%4x IO error", irq_ptr->schid.sch_no); 939 DBF_ERROR("%4x IO error", irq_ptr->schid.sch_no);
976 return; 940 qdio_set_state(irq_ptr, QDIO_IRQ_STATE_ERR);
977 case -ETIMEDOUT: 941 wake_up(&cdev->private->wait_q);
978 DBF_ERROR("%4x IO timeout", irq_ptr->schid.sch_no);
979 qdio_int_error(cdev);
980 return; 942 return;
981 default: 943 default:
982 WARN_ON(1); 944 WARN_ON(1);
@@ -1001,7 +963,6 @@ void qdio_int_handler(struct ccw_device *cdev, unsigned long intparm,
1001 case QDIO_IRQ_STATE_ACTIVE: 963 case QDIO_IRQ_STATE_ACTIVE:
1002 if (cstat & SCHN_STAT_PCI) { 964 if (cstat & SCHN_STAT_PCI) {
1003 qdio_int_handler_pci(irq_ptr); 965 qdio_int_handler_pci(irq_ptr);
1004 /* no state change so no need to wake up wait_q */
1005 return; 966 return;
1006 } 967 }
1007 if ((cstat & ~SCHN_STAT_PCI) || dstat) { 968 if ((cstat & ~SCHN_STAT_PCI) || dstat) {
diff --git a/drivers/sbus/char/jsflash.c b/drivers/sbus/char/jsflash.c
index e6d1fc8c54f1..a85ad05e8548 100644
--- a/drivers/sbus/char/jsflash.c
+++ b/drivers/sbus/char/jsflash.c
@@ -383,18 +383,22 @@ static int jsf_ioctl_program(void __user *arg)
383 return 0; 383 return 0;
384} 384}
385 385
386static int jsf_ioctl(struct inode *inode, struct file *f, unsigned int cmd, 386static long jsf_ioctl(struct file *f, unsigned int cmd, unsigned long arg)
387 unsigned long arg)
388{ 387{
388 lock_kernel();
389 int error = -ENOTTY; 389 int error = -ENOTTY;
390 void __user *argp = (void __user *)arg; 390 void __user *argp = (void __user *)arg;
391 391
392 if (!capable(CAP_SYS_ADMIN)) 392 if (!capable(CAP_SYS_ADMIN)) {
393 unlock_kernel();
393 return -EPERM; 394 return -EPERM;
395 }
394 switch (cmd) { 396 switch (cmd) {
395 case JSFLASH_IDENT: 397 case JSFLASH_IDENT:
396 if (copy_to_user(argp, &jsf0.id, JSFIDSZ)) 398 if (copy_to_user(argp, &jsf0.id, JSFIDSZ)) {
399 unlock_kernel();
397 return -EFAULT; 400 return -EFAULT;
401 }
398 break; 402 break;
399 case JSFLASH_ERASE: 403 case JSFLASH_ERASE:
400 error = jsf_ioctl_erase(arg); 404 error = jsf_ioctl_erase(arg);
@@ -404,6 +408,7 @@ static int jsf_ioctl(struct inode *inode, struct file *f, unsigned int cmd,
404 break; 408 break;
405 } 409 }
406 410
411 unlock_kernel();
407 return error; 412 return error;
408} 413}
409 414
@@ -439,7 +444,7 @@ static const struct file_operations jsf_fops = {
439 .llseek = jsf_lseek, 444 .llseek = jsf_lseek,
440 .read = jsf_read, 445 .read = jsf_read,
441 .write = jsf_write, 446 .write = jsf_write,
442 .ioctl = jsf_ioctl, 447 .unlocked_ioctl = jsf_ioctl,
443 .mmap = jsf_mmap, 448 .mmap = jsf_mmap,
444 .open = jsf_open, 449 .open = jsf_open,
445 .release = jsf_release, 450 .release = jsf_release,
diff --git a/drivers/sbus/char/uctrl.c b/drivers/sbus/char/uctrl.c
index 27993c37775d..2c56fd56ec63 100644
--- a/drivers/sbus/char/uctrl.c
+++ b/drivers/sbus/char/uctrl.c
@@ -197,9 +197,8 @@ static struct uctrl_driver {
197static void uctrl_get_event_status(struct uctrl_driver *); 197static void uctrl_get_event_status(struct uctrl_driver *);
198static void uctrl_get_external_status(struct uctrl_driver *); 198static void uctrl_get_external_status(struct uctrl_driver *);
199 199
200static int 200static long
201uctrl_ioctl(struct inode *inode, struct file *file, 201uctrl_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
202 unsigned int cmd, unsigned long arg)
203{ 202{
204 switch (cmd) { 203 switch (cmd) {
205 default: 204 default:
@@ -226,7 +225,7 @@ static irqreturn_t uctrl_interrupt(int irq, void *dev_id)
226static const struct file_operations uctrl_fops = { 225static const struct file_operations uctrl_fops = {
227 .owner = THIS_MODULE, 226 .owner = THIS_MODULE,
228 .llseek = no_llseek, 227 .llseek = no_llseek,
229 .ioctl = uctrl_ioctl, 228 .unlocked_ioctl = uctrl_ioctl,
230 .open = uctrl_open, 229 .open = uctrl_open,
231}; 230};
232 231
diff --git a/drivers/scsi/3w-9xxx.c b/drivers/scsi/3w-9xxx.c
index fdb14ec4fd47..8b7983aba8f7 100644
--- a/drivers/scsi/3w-9xxx.c
+++ b/drivers/scsi/3w-9xxx.c
@@ -2234,10 +2234,10 @@ static int twa_resume(struct pci_dev *pdev)
2234 pci_set_master(pdev); 2234 pci_set_master(pdev);
2235 pci_try_set_mwi(pdev); 2235 pci_try_set_mwi(pdev);
2236 2236
2237 if (pci_set_dma_mask(pdev, DMA_64BIT_MASK) 2237 if (pci_set_dma_mask(pdev, DMA_BIT_MASK(64))
2238 || pci_set_consistent_dma_mask(pdev, DMA_64BIT_MASK)) 2238 || pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(64)))
2239 if (pci_set_dma_mask(pdev, DMA_32BIT_MASK) 2239 if (pci_set_dma_mask(pdev, DMA_BIT_MASK(32))
2240 || pci_set_consistent_dma_mask(pdev, DMA_32BIT_MASK)) { 2240 || pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32))) {
2241 TW_PRINTK(host, TW_DRIVER, 0x40, "Failed to set dma mask during resume"); 2241 TW_PRINTK(host, TW_DRIVER, 0x40, "Failed to set dma mask during resume");
2242 retval = -ENODEV; 2242 retval = -ENODEV;
2243 goto out_disable_device; 2243 goto out_disable_device;
diff --git a/drivers/scsi/aacraid/aachba.c b/drivers/scsi/aacraid/aachba.c
index 280261c451d6..2a889853a106 100644
--- a/drivers/scsi/aacraid/aachba.c
+++ b/drivers/scsi/aacraid/aachba.c
@@ -1378,7 +1378,7 @@ int aac_get_adapter_info(struct aac_dev* dev)
1378 if (dev->nondasd_support && !dev->in_reset) 1378 if (dev->nondasd_support && !dev->in_reset)
1379 printk(KERN_INFO "%s%d: Non-DASD support enabled.\n",dev->name, dev->id); 1379 printk(KERN_INFO "%s%d: Non-DASD support enabled.\n",dev->name, dev->id);
1380 1380
1381 if (dma_get_required_mask(&dev->pdev->dev) > DMA_32BIT_MASK) 1381 if (dma_get_required_mask(&dev->pdev->dev) > DMA_BIT_MASK(32))
1382 dev->needs_dac = 1; 1382 dev->needs_dac = 1;
1383 dev->dac_support = 0; 1383 dev->dac_support = 0;
1384 if ((sizeof(dma_addr_t) > 4) && dev->needs_dac && 1384 if ((sizeof(dma_addr_t) > 4) && dev->needs_dac &&
diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c
index b1bd3fc7bae8..36fd2e75da1c 100644
--- a/drivers/scsi/lpfc/lpfc_scsi.c
+++ b/drivers/scsi/lpfc/lpfc_scsi.c
@@ -1394,7 +1394,7 @@ lpfc_parse_bg_err(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd,
1394 */ 1394 */
1395 cmd->sense_buffer[8] = 0; /* Information */ 1395 cmd->sense_buffer[8] = 0; /* Information */
1396 cmd->sense_buffer[9] = 0xa; /* Add. length */ 1396 cmd->sense_buffer[9] = 0xa; /* Add. length */
1397 do_div(bghm, cmd->device->sector_size); 1397 bghm /= cmd->device->sector_size;
1398 1398
1399 failing_sector = scsi_get_lba(cmd); 1399 failing_sector = scsi_get_lba(cmd);
1400 failing_sector += bghm; 1400 failing_sector += bghm;
diff --git a/drivers/scsi/mpt2sas/mpt2sas_base.c b/drivers/scsi/mpt2sas/mpt2sas_base.c
index 52427a8324f5..a91f5143ceac 100644
--- a/drivers/scsi/mpt2sas/mpt2sas_base.c
+++ b/drivers/scsi/mpt2sas/mpt2sas_base.c
@@ -855,9 +855,9 @@ _base_config_dma_addressing(struct MPT2SAS_ADAPTER *ioc, struct pci_dev *pdev)
855 if (sizeof(dma_addr_t) > 4) { 855 if (sizeof(dma_addr_t) > 4) {
856 const uint64_t required_mask = 856 const uint64_t required_mask =
857 dma_get_required_mask(&pdev->dev); 857 dma_get_required_mask(&pdev->dev);
858 if ((required_mask > DMA_32BIT_MASK) && !pci_set_dma_mask(pdev, 858 if ((required_mask > DMA_BIT_MASK(32)) && !pci_set_dma_mask(pdev,
859 DMA_64BIT_MASK) && !pci_set_consistent_dma_mask(pdev, 859 DMA_BIT_MASK(64)) && !pci_set_consistent_dma_mask(pdev,
860 DMA_64BIT_MASK)) { 860 DMA_BIT_MASK(64))) {
861 ioc->base_add_sg_single = &_base_add_sg_single_64; 861 ioc->base_add_sg_single = &_base_add_sg_single_64;
862 ioc->sge_size = sizeof(Mpi2SGESimple64_t); 862 ioc->sge_size = sizeof(Mpi2SGESimple64_t);
863 desc = "64"; 863 desc = "64";
@@ -865,8 +865,8 @@ _base_config_dma_addressing(struct MPT2SAS_ADAPTER *ioc, struct pci_dev *pdev)
865 } 865 }
866 } 866 }
867 867
868 if (!pci_set_dma_mask(pdev, DMA_32BIT_MASK) 868 if (!pci_set_dma_mask(pdev, DMA_BIT_MASK(32))
869 && !pci_set_consistent_dma_mask(pdev, DMA_32BIT_MASK)) { 869 && !pci_set_consistent_dma_mask(pdev, DMA_BIT_MASK(32))) {
870 ioc->base_add_sg_single = &_base_add_sg_single_32; 870 ioc->base_add_sg_single = &_base_add_sg_single_32;
871 ioc->sge_size = sizeof(Mpi2SGESimple32_t); 871 ioc->sge_size = sizeof(Mpi2SGESimple32_t);
872 desc = "32"; 872 desc = "32";
diff --git a/drivers/sh/intc.c b/drivers/sh/intc.c
index 7fb9b5c4669a..12d13d99b6f0 100644
--- a/drivers/sh/intc.c
+++ b/drivers/sh/intc.c
@@ -44,6 +44,7 @@ struct intc_handle_int {
44struct intc_desc_int { 44struct intc_desc_int {
45 struct list_head list; 45 struct list_head list;
46 struct sys_device sysdev; 46 struct sys_device sysdev;
47 pm_message_t state;
47 unsigned long *reg; 48 unsigned long *reg;
48#ifdef CONFIG_SMP 49#ifdef CONFIG_SMP
49 unsigned long *smp; 50 unsigned long *smp;
@@ -786,18 +787,44 @@ static int intc_suspend(struct sys_device *dev, pm_message_t state)
786 /* get intc controller associated with this sysdev */ 787 /* get intc controller associated with this sysdev */
787 d = container_of(dev, struct intc_desc_int, sysdev); 788 d = container_of(dev, struct intc_desc_int, sysdev);
788 789
789 /* enable wakeup irqs belonging to this intc controller */ 790 switch (state.event) {
790 for_each_irq_desc(irq, desc) { 791 case PM_EVENT_ON:
791 if ((desc->status & IRQ_WAKEUP) && (desc->chip == &d->chip)) 792 if (d->state.event != PM_EVENT_FREEZE)
792 intc_enable(irq); 793 break;
794 for_each_irq_desc(irq, desc) {
795 if (desc->chip != &d->chip)
796 continue;
797 if (desc->status & IRQ_DISABLED)
798 intc_disable(irq);
799 else
800 intc_enable(irq);
801 }
802 break;
803 case PM_EVENT_FREEZE:
804 /* nothing has to be done */
805 break;
806 case PM_EVENT_SUSPEND:
807 /* enable wakeup irqs belonging to this intc controller */
808 for_each_irq_desc(irq, desc) {
809 if ((desc->status & IRQ_WAKEUP) && (desc->chip == &d->chip))
810 intc_enable(irq);
811 }
812 break;
793 } 813 }
814 d->state = state;
794 815
795 return 0; 816 return 0;
796} 817}
797 818
819static int intc_resume(struct sys_device *dev)
820{
821 return intc_suspend(dev, PMSG_ON);
822}
823
798static struct sysdev_class intc_sysdev_class = { 824static struct sysdev_class intc_sysdev_class = {
799 .name = "intc", 825 .name = "intc",
800 .suspend = intc_suspend, 826 .suspend = intc_suspend,
827 .resume = intc_resume,
801}; 828};
802 829
803/* register this intc as sysdev to allow suspend/resume */ 830/* register this intc as sysdev to allow suspend/resume */
diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c
index 643908b74bc0..8eba98c8ed1e 100644
--- a/drivers/spi/spi.c
+++ b/drivers/spi/spi.c
@@ -658,7 +658,7 @@ int spi_write_then_read(struct spi_device *spi,
658 658
659 int status; 659 int status;
660 struct spi_message message; 660 struct spi_message message;
661 struct spi_transfer x; 661 struct spi_transfer x[2];
662 u8 *local_buf; 662 u8 *local_buf;
663 663
664 /* Use preallocated DMA-safe buffer. We can't avoid copying here, 664 /* Use preallocated DMA-safe buffer. We can't avoid copying here,
@@ -669,9 +669,15 @@ int spi_write_then_read(struct spi_device *spi,
669 return -EINVAL; 669 return -EINVAL;
670 670
671 spi_message_init(&message); 671 spi_message_init(&message);
672 memset(&x, 0, sizeof x); 672 memset(x, 0, sizeof x);
673 x.len = n_tx + n_rx; 673 if (n_tx) {
674 spi_message_add_tail(&x, &message); 674 x[0].len = n_tx;
675 spi_message_add_tail(&x[0], &message);
676 }
677 if (n_rx) {
678 x[1].len = n_rx;
679 spi_message_add_tail(&x[1], &message);
680 }
675 681
676 /* ... unless someone else is using the pre-allocated buffer */ 682 /* ... unless someone else is using the pre-allocated buffer */
677 if (!mutex_trylock(&lock)) { 683 if (!mutex_trylock(&lock)) {
@@ -682,15 +688,15 @@ int spi_write_then_read(struct spi_device *spi,
682 local_buf = buf; 688 local_buf = buf;
683 689
684 memcpy(local_buf, txbuf, n_tx); 690 memcpy(local_buf, txbuf, n_tx);
685 x.tx_buf = local_buf; 691 x[0].tx_buf = local_buf;
686 x.rx_buf = local_buf; 692 x[1].rx_buf = local_buf + n_tx;
687 693
688 /* do the i/o */ 694 /* do the i/o */
689 status = spi_sync(spi, &message); 695 status = spi_sync(spi, &message);
690 if (status == 0) 696 if (status == 0)
691 memcpy(rxbuf, x.rx_buf + n_tx, n_rx); 697 memcpy(rxbuf, x[1].rx_buf, n_rx);
692 698
693 if (x.tx_buf == buf) 699 if (x[0].tx_buf == buf)
694 mutex_unlock(&lock); 700 mutex_unlock(&lock);
695 else 701 else
696 kfree(local_buf); 702 kfree(local_buf);
diff --git a/drivers/staging/b3dfg/b3dfg.c b/drivers/staging/b3dfg/b3dfg.c
index 0348072b3ab5..75ebe338c6f2 100644
--- a/drivers/staging/b3dfg/b3dfg.c
+++ b/drivers/staging/b3dfg/b3dfg.c
@@ -1000,7 +1000,7 @@ static int __devinit b3dfg_probe(struct pci_dev *pdev,
1000 1000
1001 pci_set_master(pdev); 1001 pci_set_master(pdev);
1002 1002
1003 r = pci_set_dma_mask(pdev, DMA_32BIT_MASK); 1003 r = pci_set_dma_mask(pdev, DMA_BIT_MASK(32));
1004 if (r) { 1004 if (r) {
1005 dev_err(&pdev->dev, "no usable DMA configuration\n"); 1005 dev_err(&pdev->dev, "no usable DMA configuration\n");
1006 goto err_free_res; 1006 goto err_free_res;
diff --git a/drivers/usb/class/cdc-acm.c b/drivers/usb/class/cdc-acm.c
index 869d47cb6db3..0a69c0977e3f 100644
--- a/drivers/usb/class/cdc-acm.c
+++ b/drivers/usb/class/cdc-acm.c
@@ -546,10 +546,6 @@ static int acm_tty_open(struct tty_struct *tty, struct file *filp)
546 tty->driver_data = acm; 546 tty->driver_data = acm;
547 acm->tty = tty; 547 acm->tty = tty;
548 548
549 /* force low_latency on so that our tty_push actually forces the data through,
550 otherwise it is scheduled, and with high data rates data can get lost. */
551 tty->low_latency = 1;
552
553 if (usb_autopm_get_interface(acm->control) < 0) 549 if (usb_autopm_get_interface(acm->control) < 0)
554 goto early_bail; 550 goto early_bail;
555 else 551 else
diff --git a/drivers/usb/otg/nop-usb-xceiv.c b/drivers/usb/otg/nop-usb-xceiv.c
index 4b933f646f2e..c567168f89af 100644
--- a/drivers/usb/otg/nop-usb-xceiv.c
+++ b/drivers/usb/otg/nop-usb-xceiv.c
@@ -36,14 +36,14 @@ struct nop_usb_xceiv {
36 struct device *dev; 36 struct device *dev;
37}; 37};
38 38
39static u64 nop_xceiv_dmamask = DMA_32BIT_MASK; 39static u64 nop_xceiv_dmamask = DMA_BIT_MASK(32);
40 40
41static struct platform_device nop_xceiv_device = { 41static struct platform_device nop_xceiv_device = {
42 .name = "nop_usb_xceiv", 42 .name = "nop_usb_xceiv",
43 .id = -1, 43 .id = -1,
44 .dev = { 44 .dev = {
45 .dma_mask = &nop_xceiv_dmamask, 45 .dma_mask = &nop_xceiv_dmamask,
46 .coherent_dma_mask = DMA_32BIT_MASK, 46 .coherent_dma_mask = DMA_BIT_MASK(32),
47 .platform_data = NULL, 47 .platform_data = NULL,
48 }, 48 },
49}; 49};
diff --git a/drivers/usb/serial/ti_usb_3410_5052.c b/drivers/usb/serial/ti_usb_3410_5052.c
index 2620bf6fe5e1..9c4c700c7cc6 100644
--- a/drivers/usb/serial/ti_usb_3410_5052.c
+++ b/drivers/usb/serial/ti_usb_3410_5052.c
@@ -1215,20 +1215,22 @@ static void ti_bulk_in_callback(struct urb *urb)
1215 } 1215 }
1216 1216
1217 tty = tty_port_tty_get(&port->port); 1217 tty = tty_port_tty_get(&port->port);
1218 if (tty && urb->actual_length) { 1218 if (tty) {
1219 usb_serial_debug_data(debug, dev, __func__, 1219 if (urb->actual_length) {
1220 urb->actual_length, urb->transfer_buffer); 1220 usb_serial_debug_data(debug, dev, __func__,
1221 1221 urb->actual_length, urb->transfer_buffer);
1222 if (!tport->tp_is_open) 1222
1223 dbg("%s - port closed, dropping data", __func__); 1223 if (!tport->tp_is_open)
1224 else 1224 dbg("%s - port closed, dropping data",
1225 ti_recv(&urb->dev->dev, tty, 1225 __func__);
1226 else
1227 ti_recv(&urb->dev->dev, tty,
1226 urb->transfer_buffer, 1228 urb->transfer_buffer,
1227 urb->actual_length); 1229 urb->actual_length);
1228 1230 spin_lock(&tport->tp_lock);
1229 spin_lock(&tport->tp_lock); 1231 tport->tp_icount.rx += urb->actual_length;
1230 tport->tp_icount.rx += urb->actual_length; 1232 spin_unlock(&tport->tp_lock);
1231 spin_unlock(&tport->tp_lock); 1233 }
1232 tty_kref_put(tty); 1234 tty_kref_put(tty);
1233 } 1235 }
1234 1236
diff --git a/drivers/video/aty/radeon_base.c b/drivers/video/aty/radeon_base.c
index 16bb7e3c0310..6c37e8ee5efe 100644
--- a/drivers/video/aty/radeon_base.c
+++ b/drivers/video/aty/radeon_base.c
@@ -698,8 +698,8 @@ static void __devinit radeon_get_pllinfo(struct radeonfb_info *rinfo)
698found: 698found:
699 /* 699 /*
700 * Some methods fail to retrieve SCLK and MCLK values, we apply default 700 * Some methods fail to retrieve SCLK and MCLK values, we apply default
701 * settings in this case (200Mhz). If that really happne often, we could 701 * settings in this case (200Mhz). If that really happens often, we
702 * fetch from registers instead... 702 * could fetch from registers instead...
703 */ 703 */
704 if (rinfo->pll.mclk == 0) 704 if (rinfo->pll.mclk == 0)
705 rinfo->pll.mclk = 20000; 705 rinfo->pll.mclk = 20000;
diff --git a/drivers/video/backlight/backlight.c b/drivers/video/backlight/backlight.c
index dd37cbcaf8ce..157057c79ca3 100644
--- a/drivers/video/backlight/backlight.c
+++ b/drivers/video/backlight/backlight.c
@@ -35,8 +35,6 @@ static int fb_notifier_callback(struct notifier_block *self,
35 return 0; 35 return 0;
36 36
37 bd = container_of(self, struct backlight_device, fb_notif); 37 bd = container_of(self, struct backlight_device, fb_notif);
38 if (!lock_fb_info(evdata->info))
39 return -ENODEV;
40 mutex_lock(&bd->ops_lock); 38 mutex_lock(&bd->ops_lock);
41 if (bd->ops) 39 if (bd->ops)
42 if (!bd->ops->check_fb || 40 if (!bd->ops->check_fb ||
@@ -49,7 +47,6 @@ static int fb_notifier_callback(struct notifier_block *self,
49 backlight_update_status(bd); 47 backlight_update_status(bd);
50 } 48 }
51 mutex_unlock(&bd->ops_lock); 49 mutex_unlock(&bd->ops_lock);
52 unlock_fb_info(evdata->info);
53 return 0; 50 return 0;
54} 51}
55 52
diff --git a/drivers/video/backlight/lcd.c b/drivers/video/backlight/lcd.c
index 0bb13df0fa89..b6449470106c 100644
--- a/drivers/video/backlight/lcd.c
+++ b/drivers/video/backlight/lcd.c
@@ -40,8 +40,6 @@ static int fb_notifier_callback(struct notifier_block *self,
40 if (!ld->ops) 40 if (!ld->ops)
41 return 0; 41 return 0;
42 42
43 if (!lock_fb_info(evdata->info))
44 return -ENODEV;
45 mutex_lock(&ld->ops_lock); 43 mutex_lock(&ld->ops_lock);
46 if (!ld->ops->check_fb || ld->ops->check_fb(ld, evdata->info)) { 44 if (!ld->ops->check_fb || ld->ops->check_fb(ld, evdata->info)) {
47 if (event == FB_EVENT_BLANK) { 45 if (event == FB_EVENT_BLANK) {
@@ -53,7 +51,6 @@ static int fb_notifier_callback(struct notifier_block *self,
53 } 51 }
54 } 52 }
55 mutex_unlock(&ld->ops_lock); 53 mutex_unlock(&ld->ops_lock);
56 unlock_fb_info(evdata->info);
57 return 0; 54 return 0;
58} 55}
59 56
diff --git a/drivers/video/cirrusfb.c b/drivers/video/cirrusfb.c
index d42e385f091c..4c2bf923418c 100644
--- a/drivers/video/cirrusfb.c
+++ b/drivers/video/cirrusfb.c
@@ -567,9 +567,7 @@ static int cirrusfb_check_var(struct fb_var_screeninfo *var,
567 default: 567 default:
568 dev_dbg(info->device, 568 dev_dbg(info->device,
569 "Unsupported bpp size: %d\n", var->bits_per_pixel); 569 "Unsupported bpp size: %d\n", var->bits_per_pixel);
570 assert(false); 570 return -EINVAL;
571 /* should never occur */
572 break;
573 } 571 }
574 572
575 if (var->xres_virtual < var->xres) 573 if (var->xres_virtual < var->xres)
diff --git a/drivers/video/console/fbcon.c b/drivers/video/console/fbcon.c
index 2cd500a304f2..471a9a60376a 100644
--- a/drivers/video/console/fbcon.c
+++ b/drivers/video/console/fbcon.c
@@ -2263,9 +2263,12 @@ static void fbcon_generic_blank(struct vc_data *vc, struct fb_info *info,
2263 } 2263 }
2264 2264
2265 2265
2266 if (!lock_fb_info(info))
2267 return;
2266 event.info = info; 2268 event.info = info;
2267 event.data = &blank; 2269 event.data = &blank;
2268 fb_notifier_call_chain(FB_EVENT_CONBLANK, &event); 2270 fb_notifier_call_chain(FB_EVENT_CONBLANK, &event);
2271 unlock_fb_info(info);
2269} 2272}
2270 2273
2271static int fbcon_blank(struct vc_data *vc, int blank, int mode_switch) 2274static int fbcon_blank(struct vc_data *vc, int blank, int mode_switch)
@@ -2956,8 +2959,6 @@ static int fbcon_fb_unregistered(struct fb_info *info)
2956{ 2959{
2957 int i, idx; 2960 int i, idx;
2958 2961
2959 if (!lock_fb_info(info))
2960 return -ENODEV;
2961 idx = info->node; 2962 idx = info->node;
2962 for (i = first_fb_vc; i <= last_fb_vc; i++) { 2963 for (i = first_fb_vc; i <= last_fb_vc; i++) {
2963 if (con2fb_map[i] == idx) 2964 if (con2fb_map[i] == idx)
@@ -2985,8 +2986,6 @@ static int fbcon_fb_unregistered(struct fb_info *info)
2985 if (primary_device == idx) 2986 if (primary_device == idx)
2986 primary_device = -1; 2987 primary_device = -1;
2987 2988
2988 unlock_fb_info(info);
2989
2990 if (!num_registered_fb) 2989 if (!num_registered_fb)
2991 unregister_con_driver(&fb_con); 2990 unregister_con_driver(&fb_con);
2992 2991
@@ -3027,11 +3026,8 @@ static int fbcon_fb_registered(struct fb_info *info)
3027{ 3026{
3028 int ret = 0, i, idx; 3027 int ret = 0, i, idx;
3029 3028
3030 if (!lock_fb_info(info))
3031 return -ENODEV;
3032 idx = info->node; 3029 idx = info->node;
3033 fbcon_select_primary(info); 3030 fbcon_select_primary(info);
3034 unlock_fb_info(info);
3035 3031
3036 if (info_idx == -1) { 3032 if (info_idx == -1) {
3037 for (i = first_fb_vc; i <= last_fb_vc; i++) { 3033 for (i = first_fb_vc; i <= last_fb_vc; i++) {
@@ -3152,53 +3148,23 @@ static int fbcon_event_notify(struct notifier_block *self,
3152 3148
3153 switch(action) { 3149 switch(action) {
3154 case FB_EVENT_SUSPEND: 3150 case FB_EVENT_SUSPEND:
3155 if (!lock_fb_info(info)) {
3156 ret = -ENODEV;
3157 goto done;
3158 }
3159 fbcon_suspended(info); 3151 fbcon_suspended(info);
3160 unlock_fb_info(info);
3161 break; 3152 break;
3162 case FB_EVENT_RESUME: 3153 case FB_EVENT_RESUME:
3163 if (!lock_fb_info(info)) {
3164 ret = -ENODEV;
3165 goto done;
3166 }
3167 fbcon_resumed(info); 3154 fbcon_resumed(info);
3168 unlock_fb_info(info);
3169 break; 3155 break;
3170 case FB_EVENT_MODE_CHANGE: 3156 case FB_EVENT_MODE_CHANGE:
3171 if (!lock_fb_info(info)) {
3172 ret = -ENODEV;
3173 goto done;
3174 }
3175 fbcon_modechanged(info); 3157 fbcon_modechanged(info);
3176 unlock_fb_info(info);
3177 break; 3158 break;
3178 case FB_EVENT_MODE_CHANGE_ALL: 3159 case FB_EVENT_MODE_CHANGE_ALL:
3179 if (!lock_fb_info(info)) {
3180 ret = -ENODEV;
3181 goto done;
3182 }
3183 fbcon_set_all_vcs(info); 3160 fbcon_set_all_vcs(info);
3184 unlock_fb_info(info);
3185 break; 3161 break;
3186 case FB_EVENT_MODE_DELETE: 3162 case FB_EVENT_MODE_DELETE:
3187 mode = event->data; 3163 mode = event->data;
3188 if (!lock_fb_info(info)) {
3189 ret = -ENODEV;
3190 goto done;
3191 }
3192 ret = fbcon_mode_deleted(info, mode); 3164 ret = fbcon_mode_deleted(info, mode);
3193 unlock_fb_info(info);
3194 break; 3165 break;
3195 case FB_EVENT_FB_UNBIND: 3166 case FB_EVENT_FB_UNBIND:
3196 if (!lock_fb_info(info)) {
3197 ret = -ENODEV;
3198 goto done;
3199 }
3200 idx = info->node; 3167 idx = info->node;
3201 unlock_fb_info(info);
3202 ret = fbcon_fb_unbind(idx); 3168 ret = fbcon_fb_unbind(idx);
3203 break; 3169 break;
3204 case FB_EVENT_FB_REGISTERED: 3170 case FB_EVENT_FB_REGISTERED:
@@ -3217,29 +3183,14 @@ static int fbcon_event_notify(struct notifier_block *self,
3217 con2fb->framebuffer = con2fb_map[con2fb->console - 1]; 3183 con2fb->framebuffer = con2fb_map[con2fb->console - 1];
3218 break; 3184 break;
3219 case FB_EVENT_BLANK: 3185 case FB_EVENT_BLANK:
3220 if (!lock_fb_info(info)) {
3221 ret = -ENODEV;
3222 goto done;
3223 }
3224 fbcon_fb_blanked(info, *(int *)event->data); 3186 fbcon_fb_blanked(info, *(int *)event->data);
3225 unlock_fb_info(info);
3226 break; 3187 break;
3227 case FB_EVENT_NEW_MODELIST: 3188 case FB_EVENT_NEW_MODELIST:
3228 if (!lock_fb_info(info)) {
3229 ret = -ENODEV;
3230 goto done;
3231 }
3232 fbcon_new_modelist(info); 3189 fbcon_new_modelist(info);
3233 unlock_fb_info(info);
3234 break; 3190 break;
3235 case FB_EVENT_GET_REQ: 3191 case FB_EVENT_GET_REQ:
3236 caps = event->data; 3192 caps = event->data;
3237 if (!lock_fb_info(info)) {
3238 ret = -ENODEV;
3239 goto done;
3240 }
3241 fbcon_get_requirement(info, caps); 3193 fbcon_get_requirement(info, caps);
3242 unlock_fb_info(info);
3243 break; 3194 break;
3244 } 3195 }
3245done: 3196done:
diff --git a/drivers/video/efifb.c b/drivers/video/efifb.c
index 0c5b9a9fd56f..8dea2bc92705 100644
--- a/drivers/video/efifb.c
+++ b/drivers/video/efifb.c
@@ -210,12 +210,15 @@ static int __init efifb_probe(struct platform_device *dev)
210 unsigned int size_total; 210 unsigned int size_total;
211 int request_succeeded = 0; 211 int request_succeeded = 0;
212 212
213 printk(KERN_INFO "efifb: probing for efifb\n");
214
215 if (!screen_info.lfb_depth) 213 if (!screen_info.lfb_depth)
216 screen_info.lfb_depth = 32; 214 screen_info.lfb_depth = 32;
217 if (!screen_info.pages) 215 if (!screen_info.pages)
218 screen_info.pages = 1; 216 screen_info.pages = 1;
217 if (!screen_info.lfb_base) {
218 printk(KERN_DEBUG "efifb: invalid framebuffer address\n");
219 return -ENODEV;
220 }
221 printk(KERN_INFO "efifb: probing for efifb\n");
219 222
220 /* just assume they're all unset if any are */ 223 /* just assume they're all unset if any are */
221 if (!screen_info.blue_size) { 224 if (!screen_info.blue_size) {
diff --git a/drivers/video/fbmem.c b/drivers/video/fbmem.c
index 2ac32e6b5953..d412a1ddc12f 100644
--- a/drivers/video/fbmem.c
+++ b/drivers/video/fbmem.c
@@ -1097,8 +1097,11 @@ static long do_fb_ioctl(struct fb_info *info, unsigned int cmd,
1097 return -EINVAL; 1097 return -EINVAL;
1098 con2fb.framebuffer = -1; 1098 con2fb.framebuffer = -1;
1099 event.data = &con2fb; 1099 event.data = &con2fb;
1100 if (!lock_fb_info(info))
1101 return -ENODEV;
1100 event.info = info; 1102 event.info = info;
1101 fb_notifier_call_chain(FB_EVENT_GET_CONSOLE_MAP, &event); 1103 fb_notifier_call_chain(FB_EVENT_GET_CONSOLE_MAP, &event);
1104 unlock_fb_info(info);
1102 ret = copy_to_user(argp, &con2fb, sizeof(con2fb)) ? -EFAULT : 0; 1105 ret = copy_to_user(argp, &con2fb, sizeof(con2fb)) ? -EFAULT : 0;
1103 break; 1106 break;
1104 case FBIOPUT_CON2FBMAP: 1107 case FBIOPUT_CON2FBMAP:
@@ -1115,8 +1118,11 @@ static long do_fb_ioctl(struct fb_info *info, unsigned int cmd,
1115 break; 1118 break;
1116 } 1119 }
1117 event.data = &con2fb; 1120 event.data = &con2fb;
1121 if (!lock_fb_info(info))
1122 return -ENODEV;
1118 event.info = info; 1123 event.info = info;
1119 ret = fb_notifier_call_chain(FB_EVENT_SET_CONSOLE_MAP, &event); 1124 ret = fb_notifier_call_chain(FB_EVENT_SET_CONSOLE_MAP, &event);
1125 unlock_fb_info(info);
1120 break; 1126 break;
1121 case FBIOBLANK: 1127 case FBIOBLANK:
1122 if (!lock_fb_info(info)) 1128 if (!lock_fb_info(info))
@@ -1521,7 +1527,10 @@ register_framebuffer(struct fb_info *fb_info)
1521 registered_fb[i] = fb_info; 1527 registered_fb[i] = fb_info;
1522 1528
1523 event.info = fb_info; 1529 event.info = fb_info;
1530 if (!lock_fb_info(fb_info))
1531 return -ENODEV;
1524 fb_notifier_call_chain(FB_EVENT_FB_REGISTERED, &event); 1532 fb_notifier_call_chain(FB_EVENT_FB_REGISTERED, &event);
1533 unlock_fb_info(fb_info);
1525 return 0; 1534 return 0;
1526} 1535}
1527 1536
@@ -1555,8 +1564,12 @@ unregister_framebuffer(struct fb_info *fb_info)
1555 goto done; 1564 goto done;
1556 } 1565 }
1557 1566
1567
1568 if (!lock_fb_info(fb_info))
1569 return -ENODEV;
1558 event.info = fb_info; 1570 event.info = fb_info;
1559 ret = fb_notifier_call_chain(FB_EVENT_FB_UNBIND, &event); 1571 ret = fb_notifier_call_chain(FB_EVENT_FB_UNBIND, &event);
1572 unlock_fb_info(fb_info);
1560 1573
1561 if (ret) { 1574 if (ret) {
1562 ret = -EINVAL; 1575 ret = -EINVAL;
@@ -1590,6 +1603,8 @@ void fb_set_suspend(struct fb_info *info, int state)
1590{ 1603{
1591 struct fb_event event; 1604 struct fb_event event;
1592 1605
1606 if (!lock_fb_info(info))
1607 return;
1593 event.info = info; 1608 event.info = info;
1594 if (state) { 1609 if (state) {
1595 fb_notifier_call_chain(FB_EVENT_SUSPEND, &event); 1610 fb_notifier_call_chain(FB_EVENT_SUSPEND, &event);
@@ -1598,6 +1613,7 @@ void fb_set_suspend(struct fb_info *info, int state)
1598 info->state = FBINFO_STATE_RUNNING; 1613 info->state = FBINFO_STATE_RUNNING;
1599 fb_notifier_call_chain(FB_EVENT_RESUME, &event); 1614 fb_notifier_call_chain(FB_EVENT_RESUME, &event);
1600 } 1615 }
1616 unlock_fb_info(info);
1601} 1617}
1602 1618
1603/** 1619/**
@@ -1667,8 +1683,11 @@ int fb_new_modelist(struct fb_info *info)
1667 err = 1; 1683 err = 1;
1668 1684
1669 if (!list_empty(&info->modelist)) { 1685 if (!list_empty(&info->modelist)) {
1686 if (!lock_fb_info(info))
1687 return -ENODEV;
1670 event.info = info; 1688 event.info = info;
1671 err = fb_notifier_call_chain(FB_EVENT_NEW_MODELIST, &event); 1689 err = fb_notifier_call_chain(FB_EVENT_NEW_MODELIST, &event);
1690 unlock_fb_info(info);
1672 } 1691 }
1673 1692
1674 return err; 1693 return err;
diff --git a/drivers/video/intelfb/intelfb.h b/drivers/video/intelfb/intelfb.h
index a50bea614804..40984551c927 100644
--- a/drivers/video/intelfb/intelfb.h
+++ b/drivers/video/intelfb/intelfb.h
@@ -53,6 +53,7 @@
53#define PCI_DEVICE_ID_INTEL_830M 0x3577 53#define PCI_DEVICE_ID_INTEL_830M 0x3577
54#define PCI_DEVICE_ID_INTEL_845G 0x2562 54#define PCI_DEVICE_ID_INTEL_845G 0x2562
55#define PCI_DEVICE_ID_INTEL_85XGM 0x3582 55#define PCI_DEVICE_ID_INTEL_85XGM 0x3582
56#define PCI_DEVICE_ID_INTEL_854 0x358E
56#define PCI_DEVICE_ID_INTEL_865G 0x2572 57#define PCI_DEVICE_ID_INTEL_865G 0x2572
57#define PCI_DEVICE_ID_INTEL_915G 0x2582 58#define PCI_DEVICE_ID_INTEL_915G 0x2582
58#define PCI_DEVICE_ID_INTEL_915GM 0x2592 59#define PCI_DEVICE_ID_INTEL_915GM 0x2592
@@ -154,6 +155,7 @@ enum intel_chips {
154 INTEL_85XGM, 155 INTEL_85XGM,
155 INTEL_852GM, 156 INTEL_852GM,
156 INTEL_852GME, 157 INTEL_852GME,
158 INTEL_854,
157 INTEL_855GM, 159 INTEL_855GM,
158 INTEL_855GME, 160 INTEL_855GME,
159 INTEL_865G, 161 INTEL_865G,
diff --git a/drivers/video/intelfb/intelfb_i2c.c b/drivers/video/intelfb/intelfb_i2c.c
index b3065492bb20..487f2be47460 100644
--- a/drivers/video/intelfb/intelfb_i2c.c
+++ b/drivers/video/intelfb/intelfb_i2c.c
@@ -156,6 +156,7 @@ void intelfb_create_i2c_busses(struct intelfb_info *dinfo)
156 switch(dinfo->chipset) { 156 switch(dinfo->chipset) {
157 case INTEL_830M: 157 case INTEL_830M:
158 case INTEL_845G: 158 case INTEL_845G:
159 case INTEL_854:
159 case INTEL_855GM: 160 case INTEL_855GM:
160 case INTEL_865G: 161 case INTEL_865G:
161 dinfo->output[i].type = INTELFB_OUTPUT_DVO; 162 dinfo->output[i].type = INTELFB_OUTPUT_DVO;
diff --git a/drivers/video/intelfb/intelfbdrv.c b/drivers/video/intelfb/intelfbdrv.c
index 6d8e5415c809..ace14fe02fc4 100644
--- a/drivers/video/intelfb/intelfbdrv.c
+++ b/drivers/video/intelfb/intelfbdrv.c
@@ -182,6 +182,7 @@ static struct pci_device_id intelfb_pci_table[] __devinitdata = {
182 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_845G, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_845G }, 182 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_845G, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_845G },
183 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_85XGM, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_85XGM }, 183 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_85XGM, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_85XGM },
184 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_865G, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_865G }, 184 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_865G, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_865G },
185 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_854, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_854 },
185 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_915G, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_915G }, 186 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_915G, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_915G },
186 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_915GM, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_915GM }, 187 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_915GM, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_915GM },
187 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_945G, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_945G }, 188 { PCI_VENDOR_ID_INTEL, PCI_DEVICE_ID_INTEL_945G, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, INTELFB_CLASS_MASK, INTEL_945G },
diff --git a/drivers/video/intelfb/intelfbhw.c b/drivers/video/intelfb/intelfbhw.c
index 8b26b27c2db6..0689f97c5238 100644
--- a/drivers/video/intelfb/intelfbhw.c
+++ b/drivers/video/intelfb/intelfbhw.c
@@ -84,6 +84,11 @@ int intelfbhw_get_chipset(struct pci_dev *pdev, struct intelfb_info *dinfo)
84 dinfo->mobile = 0; 84 dinfo->mobile = 0;
85 dinfo->pll_index = PLLS_I8xx; 85 dinfo->pll_index = PLLS_I8xx;
86 return 0; 86 return 0;
87 case PCI_DEVICE_ID_INTEL_854:
88 dinfo->mobile = 1;
89 dinfo->name = "Intel(R) 854";
90 dinfo->chipset = INTEL_854;
91 return 0;
87 case PCI_DEVICE_ID_INTEL_85XGM: 92 case PCI_DEVICE_ID_INTEL_85XGM:
88 tmp = 0; 93 tmp = 0;
89 dinfo->mobile = 1; 94 dinfo->mobile = 1;
diff --git a/drivers/video/s3fb.c b/drivers/video/s3fb.c
index 4dcec48a1d78..c3fad34309ed 100644
--- a/drivers/video/s3fb.c
+++ b/drivers/video/s3fb.c
@@ -45,11 +45,11 @@ struct s3fb_info {
45static const struct svga_fb_format s3fb_formats[] = { 45static const struct svga_fb_format s3fb_formats[] = {
46 { 0, {0, 6, 0}, {0, 6, 0}, {0, 6, 0}, {0, 0, 0}, 0, 46 { 0, {0, 6, 0}, {0, 6, 0}, {0, 6, 0}, {0, 0, 0}, 0,
47 FB_TYPE_TEXT, FB_AUX_TEXT_SVGA_STEP4, FB_VISUAL_PSEUDOCOLOR, 8, 16}, 47 FB_TYPE_TEXT, FB_AUX_TEXT_SVGA_STEP4, FB_VISUAL_PSEUDOCOLOR, 8, 16},
48 { 4, {0, 6, 0}, {0, 6, 0}, {0, 6, 0}, {0, 0, 0}, 0, 48 { 4, {0, 4, 0}, {0, 4, 0}, {0, 4, 0}, {0, 0, 0}, 0,
49 FB_TYPE_PACKED_PIXELS, 0, FB_VISUAL_PSEUDOCOLOR, 8, 16}, 49 FB_TYPE_PACKED_PIXELS, 0, FB_VISUAL_PSEUDOCOLOR, 8, 16},
50 { 4, {0, 6, 0}, {0, 6, 0}, {0, 6, 0}, {0, 0, 0}, 1, 50 { 4, {0, 4, 0}, {0, 4, 0}, {0, 4, 0}, {0, 0, 0}, 1,
51 FB_TYPE_INTERLEAVED_PLANES, 1, FB_VISUAL_PSEUDOCOLOR, 8, 16}, 51 FB_TYPE_INTERLEAVED_PLANES, 1, FB_VISUAL_PSEUDOCOLOR, 8, 16},
52 { 8, {0, 6, 0}, {0, 6, 0}, {0, 6, 0}, {0, 0, 0}, 0, 52 { 8, {0, 8, 0}, {0, 8, 0}, {0, 8, 0}, {0, 0, 0}, 0,
53 FB_TYPE_PACKED_PIXELS, 0, FB_VISUAL_PSEUDOCOLOR, 4, 8}, 53 FB_TYPE_PACKED_PIXELS, 0, FB_VISUAL_PSEUDOCOLOR, 4, 8},
54 {16, {10, 5, 0}, {5, 5, 0}, {0, 5, 0}, {0, 0, 0}, 0, 54 {16, {10, 5, 0}, {5, 5, 0}, {0, 5, 0}, {0, 0, 0}, 0,
55 FB_TYPE_PACKED_PIXELS, 0, FB_VISUAL_TRUECOLOR, 2, 4}, 55 FB_TYPE_PACKED_PIXELS, 0, FB_VISUAL_TRUECOLOR, 2, 4},
diff --git a/drivers/video/sa1100fb.c b/drivers/video/sa1100fb.c
index fad58cf9ef73..10ddad8e17d6 100644
--- a/drivers/video/sa1100fb.c
+++ b/drivers/video/sa1100fb.c
@@ -199,16 +199,20 @@
199extern void (*sa1100fb_backlight_power)(int on); 199extern void (*sa1100fb_backlight_power)(int on);
200extern void (*sa1100fb_lcd_power)(int on); 200extern void (*sa1100fb_lcd_power)(int on);
201 201
202/* 202static struct sa1100fb_rgb rgb_4 = {
203 * IMHO this looks wrong. In 8BPP, length should be 8.
204 */
205static struct sa1100fb_rgb rgb_8 = {
206 .red = { .offset = 0, .length = 4, }, 203 .red = { .offset = 0, .length = 4, },
207 .green = { .offset = 0, .length = 4, }, 204 .green = { .offset = 0, .length = 4, },
208 .blue = { .offset = 0, .length = 4, }, 205 .blue = { .offset = 0, .length = 4, },
209 .transp = { .offset = 0, .length = 0, }, 206 .transp = { .offset = 0, .length = 0, },
210}; 207};
211 208
209static struct sa1100fb_rgb rgb_8 = {
210 .red = { .offset = 0, .length = 8, },
211 .green = { .offset = 0, .length = 8, },
212 .blue = { .offset = 0, .length = 8, },
213 .transp = { .offset = 0, .length = 0, },
214};
215
212static struct sa1100fb_rgb def_rgb_16 = { 216static struct sa1100fb_rgb def_rgb_16 = {
213 .red = { .offset = 11, .length = 5, }, 217 .red = { .offset = 11, .length = 5, },
214 .green = { .offset = 5, .length = 6, }, 218 .green = { .offset = 5, .length = 6, },
@@ -613,7 +617,7 @@ sa1100fb_check_var(struct fb_var_screeninfo *var, struct fb_info *info)
613 DPRINTK("var->bits_per_pixel=%d\n", var->bits_per_pixel); 617 DPRINTK("var->bits_per_pixel=%d\n", var->bits_per_pixel);
614 switch (var->bits_per_pixel) { 618 switch (var->bits_per_pixel) {
615 case 4: 619 case 4:
616 rgbidx = RGB_8; 620 rgbidx = RGB_4;
617 break; 621 break;
618 case 8: 622 case 8:
619 rgbidx = RGB_8; 623 rgbidx = RGB_8;
@@ -1382,6 +1386,7 @@ static struct sa1100fb_info * __init sa1100fb_init_fbinfo(struct device *dev)
1382 fbi->fb.monspecs = monspecs; 1386 fbi->fb.monspecs = monspecs;
1383 fbi->fb.pseudo_palette = (fbi + 1); 1387 fbi->fb.pseudo_palette = (fbi + 1);
1384 1388
1389 fbi->rgb[RGB_4] = &rgb_4;
1385 fbi->rgb[RGB_8] = &rgb_8; 1390 fbi->rgb[RGB_8] = &rgb_8;
1386 fbi->rgb[RGB_16] = &def_rgb_16; 1391 fbi->rgb[RGB_16] = &def_rgb_16;
1387 1392
diff --git a/drivers/video/sa1100fb.h b/drivers/video/sa1100fb.h
index 86831db9a042..1c3b459865d8 100644
--- a/drivers/video/sa1100fb.h
+++ b/drivers/video/sa1100fb.h
@@ -57,9 +57,10 @@ struct sa1100fb_lcd_reg {
57 unsigned long lccr3; 57 unsigned long lccr3;
58}; 58};
59 59
60#define RGB_8 (0) 60#define RGB_4 (0)
61#define RGB_16 (1) 61#define RGB_8 (1)
62#define NR_RGB 2 62#define RGB_16 (2)
63#define NR_RGB 3
63 64
64struct sa1100fb_info { 65struct sa1100fb_info {
65 struct fb_info fb; 66 struct fb_info fb;
diff --git a/drivers/video/sis/sis_main.c b/drivers/video/sis/sis_main.c
index 346d6458cf76..7e17ee95a97a 100644
--- a/drivers/video/sis/sis_main.c
+++ b/drivers/video/sis/sis_main.c
@@ -1129,7 +1129,7 @@ sisfb_bpp_to_var(struct sis_video_info *ivideo, struct fb_var_screeninfo *var)
1129 switch(var->bits_per_pixel) { 1129 switch(var->bits_per_pixel) {
1130 case 8: 1130 case 8:
1131 var->red.offset = var->green.offset = var->blue.offset = 0; 1131 var->red.offset = var->green.offset = var->blue.offset = 0;
1132 var->red.length = var->green.length = var->blue.length = 6; 1132 var->red.length = var->green.length = var->blue.length = 8;
1133 break; 1133 break;
1134 case 16: 1134 case 16:
1135 var->red.offset = 11; 1135 var->red.offset = 11;
diff --git a/drivers/video/skeletonfb.c b/drivers/video/skeletonfb.c
index a439159204a8..89158bc71da2 100644
--- a/drivers/video/skeletonfb.c
+++ b/drivers/video/skeletonfb.c
@@ -308,9 +308,11 @@ static int xxxfb_setcolreg(unsigned regno, unsigned red, unsigned green,
308 * color depth = SUM(var->{color}.length) 308 * color depth = SUM(var->{color}.length)
309 * 309 *
310 * Pseudocolor: 310 * Pseudocolor:
311 * var->{color}.offset is 0 311 * var->{color}.offset is 0 unless the palette index takes less than
312 * var->{color}.length contains width of DAC or the number of unique 312 * bits_per_pixel bits and is stored in the upper
313 * colors available (color depth) 313 * bits of the pixel value
314 * var->{color}.length is set so that 1 << length is the number of
315 * available palette entries
314 * pseudo_palette is not used 316 * pseudo_palette is not used
315 * RAMDAC[X] is programmed to (red, green, blue) 317 * RAMDAC[X] is programmed to (red, green, blue)
316 * color depth = var->{color}.length 318 * color depth = var->{color}.length
diff --git a/drivers/video/uvesafb.c b/drivers/video/uvesafb.c
index 0b370aebdbfd..421770b5e6ab 100644
--- a/drivers/video/uvesafb.c
+++ b/drivers/video/uvesafb.c
@@ -55,6 +55,7 @@ static u16 maxvf __devinitdata; /* maximum vertical frequency */
55static u16 maxhf __devinitdata; /* maximum horizontal frequency */ 55static u16 maxhf __devinitdata; /* maximum horizontal frequency */
56static u16 vbemode __devinitdata; /* force use of a specific VBE mode */ 56static u16 vbemode __devinitdata; /* force use of a specific VBE mode */
57static char *mode_option __devinitdata; 57static char *mode_option __devinitdata;
58static u8 dac_width = 6;
58 59
59static struct uvesafb_ktask *uvfb_tasks[UVESAFB_TASKS_MAX]; 60static struct uvesafb_ktask *uvfb_tasks[UVESAFB_TASKS_MAX];
60static DEFINE_MUTEX(uvfb_lock); 61static DEFINE_MUTEX(uvfb_lock);
@@ -303,22 +304,10 @@ static void uvesafb_setup_var(struct fb_var_screeninfo *var,
303 var->blue.offset = 0; 304 var->blue.offset = 0;
304 var->transp.offset = 0; 305 var->transp.offset = 0;
305 306
306 /* 307 var->red.length = 8;
307 * We're assuming that we can switch the DAC to 8 bits. If 308 var->green.length = 8;
308 * this proves to be incorrect, we'll update the fields 309 var->blue.length = 8;
309 * later in set_par(). 310 var->transp.length = 0;
310 */
311 if (par->vbe_ib.capabilities & VBE_CAP_CAN_SWITCH_DAC) {
312 var->red.length = 8;
313 var->green.length = 8;
314 var->blue.length = 8;
315 var->transp.length = 0;
316 } else {
317 var->red.length = 6;
318 var->green.length = 6;
319 var->blue.length = 6;
320 var->transp.length = 0;
321 }
322 } 311 }
323} 312}
324 313
@@ -1006,7 +995,7 @@ static int uvesafb_setcolreg(unsigned regno, unsigned red, unsigned green,
1006 struct fb_info *info) 995 struct fb_info *info)
1007{ 996{
1008 struct uvesafb_pal_entry entry; 997 struct uvesafb_pal_entry entry;
1009 int shift = 16 - info->var.green.length; 998 int shift = 16 - dac_width;
1010 int err = 0; 999 int err = 0;
1011 1000
1012 if (regno >= info->cmap.len) 1001 if (regno >= info->cmap.len)
@@ -1055,7 +1044,7 @@ static int uvesafb_setcolreg(unsigned regno, unsigned red, unsigned green,
1055static int uvesafb_setcmap(struct fb_cmap *cmap, struct fb_info *info) 1044static int uvesafb_setcmap(struct fb_cmap *cmap, struct fb_info *info)
1056{ 1045{
1057 struct uvesafb_pal_entry *entries; 1046 struct uvesafb_pal_entry *entries;
1058 int shift = 16 - info->var.green.length; 1047 int shift = 16 - dac_width;
1059 int i, err = 0; 1048 int i, err = 0;
1060 1049
1061 if (info->var.bits_per_pixel == 8) { 1050 if (info->var.bits_per_pixel == 8) {
@@ -1317,13 +1306,9 @@ setmode:
1317 err = uvesafb_exec(task); 1306 err = uvesafb_exec(task);
1318 if (err || (task->t.regs.eax & 0xffff) != 0x004f || 1307 if (err || (task->t.regs.eax & 0xffff) != 0x004f ||
1319 ((task->t.regs.ebx & 0xff00) >> 8) != 8) { 1308 ((task->t.regs.ebx & 0xff00) >> 8) != 8) {
1320 /* 1309 dac_width = 6;
1321 * We've failed to set the DAC palette format - 1310 } else {
1322 * time to correct var. 1311 dac_width = 8;
1323 */
1324 info->var.red.length = 6;
1325 info->var.green.length = 6;
1326 info->var.blue.length = 6;
1327 } 1312 }
1328 } 1313 }
1329 1314
diff --git a/drivers/video/vfb.c b/drivers/video/vfb.c
index cc919ae46571..050d432c7d95 100644
--- a/drivers/video/vfb.c
+++ b/drivers/video/vfb.c
@@ -318,13 +318,16 @@ static int vfb_setcolreg(u_int regno, u_int red, u_int green, u_int blue,
318 * {hardwarespecific} contains width of RAMDAC 318 * {hardwarespecific} contains width of RAMDAC
319 * cmap[X] is programmed to (X << red.offset) | (X << green.offset) | (X << blue.offset) 319 * cmap[X] is programmed to (X << red.offset) | (X << green.offset) | (X << blue.offset)
320 * RAMDAC[X] is programmed to (red, green, blue) 320 * RAMDAC[X] is programmed to (red, green, blue)
321 * 321 *
322 * Pseudocolor: 322 * Pseudocolor:
323 * uses offset = 0 && length = RAMDAC register width. 323 * var->{color}.offset is 0 unless the palette index takes less than
324 * var->{color}.offset is 0 324 * bits_per_pixel bits and is stored in the upper
325 * var->{color}.length contains widht of DAC 325 * bits of the pixel value
326 * var->{color}.length is set so that 1 << length is the number of available
327 * palette entries
326 * cmap is not used 328 * cmap is not used
327 * RAMDAC[X] is programmed to (red, green, blue) 329 * RAMDAC[X] is programmed to (red, green, blue)
330 *
328 * Truecolor: 331 * Truecolor:
329 * does not use DAC. Usually 3 are present. 332 * does not use DAC. Usually 3 are present.
330 * var->{color}.offset contains start of bitfield 333 * var->{color}.offset contains start of bitfield
diff --git a/drivers/xen/cpu_hotplug.c b/drivers/xen/cpu_hotplug.c
index 5f54c01c1568..bdfd584ad853 100644
--- a/drivers/xen/cpu_hotplug.c
+++ b/drivers/xen/cpu_hotplug.c
@@ -21,29 +21,41 @@ static void disable_hotplug_cpu(int cpu)
21 set_cpu_present(cpu, false); 21 set_cpu_present(cpu, false);
22} 22}
23 23
24static void vcpu_hotplug(unsigned int cpu) 24static int vcpu_online(unsigned int cpu)
25{ 25{
26 int err; 26 int err;
27 char dir[32], state[32]; 27 char dir[32], state[32];
28 28
29 if (!cpu_possible(cpu))
30 return;
31
32 sprintf(dir, "cpu/%u", cpu); 29 sprintf(dir, "cpu/%u", cpu);
33 err = xenbus_scanf(XBT_NIL, dir, "availability", "%s", state); 30 err = xenbus_scanf(XBT_NIL, dir, "availability", "%s", state);
34 if (err != 1) { 31 if (err != 1) {
35 printk(KERN_ERR "XENBUS: Unable to read cpu state\n"); 32 printk(KERN_ERR "XENBUS: Unable to read cpu state\n");
36 return; 33 return err;
37 } 34 }
38 35
39 if (strcmp(state, "online") == 0) { 36 if (strcmp(state, "online") == 0)
37 return 1;
38 else if (strcmp(state, "offline") == 0)
39 return 0;
40
41 printk(KERN_ERR "XENBUS: unknown state(%s) on CPU%d\n", state, cpu);
42 return -EINVAL;
43}
44static void vcpu_hotplug(unsigned int cpu)
45{
46 if (!cpu_possible(cpu))
47 return;
48
49 switch (vcpu_online(cpu)) {
50 case 1:
40 enable_hotplug_cpu(cpu); 51 enable_hotplug_cpu(cpu);
41 } else if (strcmp(state, "offline") == 0) { 52 break;
53 case 0:
42 (void)cpu_down(cpu); 54 (void)cpu_down(cpu);
43 disable_hotplug_cpu(cpu); 55 disable_hotplug_cpu(cpu);
44 } else { 56 break;
45 printk(KERN_ERR "XENBUS: unknown state(%s) on CPU%d\n", 57 default:
46 state, cpu); 58 break;
47 } 59 }
48} 60}
49 61
@@ -64,12 +76,20 @@ static void handle_vcpu_hotplug_event(struct xenbus_watch *watch,
64static int setup_cpu_watcher(struct notifier_block *notifier, 76static int setup_cpu_watcher(struct notifier_block *notifier,
65 unsigned long event, void *data) 77 unsigned long event, void *data)
66{ 78{
79 int cpu;
67 static struct xenbus_watch cpu_watch = { 80 static struct xenbus_watch cpu_watch = {
68 .node = "cpu", 81 .node = "cpu",
69 .callback = handle_vcpu_hotplug_event}; 82 .callback = handle_vcpu_hotplug_event};
70 83
71 (void)register_xenbus_watch(&cpu_watch); 84 (void)register_xenbus_watch(&cpu_watch);
72 85
86 for_each_possible_cpu(cpu) {
87 if (vcpu_online(cpu) == 0) {
88 (void)cpu_down(cpu);
89 cpu_clear(cpu, cpu_present_map);
90 }
91 }
92
73 return NOTIFY_DONE; 93 return NOTIFY_DONE;
74} 94}
75 95
diff --git a/drivers/xen/manage.c b/drivers/xen/manage.c
index 0d61db1e7b49..4b5b84837ee1 100644
--- a/drivers/xen/manage.c
+++ b/drivers/xen/manage.c
@@ -62,14 +62,15 @@ static int xen_suspend(void *data)
62 gnttab_resume(); 62 gnttab_resume();
63 xen_mm_unpin_all(); 63 xen_mm_unpin_all();
64 64
65 sysdev_resume();
66
67 if (!*cancelled) { 65 if (!*cancelled) {
68 xen_irq_resume(); 66 xen_irq_resume();
69 xen_console_resume(); 67 xen_console_resume();
70 xen_timer_resume(); 68 xen_timer_resume();
71 } 69 }
72 70
71 sysdev_resume();
72 device_power_up(PMSG_RESUME);
73
73 return 0; 74 return 0;
74} 75}
75 76
diff --git a/fs/bio.c b/fs/bio.c
index e0c9e545bbfa..cd42bb882f30 100644
--- a/fs/bio.c
+++ b/fs/bio.c
@@ -348,6 +348,24 @@ err:
348 return NULL; 348 return NULL;
349} 349}
350 350
351/**
352 * bio_alloc - allocate a bio for I/O
353 * @gfp_mask: the GFP_ mask given to the slab allocator
354 * @nr_iovecs: number of iovecs to pre-allocate
355 *
356 * Description:
357 * bio_alloc will allocate a bio and associated bio_vec array that can hold
358 * at least @nr_iovecs entries. Allocations will be done from the
359 * fs_bio_set. Also see @bio_alloc_bioset.
360 *
361 * If %__GFP_WAIT is set, then bio_alloc will always be able to allocate
362 * a bio. This is due to the mempool guarantees. To make this work, callers
363 * must never allocate more than 1 bio at the time from this pool. Callers
364 * that need to allocate more than 1 bio must always submit the previously
365 * allocate bio for IO before attempting to allocate a new one. Failure to
366 * do so can cause livelocks under memory pressure.
367 *
368 **/
351struct bio *bio_alloc(gfp_t gfp_mask, int nr_iovecs) 369struct bio *bio_alloc(gfp_t gfp_mask, int nr_iovecs)
352{ 370{
353 struct bio *bio = bio_alloc_bioset(gfp_mask, nr_iovecs, fs_bio_set); 371 struct bio *bio = bio_alloc_bioset(gfp_mask, nr_iovecs, fs_bio_set);
diff --git a/fs/buffer.c b/fs/buffer.c
index 13edf7ad3ff1..b3e5be7514f5 100644
--- a/fs/buffer.c
+++ b/fs/buffer.c
@@ -360,7 +360,7 @@ still_busy:
360 * Completion handler for block_write_full_page() - pages which are unlocked 360 * Completion handler for block_write_full_page() - pages which are unlocked
361 * during I/O, and which have PageWriteback cleared upon I/O completion. 361 * during I/O, and which have PageWriteback cleared upon I/O completion.
362 */ 362 */
363static void end_buffer_async_write(struct buffer_head *bh, int uptodate) 363void end_buffer_async_write(struct buffer_head *bh, int uptodate)
364{ 364{
365 char b[BDEVNAME_SIZE]; 365 char b[BDEVNAME_SIZE];
366 unsigned long flags; 366 unsigned long flags;
@@ -438,11 +438,17 @@ static void mark_buffer_async_read(struct buffer_head *bh)
438 set_buffer_async_read(bh); 438 set_buffer_async_read(bh);
439} 439}
440 440
441void mark_buffer_async_write(struct buffer_head *bh) 441void mark_buffer_async_write_endio(struct buffer_head *bh,
442 bh_end_io_t *handler)
442{ 443{
443 bh->b_end_io = end_buffer_async_write; 444 bh->b_end_io = handler;
444 set_buffer_async_write(bh); 445 set_buffer_async_write(bh);
445} 446}
447
448void mark_buffer_async_write(struct buffer_head *bh)
449{
450 mark_buffer_async_write_endio(bh, end_buffer_async_write);
451}
446EXPORT_SYMBOL(mark_buffer_async_write); 452EXPORT_SYMBOL(mark_buffer_async_write);
447 453
448 454
@@ -547,7 +553,7 @@ repeat:
547 return err; 553 return err;
548} 554}
549 555
550void do_thaw_all(unsigned long unused) 556void do_thaw_all(struct work_struct *work)
551{ 557{
552 struct super_block *sb; 558 struct super_block *sb;
553 char b[BDEVNAME_SIZE]; 559 char b[BDEVNAME_SIZE];
@@ -567,6 +573,7 @@ restart:
567 goto restart; 573 goto restart;
568 } 574 }
569 spin_unlock(&sb_lock); 575 spin_unlock(&sb_lock);
576 kfree(work);
570 printk(KERN_WARNING "Emergency Thaw complete\n"); 577 printk(KERN_WARNING "Emergency Thaw complete\n");
571} 578}
572 579
@@ -577,7 +584,13 @@ restart:
577 */ 584 */
578void emergency_thaw_all(void) 585void emergency_thaw_all(void)
579{ 586{
580 pdflush_operation(do_thaw_all, 0); 587 struct work_struct *work;
588
589 work = kmalloc(sizeof(*work), GFP_ATOMIC);
590 if (work) {
591 INIT_WORK(work, do_thaw_all);
592 schedule_work(work);
593 }
581} 594}
582 595
583/** 596/**
@@ -1608,7 +1621,8 @@ EXPORT_SYMBOL(unmap_underlying_metadata);
1608 * unplugging the device queue. 1621 * unplugging the device queue.
1609 */ 1622 */
1610static int __block_write_full_page(struct inode *inode, struct page *page, 1623static int __block_write_full_page(struct inode *inode, struct page *page,
1611 get_block_t *get_block, struct writeback_control *wbc) 1624 get_block_t *get_block, struct writeback_control *wbc,
1625 bh_end_io_t *handler)
1612{ 1626{
1613 int err; 1627 int err;
1614 sector_t block; 1628 sector_t block;
@@ -1693,7 +1707,7 @@ static int __block_write_full_page(struct inode *inode, struct page *page,
1693 continue; 1707 continue;
1694 } 1708 }
1695 if (test_clear_buffer_dirty(bh)) { 1709 if (test_clear_buffer_dirty(bh)) {
1696 mark_buffer_async_write(bh); 1710 mark_buffer_async_write_endio(bh, handler);
1697 } else { 1711 } else {
1698 unlock_buffer(bh); 1712 unlock_buffer(bh);
1699 } 1713 }
@@ -1746,7 +1760,7 @@ recover:
1746 if (buffer_mapped(bh) && buffer_dirty(bh) && 1760 if (buffer_mapped(bh) && buffer_dirty(bh) &&
1747 !buffer_delay(bh)) { 1761 !buffer_delay(bh)) {
1748 lock_buffer(bh); 1762 lock_buffer(bh);
1749 mark_buffer_async_write(bh); 1763 mark_buffer_async_write_endio(bh, handler);
1750 } else { 1764 } else {
1751 /* 1765 /*
1752 * The buffer may have been set dirty during 1766 * The buffer may have been set dirty during
@@ -2672,7 +2686,8 @@ int nobh_writepage(struct page *page, get_block_t *get_block,
2672out: 2686out:
2673 ret = mpage_writepage(page, get_block, wbc); 2687 ret = mpage_writepage(page, get_block, wbc);
2674 if (ret == -EAGAIN) 2688 if (ret == -EAGAIN)
2675 ret = __block_write_full_page(inode, page, get_block, wbc); 2689 ret = __block_write_full_page(inode, page, get_block, wbc,
2690 end_buffer_async_write);
2676 return ret; 2691 return ret;
2677} 2692}
2678EXPORT_SYMBOL(nobh_writepage); 2693EXPORT_SYMBOL(nobh_writepage);
@@ -2830,9 +2845,10 @@ out:
2830 2845
2831/* 2846/*
2832 * The generic ->writepage function for buffer-backed address_spaces 2847 * The generic ->writepage function for buffer-backed address_spaces
2848 * this form passes in the end_io handler used to finish the IO.
2833 */ 2849 */
2834int block_write_full_page(struct page *page, get_block_t *get_block, 2850int block_write_full_page_endio(struct page *page, get_block_t *get_block,
2835 struct writeback_control *wbc) 2851 struct writeback_control *wbc, bh_end_io_t *handler)
2836{ 2852{
2837 struct inode * const inode = page->mapping->host; 2853 struct inode * const inode = page->mapping->host;
2838 loff_t i_size = i_size_read(inode); 2854 loff_t i_size = i_size_read(inode);
@@ -2841,7 +2857,8 @@ int block_write_full_page(struct page *page, get_block_t *get_block,
2841 2857
2842 /* Is the page fully inside i_size? */ 2858 /* Is the page fully inside i_size? */
2843 if (page->index < end_index) 2859 if (page->index < end_index)
2844 return __block_write_full_page(inode, page, get_block, wbc); 2860 return __block_write_full_page(inode, page, get_block, wbc,
2861 handler);
2845 2862
2846 /* Is the page fully outside i_size? (truncate in progress) */ 2863 /* Is the page fully outside i_size? (truncate in progress) */
2847 offset = i_size & (PAGE_CACHE_SIZE-1); 2864 offset = i_size & (PAGE_CACHE_SIZE-1);
@@ -2864,9 +2881,20 @@ int block_write_full_page(struct page *page, get_block_t *get_block,
2864 * writes to that region are not written out to the file." 2881 * writes to that region are not written out to the file."
2865 */ 2882 */
2866 zero_user_segment(page, offset, PAGE_CACHE_SIZE); 2883 zero_user_segment(page, offset, PAGE_CACHE_SIZE);
2867 return __block_write_full_page(inode, page, get_block, wbc); 2884 return __block_write_full_page(inode, page, get_block, wbc, handler);
2868} 2885}
2869 2886
2887/*
2888 * The generic ->writepage function for buffer-backed address_spaces
2889 */
2890int block_write_full_page(struct page *page, get_block_t *get_block,
2891 struct writeback_control *wbc)
2892{
2893 return block_write_full_page_endio(page, get_block, wbc,
2894 end_buffer_async_write);
2895}
2896
2897
2870sector_t generic_block_bmap(struct address_space *mapping, sector_t block, 2898sector_t generic_block_bmap(struct address_space *mapping, sector_t block,
2871 get_block_t *get_block) 2899 get_block_t *get_block)
2872{ 2900{
@@ -3335,9 +3363,11 @@ EXPORT_SYMBOL(block_read_full_page);
3335EXPORT_SYMBOL(block_sync_page); 3363EXPORT_SYMBOL(block_sync_page);
3336EXPORT_SYMBOL(block_truncate_page); 3364EXPORT_SYMBOL(block_truncate_page);
3337EXPORT_SYMBOL(block_write_full_page); 3365EXPORT_SYMBOL(block_write_full_page);
3366EXPORT_SYMBOL(block_write_full_page_endio);
3338EXPORT_SYMBOL(cont_write_begin); 3367EXPORT_SYMBOL(cont_write_begin);
3339EXPORT_SYMBOL(end_buffer_read_sync); 3368EXPORT_SYMBOL(end_buffer_read_sync);
3340EXPORT_SYMBOL(end_buffer_write_sync); 3369EXPORT_SYMBOL(end_buffer_write_sync);
3370EXPORT_SYMBOL(end_buffer_async_write);
3341EXPORT_SYMBOL(file_fsync); 3371EXPORT_SYMBOL(file_fsync);
3342EXPORT_SYMBOL(generic_block_bmap); 3372EXPORT_SYMBOL(generic_block_bmap);
3343EXPORT_SYMBOL(generic_cont_expand_simple); 3373EXPORT_SYMBOL(generic_cont_expand_simple);
diff --git a/fs/direct-io.c b/fs/direct-io.c
index da258e7249cc..05763bbc2050 100644
--- a/fs/direct-io.c
+++ b/fs/direct-io.c
@@ -307,8 +307,6 @@ dio_bio_alloc(struct dio *dio, struct block_device *bdev,
307 struct bio *bio; 307 struct bio *bio;
308 308
309 bio = bio_alloc(GFP_KERNEL, nr_vecs); 309 bio = bio_alloc(GFP_KERNEL, nr_vecs);
310 if (bio == NULL)
311 return -ENOMEM;
312 310
313 bio->bi_bdev = bdev; 311 bio->bi_bdev = bdev;
314 bio->bi_sector = first_sector; 312 bio->bi_sector = first_sector;
diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
index b43b95563663..acf678831103 100644
--- a/fs/ext2/inode.c
+++ b/fs/ext2/inode.c
@@ -590,9 +590,8 @@ static int ext2_get_blocks(struct inode *inode,
590 590
591 if (depth == 0) 591 if (depth == 0)
592 return (err); 592 return (err);
593reread:
594 partial = ext2_get_branch(inode, depth, offsets, chain, &err);
595 593
594 partial = ext2_get_branch(inode, depth, offsets, chain, &err);
596 /* Simplest case - block found, no allocation needed */ 595 /* Simplest case - block found, no allocation needed */
597 if (!partial) { 596 if (!partial) {
598 first_block = le32_to_cpu(chain[depth - 1].key); 597 first_block = le32_to_cpu(chain[depth - 1].key);
@@ -602,15 +601,16 @@ reread:
602 while (count < maxblocks && count <= blocks_to_boundary) { 601 while (count < maxblocks && count <= blocks_to_boundary) {
603 ext2_fsblk_t blk; 602 ext2_fsblk_t blk;
604 603
605 if (!verify_chain(chain, partial)) { 604 if (!verify_chain(chain, chain + depth - 1)) {
606 /* 605 /*
607 * Indirect block might be removed by 606 * Indirect block might be removed by
608 * truncate while we were reading it. 607 * truncate while we were reading it.
609 * Handling of that case: forget what we've 608 * Handling of that case: forget what we've
610 * got now, go to reread. 609 * got now, go to reread.
611 */ 610 */
611 err = -EAGAIN;
612 count = 0; 612 count = 0;
613 goto changed; 613 break;
614 } 614 }
615 blk = le32_to_cpu(*(chain[depth-1].p + count)); 615 blk = le32_to_cpu(*(chain[depth-1].p + count));
616 if (blk == first_block + count) 616 if (blk == first_block + count)
@@ -618,7 +618,8 @@ reread:
618 else 618 else
619 break; 619 break;
620 } 620 }
621 goto got_it; 621 if (err != -EAGAIN)
622 goto got_it;
622 } 623 }
623 624
624 /* Next simple case - plain lookup or failed read of indirect block */ 625 /* Next simple case - plain lookup or failed read of indirect block */
@@ -626,6 +627,33 @@ reread:
626 goto cleanup; 627 goto cleanup;
627 628
628 mutex_lock(&ei->truncate_mutex); 629 mutex_lock(&ei->truncate_mutex);
630 /*
631 * If the indirect block is missing while we are reading
632 * the chain(ext3_get_branch() returns -EAGAIN err), or
633 * if the chain has been changed after we grab the semaphore,
634 * (either because another process truncated this branch, or
635 * another get_block allocated this branch) re-grab the chain to see if
636 * the request block has been allocated or not.
637 *
638 * Since we already block the truncate/other get_block
639 * at this point, we will have the current copy of the chain when we
640 * splice the branch into the tree.
641 */
642 if (err == -EAGAIN || !verify_chain(chain, partial)) {
643 while (partial > chain) {
644 brelse(partial->bh);
645 partial--;
646 }
647 partial = ext2_get_branch(inode, depth, offsets, chain, &err);
648 if (!partial) {
649 count++;
650 mutex_unlock(&ei->truncate_mutex);
651 if (err)
652 goto cleanup;
653 clear_buffer_new(bh_result);
654 goto got_it;
655 }
656 }
629 657
630 /* 658 /*
631 * Okay, we need to do block allocation. Lazily initialize the block 659 * Okay, we need to do block allocation. Lazily initialize the block
@@ -683,12 +711,6 @@ cleanup:
683 partial--; 711 partial--;
684 } 712 }
685 return err; 713 return err;
686changed:
687 while (partial > chain) {
688 brelse(partial->bh);
689 partial--;
690 }
691 goto reread;
692} 714}
693 715
694int ext2_get_block(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create) 716int ext2_get_block(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create)
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index 6132353dcf62..2a1cb0979768 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -2416,8 +2416,6 @@ static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex)
2416 len = ee_len; 2416 len = ee_len;
2417 2417
2418 bio = bio_alloc(GFP_NOIO, len); 2418 bio = bio_alloc(GFP_NOIO, len);
2419 if (!bio)
2420 return -ENOMEM;
2421 bio->bi_sector = ee_pblock; 2419 bio->bi_sector = ee_pblock;
2422 bio->bi_bdev = inode->i_sb->s_bdev; 2420 bio->bi_bdev = inode->i_sb->s_bdev;
2423 2421
diff --git a/fs/fuse/file.c b/fs/fuse/file.c
index 2b25133524a3..06f30e965676 100644
--- a/fs/fuse/file.c
+++ b/fs/fuse/file.c
@@ -938,9 +938,9 @@ static void fuse_release_user_pages(struct fuse_req *req, int write)
938} 938}
939 939
940static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf, 940static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf,
941 unsigned *nbytesp, int write) 941 size_t *nbytesp, int write)
942{ 942{
943 unsigned nbytes = *nbytesp; 943 size_t nbytes = *nbytesp;
944 unsigned long user_addr = (unsigned long) buf; 944 unsigned long user_addr = (unsigned long) buf;
945 unsigned offset = user_addr & ~PAGE_MASK; 945 unsigned offset = user_addr & ~PAGE_MASK;
946 int npages; 946 int npages;
@@ -955,7 +955,7 @@ static int fuse_get_user_pages(struct fuse_req *req, const char __user *buf,
955 return 0; 955 return 0;
956 } 956 }
957 957
958 nbytes = min(nbytes, (unsigned) FUSE_MAX_PAGES_PER_REQ << PAGE_SHIFT); 958 nbytes = min_t(size_t, nbytes, FUSE_MAX_PAGES_PER_REQ << PAGE_SHIFT);
959 npages = (nbytes + offset + PAGE_SIZE - 1) >> PAGE_SHIFT; 959 npages = (nbytes + offset + PAGE_SIZE - 1) >> PAGE_SHIFT;
960 npages = clamp(npages, 1, FUSE_MAX_PAGES_PER_REQ); 960 npages = clamp(npages, 1, FUSE_MAX_PAGES_PER_REQ);
961 down_read(&current->mm->mmap_sem); 961 down_read(&current->mm->mmap_sem);
@@ -1298,6 +1298,8 @@ static int fuse_direct_mmap(struct file *file, struct vm_area_struct *vma)
1298 if (vma->vm_flags & VM_MAYSHARE) 1298 if (vma->vm_flags & VM_MAYSHARE)
1299 return -ENODEV; 1299 return -ENODEV;
1300 1300
1301 invalidate_inode_pages2(file->f_mapping);
1302
1301 return generic_file_mmap(file, vma); 1303 return generic_file_mmap(file, vma);
1302} 1304}
1303 1305
diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 3984e47d1d33..1afd9f26bcb1 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -597,7 +597,6 @@ __acquires(&gl->gl_spin)
597 597
598 GLOCK_BUG_ON(gl, test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags)); 598 GLOCK_BUG_ON(gl, test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags));
599 599
600 down_read(&gfs2_umount_flush_sem);
601 if (test_bit(GLF_DEMOTE, &gl->gl_flags) && 600 if (test_bit(GLF_DEMOTE, &gl->gl_flags) &&
602 gl->gl_demote_state != gl->gl_state) { 601 gl->gl_demote_state != gl->gl_state) {
603 if (find_first_holder(gl)) 602 if (find_first_holder(gl))
@@ -614,15 +613,14 @@ __acquires(&gl->gl_spin)
614 if (ret == 0) 613 if (ret == 0)
615 goto out_unlock; 614 goto out_unlock;
616 if (ret == 2) 615 if (ret == 2)
617 goto out_sem; 616 goto out;
618 gh = find_first_waiter(gl); 617 gh = find_first_waiter(gl);
619 gl->gl_target = gh->gh_state; 618 gl->gl_target = gh->gh_state;
620 if (!(gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB))) 619 if (!(gh->gh_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB)))
621 do_error(gl, 0); /* Fail queued try locks */ 620 do_error(gl, 0); /* Fail queued try locks */
622 } 621 }
623 do_xmote(gl, gh, gl->gl_target); 622 do_xmote(gl, gh, gl->gl_target);
624out_sem: 623out:
625 up_read(&gfs2_umount_flush_sem);
626 return; 624 return;
627 625
628out_sched: 626out_sched:
@@ -631,7 +629,7 @@ out_sched:
631 gfs2_glock_put(gl); 629 gfs2_glock_put(gl);
632out_unlock: 630out_unlock:
633 clear_bit(GLF_LOCK, &gl->gl_flags); 631 clear_bit(GLF_LOCK, &gl->gl_flags);
634 goto out_sem; 632 goto out;
635} 633}
636 634
637static void glock_work_func(struct work_struct *work) 635static void glock_work_func(struct work_struct *work)
@@ -641,6 +639,7 @@ static void glock_work_func(struct work_struct *work)
641 639
642 if (test_and_clear_bit(GLF_REPLY_PENDING, &gl->gl_flags)) 640 if (test_and_clear_bit(GLF_REPLY_PENDING, &gl->gl_flags))
643 finish_xmote(gl, gl->gl_reply); 641 finish_xmote(gl, gl->gl_reply);
642 down_read(&gfs2_umount_flush_sem);
644 spin_lock(&gl->gl_spin); 643 spin_lock(&gl->gl_spin);
645 if (test_and_clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) && 644 if (test_and_clear_bit(GLF_PENDING_DEMOTE, &gl->gl_flags) &&
646 gl->gl_state != LM_ST_UNLOCKED && 645 gl->gl_state != LM_ST_UNLOCKED &&
@@ -653,6 +652,7 @@ static void glock_work_func(struct work_struct *work)
653 } 652 }
654 run_queue(gl, 0); 653 run_queue(gl, 0);
655 spin_unlock(&gl->gl_spin); 654 spin_unlock(&gl->gl_spin);
655 up_read(&gfs2_umount_flush_sem);
656 if (!delay || 656 if (!delay ||
657 queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0) 657 queue_delayed_work(glock_workqueue, &gl->gl_work, delay) == 0)
658 gfs2_glock_put(gl); 658 gfs2_glock_put(gl);
diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c
index 7b277d449155..5a31d426116f 100644
--- a/fs/gfs2/inode.c
+++ b/fs/gfs2/inode.c
@@ -137,15 +137,15 @@ void gfs2_set_iop(struct inode *inode)
137 if (S_ISREG(mode)) { 137 if (S_ISREG(mode)) {
138 inode->i_op = &gfs2_file_iops; 138 inode->i_op = &gfs2_file_iops;
139 if (gfs2_localflocks(sdp)) 139 if (gfs2_localflocks(sdp))
140 inode->i_fop = gfs2_file_fops_nolock; 140 inode->i_fop = &gfs2_file_fops_nolock;
141 else 141 else
142 inode->i_fop = gfs2_file_fops; 142 inode->i_fop = &gfs2_file_fops;
143 } else if (S_ISDIR(mode)) { 143 } else if (S_ISDIR(mode)) {
144 inode->i_op = &gfs2_dir_iops; 144 inode->i_op = &gfs2_dir_iops;
145 if (gfs2_localflocks(sdp)) 145 if (gfs2_localflocks(sdp))
146 inode->i_fop = gfs2_dir_fops_nolock; 146 inode->i_fop = &gfs2_dir_fops_nolock;
147 else 147 else
148 inode->i_fop = gfs2_dir_fops; 148 inode->i_fop = &gfs2_dir_fops;
149 } else if (S_ISLNK(mode)) { 149 } else if (S_ISLNK(mode)) {
150 inode->i_op = &gfs2_symlink_iops; 150 inode->i_op = &gfs2_symlink_iops;
151 } else { 151 } else {
diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h
index dca4fee3078b..c30be2b66580 100644
--- a/fs/gfs2/inode.h
+++ b/fs/gfs2/inode.h
@@ -101,21 +101,23 @@ void gfs2_dinode_print(const struct gfs2_inode *ip);
101extern const struct inode_operations gfs2_file_iops; 101extern const struct inode_operations gfs2_file_iops;
102extern const struct inode_operations gfs2_dir_iops; 102extern const struct inode_operations gfs2_dir_iops;
103extern const struct inode_operations gfs2_symlink_iops; 103extern const struct inode_operations gfs2_symlink_iops;
104extern const struct file_operations *gfs2_file_fops_nolock; 104extern const struct file_operations gfs2_file_fops_nolock;
105extern const struct file_operations *gfs2_dir_fops_nolock; 105extern const struct file_operations gfs2_dir_fops_nolock;
106 106
107extern void gfs2_set_inode_flags(struct inode *inode); 107extern void gfs2_set_inode_flags(struct inode *inode);
108 108
109#ifdef CONFIG_GFS2_FS_LOCKING_DLM 109#ifdef CONFIG_GFS2_FS_LOCKING_DLM
110extern const struct file_operations *gfs2_file_fops; 110extern const struct file_operations gfs2_file_fops;
111extern const struct file_operations *gfs2_dir_fops; 111extern const struct file_operations gfs2_dir_fops;
112
112static inline int gfs2_localflocks(const struct gfs2_sbd *sdp) 113static inline int gfs2_localflocks(const struct gfs2_sbd *sdp)
113{ 114{
114 return sdp->sd_args.ar_localflocks; 115 return sdp->sd_args.ar_localflocks;
115} 116}
116#else /* Single node only */ 117#else /* Single node only */
117#define gfs2_file_fops NULL 118#define gfs2_file_fops gfs2_file_fops_nolock
118#define gfs2_dir_fops NULL 119#define gfs2_dir_fops gfs2_dir_fops_nolock
120
119static inline int gfs2_localflocks(const struct gfs2_sbd *sdp) 121static inline int gfs2_localflocks(const struct gfs2_sbd *sdp)
120{ 122{
121 return 1; 123 return 1;
diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c
index 70b9b8548945..101caf3ee861 100644
--- a/fs/gfs2/ops_file.c
+++ b/fs/gfs2/ops_file.c
@@ -705,7 +705,7 @@ static int gfs2_flock(struct file *file, int cmd, struct file_lock *fl)
705 } 705 }
706} 706}
707 707
708const struct file_operations *gfs2_file_fops = &(const struct file_operations){ 708const struct file_operations gfs2_file_fops = {
709 .llseek = gfs2_llseek, 709 .llseek = gfs2_llseek,
710 .read = do_sync_read, 710 .read = do_sync_read,
711 .aio_read = generic_file_aio_read, 711 .aio_read = generic_file_aio_read,
@@ -723,7 +723,7 @@ const struct file_operations *gfs2_file_fops = &(const struct file_operations){
723 .setlease = gfs2_setlease, 723 .setlease = gfs2_setlease,
724}; 724};
725 725
726const struct file_operations *gfs2_dir_fops = &(const struct file_operations){ 726const struct file_operations gfs2_dir_fops = {
727 .readdir = gfs2_readdir, 727 .readdir = gfs2_readdir,
728 .unlocked_ioctl = gfs2_ioctl, 728 .unlocked_ioctl = gfs2_ioctl,
729 .open = gfs2_open, 729 .open = gfs2_open,
@@ -735,7 +735,7 @@ const struct file_operations *gfs2_dir_fops = &(const struct file_operations){
735 735
736#endif /* CONFIG_GFS2_FS_LOCKING_DLM */ 736#endif /* CONFIG_GFS2_FS_LOCKING_DLM */
737 737
738const struct file_operations *gfs2_file_fops_nolock = &(const struct file_operations){ 738const struct file_operations gfs2_file_fops_nolock = {
739 .llseek = gfs2_llseek, 739 .llseek = gfs2_llseek,
740 .read = do_sync_read, 740 .read = do_sync_read,
741 .aio_read = generic_file_aio_read, 741 .aio_read = generic_file_aio_read,
@@ -751,7 +751,7 @@ const struct file_operations *gfs2_file_fops_nolock = &(const struct file_operat
751 .setlease = generic_setlease, 751 .setlease = generic_setlease,
752}; 752};
753 753
754const struct file_operations *gfs2_dir_fops_nolock = &(const struct file_operations){ 754const struct file_operations gfs2_dir_fops_nolock = {
755 .readdir = gfs2_readdir, 755 .readdir = gfs2_readdir,
756 .unlocked_ioctl = gfs2_ioctl, 756 .unlocked_ioctl = gfs2_ioctl,
757 .open = gfs2_open, 757 .open = gfs2_open,
diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c
index 51883b3ad89c..650a730707b7 100644
--- a/fs/gfs2/ops_fstype.c
+++ b/fs/gfs2/ops_fstype.c
@@ -272,11 +272,6 @@ static int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector)
272 lock_page(page); 272 lock_page(page);
273 273
274 bio = bio_alloc(GFP_NOFS, 1); 274 bio = bio_alloc(GFP_NOFS, 1);
275 if (unlikely(!bio)) {
276 __free_page(page);
277 return -ENOBUFS;
278 }
279
280 bio->bi_sector = sector * (sb->s_blocksize >> 9); 275 bio->bi_sector = sector * (sb->s_blocksize >> 9);
281 bio->bi_bdev = sb->s_bdev; 276 bio->bi_bdev = sb->s_bdev;
282 bio_add_page(bio, page, PAGE_SIZE, 0); 277 bio_add_page(bio, page, PAGE_SIZE, 0);
diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c
index abd5429ae285..1c70fa5168d6 100644
--- a/fs/gfs2/ops_inode.c
+++ b/fs/gfs2/ops_inode.c
@@ -371,6 +371,7 @@ static int gfs2_symlink(struct inode *dir, struct dentry *dentry,
371 ip = ghs[1].gh_gl->gl_object; 371 ip = ghs[1].gh_gl->gl_object;
372 372
373 ip->i_disksize = size; 373 ip->i_disksize = size;
374 i_size_write(inode, size);
374 375
375 error = gfs2_meta_inode_buffer(ip, &dibh); 376 error = gfs2_meta_inode_buffer(ip, &dibh);
376 377
diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c
index 8d53f66b5bcc..152e6c4a0dca 100644
--- a/fs/gfs2/quota.c
+++ b/fs/gfs2/quota.c
@@ -81,7 +81,7 @@ struct gfs2_quota_change_host {
81 81
82static LIST_HEAD(qd_lru_list); 82static LIST_HEAD(qd_lru_list);
83static atomic_t qd_lru_count = ATOMIC_INIT(0); 83static atomic_t qd_lru_count = ATOMIC_INIT(0);
84static spinlock_t qd_lru_lock = SPIN_LOCK_UNLOCKED; 84static DEFINE_SPINLOCK(qd_lru_lock);
85 85
86int gfs2_shrink_qd_memory(int nr, gfp_t gfp_mask) 86int gfs2_shrink_qd_memory(int nr, gfp_t gfp_mask)
87{ 87{
@@ -1364,7 +1364,7 @@ int gfs2_quotad(void *data)
1364 refrigerator(); 1364 refrigerator();
1365 t = min(quotad_timeo, statfs_timeo); 1365 t = min(quotad_timeo, statfs_timeo);
1366 1366
1367 prepare_to_wait(&sdp->sd_quota_wait, &wait, TASK_UNINTERRUPTIBLE); 1367 prepare_to_wait(&sdp->sd_quota_wait, &wait, TASK_INTERRUPTIBLE);
1368 spin_lock(&sdp->sd_trunc_lock); 1368 spin_lock(&sdp->sd_trunc_lock);
1369 empty = list_empty(&sdp->sd_trunc_list); 1369 empty = list_empty(&sdp->sd_trunc_list);
1370 spin_unlock(&sdp->sd_trunc_lock); 1370 spin_unlock(&sdp->sd_trunc_lock);
diff --git a/fs/hfs/inode.c b/fs/hfs/inode.c
index 9435dda8f1e0..a1cbff2b4d99 100644
--- a/fs/hfs/inode.c
+++ b/fs/hfs/inode.c
@@ -70,6 +70,10 @@ static int hfs_releasepage(struct page *page, gfp_t mask)
70 BUG(); 70 BUG();
71 return 0; 71 return 0;
72 } 72 }
73
74 if (!tree)
75 return 0;
76
73 if (tree->node_size >= PAGE_CACHE_SIZE) { 77 if (tree->node_size >= PAGE_CACHE_SIZE) {
74 nidx = page->index >> (tree->node_size_shift - PAGE_CACHE_SHIFT); 78 nidx = page->index >> (tree->node_size_shift - PAGE_CACHE_SHIFT);
75 spin_lock(&tree->hash_lock); 79 spin_lock(&tree->hash_lock);
diff --git a/fs/hfs/mdb.c b/fs/hfs/mdb.c
index 36ca2e1a4fa3..7b6165f25fbe 100644
--- a/fs/hfs/mdb.c
+++ b/fs/hfs/mdb.c
@@ -349,6 +349,7 @@ void hfs_mdb_put(struct super_block *sb)
349 if (HFS_SB(sb)->nls_disk) 349 if (HFS_SB(sb)->nls_disk)
350 unload_nls(HFS_SB(sb)->nls_disk); 350 unload_nls(HFS_SB(sb)->nls_disk);
351 351
352 free_pages((unsigned long)HFS_SB(sb)->bitmap, PAGE_SIZE < 8192 ? 1 : 0);
352 kfree(HFS_SB(sb)); 353 kfree(HFS_SB(sb));
353 sb->s_fs_info = NULL; 354 sb->s_fs_info = NULL;
354} 355}
diff --git a/fs/inode.c b/fs/inode.c
index d06d6d268de9..6ad14a1cd8c9 100644
--- a/fs/inode.c
+++ b/fs/inode.c
@@ -1470,42 +1470,6 @@ static void __wait_on_freeing_inode(struct inode *inode)
1470 spin_lock(&inode_lock); 1470 spin_lock(&inode_lock);
1471} 1471}
1472 1472
1473/*
1474 * We rarely want to lock two inodes that do not have a parent/child
1475 * relationship (such as directory, child inode) simultaneously. The
1476 * vast majority of file systems should be able to get along fine
1477 * without this. Do not use these functions except as a last resort.
1478 */
1479void inode_double_lock(struct inode *inode1, struct inode *inode2)
1480{
1481 if (inode1 == NULL || inode2 == NULL || inode1 == inode2) {
1482 if (inode1)
1483 mutex_lock(&inode1->i_mutex);
1484 else if (inode2)
1485 mutex_lock(&inode2->i_mutex);
1486 return;
1487 }
1488
1489 if (inode1 < inode2) {
1490 mutex_lock_nested(&inode1->i_mutex, I_MUTEX_PARENT);
1491 mutex_lock_nested(&inode2->i_mutex, I_MUTEX_CHILD);
1492 } else {
1493 mutex_lock_nested(&inode2->i_mutex, I_MUTEX_PARENT);
1494 mutex_lock_nested(&inode1->i_mutex, I_MUTEX_CHILD);
1495 }
1496}
1497EXPORT_SYMBOL(inode_double_lock);
1498
1499void inode_double_unlock(struct inode *inode1, struct inode *inode2)
1500{
1501 if (inode1)
1502 mutex_unlock(&inode1->i_mutex);
1503
1504 if (inode2 && inode2 != inode1)
1505 mutex_unlock(&inode2->i_mutex);
1506}
1507EXPORT_SYMBOL(inode_double_unlock);
1508
1509static __initdata unsigned long ihash_entries; 1473static __initdata unsigned long ihash_entries;
1510static int __init set_ihash_entries(char *str) 1474static int __init set_ihash_entries(char *str)
1511{ 1475{
diff --git a/fs/jbd/revoke.c b/fs/jbd/revoke.c
index c7bd649bbbdc..3e9afc2a91d2 100644
--- a/fs/jbd/revoke.c
+++ b/fs/jbd/revoke.c
@@ -55,6 +55,25 @@
55 * need do nothing. 55 * need do nothing.
56 * RevokeValid set, Revoked set: 56 * RevokeValid set, Revoked set:
57 * buffer has been revoked. 57 * buffer has been revoked.
58 *
59 * Locking rules:
60 * We keep two hash tables of revoke records. One hashtable belongs to the
61 * running transaction (is pointed to by journal->j_revoke), the other one
62 * belongs to the committing transaction. Accesses to the second hash table
63 * happen only from the kjournald and no other thread touches this table. Also
64 * journal_switch_revoke_table() which switches which hashtable belongs to the
65 * running and which to the committing transaction is called only from
66 * kjournald. Therefore we need no locks when accessing the hashtable belonging
67 * to the committing transaction.
68 *
69 * All users operating on the hash table belonging to the running transaction
70 * have a handle to the transaction. Therefore they are safe from kjournald
71 * switching hash tables under them. For operations on the lists of entries in
72 * the hash table j_revoke_lock is used.
73 *
74 * Finally, also replay code uses the hash tables but at this moment noone else
75 * can touch them (filesystem isn't mounted yet) and hence no locking is
76 * needed.
58 */ 77 */
59 78
60#ifndef __KERNEL__ 79#ifndef __KERNEL__
@@ -402,8 +421,6 @@ int journal_revoke(handle_t *handle, unsigned long blocknr,
402 * the second time we would still have a pending revoke to cancel. So, 421 * the second time we would still have a pending revoke to cancel. So,
403 * do not trust the Revoked bit on buffers unless RevokeValid is also 422 * do not trust the Revoked bit on buffers unless RevokeValid is also
404 * set. 423 * set.
405 *
406 * The caller must have the journal locked.
407 */ 424 */
408int journal_cancel_revoke(handle_t *handle, struct journal_head *jh) 425int journal_cancel_revoke(handle_t *handle, struct journal_head *jh)
409{ 426{
@@ -481,10 +498,7 @@ void journal_switch_revoke_table(journal_t *journal)
481/* 498/*
482 * Write revoke records to the journal for all entries in the current 499 * Write revoke records to the journal for all entries in the current
483 * revoke hash, deleting the entries as we go. 500 * revoke hash, deleting the entries as we go.
484 *
485 * Called with the journal lock held.
486 */ 501 */
487
488void journal_write_revoke_records(journal_t *journal, 502void journal_write_revoke_records(journal_t *journal,
489 transaction_t *transaction) 503 transaction_t *transaction)
490{ 504{
diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c
index 24638e059bf3..064279e33bbb 100644
--- a/fs/nilfs2/bmap.c
+++ b/fs/nilfs2/bmap.c
@@ -688,6 +688,8 @@ static const struct nilfs_bmap_ptr_operations nilfs_bmap_ptr_ops_gc = {
688 .bpop_translate = NULL, 688 .bpop_translate = NULL,
689}; 689};
690 690
691static struct lock_class_key nilfs_bmap_dat_lock_key;
692
691/** 693/**
692 * nilfs_bmap_read - read a bmap from an inode 694 * nilfs_bmap_read - read a bmap from an inode
693 * @bmap: bmap 695 * @bmap: bmap
@@ -715,6 +717,7 @@ int nilfs_bmap_read(struct nilfs_bmap *bmap, struct nilfs_inode *raw_inode)
715 bmap->b_pops = &nilfs_bmap_ptr_ops_p; 717 bmap->b_pops = &nilfs_bmap_ptr_ops_p;
716 bmap->b_last_allocated_key = 0; /* XXX: use macro */ 718 bmap->b_last_allocated_key = 0; /* XXX: use macro */
717 bmap->b_last_allocated_ptr = NILFS_BMAP_NEW_PTR_INIT; 719 bmap->b_last_allocated_ptr = NILFS_BMAP_NEW_PTR_INIT;
720 lockdep_set_class(&bmap->b_sem, &nilfs_bmap_dat_lock_key);
718 break; 721 break;
719 case NILFS_CPFILE_INO: 722 case NILFS_CPFILE_INO:
720 case NILFS_SUFILE_INO: 723 case NILFS_SUFILE_INO:
@@ -772,6 +775,7 @@ void nilfs_bmap_init_gcdat(struct nilfs_bmap *gcbmap, struct nilfs_bmap *bmap)
772{ 775{
773 memcpy(gcbmap, bmap, sizeof(union nilfs_bmap_union)); 776 memcpy(gcbmap, bmap, sizeof(union nilfs_bmap_union));
774 init_rwsem(&gcbmap->b_sem); 777 init_rwsem(&gcbmap->b_sem);
778 lockdep_set_class(&bmap->b_sem, &nilfs_bmap_dat_lock_key);
775 gcbmap->b_inode = &NILFS_BMAP_I(gcbmap)->vfs_inode; 779 gcbmap->b_inode = &NILFS_BMAP_I(gcbmap)->vfs_inode;
776} 780}
777 781
@@ -779,5 +783,6 @@ void nilfs_bmap_commit_gcdat(struct nilfs_bmap *gcbmap, struct nilfs_bmap *bmap)
779{ 783{
780 memcpy(bmap, gcbmap, sizeof(union nilfs_bmap_union)); 784 memcpy(bmap, gcbmap, sizeof(union nilfs_bmap_union));
781 init_rwsem(&bmap->b_sem); 785 init_rwsem(&bmap->b_sem);
786 lockdep_set_class(&bmap->b_sem, &nilfs_bmap_dat_lock_key);
782 bmap->b_inode = &NILFS_BMAP_I(bmap)->vfs_inode; 787 bmap->b_inode = &NILFS_BMAP_I(bmap)->vfs_inode;
783} 788}
diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h
index 7558c977db02..3d0c18a16db1 100644
--- a/fs/nilfs2/nilfs.h
+++ b/fs/nilfs2/nilfs.h
@@ -35,11 +35,6 @@
35#include "bmap_union.h" 35#include "bmap_union.h"
36 36
37/* 37/*
38 * NILFS filesystem version
39 */
40#define NILFS_VERSION "2.0.5"
41
42/*
43 * nilfs inode data in memory 38 * nilfs inode data in memory
44 */ 39 */
45struct nilfs_inode_info { 40struct nilfs_inode_info {
diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c
index 6ade0963fc1d..4fc081e47d70 100644
--- a/fs/nilfs2/recovery.c
+++ b/fs/nilfs2/recovery.c
@@ -413,7 +413,6 @@ static int nilfs_prepare_segment_for_recovery(struct the_nilfs *nilfs,
413 struct nilfs_segment_entry *ent, *n; 413 struct nilfs_segment_entry *ent, *n;
414 struct inode *sufile = nilfs->ns_sufile; 414 struct inode *sufile = nilfs->ns_sufile;
415 __u64 segnum[4]; 415 __u64 segnum[4];
416 time_t mtime;
417 int err; 416 int err;
418 int i; 417 int i;
419 418
@@ -442,24 +441,13 @@ static int nilfs_prepare_segment_for_recovery(struct the_nilfs *nilfs,
442 * Collecting segments written after the latest super root. 441 * Collecting segments written after the latest super root.
443 * These are marked dirty to avoid being reallocated in the next write. 442 * These are marked dirty to avoid being reallocated in the next write.
444 */ 443 */
445 mtime = get_seconds();
446 list_for_each_entry_safe(ent, n, head, list) { 444 list_for_each_entry_safe(ent, n, head, list) {
447 if (ent->segnum == segnum[0]) { 445 if (ent->segnum != segnum[0]) {
448 list_del(&ent->list); 446 err = nilfs_sufile_scrap(sufile, ent->segnum);
449 nilfs_free_segment_entry(ent); 447 if (unlikely(err))
450 continue; 448 goto failed;
451 }
452 err = nilfs_open_segment_entry(ent, sufile);
453 if (unlikely(err))
454 goto failed;
455 if (!nilfs_segment_usage_dirty(ent->raw_su)) {
456 /* make the segment garbage */
457 ent->raw_su->su_nblocks = cpu_to_le32(0);
458 ent->raw_su->su_lastmod = cpu_to_le32(mtime);
459 nilfs_segment_usage_set_dirty(ent->raw_su);
460 } 449 }
461 list_del(&ent->list); 450 list_del(&ent->list);
462 nilfs_close_segment_entry(ent, sufile);
463 nilfs_free_segment_entry(ent); 451 nilfs_free_segment_entry(ent);
464 } 452 }
465 453
diff --git a/fs/nilfs2/sufile.c b/fs/nilfs2/sufile.c
index c774cf397e2f..98e68677f045 100644
--- a/fs/nilfs2/sufile.c
+++ b/fs/nilfs2/sufile.c
@@ -93,6 +93,52 @@ nilfs_sufile_get_segment_usage_block(struct inode *sufile, __u64 segnum,
93 create, NULL, bhp); 93 create, NULL, bhp);
94} 94}
95 95
96static void nilfs_sufile_mod_counter(struct buffer_head *header_bh,
97 u64 ncleanadd, u64 ndirtyadd)
98{
99 struct nilfs_sufile_header *header;
100 void *kaddr;
101
102 kaddr = kmap_atomic(header_bh->b_page, KM_USER0);
103 header = kaddr + bh_offset(header_bh);
104 le64_add_cpu(&header->sh_ncleansegs, ncleanadd);
105 le64_add_cpu(&header->sh_ndirtysegs, ndirtyadd);
106 kunmap_atomic(kaddr, KM_USER0);
107
108 nilfs_mdt_mark_buffer_dirty(header_bh);
109}
110
111int nilfs_sufile_update(struct inode *sufile, __u64 segnum, int create,
112 void (*dofunc)(struct inode *, __u64,
113 struct buffer_head *,
114 struct buffer_head *))
115{
116 struct buffer_head *header_bh, *bh;
117 int ret;
118
119 if (unlikely(segnum >= nilfs_sufile_get_nsegments(sufile))) {
120 printk(KERN_WARNING "%s: invalid segment number: %llu\n",
121 __func__, (unsigned long long)segnum);
122 return -EINVAL;
123 }
124 down_write(&NILFS_MDT(sufile)->mi_sem);
125
126 ret = nilfs_sufile_get_header_block(sufile, &header_bh);
127 if (ret < 0)
128 goto out_sem;
129
130 ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, create, &bh);
131 if (!ret) {
132 dofunc(sufile, segnum, header_bh, bh);
133 brelse(bh);
134 }
135 brelse(header_bh);
136
137 out_sem:
138 up_write(&NILFS_MDT(sufile)->mi_sem);
139 return ret;
140}
141
96/** 142/**
97 * nilfs_sufile_alloc - allocate a segment 143 * nilfs_sufile_alloc - allocate a segment
98 * @sufile: inode of segment usage file 144 * @sufile: inode of segment usage file
@@ -113,7 +159,6 @@ nilfs_sufile_get_segment_usage_block(struct inode *sufile, __u64 segnum,
113int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump) 159int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump)
114{ 160{
115 struct buffer_head *header_bh, *su_bh; 161 struct buffer_head *header_bh, *su_bh;
116 struct the_nilfs *nilfs;
117 struct nilfs_sufile_header *header; 162 struct nilfs_sufile_header *header;
118 struct nilfs_segment_usage *su; 163 struct nilfs_segment_usage *su;
119 size_t susz = NILFS_MDT(sufile)->mi_entry_size; 164 size_t susz = NILFS_MDT(sufile)->mi_entry_size;
@@ -124,8 +169,6 @@ int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump)
124 169
125 down_write(&NILFS_MDT(sufile)->mi_sem); 170 down_write(&NILFS_MDT(sufile)->mi_sem);
126 171
127 nilfs = NILFS_MDT(sufile)->mi_nilfs;
128
129 ret = nilfs_sufile_get_header_block(sufile, &header_bh); 172 ret = nilfs_sufile_get_header_block(sufile, &header_bh);
130 if (ret < 0) 173 if (ret < 0)
131 goto out_sem; 174 goto out_sem;
@@ -192,165 +235,84 @@ int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump)
192 return ret; 235 return ret;
193} 236}
194 237
195/** 238void nilfs_sufile_do_cancel_free(struct inode *sufile, __u64 segnum,
196 * nilfs_sufile_cancel_free - 239 struct buffer_head *header_bh,
197 * @sufile: inode of segment usage file 240 struct buffer_head *su_bh)
198 * @segnum: segment number
199 *
200 * Description:
201 *
202 * Return Value: On success, 0 is returned. On error, one of the following
203 * negative error codes is returned.
204 *
205 * %-EIO - I/O error.
206 *
207 * %-ENOMEM - Insufficient amount of memory available.
208 */
209int nilfs_sufile_cancel_free(struct inode *sufile, __u64 segnum)
210{ 241{
211 struct buffer_head *header_bh, *su_bh;
212 struct the_nilfs *nilfs;
213 struct nilfs_sufile_header *header;
214 struct nilfs_segment_usage *su; 242 struct nilfs_segment_usage *su;
215 void *kaddr; 243 void *kaddr;
216 int ret;
217
218 down_write(&NILFS_MDT(sufile)->mi_sem);
219
220 nilfs = NILFS_MDT(sufile)->mi_nilfs;
221
222 ret = nilfs_sufile_get_header_block(sufile, &header_bh);
223 if (ret < 0)
224 goto out_sem;
225
226 ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 0, &su_bh);
227 if (ret < 0)
228 goto out_header;
229 244
230 kaddr = kmap_atomic(su_bh->b_page, KM_USER0); 245 kaddr = kmap_atomic(su_bh->b_page, KM_USER0);
231 su = nilfs_sufile_block_get_segment_usage( 246 su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr);
232 sufile, segnum, su_bh, kaddr);
233 if (unlikely(!nilfs_segment_usage_clean(su))) { 247 if (unlikely(!nilfs_segment_usage_clean(su))) {
234 printk(KERN_WARNING "%s: segment %llu must be clean\n", 248 printk(KERN_WARNING "%s: segment %llu must be clean\n",
235 __func__, (unsigned long long)segnum); 249 __func__, (unsigned long long)segnum);
236 kunmap_atomic(kaddr, KM_USER0); 250 kunmap_atomic(kaddr, KM_USER0);
237 goto out_su_bh; 251 return;
238 } 252 }
239 nilfs_segment_usage_set_dirty(su); 253 nilfs_segment_usage_set_dirty(su);
240 kunmap_atomic(kaddr, KM_USER0); 254 kunmap_atomic(kaddr, KM_USER0);
241 255
242 kaddr = kmap_atomic(header_bh->b_page, KM_USER0); 256 nilfs_sufile_mod_counter(header_bh, -1, 1);
243 header = nilfs_sufile_block_get_header(sufile, header_bh, kaddr);
244 le64_add_cpu(&header->sh_ncleansegs, -1);
245 le64_add_cpu(&header->sh_ndirtysegs, 1);
246 kunmap_atomic(kaddr, KM_USER0);
247
248 nilfs_mdt_mark_buffer_dirty(header_bh);
249 nilfs_mdt_mark_buffer_dirty(su_bh); 257 nilfs_mdt_mark_buffer_dirty(su_bh);
250 nilfs_mdt_mark_dirty(sufile); 258 nilfs_mdt_mark_dirty(sufile);
251
252 out_su_bh:
253 brelse(su_bh);
254 out_header:
255 brelse(header_bh);
256 out_sem:
257 up_write(&NILFS_MDT(sufile)->mi_sem);
258 return ret;
259} 259}
260 260
261/** 261void nilfs_sufile_do_scrap(struct inode *sufile, __u64 segnum,
262 * nilfs_sufile_freev - free segments 262 struct buffer_head *header_bh,
263 * @sufile: inode of segment usage file 263 struct buffer_head *su_bh)
264 * @segnum: array of segment numbers
265 * @nsegs: number of segments
266 *
267 * Description: nilfs_sufile_freev() frees segments specified by @segnum and
268 * @nsegs, which must have been returned by a previous call to
269 * nilfs_sufile_alloc().
270 *
271 * Return Value: On success, 0 is returned. On error, one of the following
272 * negative error codes is returned.
273 *
274 * %-EIO - I/O error.
275 *
276 * %-ENOMEM - Insufficient amount of memory available.
277 */
278#define NILFS_SUFILE_FREEV_PREALLOC 16
279int nilfs_sufile_freev(struct inode *sufile, __u64 *segnum, size_t nsegs)
280{ 264{
281 struct buffer_head *header_bh, **su_bh,
282 *su_bh_prealloc[NILFS_SUFILE_FREEV_PREALLOC];
283 struct the_nilfs *nilfs;
284 struct nilfs_sufile_header *header;
285 struct nilfs_segment_usage *su; 265 struct nilfs_segment_usage *su;
286 void *kaddr; 266 void *kaddr;
287 int ret, i; 267 int clean, dirty;
288 268
289 down_write(&NILFS_MDT(sufile)->mi_sem); 269 kaddr = kmap_atomic(su_bh->b_page, KM_USER0);
290 270 su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr);
291 nilfs = NILFS_MDT(sufile)->mi_nilfs; 271 if (su->su_flags == cpu_to_le32(1UL << NILFS_SEGMENT_USAGE_DIRTY) &&
292 272 su->su_nblocks == cpu_to_le32(0)) {
293 /* prepare resources */
294 if (nsegs <= NILFS_SUFILE_FREEV_PREALLOC)
295 su_bh = su_bh_prealloc;
296 else {
297 su_bh = kmalloc(sizeof(*su_bh) * nsegs, GFP_NOFS);
298 if (su_bh == NULL) {
299 ret = -ENOMEM;
300 goto out_sem;
301 }
302 }
303
304 ret = nilfs_sufile_get_header_block(sufile, &header_bh);
305 if (ret < 0)
306 goto out_su_bh;
307 for (i = 0; i < nsegs; i++) {
308 ret = nilfs_sufile_get_segment_usage_block(sufile, segnum[i],
309 0, &su_bh[i]);
310 if (ret < 0)
311 goto out_bh;
312 }
313
314 /* free segments */
315 for (i = 0; i < nsegs; i++) {
316 kaddr = kmap_atomic(su_bh[i]->b_page, KM_USER0);
317 su = nilfs_sufile_block_get_segment_usage(
318 sufile, segnum[i], su_bh[i], kaddr);
319 WARN_ON(nilfs_segment_usage_error(su));
320 nilfs_segment_usage_set_clean(su);
321 kunmap_atomic(kaddr, KM_USER0); 273 kunmap_atomic(kaddr, KM_USER0);
322 nilfs_mdt_mark_buffer_dirty(su_bh[i]); 274 return;
323 } 275 }
324 kaddr = kmap_atomic(header_bh->b_page, KM_USER0); 276 clean = nilfs_segment_usage_clean(su);
325 header = nilfs_sufile_block_get_header(sufile, header_bh, kaddr); 277 dirty = nilfs_segment_usage_dirty(su);
326 le64_add_cpu(&header->sh_ncleansegs, nsegs); 278
327 le64_add_cpu(&header->sh_ndirtysegs, -(u64)nsegs); 279 /* make the segment garbage */
280 su->su_lastmod = cpu_to_le64(0);
281 su->su_nblocks = cpu_to_le32(0);
282 su->su_flags = cpu_to_le32(1UL << NILFS_SEGMENT_USAGE_DIRTY);
328 kunmap_atomic(kaddr, KM_USER0); 283 kunmap_atomic(kaddr, KM_USER0);
329 nilfs_mdt_mark_buffer_dirty(header_bh); 284
285 nilfs_sufile_mod_counter(header_bh, clean ? (u64)-1 : 0, dirty ? 0 : 1);
286 nilfs_mdt_mark_buffer_dirty(su_bh);
330 nilfs_mdt_mark_dirty(sufile); 287 nilfs_mdt_mark_dirty(sufile);
288}
331 289
332 out_bh: 290void nilfs_sufile_do_free(struct inode *sufile, __u64 segnum,
333 for (i--; i >= 0; i--) 291 struct buffer_head *header_bh,
334 brelse(su_bh[i]); 292 struct buffer_head *su_bh)
335 brelse(header_bh); 293{
294 struct nilfs_segment_usage *su;
295 void *kaddr;
296 int sudirty;
336 297
337 out_su_bh: 298 kaddr = kmap_atomic(su_bh->b_page, KM_USER0);
338 if (su_bh != su_bh_prealloc) 299 su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr);
339 kfree(su_bh); 300 if (nilfs_segment_usage_clean(su)) {
301 printk(KERN_WARNING "%s: segment %llu is already clean\n",
302 __func__, (unsigned long long)segnum);
303 kunmap_atomic(kaddr, KM_USER0);
304 return;
305 }
306 WARN_ON(nilfs_segment_usage_error(su));
307 WARN_ON(!nilfs_segment_usage_dirty(su));
340 308
341 out_sem: 309 sudirty = nilfs_segment_usage_dirty(su);
342 up_write(&NILFS_MDT(sufile)->mi_sem); 310 nilfs_segment_usage_set_clean(su);
343 return ret; 311 kunmap_atomic(kaddr, KM_USER0);
344} 312 nilfs_mdt_mark_buffer_dirty(su_bh);
345 313
346/** 314 nilfs_sufile_mod_counter(header_bh, 1, sudirty ? (u64)-1 : 0);
347 * nilfs_sufile_free - 315 nilfs_mdt_mark_dirty(sufile);
348 * @sufile:
349 * @segnum:
350 */
351int nilfs_sufile_free(struct inode *sufile, __u64 segnum)
352{
353 return nilfs_sufile_freev(sufile, &segnum, 1);
354} 316}
355 317
356/** 318/**
@@ -500,72 +462,28 @@ int nilfs_sufile_get_ncleansegs(struct inode *sufile, unsigned long *nsegsp)
500 return ret; 462 return ret;
501} 463}
502 464
503/** 465void nilfs_sufile_do_set_error(struct inode *sufile, __u64 segnum,
504 * nilfs_sufile_set_error - mark a segment as erroneous 466 struct buffer_head *header_bh,
505 * @sufile: inode of segment usage file 467 struct buffer_head *su_bh)
506 * @segnum: segment number
507 *
508 * Description: nilfs_sufile_set_error() marks the segment specified by
509 * @segnum as erroneous. The error segment will never be used again.
510 *
511 * Return Value: On success, 0 is returned. On error, one of the following
512 * negative error codes is returned.
513 *
514 * %-EIO - I/O error.
515 *
516 * %-ENOMEM - Insufficient amount of memory available.
517 *
518 * %-EINVAL - Invalid segment usage number.
519 */
520int nilfs_sufile_set_error(struct inode *sufile, __u64 segnum)
521{ 468{
522 struct buffer_head *header_bh, *su_bh;
523 struct nilfs_segment_usage *su; 469 struct nilfs_segment_usage *su;
524 struct nilfs_sufile_header *header;
525 void *kaddr; 470 void *kaddr;
526 int ret; 471 int suclean;
527
528 if (unlikely(segnum >= nilfs_sufile_get_nsegments(sufile))) {
529 printk(KERN_WARNING "%s: invalid segment number: %llu\n",
530 __func__, (unsigned long long)segnum);
531 return -EINVAL;
532 }
533 down_write(&NILFS_MDT(sufile)->mi_sem);
534
535 ret = nilfs_sufile_get_header_block(sufile, &header_bh);
536 if (ret < 0)
537 goto out_sem;
538 ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 0, &su_bh);
539 if (ret < 0)
540 goto out_header;
541 472
542 kaddr = kmap_atomic(su_bh->b_page, KM_USER0); 473 kaddr = kmap_atomic(su_bh->b_page, KM_USER0);
543 su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr); 474 su = nilfs_sufile_block_get_segment_usage(sufile, segnum, su_bh, kaddr);
544 if (nilfs_segment_usage_error(su)) { 475 if (nilfs_segment_usage_error(su)) {
545 kunmap_atomic(kaddr, KM_USER0); 476 kunmap_atomic(kaddr, KM_USER0);
546 brelse(su_bh); 477 return;
547 goto out_header;
548 } 478 }
549 479 suclean = nilfs_segment_usage_clean(su);
550 nilfs_segment_usage_set_error(su); 480 nilfs_segment_usage_set_error(su);
551 kunmap_atomic(kaddr, KM_USER0); 481 kunmap_atomic(kaddr, KM_USER0);
552 brelse(su_bh);
553 482
554 kaddr = kmap_atomic(header_bh->b_page, KM_USER0); 483 if (suclean)
555 header = nilfs_sufile_block_get_header(sufile, header_bh, kaddr); 484 nilfs_sufile_mod_counter(header_bh, -1, 0);
556 le64_add_cpu(&header->sh_ndirtysegs, -1);
557 kunmap_atomic(kaddr, KM_USER0);
558 nilfs_mdt_mark_buffer_dirty(header_bh);
559 nilfs_mdt_mark_buffer_dirty(su_bh); 485 nilfs_mdt_mark_buffer_dirty(su_bh);
560 nilfs_mdt_mark_dirty(sufile); 486 nilfs_mdt_mark_dirty(sufile);
561 brelse(su_bh);
562
563 out_header:
564 brelse(header_bh);
565
566 out_sem:
567 up_write(&NILFS_MDT(sufile)->mi_sem);
568 return ret;
569} 487}
570 488
571/** 489/**
@@ -625,7 +543,7 @@ ssize_t nilfs_sufile_get_suinfo(struct inode *sufile, __u64 segnum,
625 si[i + j].sui_nblocks = le32_to_cpu(su->su_nblocks); 543 si[i + j].sui_nblocks = le32_to_cpu(su->su_nblocks);
626 si[i + j].sui_flags = le32_to_cpu(su->su_flags) & 544 si[i + j].sui_flags = le32_to_cpu(su->su_flags) &
627 ~(1UL << NILFS_SEGMENT_USAGE_ACTIVE); 545 ~(1UL << NILFS_SEGMENT_USAGE_ACTIVE);
628 if (nilfs_segment_is_active(nilfs, segnum + i + j)) 546 if (nilfs_segment_is_active(nilfs, segnum + j))
629 si[i + j].sui_flags |= 547 si[i + j].sui_flags |=
630 (1UL << NILFS_SEGMENT_USAGE_ACTIVE); 548 (1UL << NILFS_SEGMENT_USAGE_ACTIVE);
631 } 549 }
diff --git a/fs/nilfs2/sufile.h b/fs/nilfs2/sufile.h
index d595f33a768d..a2e2efd4ade1 100644
--- a/fs/nilfs2/sufile.h
+++ b/fs/nilfs2/sufile.h
@@ -36,9 +36,6 @@ static inline unsigned long nilfs_sufile_get_nsegments(struct inode *sufile)
36} 36}
37 37
38int nilfs_sufile_alloc(struct inode *, __u64 *); 38int nilfs_sufile_alloc(struct inode *, __u64 *);
39int nilfs_sufile_cancel_free(struct inode *, __u64);
40int nilfs_sufile_freev(struct inode *, __u64 *, size_t);
41int nilfs_sufile_free(struct inode *, __u64);
42int nilfs_sufile_get_segment_usage(struct inode *, __u64, 39int nilfs_sufile_get_segment_usage(struct inode *, __u64,
43 struct nilfs_segment_usage **, 40 struct nilfs_segment_usage **,
44 struct buffer_head **); 41 struct buffer_head **);
@@ -46,9 +43,83 @@ void nilfs_sufile_put_segment_usage(struct inode *, __u64,
46 struct buffer_head *); 43 struct buffer_head *);
47int nilfs_sufile_get_stat(struct inode *, struct nilfs_sustat *); 44int nilfs_sufile_get_stat(struct inode *, struct nilfs_sustat *);
48int nilfs_sufile_get_ncleansegs(struct inode *, unsigned long *); 45int nilfs_sufile_get_ncleansegs(struct inode *, unsigned long *);
49int nilfs_sufile_set_error(struct inode *, __u64);
50ssize_t nilfs_sufile_get_suinfo(struct inode *, __u64, struct nilfs_suinfo *, 46ssize_t nilfs_sufile_get_suinfo(struct inode *, __u64, struct nilfs_suinfo *,
51 size_t); 47 size_t);
52 48
49int nilfs_sufile_update(struct inode *, __u64, int,
50 void (*dofunc)(struct inode *, __u64,
51 struct buffer_head *,
52 struct buffer_head *));
53void nilfs_sufile_do_cancel_free(struct inode *, __u64, struct buffer_head *,
54 struct buffer_head *);
55void nilfs_sufile_do_scrap(struct inode *, __u64, struct buffer_head *,
56 struct buffer_head *);
57void nilfs_sufile_do_free(struct inode *, __u64, struct buffer_head *,
58 struct buffer_head *);
59void nilfs_sufile_do_set_error(struct inode *, __u64, struct buffer_head *,
60 struct buffer_head *);
61
62/**
63 * nilfs_sufile_cancel_free -
64 * @sufile: inode of segment usage file
65 * @segnum: segment number
66 *
67 * Description:
68 *
69 * Return Value: On success, 0 is returned. On error, one of the following
70 * negative error codes is returned.
71 *
72 * %-EIO - I/O error.
73 *
74 * %-ENOMEM - Insufficient amount of memory available.
75 */
76static inline int nilfs_sufile_cancel_free(struct inode *sufile, __u64 segnum)
77{
78 return nilfs_sufile_update(sufile, segnum, 0,
79 nilfs_sufile_do_cancel_free);
80}
81
82/**
83 * nilfs_sufile_scrap - make a segment garbage
84 * @sufile: inode of segment usage file
85 * @segnum: segment number to be freed
86 */
87static inline int nilfs_sufile_scrap(struct inode *sufile, __u64 segnum)
88{
89 return nilfs_sufile_update(sufile, segnum, 1, nilfs_sufile_do_scrap);
90}
91
92/**
93 * nilfs_sufile_free - free segment
94 * @sufile: inode of segment usage file
95 * @segnum: segment number to be freed
96 */
97static inline int nilfs_sufile_free(struct inode *sufile, __u64 segnum)
98{
99 return nilfs_sufile_update(sufile, segnum, 0, nilfs_sufile_do_free);
100}
101
102/**
103 * nilfs_sufile_set_error - mark a segment as erroneous
104 * @sufile: inode of segment usage file
105 * @segnum: segment number
106 *
107 * Description: nilfs_sufile_set_error() marks the segment specified by
108 * @segnum as erroneous. The error segment will never be used again.
109 *
110 * Return Value: On success, 0 is returned. On error, one of the following
111 * negative error codes is returned.
112 *
113 * %-EIO - I/O error.
114 *
115 * %-ENOMEM - Insufficient amount of memory available.
116 *
117 * %-EINVAL - Invalid segment usage number.
118 */
119static inline int nilfs_sufile_set_error(struct inode *sufile, __u64 segnum)
120{
121 return nilfs_sufile_update(sufile, segnum, 0,
122 nilfs_sufile_do_set_error);
123}
53 124
54#endif /* _NILFS_SUFILE_H */ 125#endif /* _NILFS_SUFILE_H */
diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c
index e117e1ea9bff..6989b03e97ab 100644
--- a/fs/nilfs2/super.c
+++ b/fs/nilfs2/super.c
@@ -63,7 +63,6 @@
63MODULE_AUTHOR("NTT Corp."); 63MODULE_AUTHOR("NTT Corp.");
64MODULE_DESCRIPTION("A New Implementation of the Log-structured Filesystem " 64MODULE_DESCRIPTION("A New Implementation of the Log-structured Filesystem "
65 "(NILFS)"); 65 "(NILFS)");
66MODULE_VERSION(NILFS_VERSION);
67MODULE_LICENSE("GPL"); 66MODULE_LICENSE("GPL");
68 67
69static int nilfs_remount(struct super_block *sb, int *flags, char *data); 68static int nilfs_remount(struct super_block *sb, int *flags, char *data);
@@ -476,11 +475,12 @@ static int nilfs_statfs(struct dentry *dentry, struct kstatfs *buf)
476{ 475{
477 struct super_block *sb = dentry->d_sb; 476 struct super_block *sb = dentry->d_sb;
478 struct nilfs_sb_info *sbi = NILFS_SB(sb); 477 struct nilfs_sb_info *sbi = NILFS_SB(sb);
478 struct the_nilfs *nilfs = sbi->s_nilfs;
479 u64 id = huge_encode_dev(sb->s_bdev->bd_dev);
479 unsigned long long blocks; 480 unsigned long long blocks;
480 unsigned long overhead; 481 unsigned long overhead;
481 unsigned long nrsvblocks; 482 unsigned long nrsvblocks;
482 sector_t nfreeblocks; 483 sector_t nfreeblocks;
483 struct the_nilfs *nilfs = sbi->s_nilfs;
484 int err; 484 int err;
485 485
486 /* 486 /*
@@ -514,6 +514,9 @@ static int nilfs_statfs(struct dentry *dentry, struct kstatfs *buf)
514 buf->f_files = atomic_read(&sbi->s_inodes_count); 514 buf->f_files = atomic_read(&sbi->s_inodes_count);
515 buf->f_ffree = 0; /* nilfs_count_free_inodes(sb); */ 515 buf->f_ffree = 0; /* nilfs_count_free_inodes(sb); */
516 buf->f_namelen = NILFS_NAME_LEN; 516 buf->f_namelen = NILFS_NAME_LEN;
517 buf->f_fsid.val[0] = (u32)id;
518 buf->f_fsid.val[1] = (u32)(id >> 32);
519
517 return 0; 520 return 0;
518} 521}
519 522
diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c
index 33400cf0bbe2..7f65b3be4aa9 100644
--- a/fs/nilfs2/the_nilfs.c
+++ b/fs/nilfs2/the_nilfs.c
@@ -115,6 +115,7 @@ void put_nilfs(struct the_nilfs *nilfs)
115static int nilfs_load_super_root(struct the_nilfs *nilfs, 115static int nilfs_load_super_root(struct the_nilfs *nilfs,
116 struct nilfs_sb_info *sbi, sector_t sr_block) 116 struct nilfs_sb_info *sbi, sector_t sr_block)
117{ 117{
118 static struct lock_class_key dat_lock_key;
118 struct buffer_head *bh_sr; 119 struct buffer_head *bh_sr;
119 struct nilfs_super_root *raw_sr; 120 struct nilfs_super_root *raw_sr;
120 struct nilfs_super_block **sbp = nilfs->ns_sbp; 121 struct nilfs_super_block **sbp = nilfs->ns_sbp;
@@ -163,6 +164,9 @@ static int nilfs_load_super_root(struct the_nilfs *nilfs,
163 if (unlikely(err)) 164 if (unlikely(err))
164 goto failed_sufile; 165 goto failed_sufile;
165 166
167 lockdep_set_class(&NILFS_MDT(nilfs->ns_dat)->mi_sem, &dat_lock_key);
168 lockdep_set_class(&NILFS_MDT(nilfs->ns_gc_dat)->mi_sem, &dat_lock_key);
169
166 nilfs_mdt_set_shadow(nilfs->ns_dat, nilfs->ns_gc_dat); 170 nilfs_mdt_set_shadow(nilfs->ns_dat, nilfs->ns_gc_dat);
167 nilfs_mdt_set_entry_size(nilfs->ns_cpfile, checkpoint_size, 171 nilfs_mdt_set_entry_size(nilfs->ns_cpfile, checkpoint_size,
168 sizeof(struct nilfs_cpfile_header)); 172 sizeof(struct nilfs_cpfile_header));
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 8672b9536039..c2a87c885b73 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -1912,6 +1912,22 @@ out_sems:
1912 return written ? written : ret; 1912 return written ? written : ret;
1913} 1913}
1914 1914
1915static int ocfs2_splice_to_file(struct pipe_inode_info *pipe,
1916 struct file *out,
1917 struct splice_desc *sd)
1918{
1919 int ret;
1920
1921 ret = ocfs2_prepare_inode_for_write(out->f_path.dentry, &sd->pos,
1922 sd->total_len, 0, NULL);
1923 if (ret < 0) {
1924 mlog_errno(ret);
1925 return ret;
1926 }
1927
1928 return splice_from_pipe_feed(pipe, sd, pipe_to_file);
1929}
1930
1915static ssize_t ocfs2_file_splice_write(struct pipe_inode_info *pipe, 1931static ssize_t ocfs2_file_splice_write(struct pipe_inode_info *pipe,
1916 struct file *out, 1932 struct file *out,
1917 loff_t *ppos, 1933 loff_t *ppos,
@@ -1919,38 +1935,76 @@ static ssize_t ocfs2_file_splice_write(struct pipe_inode_info *pipe,
1919 unsigned int flags) 1935 unsigned int flags)
1920{ 1936{
1921 int ret; 1937 int ret;
1922 struct inode *inode = out->f_path.dentry->d_inode; 1938 struct address_space *mapping = out->f_mapping;
1939 struct inode *inode = mapping->host;
1940 struct splice_desc sd = {
1941 .total_len = len,
1942 .flags = flags,
1943 .pos = *ppos,
1944 .u.file = out,
1945 };
1923 1946
1924 mlog_entry("(0x%p, 0x%p, %u, '%.*s')\n", out, pipe, 1947 mlog_entry("(0x%p, 0x%p, %u, '%.*s')\n", out, pipe,
1925 (unsigned int)len, 1948 (unsigned int)len,
1926 out->f_path.dentry->d_name.len, 1949 out->f_path.dentry->d_name.len,
1927 out->f_path.dentry->d_name.name); 1950 out->f_path.dentry->d_name.name);
1928 1951
1929 mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT); 1952 if (pipe->inode)
1953 mutex_lock_nested(&pipe->inode->i_mutex, I_MUTEX_PARENT);
1930 1954
1931 ret = ocfs2_rw_lock(inode, 1); 1955 splice_from_pipe_begin(&sd);
1932 if (ret < 0) { 1956 do {
1933 mlog_errno(ret); 1957 ret = splice_from_pipe_next(pipe, &sd);
1934 goto out; 1958 if (ret <= 0)
1935 } 1959 break;
1936 1960
1937 ret = ocfs2_prepare_inode_for_write(out->f_path.dentry, ppos, len, 0, 1961 mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
1938 NULL); 1962 ret = ocfs2_rw_lock(inode, 1);
1939 if (ret < 0) { 1963 if (ret < 0)
1940 mlog_errno(ret); 1964 mlog_errno(ret);
1941 goto out_unlock; 1965 else {
1942 } 1966 ret = ocfs2_splice_to_file(pipe, out, &sd);
1967 ocfs2_rw_unlock(inode, 1);
1968 }
1969 mutex_unlock(&inode->i_mutex);
1970 } while (ret > 0);
1971 splice_from_pipe_end(pipe, &sd);
1943 1972
1944 if (pipe->inode) 1973 if (pipe->inode)
1945 mutex_lock_nested(&pipe->inode->i_mutex, I_MUTEX_CHILD);
1946 ret = generic_file_splice_write_nolock(pipe, out, ppos, len, flags);
1947 if (pipe->inode)
1948 mutex_unlock(&pipe->inode->i_mutex); 1974 mutex_unlock(&pipe->inode->i_mutex);
1949 1975
1950out_unlock: 1976 if (sd.num_spliced)
1951 ocfs2_rw_unlock(inode, 1); 1977 ret = sd.num_spliced;
1952out: 1978
1953 mutex_unlock(&inode->i_mutex); 1979 if (ret > 0) {
1980 unsigned long nr_pages;
1981
1982 *ppos += ret;
1983 nr_pages = (ret + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
1984
1985 /*
1986 * If file or inode is SYNC and we actually wrote some data,
1987 * sync it.
1988 */
1989 if (unlikely((out->f_flags & O_SYNC) || IS_SYNC(inode))) {
1990 int err;
1991
1992 mutex_lock(&inode->i_mutex);
1993 err = ocfs2_rw_lock(inode, 1);
1994 if (err < 0) {
1995 mlog_errno(err);
1996 } else {
1997 err = generic_osync_inode(inode, mapping,
1998 OSYNC_METADATA|OSYNC_DATA);
1999 ocfs2_rw_unlock(inode, 1);
2000 }
2001 mutex_unlock(&inode->i_mutex);
2002
2003 if (err)
2004 ret = err;
2005 }
2006 balance_dirty_pages_ratelimited_nr(mapping, nr_pages);
2007 }
1954 2008
1955 mlog_exit(ret); 2009 mlog_exit(ret);
1956 return ret; 2010 return ret;
diff --git a/fs/pipe.c b/fs/pipe.c
index 4af7aa521813..13414ec45b8d 100644
--- a/fs/pipe.c
+++ b/fs/pipe.c
@@ -37,6 +37,42 @@
37 * -- Manfred Spraul <manfred@colorfullife.com> 2002-05-09 37 * -- Manfred Spraul <manfred@colorfullife.com> 2002-05-09
38 */ 38 */
39 39
40static void pipe_lock_nested(struct pipe_inode_info *pipe, int subclass)
41{
42 if (pipe->inode)
43 mutex_lock_nested(&pipe->inode->i_mutex, subclass);
44}
45
46void pipe_lock(struct pipe_inode_info *pipe)
47{
48 /*
49 * pipe_lock() nests non-pipe inode locks (for writing to a file)
50 */
51 pipe_lock_nested(pipe, I_MUTEX_PARENT);
52}
53EXPORT_SYMBOL(pipe_lock);
54
55void pipe_unlock(struct pipe_inode_info *pipe)
56{
57 if (pipe->inode)
58 mutex_unlock(&pipe->inode->i_mutex);
59}
60EXPORT_SYMBOL(pipe_unlock);
61
62void pipe_double_lock(struct pipe_inode_info *pipe1,
63 struct pipe_inode_info *pipe2)
64{
65 BUG_ON(pipe1 == pipe2);
66
67 if (pipe1 < pipe2) {
68 pipe_lock_nested(pipe1, I_MUTEX_PARENT);
69 pipe_lock_nested(pipe2, I_MUTEX_CHILD);
70 } else {
71 pipe_lock_nested(pipe2, I_MUTEX_CHILD);
72 pipe_lock_nested(pipe1, I_MUTEX_PARENT);
73 }
74}
75
40/* Drop the inode semaphore and wait for a pipe event, atomically */ 76/* Drop the inode semaphore and wait for a pipe event, atomically */
41void pipe_wait(struct pipe_inode_info *pipe) 77void pipe_wait(struct pipe_inode_info *pipe)
42{ 78{
@@ -47,12 +83,10 @@ void pipe_wait(struct pipe_inode_info *pipe)
47 * is considered a noninteractive wait: 83 * is considered a noninteractive wait:
48 */ 84 */
49 prepare_to_wait(&pipe->wait, &wait, TASK_INTERRUPTIBLE); 85 prepare_to_wait(&pipe->wait, &wait, TASK_INTERRUPTIBLE);
50 if (pipe->inode) 86 pipe_unlock(pipe);
51 mutex_unlock(&pipe->inode->i_mutex);
52 schedule(); 87 schedule();
53 finish_wait(&pipe->wait, &wait); 88 finish_wait(&pipe->wait, &wait);
54 if (pipe->inode) 89 pipe_lock(pipe);
55 mutex_lock(&pipe->inode->i_mutex);
56} 90}
57 91
58static int 92static int
diff --git a/fs/splice.c b/fs/splice.c
index c18aa7e03e2b..5384a90665d0 100644
--- a/fs/splice.c
+++ b/fs/splice.c
@@ -182,8 +182,7 @@ ssize_t splice_to_pipe(struct pipe_inode_info *pipe,
182 do_wakeup = 0; 182 do_wakeup = 0;
183 page_nr = 0; 183 page_nr = 0;
184 184
185 if (pipe->inode) 185 pipe_lock(pipe);
186 mutex_lock(&pipe->inode->i_mutex);
187 186
188 for (;;) { 187 for (;;) {
189 if (!pipe->readers) { 188 if (!pipe->readers) {
@@ -245,15 +244,13 @@ ssize_t splice_to_pipe(struct pipe_inode_info *pipe,
245 pipe->waiting_writers--; 244 pipe->waiting_writers--;
246 } 245 }
247 246
248 if (pipe->inode) { 247 pipe_unlock(pipe);
249 mutex_unlock(&pipe->inode->i_mutex);
250 248
251 if (do_wakeup) { 249 if (do_wakeup) {
252 smp_mb(); 250 smp_mb();
253 if (waitqueue_active(&pipe->wait)) 251 if (waitqueue_active(&pipe->wait))
254 wake_up_interruptible(&pipe->wait); 252 wake_up_interruptible(&pipe->wait);
255 kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); 253 kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN);
256 }
257 } 254 }
258 255
259 while (page_nr < spd_pages) 256 while (page_nr < spd_pages)
@@ -555,8 +552,8 @@ static int pipe_to_sendpage(struct pipe_inode_info *pipe,
555 * SPLICE_F_MOVE isn't set, or we cannot move the page, we simply create 552 * SPLICE_F_MOVE isn't set, or we cannot move the page, we simply create
556 * a new page in the output file page cache and fill/dirty that. 553 * a new page in the output file page cache and fill/dirty that.
557 */ 554 */
558static int pipe_to_file(struct pipe_inode_info *pipe, struct pipe_buffer *buf, 555int pipe_to_file(struct pipe_inode_info *pipe, struct pipe_buffer *buf,
559 struct splice_desc *sd) 556 struct splice_desc *sd)
560{ 557{
561 struct file *file = sd->u.file; 558 struct file *file = sd->u.file;
562 struct address_space *mapping = file->f_mapping; 559 struct address_space *mapping = file->f_mapping;
@@ -600,108 +597,178 @@ static int pipe_to_file(struct pipe_inode_info *pipe, struct pipe_buffer *buf,
600out: 597out:
601 return ret; 598 return ret;
602} 599}
600EXPORT_SYMBOL(pipe_to_file);
601
602static void wakeup_pipe_writers(struct pipe_inode_info *pipe)
603{
604 smp_mb();
605 if (waitqueue_active(&pipe->wait))
606 wake_up_interruptible(&pipe->wait);
607 kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT);
608}
603 609
604/** 610/**
605 * __splice_from_pipe - splice data from a pipe to given actor 611 * splice_from_pipe_feed - feed available data from a pipe to a file
606 * @pipe: pipe to splice from 612 * @pipe: pipe to splice from
607 * @sd: information to @actor 613 * @sd: information to @actor
608 * @actor: handler that splices the data 614 * @actor: handler that splices the data
609 * 615 *
610 * Description: 616 * Description:
611 * This function does little more than loop over the pipe and call 617
612 * @actor to do the actual moving of a single struct pipe_buffer to 618 * This function loops over the pipe and calls @actor to do the
613 * the desired destination. See pipe_to_file, pipe_to_sendpage, or 619 * actual moving of a single struct pipe_buffer to the desired
614 * pipe_to_user. 620 * destination. It returns when there's no more buffers left in
621 * the pipe or if the requested number of bytes (@sd->total_len)
622 * have been copied. It returns a positive number (one) if the
623 * pipe needs to be filled with more data, zero if the required
624 * number of bytes have been copied and -errno on error.
615 * 625 *
626 * This, together with splice_from_pipe_{begin,end,next}, may be
627 * used to implement the functionality of __splice_from_pipe() when
628 * locking is required around copying the pipe buffers to the
629 * destination.
616 */ 630 */
617ssize_t __splice_from_pipe(struct pipe_inode_info *pipe, struct splice_desc *sd, 631int splice_from_pipe_feed(struct pipe_inode_info *pipe, struct splice_desc *sd,
618 splice_actor *actor) 632 splice_actor *actor)
619{ 633{
620 int ret, do_wakeup, err; 634 int ret;
621
622 ret = 0;
623 do_wakeup = 0;
624
625 for (;;) {
626 if (pipe->nrbufs) {
627 struct pipe_buffer *buf = pipe->bufs + pipe->curbuf;
628 const struct pipe_buf_operations *ops = buf->ops;
629 635
630 sd->len = buf->len; 636 while (pipe->nrbufs) {
631 if (sd->len > sd->total_len) 637 struct pipe_buffer *buf = pipe->bufs + pipe->curbuf;
632 sd->len = sd->total_len; 638 const struct pipe_buf_operations *ops = buf->ops;
633 639
634 err = actor(pipe, buf, sd); 640 sd->len = buf->len;
635 if (err <= 0) { 641 if (sd->len > sd->total_len)
636 if (!ret && err != -ENODATA) 642 sd->len = sd->total_len;
637 ret = err;
638 643
639 break; 644 ret = actor(pipe, buf, sd);
640 } 645 if (ret <= 0) {
646 if (ret == -ENODATA)
647 ret = 0;
648 return ret;
649 }
650 buf->offset += ret;
651 buf->len -= ret;
641 652
642 ret += err; 653 sd->num_spliced += ret;
643 buf->offset += err; 654 sd->len -= ret;
644 buf->len -= err; 655 sd->pos += ret;
656 sd->total_len -= ret;
645 657
646 sd->len -= err; 658 if (!buf->len) {
647 sd->pos += err; 659 buf->ops = NULL;
648 sd->total_len -= err; 660 ops->release(pipe, buf);
649 if (sd->len) 661 pipe->curbuf = (pipe->curbuf + 1) & (PIPE_BUFFERS - 1);
650 continue; 662 pipe->nrbufs--;
663 if (pipe->inode)
664 sd->need_wakeup = true;
665 }
651 666
652 if (!buf->len) { 667 if (!sd->total_len)
653 buf->ops = NULL; 668 return 0;
654 ops->release(pipe, buf); 669 }
655 pipe->curbuf = (pipe->curbuf + 1) & (PIPE_BUFFERS - 1);
656 pipe->nrbufs--;
657 if (pipe->inode)
658 do_wakeup = 1;
659 }
660 670
661 if (!sd->total_len) 671 return 1;
662 break; 672}
663 } 673EXPORT_SYMBOL(splice_from_pipe_feed);
664 674
665 if (pipe->nrbufs) 675/**
666 continue; 676 * splice_from_pipe_next - wait for some data to splice from
677 * @pipe: pipe to splice from
678 * @sd: information about the splice operation
679 *
680 * Description:
681 * This function will wait for some data and return a positive
682 * value (one) if pipe buffers are available. It will return zero
683 * or -errno if no more data needs to be spliced.
684 */
685int splice_from_pipe_next(struct pipe_inode_info *pipe, struct splice_desc *sd)
686{
687 while (!pipe->nrbufs) {
667 if (!pipe->writers) 688 if (!pipe->writers)
668 break; 689 return 0;
669 if (!pipe->waiting_writers) {
670 if (ret)
671 break;
672 }
673 690
674 if (sd->flags & SPLICE_F_NONBLOCK) { 691 if (!pipe->waiting_writers && sd->num_spliced)
675 if (!ret) 692 return 0;
676 ret = -EAGAIN;
677 break;
678 }
679 693
680 if (signal_pending(current)) { 694 if (sd->flags & SPLICE_F_NONBLOCK)
681 if (!ret) 695 return -EAGAIN;
682 ret = -ERESTARTSYS;
683 break;
684 }
685 696
686 if (do_wakeup) { 697 if (signal_pending(current))
687 smp_mb(); 698 return -ERESTARTSYS;
688 if (waitqueue_active(&pipe->wait)) 699
689 wake_up_interruptible_sync(&pipe->wait); 700 if (sd->need_wakeup) {
690 kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); 701 wakeup_pipe_writers(pipe);
691 do_wakeup = 0; 702 sd->need_wakeup = false;
692 } 703 }
693 704
694 pipe_wait(pipe); 705 pipe_wait(pipe);
695 } 706 }
696 707
697 if (do_wakeup) { 708 return 1;
698 smp_mb(); 709}
699 if (waitqueue_active(&pipe->wait)) 710EXPORT_SYMBOL(splice_from_pipe_next);
700 wake_up_interruptible(&pipe->wait);
701 kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT);
702 }
703 711
704 return ret; 712/**
713 * splice_from_pipe_begin - start splicing from pipe
714 * @pipe: pipe to splice from
715 *
716 * Description:
717 * This function should be called before a loop containing
718 * splice_from_pipe_next() and splice_from_pipe_feed() to
719 * initialize the necessary fields of @sd.
720 */
721void splice_from_pipe_begin(struct splice_desc *sd)
722{
723 sd->num_spliced = 0;
724 sd->need_wakeup = false;
725}
726EXPORT_SYMBOL(splice_from_pipe_begin);
727
728/**
729 * splice_from_pipe_end - finish splicing from pipe
730 * @pipe: pipe to splice from
731 * @sd: information about the splice operation
732 *
733 * Description:
734 * This function will wake up pipe writers if necessary. It should
735 * be called after a loop containing splice_from_pipe_next() and
736 * splice_from_pipe_feed().
737 */
738void splice_from_pipe_end(struct pipe_inode_info *pipe, struct splice_desc *sd)
739{
740 if (sd->need_wakeup)
741 wakeup_pipe_writers(pipe);
742}
743EXPORT_SYMBOL(splice_from_pipe_end);
744
745/**
746 * __splice_from_pipe - splice data from a pipe to given actor
747 * @pipe: pipe to splice from
748 * @sd: information to @actor
749 * @actor: handler that splices the data
750 *
751 * Description:
752 * This function does little more than loop over the pipe and call
753 * @actor to do the actual moving of a single struct pipe_buffer to
754 * the desired destination. See pipe_to_file, pipe_to_sendpage, or
755 * pipe_to_user.
756 *
757 */
758ssize_t __splice_from_pipe(struct pipe_inode_info *pipe, struct splice_desc *sd,
759 splice_actor *actor)
760{
761 int ret;
762
763 splice_from_pipe_begin(sd);
764 do {
765 ret = splice_from_pipe_next(pipe, sd);
766 if (ret > 0)
767 ret = splice_from_pipe_feed(pipe, sd, actor);
768 } while (ret > 0);
769 splice_from_pipe_end(pipe, sd);
770
771 return sd->num_spliced ? sd->num_spliced : ret;
705} 772}
706EXPORT_SYMBOL(__splice_from_pipe); 773EXPORT_SYMBOL(__splice_from_pipe);
707 774
@@ -715,7 +782,7 @@ EXPORT_SYMBOL(__splice_from_pipe);
715 * @actor: handler that splices the data 782 * @actor: handler that splices the data
716 * 783 *
717 * Description: 784 * Description:
718 * See __splice_from_pipe. This function locks the input and output inodes, 785 * See __splice_from_pipe. This function locks the pipe inode,
719 * otherwise it's identical to __splice_from_pipe(). 786 * otherwise it's identical to __splice_from_pipe().
720 * 787 *
721 */ 788 */
@@ -724,7 +791,6 @@ ssize_t splice_from_pipe(struct pipe_inode_info *pipe, struct file *out,
724 splice_actor *actor) 791 splice_actor *actor)
725{ 792{
726 ssize_t ret; 793 ssize_t ret;
727 struct inode *inode = out->f_mapping->host;
728 struct splice_desc sd = { 794 struct splice_desc sd = {
729 .total_len = len, 795 .total_len = len,
730 .flags = flags, 796 .flags = flags,
@@ -732,30 +798,15 @@ ssize_t splice_from_pipe(struct pipe_inode_info *pipe, struct file *out,
732 .u.file = out, 798 .u.file = out,
733 }; 799 };
734 800
735 /* 801 pipe_lock(pipe);
736 * The actor worker might be calling ->write_begin and
737 * ->write_end. Most of the time, these expect i_mutex to
738 * be held. Since this may result in an ABBA deadlock with
739 * pipe->inode, we have to order lock acquiry here.
740 *
741 * Outer lock must be inode->i_mutex, as pipe_wait() will
742 * release and reacquire pipe->inode->i_mutex, AND inode must
743 * never be a pipe.
744 */
745 WARN_ON(S_ISFIFO(inode->i_mode));
746 mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
747 if (pipe->inode)
748 mutex_lock_nested(&pipe->inode->i_mutex, I_MUTEX_CHILD);
749 ret = __splice_from_pipe(pipe, &sd, actor); 802 ret = __splice_from_pipe(pipe, &sd, actor);
750 if (pipe->inode) 803 pipe_unlock(pipe);
751 mutex_unlock(&pipe->inode->i_mutex);
752 mutex_unlock(&inode->i_mutex);
753 804
754 return ret; 805 return ret;
755} 806}
756 807
757/** 808/**
758 * generic_file_splice_write_nolock - generic_file_splice_write without mutexes 809 * generic_file_splice_write - splice data from a pipe to a file
759 * @pipe: pipe info 810 * @pipe: pipe info
760 * @out: file to write to 811 * @out: file to write to
761 * @ppos: position in @out 812 * @ppos: position in @out
@@ -764,13 +815,12 @@ ssize_t splice_from_pipe(struct pipe_inode_info *pipe, struct file *out,
764 * 815 *
765 * Description: 816 * Description:
766 * Will either move or copy pages (determined by @flags options) from 817 * Will either move or copy pages (determined by @flags options) from
767 * the given pipe inode to the given file. The caller is responsible 818 * the given pipe inode to the given file.
768 * for acquiring i_mutex on both inodes.
769 * 819 *
770 */ 820 */
771ssize_t 821ssize_t
772generic_file_splice_write_nolock(struct pipe_inode_info *pipe, struct file *out, 822generic_file_splice_write(struct pipe_inode_info *pipe, struct file *out,
773 loff_t *ppos, size_t len, unsigned int flags) 823 loff_t *ppos, size_t len, unsigned int flags)
774{ 824{
775 struct address_space *mapping = out->f_mapping; 825 struct address_space *mapping = out->f_mapping;
776 struct inode *inode = mapping->host; 826 struct inode *inode = mapping->host;
@@ -781,76 +831,28 @@ generic_file_splice_write_nolock(struct pipe_inode_info *pipe, struct file *out,
781 .u.file = out, 831 .u.file = out,
782 }; 832 };
783 ssize_t ret; 833 ssize_t ret;
784 int err;
785
786 err = file_remove_suid(out);
787 if (unlikely(err))
788 return err;
789
790 ret = __splice_from_pipe(pipe, &sd, pipe_to_file);
791 if (ret > 0) {
792 unsigned long nr_pages;
793 834
794 *ppos += ret; 835 pipe_lock(pipe);
795 nr_pages = (ret + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT;
796 836
797 /* 837 splice_from_pipe_begin(&sd);
798 * If file or inode is SYNC and we actually wrote some data, 838 do {
799 * sync it. 839 ret = splice_from_pipe_next(pipe, &sd);
800 */ 840 if (ret <= 0)
801 if (unlikely((out->f_flags & O_SYNC) || IS_SYNC(inode))) { 841 break;
802 err = generic_osync_inode(inode, mapping,
803 OSYNC_METADATA|OSYNC_DATA);
804
805 if (err)
806 ret = err;
807 }
808 balance_dirty_pages_ratelimited_nr(mapping, nr_pages);
809 }
810 842
811 return ret; 843 mutex_lock_nested(&inode->i_mutex, I_MUTEX_CHILD);
812} 844 ret = file_remove_suid(out);
845 if (!ret)
846 ret = splice_from_pipe_feed(pipe, &sd, pipe_to_file);
847 mutex_unlock(&inode->i_mutex);
848 } while (ret > 0);
849 splice_from_pipe_end(pipe, &sd);
813 850
814EXPORT_SYMBOL(generic_file_splice_write_nolock); 851 pipe_unlock(pipe);
815 852
816/** 853 if (sd.num_spliced)
817 * generic_file_splice_write - splice data from a pipe to a file 854 ret = sd.num_spliced;
818 * @pipe: pipe info
819 * @out: file to write to
820 * @ppos: position in @out
821 * @len: number of bytes to splice
822 * @flags: splice modifier flags
823 *
824 * Description:
825 * Will either move or copy pages (determined by @flags options) from
826 * the given pipe inode to the given file.
827 *
828 */
829ssize_t
830generic_file_splice_write(struct pipe_inode_info *pipe, struct file *out,
831 loff_t *ppos, size_t len, unsigned int flags)
832{
833 struct address_space *mapping = out->f_mapping;
834 struct inode *inode = mapping->host;
835 struct splice_desc sd = {
836 .total_len = len,
837 .flags = flags,
838 .pos = *ppos,
839 .u.file = out,
840 };
841 ssize_t ret;
842 855
843 WARN_ON(S_ISFIFO(inode->i_mode));
844 mutex_lock_nested(&inode->i_mutex, I_MUTEX_PARENT);
845 ret = file_remove_suid(out);
846 if (likely(!ret)) {
847 if (pipe->inode)
848 mutex_lock_nested(&pipe->inode->i_mutex, I_MUTEX_CHILD);
849 ret = __splice_from_pipe(pipe, &sd, pipe_to_file);
850 if (pipe->inode)
851 mutex_unlock(&pipe->inode->i_mutex);
852 }
853 mutex_unlock(&inode->i_mutex);
854 if (ret > 0) { 856 if (ret > 0) {
855 unsigned long nr_pages; 857 unsigned long nr_pages;
856 858
@@ -1339,8 +1341,7 @@ static long vmsplice_to_user(struct file *file, const struct iovec __user *iov,
1339 if (!pipe) 1341 if (!pipe)
1340 return -EBADF; 1342 return -EBADF;
1341 1343
1342 if (pipe->inode) 1344 pipe_lock(pipe);
1343 mutex_lock(&pipe->inode->i_mutex);
1344 1345
1345 error = ret = 0; 1346 error = ret = 0;
1346 while (nr_segs) { 1347 while (nr_segs) {
@@ -1395,8 +1396,7 @@ static long vmsplice_to_user(struct file *file, const struct iovec __user *iov,
1395 iov++; 1396 iov++;
1396 } 1397 }
1397 1398
1398 if (pipe->inode) 1399 pipe_unlock(pipe);
1399 mutex_unlock(&pipe->inode->i_mutex);
1400 1400
1401 if (!ret) 1401 if (!ret)
1402 ret = error; 1402 ret = error;
@@ -1524,7 +1524,7 @@ static int link_ipipe_prep(struct pipe_inode_info *pipe, unsigned int flags)
1524 return 0; 1524 return 0;
1525 1525
1526 ret = 0; 1526 ret = 0;
1527 mutex_lock(&pipe->inode->i_mutex); 1527 pipe_lock(pipe);
1528 1528
1529 while (!pipe->nrbufs) { 1529 while (!pipe->nrbufs) {
1530 if (signal_pending(current)) { 1530 if (signal_pending(current)) {
@@ -1542,7 +1542,7 @@ static int link_ipipe_prep(struct pipe_inode_info *pipe, unsigned int flags)
1542 pipe_wait(pipe); 1542 pipe_wait(pipe);
1543 } 1543 }
1544 1544
1545 mutex_unlock(&pipe->inode->i_mutex); 1545 pipe_unlock(pipe);
1546 return ret; 1546 return ret;
1547} 1547}
1548 1548
@@ -1562,7 +1562,7 @@ static int link_opipe_prep(struct pipe_inode_info *pipe, unsigned int flags)
1562 return 0; 1562 return 0;
1563 1563
1564 ret = 0; 1564 ret = 0;
1565 mutex_lock(&pipe->inode->i_mutex); 1565 pipe_lock(pipe);
1566 1566
1567 while (pipe->nrbufs >= PIPE_BUFFERS) { 1567 while (pipe->nrbufs >= PIPE_BUFFERS) {
1568 if (!pipe->readers) { 1568 if (!pipe->readers) {
@@ -1583,7 +1583,7 @@ static int link_opipe_prep(struct pipe_inode_info *pipe, unsigned int flags)
1583 pipe->waiting_writers--; 1583 pipe->waiting_writers--;
1584 } 1584 }
1585 1585
1586 mutex_unlock(&pipe->inode->i_mutex); 1586 pipe_unlock(pipe);
1587 return ret; 1587 return ret;
1588} 1588}
1589 1589
@@ -1599,10 +1599,10 @@ static int link_pipe(struct pipe_inode_info *ipipe,
1599 1599
1600 /* 1600 /*
1601 * Potential ABBA deadlock, work around it by ordering lock 1601 * Potential ABBA deadlock, work around it by ordering lock
1602 * grabbing by inode address. Otherwise two different processes 1602 * grabbing by pipe info address. Otherwise two different processes
1603 * could deadlock (one doing tee from A -> B, the other from B -> A). 1603 * could deadlock (one doing tee from A -> B, the other from B -> A).
1604 */ 1604 */
1605 inode_double_lock(ipipe->inode, opipe->inode); 1605 pipe_double_lock(ipipe, opipe);
1606 1606
1607 do { 1607 do {
1608 if (!opipe->readers) { 1608 if (!opipe->readers) {
@@ -1653,7 +1653,8 @@ static int link_pipe(struct pipe_inode_info *ipipe,
1653 if (!ret && ipipe->waiting_writers && (flags & SPLICE_F_NONBLOCK)) 1653 if (!ret && ipipe->waiting_writers && (flags & SPLICE_F_NONBLOCK))
1654 ret = -EAGAIN; 1654 ret = -EAGAIN;
1655 1655
1656 inode_double_unlock(ipipe->inode, opipe->inode); 1656 pipe_unlock(ipipe);
1657 pipe_unlock(opipe);
1657 1658
1658 /* 1659 /*
1659 * If we put data in the output pipe, wakeup any potential readers. 1660 * If we put data in the output pipe, wakeup any potential readers.
diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c
index c13f67300fe7..7ec89fc05b2b 100644
--- a/fs/xfs/linux-2.6/xfs_aops.c
+++ b/fs/xfs/linux-2.6/xfs_aops.c
@@ -153,23 +153,6 @@ xfs_find_bdev_for_inode(
153} 153}
154 154
155/* 155/*
156 * Schedule IO completion handling on a xfsdatad if this was
157 * the final hold on this ioend. If we are asked to wait,
158 * flush the workqueue.
159 */
160STATIC void
161xfs_finish_ioend(
162 xfs_ioend_t *ioend,
163 int wait)
164{
165 if (atomic_dec_and_test(&ioend->io_remaining)) {
166 queue_work(xfsdatad_workqueue, &ioend->io_work);
167 if (wait)
168 flush_workqueue(xfsdatad_workqueue);
169 }
170}
171
172/*
173 * We're now finished for good with this ioend structure. 156 * We're now finished for good with this ioend structure.
174 * Update the page state via the associated buffer_heads, 157 * Update the page state via the associated buffer_heads,
175 * release holds on the inode and bio, and finally free 158 * release holds on the inode and bio, and finally free
@@ -310,6 +293,27 @@ xfs_end_bio_read(
310} 293}
311 294
312/* 295/*
296 * Schedule IO completion handling on a xfsdatad if this was
297 * the final hold on this ioend. If we are asked to wait,
298 * flush the workqueue.
299 */
300STATIC void
301xfs_finish_ioend(
302 xfs_ioend_t *ioend,
303 int wait)
304{
305 if (atomic_dec_and_test(&ioend->io_remaining)) {
306 struct workqueue_struct *wq = xfsdatad_workqueue;
307 if (ioend->io_work.func == xfs_end_bio_unwritten)
308 wq = xfsconvertd_workqueue;
309
310 queue_work(wq, &ioend->io_work);
311 if (wait)
312 flush_workqueue(wq);
313 }
314}
315
316/*
313 * Allocate and initialise an IO completion structure. 317 * Allocate and initialise an IO completion structure.
314 * We need to track unwritten extent write completion here initially. 318 * We need to track unwritten extent write completion here initially.
315 * We'll need to extend this for updating the ondisk inode size later 319 * We'll need to extend this for updating the ondisk inode size later
diff --git a/fs/xfs/linux-2.6/xfs_aops.h b/fs/xfs/linux-2.6/xfs_aops.h
index 1dd528849755..221b3e66ceef 100644
--- a/fs/xfs/linux-2.6/xfs_aops.h
+++ b/fs/xfs/linux-2.6/xfs_aops.h
@@ -19,6 +19,7 @@
19#define __XFS_AOPS_H__ 19#define __XFS_AOPS_H__
20 20
21extern struct workqueue_struct *xfsdatad_workqueue; 21extern struct workqueue_struct *xfsdatad_workqueue;
22extern struct workqueue_struct *xfsconvertd_workqueue;
22extern mempool_t *xfs_ioend_pool; 23extern mempool_t *xfs_ioend_pool;
23 24
24/* 25/*
diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c
index aa1016bb9134..e28800a9f2b5 100644
--- a/fs/xfs/linux-2.6/xfs_buf.c
+++ b/fs/xfs/linux-2.6/xfs_buf.c
@@ -51,6 +51,7 @@ static struct shrinker xfs_buf_shake = {
51 51
52static struct workqueue_struct *xfslogd_workqueue; 52static struct workqueue_struct *xfslogd_workqueue;
53struct workqueue_struct *xfsdatad_workqueue; 53struct workqueue_struct *xfsdatad_workqueue;
54struct workqueue_struct *xfsconvertd_workqueue;
54 55
55#ifdef XFS_BUF_TRACE 56#ifdef XFS_BUF_TRACE
56void 57void
@@ -1775,6 +1776,7 @@ xfs_flush_buftarg(
1775 xfs_buf_t *bp, *n; 1776 xfs_buf_t *bp, *n;
1776 int pincount = 0; 1777 int pincount = 0;
1777 1778
1779 xfs_buf_runall_queues(xfsconvertd_workqueue);
1778 xfs_buf_runall_queues(xfsdatad_workqueue); 1780 xfs_buf_runall_queues(xfsdatad_workqueue);
1779 xfs_buf_runall_queues(xfslogd_workqueue); 1781 xfs_buf_runall_queues(xfslogd_workqueue);
1780 1782
@@ -1831,9 +1833,15 @@ xfs_buf_init(void)
1831 if (!xfsdatad_workqueue) 1833 if (!xfsdatad_workqueue)
1832 goto out_destroy_xfslogd_workqueue; 1834 goto out_destroy_xfslogd_workqueue;
1833 1835
1836 xfsconvertd_workqueue = create_workqueue("xfsconvertd");
1837 if (!xfsconvertd_workqueue)
1838 goto out_destroy_xfsdatad_workqueue;
1839
1834 register_shrinker(&xfs_buf_shake); 1840 register_shrinker(&xfs_buf_shake);
1835 return 0; 1841 return 0;
1836 1842
1843 out_destroy_xfsdatad_workqueue:
1844 destroy_workqueue(xfsdatad_workqueue);
1837 out_destroy_xfslogd_workqueue: 1845 out_destroy_xfslogd_workqueue:
1838 destroy_workqueue(xfslogd_workqueue); 1846 destroy_workqueue(xfslogd_workqueue);
1839 out_free_buf_zone: 1847 out_free_buf_zone:
@@ -1849,6 +1857,7 @@ void
1849xfs_buf_terminate(void) 1857xfs_buf_terminate(void)
1850{ 1858{
1851 unregister_shrinker(&xfs_buf_shake); 1859 unregister_shrinker(&xfs_buf_shake);
1860 destroy_workqueue(xfsconvertd_workqueue);
1852 destroy_workqueue(xfsdatad_workqueue); 1861 destroy_workqueue(xfsdatad_workqueue);
1853 destroy_workqueue(xfslogd_workqueue); 1862 destroy_workqueue(xfslogd_workqueue);
1854 kmem_zone_destroy(xfs_buf_zone); 1863 kmem_zone_destroy(xfs_buf_zone);
diff --git a/fs/xfs/linux-2.6/xfs_fs_subr.c b/fs/xfs/linux-2.6/xfs_fs_subr.c
index 5aeb77776961..08be36d7326c 100644
--- a/fs/xfs/linux-2.6/xfs_fs_subr.c
+++ b/fs/xfs/linux-2.6/xfs_fs_subr.c
@@ -74,14 +74,14 @@ xfs_flush_pages(
74 74
75 if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) { 75 if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) {
76 xfs_iflags_clear(ip, XFS_ITRUNCATED); 76 xfs_iflags_clear(ip, XFS_ITRUNCATED);
77 ret = filemap_fdatawrite(mapping); 77 ret = -filemap_fdatawrite(mapping);
78 if (flags & XFS_B_ASYNC)
79 return -ret;
80 ret2 = filemap_fdatawait(mapping);
81 if (!ret)
82 ret = ret2;
83 } 78 }
84 return -ret; 79 if (flags & XFS_B_ASYNC)
80 return ret;
81 ret2 = xfs_wait_on_pages(ip, first, last);
82 if (!ret)
83 ret = ret2;
84 return ret;
85} 85}
86 86
87int 87int
diff --git a/fs/xfs/linux-2.6/xfs_lrw.c b/fs/xfs/linux-2.6/xfs_lrw.c
index 7e90daa0d1d1..9142192ccbe6 100644
--- a/fs/xfs/linux-2.6/xfs_lrw.c
+++ b/fs/xfs/linux-2.6/xfs_lrw.c
@@ -751,10 +751,26 @@ start:
751 goto relock; 751 goto relock;
752 } 752 }
753 } else { 753 } else {
754 int enospc = 0;
755 ssize_t ret2 = 0;
756
757write_retry:
754 xfs_rw_enter_trace(XFS_WRITE_ENTER, xip, (void *)iovp, segs, 758 xfs_rw_enter_trace(XFS_WRITE_ENTER, xip, (void *)iovp, segs,
755 *offset, ioflags); 759 *offset, ioflags);
756 ret = generic_file_buffered_write(iocb, iovp, segs, 760 ret2 = generic_file_buffered_write(iocb, iovp, segs,
757 pos, offset, count, ret); 761 pos, offset, count, ret);
762 /*
763 * if we just got an ENOSPC, flush the inode now we
764 * aren't holding any page locks and retry *once*
765 */
766 if (ret2 == -ENOSPC && !enospc) {
767 error = xfs_flush_pages(xip, 0, -1, 0, FI_NONE);
768 if (error)
769 goto out_unlock_internal;
770 enospc = 1;
771 goto write_retry;
772 }
773 ret = ret2;
758 } 774 }
759 775
760 current->backing_dev_info = NULL; 776 current->backing_dev_info = NULL;
diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c
index a608e72fa405..f7ba76633c29 100644
--- a/fs/xfs/linux-2.6/xfs_sync.c
+++ b/fs/xfs/linux-2.6/xfs_sync.c
@@ -62,12 +62,6 @@ xfs_sync_inodes_ag(
62 uint32_t first_index = 0; 62 uint32_t first_index = 0;
63 int error = 0; 63 int error = 0;
64 int last_error = 0; 64 int last_error = 0;
65 int fflag = XFS_B_ASYNC;
66
67 if (flags & SYNC_DELWRI)
68 fflag = XFS_B_DELWRI;
69 if (flags & SYNC_WAIT)
70 fflag = 0; /* synchronous overrides all */
71 65
72 do { 66 do {
73 struct inode *inode; 67 struct inode *inode;
@@ -128,11 +122,23 @@ xfs_sync_inodes_ag(
128 * If we have to flush data or wait for I/O completion 122 * If we have to flush data or wait for I/O completion
129 * we need to hold the iolock. 123 * we need to hold the iolock.
130 */ 124 */
131 if ((flags & SYNC_DELWRI) && VN_DIRTY(inode)) { 125 if (flags & SYNC_DELWRI) {
132 xfs_ilock(ip, XFS_IOLOCK_SHARED); 126 if (VN_DIRTY(inode)) {
133 lock_flags |= XFS_IOLOCK_SHARED; 127 if (flags & SYNC_TRYLOCK) {
134 error = xfs_flush_pages(ip, 0, -1, fflag, FI_NONE); 128 if (xfs_ilock_nowait(ip, XFS_IOLOCK_SHARED))
135 if (flags & SYNC_IOWAIT) 129 lock_flags |= XFS_IOLOCK_SHARED;
130 } else {
131 xfs_ilock(ip, XFS_IOLOCK_SHARED);
132 lock_flags |= XFS_IOLOCK_SHARED;
133 }
134 if (lock_flags & XFS_IOLOCK_SHARED) {
135 error = xfs_flush_pages(ip, 0, -1,
136 (flags & SYNC_WAIT) ? 0
137 : XFS_B_ASYNC,
138 FI_NONE);
139 }
140 }
141 if (VN_CACHED(inode) && (flags & SYNC_IOWAIT))
136 xfs_ioend_wait(ip); 142 xfs_ioend_wait(ip);
137 } 143 }
138 xfs_ilock(ip, XFS_ILOCK_SHARED); 144 xfs_ilock(ip, XFS_ILOCK_SHARED);
@@ -398,15 +404,17 @@ STATIC void
398xfs_syncd_queue_work( 404xfs_syncd_queue_work(
399 struct xfs_mount *mp, 405 struct xfs_mount *mp,
400 void *data, 406 void *data,
401 void (*syncer)(struct xfs_mount *, void *)) 407 void (*syncer)(struct xfs_mount *, void *),
408 struct completion *completion)
402{ 409{
403 struct bhv_vfs_sync_work *work; 410 struct xfs_sync_work *work;
404 411
405 work = kmem_alloc(sizeof(struct bhv_vfs_sync_work), KM_SLEEP); 412 work = kmem_alloc(sizeof(struct xfs_sync_work), KM_SLEEP);
406 INIT_LIST_HEAD(&work->w_list); 413 INIT_LIST_HEAD(&work->w_list);
407 work->w_syncer = syncer; 414 work->w_syncer = syncer;
408 work->w_data = data; 415 work->w_data = data;
409 work->w_mount = mp; 416 work->w_mount = mp;
417 work->w_completion = completion;
410 spin_lock(&mp->m_sync_lock); 418 spin_lock(&mp->m_sync_lock);
411 list_add_tail(&work->w_list, &mp->m_sync_list); 419 list_add_tail(&work->w_list, &mp->m_sync_list);
412 spin_unlock(&mp->m_sync_lock); 420 spin_unlock(&mp->m_sync_lock);
@@ -420,49 +428,26 @@ xfs_syncd_queue_work(
420 * heads, looking about for more room... 428 * heads, looking about for more room...
421 */ 429 */
422STATIC void 430STATIC void
423xfs_flush_inode_work( 431xfs_flush_inodes_work(
424 struct xfs_mount *mp,
425 void *arg)
426{
427 struct inode *inode = arg;
428 filemap_flush(inode->i_mapping);
429 iput(inode);
430}
431
432void
433xfs_flush_inode(
434 xfs_inode_t *ip)
435{
436 struct inode *inode = VFS_I(ip);
437
438 igrab(inode);
439 xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inode_work);
440 delay(msecs_to_jiffies(500));
441}
442
443/*
444 * This is the "bigger hammer" version of xfs_flush_inode_work...
445 * (IOW, "If at first you don't succeed, use a Bigger Hammer").
446 */
447STATIC void
448xfs_flush_device_work(
449 struct xfs_mount *mp, 432 struct xfs_mount *mp,
450 void *arg) 433 void *arg)
451{ 434{
452 struct inode *inode = arg; 435 struct inode *inode = arg;
453 sync_blockdev(mp->m_super->s_bdev); 436 xfs_sync_inodes(mp, SYNC_DELWRI | SYNC_TRYLOCK);
437 xfs_sync_inodes(mp, SYNC_DELWRI | SYNC_TRYLOCK | SYNC_IOWAIT);
454 iput(inode); 438 iput(inode);
455} 439}
456 440
457void 441void
458xfs_flush_device( 442xfs_flush_inodes(
459 xfs_inode_t *ip) 443 xfs_inode_t *ip)
460{ 444{
461 struct inode *inode = VFS_I(ip); 445 struct inode *inode = VFS_I(ip);
446 DECLARE_COMPLETION_ONSTACK(completion);
462 447
463 igrab(inode); 448 igrab(inode);
464 xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_device_work); 449 xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inodes_work, &completion);
465 delay(msecs_to_jiffies(500)); 450 wait_for_completion(&completion);
466 xfs_log_force(ip->i_mount, (xfs_lsn_t)0, XFS_LOG_FORCE|XFS_LOG_SYNC); 451 xfs_log_force(ip->i_mount, (xfs_lsn_t)0, XFS_LOG_FORCE|XFS_LOG_SYNC);
467} 452}
468 453
@@ -497,7 +482,7 @@ xfssyncd(
497{ 482{
498 struct xfs_mount *mp = arg; 483 struct xfs_mount *mp = arg;
499 long timeleft; 484 long timeleft;
500 bhv_vfs_sync_work_t *work, *n; 485 xfs_sync_work_t *work, *n;
501 LIST_HEAD (tmp); 486 LIST_HEAD (tmp);
502 487
503 set_freezable(); 488 set_freezable();
@@ -532,6 +517,8 @@ xfssyncd(
532 list_del(&work->w_list); 517 list_del(&work->w_list);
533 if (work == &mp->m_sync_work) 518 if (work == &mp->m_sync_work)
534 continue; 519 continue;
520 if (work->w_completion)
521 complete(work->w_completion);
535 kmem_free(work); 522 kmem_free(work);
536 } 523 }
537 } 524 }
@@ -545,6 +532,7 @@ xfs_syncd_init(
545{ 532{
546 mp->m_sync_work.w_syncer = xfs_sync_worker; 533 mp->m_sync_work.w_syncer = xfs_sync_worker;
547 mp->m_sync_work.w_mount = mp; 534 mp->m_sync_work.w_mount = mp;
535 mp->m_sync_work.w_completion = NULL;
548 mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd"); 536 mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd");
549 if (IS_ERR(mp->m_sync_task)) 537 if (IS_ERR(mp->m_sync_task))
550 return -PTR_ERR(mp->m_sync_task); 538 return -PTR_ERR(mp->m_sync_task);
diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h
index 04f058c848ae..308d5bf6dfbd 100644
--- a/fs/xfs/linux-2.6/xfs_sync.h
+++ b/fs/xfs/linux-2.6/xfs_sync.h
@@ -21,18 +21,20 @@
21struct xfs_mount; 21struct xfs_mount;
22struct xfs_perag; 22struct xfs_perag;
23 23
24typedef struct bhv_vfs_sync_work { 24typedef struct xfs_sync_work {
25 struct list_head w_list; 25 struct list_head w_list;
26 struct xfs_mount *w_mount; 26 struct xfs_mount *w_mount;
27 void *w_data; /* syncer routine argument */ 27 void *w_data; /* syncer routine argument */
28 void (*w_syncer)(struct xfs_mount *, void *); 28 void (*w_syncer)(struct xfs_mount *, void *);
29} bhv_vfs_sync_work_t; 29 struct completion *w_completion;
30} xfs_sync_work_t;
30 31
31#define SYNC_ATTR 0x0001 /* sync attributes */ 32#define SYNC_ATTR 0x0001 /* sync attributes */
32#define SYNC_DELWRI 0x0002 /* look at delayed writes */ 33#define SYNC_DELWRI 0x0002 /* look at delayed writes */
33#define SYNC_WAIT 0x0004 /* wait for i/o to complete */ 34#define SYNC_WAIT 0x0004 /* wait for i/o to complete */
34#define SYNC_BDFLUSH 0x0008 /* BDFLUSH is calling -- don't block */ 35#define SYNC_BDFLUSH 0x0008 /* BDFLUSH is calling -- don't block */
35#define SYNC_IOWAIT 0x0010 /* wait for all I/O to complete */ 36#define SYNC_IOWAIT 0x0010 /* wait for all I/O to complete */
37#define SYNC_TRYLOCK 0x0020 /* only try to lock inodes */
36 38
37int xfs_syncd_init(struct xfs_mount *mp); 39int xfs_syncd_init(struct xfs_mount *mp);
38void xfs_syncd_stop(struct xfs_mount *mp); 40void xfs_syncd_stop(struct xfs_mount *mp);
@@ -43,8 +45,7 @@ int xfs_sync_fsdata(struct xfs_mount *mp, int flags);
43int xfs_quiesce_data(struct xfs_mount *mp); 45int xfs_quiesce_data(struct xfs_mount *mp);
44void xfs_quiesce_attr(struct xfs_mount *mp); 46void xfs_quiesce_attr(struct xfs_mount *mp);
45 47
46void xfs_flush_inode(struct xfs_inode *ip); 48void xfs_flush_inodes(struct xfs_inode *ip);
47void xfs_flush_device(struct xfs_inode *ip);
48 49
49int xfs_reclaim_inode(struct xfs_inode *ip, int locked, int sync_mode); 50int xfs_reclaim_inode(struct xfs_inode *ip, int locked, int sync_mode);
50int xfs_reclaim_inodes(struct xfs_mount *mp, int noblock, int mode); 51int xfs_reclaim_inodes(struct xfs_mount *mp, int noblock, int mode);
diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c
index 478e587087fe..89b81eedce6a 100644
--- a/fs/xfs/xfs_iget.c
+++ b/fs/xfs/xfs_iget.c
@@ -69,15 +69,6 @@ xfs_inode_alloc(
69 ASSERT(!spin_is_locked(&ip->i_flags_lock)); 69 ASSERT(!spin_is_locked(&ip->i_flags_lock));
70 ASSERT(completion_done(&ip->i_flush)); 70 ASSERT(completion_done(&ip->i_flush));
71 71
72 /*
73 * initialise the VFS inode here to get failures
74 * out of the way early.
75 */
76 if (!inode_init_always(mp->m_super, VFS_I(ip))) {
77 kmem_zone_free(xfs_inode_zone, ip);
78 return NULL;
79 }
80
81 /* initialise the xfs inode */ 72 /* initialise the xfs inode */
82 ip->i_ino = ino; 73 ip->i_ino = ino;
83 ip->i_mount = mp; 74 ip->i_mount = mp;
@@ -113,6 +104,20 @@ xfs_inode_alloc(
113#ifdef XFS_DIR2_TRACE 104#ifdef XFS_DIR2_TRACE
114 ip->i_dir_trace = ktrace_alloc(XFS_DIR2_KTRACE_SIZE, KM_NOFS); 105 ip->i_dir_trace = ktrace_alloc(XFS_DIR2_KTRACE_SIZE, KM_NOFS);
115#endif 106#endif
107 /*
108 * Now initialise the VFS inode. We do this after the xfs_inode
109 * initialisation as internal failures will result in ->destroy_inode
110 * being called and that will pass down through the reclaim path and
111 * free the XFS inode. This path requires the XFS inode to already be
112 * initialised. Hence if this call fails, the xfs_inode has already
113 * been freed and we should not reference it at all in the error
114 * handling.
115 */
116 if (!inode_init_always(mp->m_super, VFS_I(ip)))
117 return NULL;
118
119 /* prevent anyone from using this yet */
120 VFS_I(ip)->i_state = I_NEW|I_LOCK;
116 121
117 return ip; 122 return ip;
118} 123}
diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c
index 08ce72316bfe..5aaa2d7ec155 100644
--- a/fs/xfs/xfs_iomap.c
+++ b/fs/xfs/xfs_iomap.c
@@ -338,38 +338,6 @@ xfs_iomap_eof_align_last_fsb(
338} 338}
339 339
340STATIC int 340STATIC int
341xfs_flush_space(
342 xfs_inode_t *ip,
343 int *fsynced,
344 int *ioflags)
345{
346 switch (*fsynced) {
347 case 0:
348 if (ip->i_delayed_blks) {
349 xfs_iunlock(ip, XFS_ILOCK_EXCL);
350 xfs_flush_inode(ip);
351 xfs_ilock(ip, XFS_ILOCK_EXCL);
352 *fsynced = 1;
353 } else {
354 *ioflags |= BMAPI_SYNC;
355 *fsynced = 2;
356 }
357 return 0;
358 case 1:
359 *fsynced = 2;
360 *ioflags |= BMAPI_SYNC;
361 return 0;
362 case 2:
363 xfs_iunlock(ip, XFS_ILOCK_EXCL);
364 xfs_flush_device(ip);
365 xfs_ilock(ip, XFS_ILOCK_EXCL);
366 *fsynced = 3;
367 return 0;
368 }
369 return 1;
370}
371
372STATIC int
373xfs_cmn_err_fsblock_zero( 341xfs_cmn_err_fsblock_zero(
374 xfs_inode_t *ip, 342 xfs_inode_t *ip,
375 xfs_bmbt_irec_t *imap) 343 xfs_bmbt_irec_t *imap)
@@ -538,15 +506,9 @@ error_out:
538} 506}
539 507
540/* 508/*
541 * If the caller is doing a write at the end of the file, 509 * If the caller is doing a write at the end of the file, then extend the
542 * then extend the allocation out to the file system's write 510 * allocation out to the file system's write iosize. We clean up any extra
543 * iosize. We clean up any extra space left over when the 511 * space left over when the file is closed in xfs_inactive().
544 * file is closed in xfs_inactive().
545 *
546 * For sync writes, we are flushing delayed allocate space to
547 * try to make additional space available for allocation near
548 * the filesystem full boundary - preallocation hurts in that
549 * situation, of course.
550 */ 512 */
551STATIC int 513STATIC int
552xfs_iomap_eof_want_preallocate( 514xfs_iomap_eof_want_preallocate(
@@ -565,7 +527,7 @@ xfs_iomap_eof_want_preallocate(
565 int n, error, imaps; 527 int n, error, imaps;
566 528
567 *prealloc = 0; 529 *prealloc = 0;
568 if ((ioflag & BMAPI_SYNC) || (offset + count) <= ip->i_size) 530 if ((offset + count) <= ip->i_size)
569 return 0; 531 return 0;
570 532
571 /* 533 /*
@@ -611,7 +573,7 @@ xfs_iomap_write_delay(
611 xfs_extlen_t extsz; 573 xfs_extlen_t extsz;
612 int nimaps; 574 int nimaps;
613 xfs_bmbt_irec_t imap[XFS_WRITE_IMAPS]; 575 xfs_bmbt_irec_t imap[XFS_WRITE_IMAPS];
614 int prealloc, fsynced = 0; 576 int prealloc, flushed = 0;
615 int error; 577 int error;
616 578
617 ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); 579 ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
@@ -627,12 +589,12 @@ xfs_iomap_write_delay(
627 extsz = xfs_get_extsz_hint(ip); 589 extsz = xfs_get_extsz_hint(ip);
628 offset_fsb = XFS_B_TO_FSBT(mp, offset); 590 offset_fsb = XFS_B_TO_FSBT(mp, offset);
629 591
630retry:
631 error = xfs_iomap_eof_want_preallocate(mp, ip, offset, count, 592 error = xfs_iomap_eof_want_preallocate(mp, ip, offset, count,
632 ioflag, imap, XFS_WRITE_IMAPS, &prealloc); 593 ioflag, imap, XFS_WRITE_IMAPS, &prealloc);
633 if (error) 594 if (error)
634 return error; 595 return error;
635 596
597retry:
636 if (prealloc) { 598 if (prealloc) {
637 aligned_offset = XFS_WRITEIO_ALIGN(mp, (offset + count - 1)); 599 aligned_offset = XFS_WRITEIO_ALIGN(mp, (offset + count - 1));
638 ioalign = XFS_B_TO_FSBT(mp, aligned_offset); 600 ioalign = XFS_B_TO_FSBT(mp, aligned_offset);
@@ -659,15 +621,22 @@ retry:
659 621
660 /* 622 /*
661 * If bmapi returned us nothing, and if we didn't get back EDQUOT, 623 * If bmapi returned us nothing, and if we didn't get back EDQUOT,
662 * then we must have run out of space - flush delalloc, and retry.. 624 * then we must have run out of space - flush all other inodes with
625 * delalloc blocks and retry without EOF preallocation.
663 */ 626 */
664 if (nimaps == 0) { 627 if (nimaps == 0) {
665 xfs_iomap_enter_trace(XFS_IOMAP_WRITE_NOSPACE, 628 xfs_iomap_enter_trace(XFS_IOMAP_WRITE_NOSPACE,
666 ip, offset, count); 629 ip, offset, count);
667 if (xfs_flush_space(ip, &fsynced, &ioflag)) 630 if (flushed)
668 return XFS_ERROR(ENOSPC); 631 return XFS_ERROR(ENOSPC);
669 632
633 xfs_iunlock(ip, XFS_ILOCK_EXCL);
634 xfs_flush_inodes(ip);
635 xfs_ilock(ip, XFS_ILOCK_EXCL);
636
637 flushed = 1;
670 error = 0; 638 error = 0;
639 prealloc = 0;
671 goto retry; 640 goto retry;
672 } 641 }
673 642
diff --git a/fs/xfs/xfs_iomap.h b/fs/xfs/xfs_iomap.h
index a1cc1322fc0f..fdcf7b82747f 100644
--- a/fs/xfs/xfs_iomap.h
+++ b/fs/xfs/xfs_iomap.h
@@ -40,8 +40,7 @@ typedef enum {
40 BMAPI_IGNSTATE = (1 << 4), /* ignore unwritten state on read */ 40 BMAPI_IGNSTATE = (1 << 4), /* ignore unwritten state on read */
41 BMAPI_DIRECT = (1 << 5), /* direct instead of buffered write */ 41 BMAPI_DIRECT = (1 << 5), /* direct instead of buffered write */
42 BMAPI_MMAP = (1 << 6), /* allocate for mmap write */ 42 BMAPI_MMAP = (1 << 6), /* allocate for mmap write */
43 BMAPI_SYNC = (1 << 7), /* sync write to flush delalloc space */ 43 BMAPI_TRYLOCK = (1 << 7), /* non-blocking request */
44 BMAPI_TRYLOCK = (1 << 8), /* non-blocking request */
45} bmapi_flags_t; 44} bmapi_flags_t;
46 45
47 46
diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c
index f76c6d7cea21..3750f04ede0b 100644
--- a/fs/xfs/xfs_log.c
+++ b/fs/xfs/xfs_log.c
@@ -562,9 +562,8 @@ xfs_log_mount(
562 } 562 }
563 563
564 mp->m_log = xlog_alloc_log(mp, log_target, blk_offset, num_bblks); 564 mp->m_log = xlog_alloc_log(mp, log_target, blk_offset, num_bblks);
565 if (!mp->m_log) { 565 if (IS_ERR(mp->m_log)) {
566 cmn_err(CE_WARN, "XFS: Log allocation failed: No memory!"); 566 error = -PTR_ERR(mp->m_log);
567 error = ENOMEM;
568 goto out; 567 goto out;
569 } 568 }
570 569
@@ -1180,10 +1179,13 @@ xlog_alloc_log(xfs_mount_t *mp,
1180 xfs_buf_t *bp; 1179 xfs_buf_t *bp;
1181 int i; 1180 int i;
1182 int iclogsize; 1181 int iclogsize;
1182 int error = ENOMEM;
1183 1183
1184 log = kmem_zalloc(sizeof(xlog_t), KM_MAYFAIL); 1184 log = kmem_zalloc(sizeof(xlog_t), KM_MAYFAIL);
1185 if (!log) 1185 if (!log) {
1186 return NULL; 1186 xlog_warn("XFS: Log allocation failed: No memory!");
1187 goto out;
1188 }
1187 1189
1188 log->l_mp = mp; 1190 log->l_mp = mp;
1189 log->l_targ = log_target; 1191 log->l_targ = log_target;
@@ -1201,19 +1203,35 @@ xlog_alloc_log(xfs_mount_t *mp,
1201 log->l_grant_reserve_cycle = 1; 1203 log->l_grant_reserve_cycle = 1;
1202 log->l_grant_write_cycle = 1; 1204 log->l_grant_write_cycle = 1;
1203 1205
1206 error = EFSCORRUPTED;
1204 if (xfs_sb_version_hassector(&mp->m_sb)) { 1207 if (xfs_sb_version_hassector(&mp->m_sb)) {
1205 log->l_sectbb_log = mp->m_sb.sb_logsectlog - BBSHIFT; 1208 log->l_sectbb_log = mp->m_sb.sb_logsectlog - BBSHIFT;
1206 ASSERT(log->l_sectbb_log <= mp->m_sectbb_log); 1209 if (log->l_sectbb_log < 0 ||
1210 log->l_sectbb_log > mp->m_sectbb_log) {
1211 xlog_warn("XFS: Log sector size (0x%x) out of range.",
1212 log->l_sectbb_log);
1213 goto out_free_log;
1214 }
1215
1207 /* for larger sector sizes, must have v2 or external log */ 1216 /* for larger sector sizes, must have v2 or external log */
1208 ASSERT(log->l_sectbb_log == 0 || 1217 if (log->l_sectbb_log != 0 &&
1209 log->l_logBBstart == 0 || 1218 (log->l_logBBstart != 0 &&
1210 xfs_sb_version_haslogv2(&mp->m_sb)); 1219 !xfs_sb_version_haslogv2(&mp->m_sb))) {
1211 ASSERT(mp->m_sb.sb_logsectlog >= BBSHIFT); 1220 xlog_warn("XFS: log sector size (0x%x) invalid "
1221 "for configuration.", log->l_sectbb_log);
1222 goto out_free_log;
1223 }
1224 if (mp->m_sb.sb_logsectlog < BBSHIFT) {
1225 xlog_warn("XFS: Log sector log (0x%x) too small.",
1226 mp->m_sb.sb_logsectlog);
1227 goto out_free_log;
1228 }
1212 } 1229 }
1213 log->l_sectbb_mask = (1 << log->l_sectbb_log) - 1; 1230 log->l_sectbb_mask = (1 << log->l_sectbb_log) - 1;
1214 1231
1215 xlog_get_iclog_buffer_size(mp, log); 1232 xlog_get_iclog_buffer_size(mp, log);
1216 1233
1234 error = ENOMEM;
1217 bp = xfs_buf_get_empty(log->l_iclog_size, mp->m_logdev_targp); 1235 bp = xfs_buf_get_empty(log->l_iclog_size, mp->m_logdev_targp);
1218 if (!bp) 1236 if (!bp)
1219 goto out_free_log; 1237 goto out_free_log;
@@ -1313,7 +1331,8 @@ out_free_iclog:
1313 xfs_buf_free(log->l_xbuf); 1331 xfs_buf_free(log->l_xbuf);
1314out_free_log: 1332out_free_log:
1315 kmem_free(log); 1333 kmem_free(log);
1316 return NULL; 1334out:
1335 return ERR_PTR(-error);
1317} /* xlog_alloc_log */ 1336} /* xlog_alloc_log */
1318 1337
1319 1338
@@ -2541,18 +2560,19 @@ redo:
2541 xlog_ins_ticketq(&log->l_reserve_headq, tic); 2560 xlog_ins_ticketq(&log->l_reserve_headq, tic);
2542 xlog_trace_loggrant(log, tic, 2561 xlog_trace_loggrant(log, tic,
2543 "xlog_grant_log_space: sleep 2"); 2562 "xlog_grant_log_space: sleep 2");
2563 spin_unlock(&log->l_grant_lock);
2564 xlog_grant_push_ail(log->l_mp, need_bytes);
2565 spin_lock(&log->l_grant_lock);
2566
2544 XFS_STATS_INC(xs_sleep_logspace); 2567 XFS_STATS_INC(xs_sleep_logspace);
2545 sv_wait(&tic->t_wait, PINOD|PLTWAIT, &log->l_grant_lock, s); 2568 sv_wait(&tic->t_wait, PINOD|PLTWAIT, &log->l_grant_lock, s);
2546 2569
2547 if (XLOG_FORCED_SHUTDOWN(log)) { 2570 spin_lock(&log->l_grant_lock);
2548 spin_lock(&log->l_grant_lock); 2571 if (XLOG_FORCED_SHUTDOWN(log))
2549 goto error_return; 2572 goto error_return;
2550 }
2551 2573
2552 xlog_trace_loggrant(log, tic, 2574 xlog_trace_loggrant(log, tic,
2553 "xlog_grant_log_space: wake 2"); 2575 "xlog_grant_log_space: wake 2");
2554 xlog_grant_push_ail(log->l_mp, need_bytes);
2555 spin_lock(&log->l_grant_lock);
2556 goto redo; 2576 goto redo;
2557 } else if (tic->t_flags & XLOG_TIC_IN_Q) 2577 } else if (tic->t_flags & XLOG_TIC_IN_Q)
2558 xlog_del_ticketq(&log->l_reserve_headq, tic); 2578 xlog_del_ticketq(&log->l_reserve_headq, tic);
@@ -2631,7 +2651,7 @@ xlog_regrant_write_log_space(xlog_t *log,
2631 * for more free space, otherwise try to get some space for 2651 * for more free space, otherwise try to get some space for
2632 * this transaction. 2652 * this transaction.
2633 */ 2653 */
2634 2654 need_bytes = tic->t_unit_res;
2635 if ((ntic = log->l_write_headq)) { 2655 if ((ntic = log->l_write_headq)) {
2636 free_bytes = xlog_space_left(log, log->l_grant_write_cycle, 2656 free_bytes = xlog_space_left(log, log->l_grant_write_cycle,
2637 log->l_grant_write_bytes); 2657 log->l_grant_write_bytes);
@@ -2651,26 +2671,25 @@ xlog_regrant_write_log_space(xlog_t *log,
2651 2671
2652 xlog_trace_loggrant(log, tic, 2672 xlog_trace_loggrant(log, tic,
2653 "xlog_regrant_write_log_space: sleep 1"); 2673 "xlog_regrant_write_log_space: sleep 1");
2674 spin_unlock(&log->l_grant_lock);
2675 xlog_grant_push_ail(log->l_mp, need_bytes);
2676 spin_lock(&log->l_grant_lock);
2677
2654 XFS_STATS_INC(xs_sleep_logspace); 2678 XFS_STATS_INC(xs_sleep_logspace);
2655 sv_wait(&tic->t_wait, PINOD|PLTWAIT, 2679 sv_wait(&tic->t_wait, PINOD|PLTWAIT,
2656 &log->l_grant_lock, s); 2680 &log->l_grant_lock, s);
2657 2681
2658 /* If we're shutting down, this tic is already 2682 /* If we're shutting down, this tic is already
2659 * off the queue */ 2683 * off the queue */
2660 if (XLOG_FORCED_SHUTDOWN(log)) { 2684 spin_lock(&log->l_grant_lock);
2661 spin_lock(&log->l_grant_lock); 2685 if (XLOG_FORCED_SHUTDOWN(log))
2662 goto error_return; 2686 goto error_return;
2663 }
2664 2687
2665 xlog_trace_loggrant(log, tic, 2688 xlog_trace_loggrant(log, tic,
2666 "xlog_regrant_write_log_space: wake 1"); 2689 "xlog_regrant_write_log_space: wake 1");
2667 xlog_grant_push_ail(log->l_mp, tic->t_unit_res);
2668 spin_lock(&log->l_grant_lock);
2669 } 2690 }
2670 } 2691 }
2671 2692
2672 need_bytes = tic->t_unit_res;
2673
2674redo: 2693redo:
2675 if (XLOG_FORCED_SHUTDOWN(log)) 2694 if (XLOG_FORCED_SHUTDOWN(log))
2676 goto error_return; 2695 goto error_return;
@@ -2680,19 +2699,20 @@ redo:
2680 if (free_bytes < need_bytes) { 2699 if (free_bytes < need_bytes) {
2681 if ((tic->t_flags & XLOG_TIC_IN_Q) == 0) 2700 if ((tic->t_flags & XLOG_TIC_IN_Q) == 0)
2682 xlog_ins_ticketq(&log->l_write_headq, tic); 2701 xlog_ins_ticketq(&log->l_write_headq, tic);
2702 spin_unlock(&log->l_grant_lock);
2703 xlog_grant_push_ail(log->l_mp, need_bytes);
2704 spin_lock(&log->l_grant_lock);
2705
2683 XFS_STATS_INC(xs_sleep_logspace); 2706 XFS_STATS_INC(xs_sleep_logspace);
2684 sv_wait(&tic->t_wait, PINOD|PLTWAIT, &log->l_grant_lock, s); 2707 sv_wait(&tic->t_wait, PINOD|PLTWAIT, &log->l_grant_lock, s);
2685 2708
2686 /* If we're shutting down, this tic is already off the queue */ 2709 /* If we're shutting down, this tic is already off the queue */
2687 if (XLOG_FORCED_SHUTDOWN(log)) { 2710 spin_lock(&log->l_grant_lock);
2688 spin_lock(&log->l_grant_lock); 2711 if (XLOG_FORCED_SHUTDOWN(log))
2689 goto error_return; 2712 goto error_return;
2690 }
2691 2713
2692 xlog_trace_loggrant(log, tic, 2714 xlog_trace_loggrant(log, tic,
2693 "xlog_regrant_write_log_space: wake 2"); 2715 "xlog_regrant_write_log_space: wake 2");
2694 xlog_grant_push_ail(log->l_mp, need_bytes);
2695 spin_lock(&log->l_grant_lock);
2696 goto redo; 2716 goto redo;
2697 } else if (tic->t_flags & XLOG_TIC_IN_Q) 2717 } else if (tic->t_flags & XLOG_TIC_IN_Q)
2698 xlog_del_ticketq(&log->l_write_headq, tic); 2718 xlog_del_ticketq(&log->l_write_headq, tic);
diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h
index 7af44adffc8f..d6a64392f983 100644
--- a/fs/xfs/xfs_mount.h
+++ b/fs/xfs/xfs_mount.h
@@ -313,7 +313,7 @@ typedef struct xfs_mount {
313#endif 313#endif
314 struct xfs_mru_cache *m_filestream; /* per-mount filestream data */ 314 struct xfs_mru_cache *m_filestream; /* per-mount filestream data */
315 struct task_struct *m_sync_task; /* generalised sync thread */ 315 struct task_struct *m_sync_task; /* generalised sync thread */
316 bhv_vfs_sync_work_t m_sync_work; /* work item for VFS_SYNC */ 316 xfs_sync_work_t m_sync_work; /* work item for VFS_SYNC */
317 struct list_head m_sync_list; /* sync thread work item list */ 317 struct list_head m_sync_list; /* sync thread work item list */
318 spinlock_t m_sync_lock; /* work item list lock */ 318 spinlock_t m_sync_lock; /* work item list lock */
319 int m_sync_seq; /* sync thread generation no. */ 319 int m_sync_seq; /* sync thread generation no. */
diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c
index 7394c7af5de5..19cf90a9c762 100644
--- a/fs/xfs/xfs_vnodeops.c
+++ b/fs/xfs/xfs_vnodeops.c
@@ -1457,6 +1457,13 @@ xfs_create(
1457 error = xfs_trans_reserve(tp, resblks, log_res, 0, 1457 error = xfs_trans_reserve(tp, resblks, log_res, 0,
1458 XFS_TRANS_PERM_LOG_RES, log_count); 1458 XFS_TRANS_PERM_LOG_RES, log_count);
1459 if (error == ENOSPC) { 1459 if (error == ENOSPC) {
1460 /* flush outstanding delalloc blocks and retry */
1461 xfs_flush_inodes(dp);
1462 error = xfs_trans_reserve(tp, resblks, XFS_CREATE_LOG_RES(mp), 0,
1463 XFS_TRANS_PERM_LOG_RES, XFS_CREATE_LOG_COUNT);
1464 }
1465 if (error == ENOSPC) {
1466 /* No space at all so try a "no-allocation" reservation */
1460 resblks = 0; 1467 resblks = 0;
1461 error = xfs_trans_reserve(tp, 0, log_res, 0, 1468 error = xfs_trans_reserve(tp, 0, log_res, 0,
1462 XFS_TRANS_PERM_LOG_RES, log_count); 1469 XFS_TRANS_PERM_LOG_RES, log_count);
diff --git a/include/asm-generic/bug.h b/include/asm-generic/bug.h
index 37b82cb96c89..e727fe0d1451 100644
--- a/include/asm-generic/bug.h
+++ b/include/asm-generic/bug.h
@@ -88,7 +88,7 @@ extern void warn_slowpath(const char *file, const int line,
88 88
89#else /* !CONFIG_BUG */ 89#else /* !CONFIG_BUG */
90#ifndef HAVE_ARCH_BUG 90#ifndef HAVE_ARCH_BUG
91#define BUG() 91#define BUG() do {} while(0)
92#endif 92#endif
93 93
94#ifndef HAVE_ARCH_BUG_ON 94#ifndef HAVE_ARCH_BUG_ON
diff --git a/include/asm-generic/siginfo.h b/include/asm-generic/siginfo.h
index 35752dadd6df..c840719a8c59 100644
--- a/include/asm-generic/siginfo.h
+++ b/include/asm-generic/siginfo.h
@@ -201,7 +201,7 @@ typedef struct siginfo {
201#define TRAP_TRACE (__SI_FAULT|2) /* process trace trap */ 201#define TRAP_TRACE (__SI_FAULT|2) /* process trace trap */
202#define TRAP_BRANCH (__SI_FAULT|3) /* process taken branch trap */ 202#define TRAP_BRANCH (__SI_FAULT|3) /* process taken branch trap */
203#define TRAP_HWBKPT (__SI_FAULT|4) /* hardware breakpoint/watchpoint */ 203#define TRAP_HWBKPT (__SI_FAULT|4) /* hardware breakpoint/watchpoint */
204#define NSIGTRAP 2 204#define NSIGTRAP 4
205 205
206/* 206/*
207 * SIGCHLD si_codes 207 * SIGCHLD si_codes
diff --git a/include/drm/drm_pciids.h b/include/drm/drm_pciids.h
index 2df74eb09563..9477af01a639 100644
--- a/include/drm/drm_pciids.h
+++ b/include/drm/drm_pciids.h
@@ -472,6 +472,7 @@
472 {0x8086, 0x2562, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \ 472 {0x8086, 0x2562, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \
473 {0x8086, 0x3582, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \ 473 {0x8086, 0x3582, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \
474 {0x8086, 0x2572, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \ 474 {0x8086, 0x2572, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \
475 {0x8086, 0x358e, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 0}, \
475 {0, 0, 0} 476 {0, 0, 0}
476 477
477#define gamma_PCI_IDS \ 478#define gamma_PCI_IDS \
@@ -533,4 +534,5 @@
533 {0x8086, 0x2e22, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, 0xffff00, 0}, \ 534 {0x8086, 0x2e22, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, 0xffff00, 0}, \
534 {0x8086, 0xa001, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, 0xffff00, 0}, \ 535 {0x8086, 0xa001, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, 0xffff00, 0}, \
535 {0x8086, 0xa011, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, 0xffff00, 0}, \ 536 {0x8086, 0xa011, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, 0xffff00, 0}, \
537 {0x8086, 0x35e8, PCI_ANY_ID, PCI_ANY_ID, PCI_CLASS_DISPLAY_VGA << 8, 0xffff00, 0}, \
536 {0, 0, 0} 538 {0, 0, 0}
diff --git a/include/drm/i915_drm.h b/include/drm/i915_drm.h
index 67e3353a56d6..95962fa8398a 100644
--- a/include/drm/i915_drm.h
+++ b/include/drm/i915_drm.h
@@ -594,6 +594,9 @@ struct drm_i915_gem_busy {
594#define I915_BIT_6_SWIZZLE_9_10_11 4 594#define I915_BIT_6_SWIZZLE_9_10_11 4
595/* Not seen by userland */ 595/* Not seen by userland */
596#define I915_BIT_6_SWIZZLE_UNKNOWN 5 596#define I915_BIT_6_SWIZZLE_UNKNOWN 5
597/* Seen by userland. */
598#define I915_BIT_6_SWIZZLE_9_17 6
599#define I915_BIT_6_SWIZZLE_9_10_17 7
597 600
598struct drm_i915_gem_set_tiling { 601struct drm_i915_gem_set_tiling {
599 /** Handle of the buffer to have its tiling state updated */ 602 /** Handle of the buffer to have its tiling state updated */
diff --git a/include/linux/bio.h b/include/linux/bio.h
index b900d2c67d29..b89cf2d82898 100644
--- a/include/linux/bio.h
+++ b/include/linux/bio.h
@@ -504,6 +504,115 @@ static inline int bio_has_data(struct bio *bio)
504 return bio && bio->bi_io_vec != NULL; 504 return bio && bio->bi_io_vec != NULL;
505} 505}
506 506
507/*
508 * BIO list managment for use by remapping drivers (e.g. DM or MD).
509 *
510 * A bio_list anchors a singly-linked list of bios chained through the bi_next
511 * member of the bio. The bio_list also caches the last list member to allow
512 * fast access to the tail.
513 */
514struct bio_list {
515 struct bio *head;
516 struct bio *tail;
517};
518
519static inline int bio_list_empty(const struct bio_list *bl)
520{
521 return bl->head == NULL;
522}
523
524static inline void bio_list_init(struct bio_list *bl)
525{
526 bl->head = bl->tail = NULL;
527}
528
529#define bio_list_for_each(bio, bl) \
530 for (bio = (bl)->head; bio; bio = bio->bi_next)
531
532static inline unsigned bio_list_size(const struct bio_list *bl)
533{
534 unsigned sz = 0;
535 struct bio *bio;
536
537 bio_list_for_each(bio, bl)
538 sz++;
539
540 return sz;
541}
542
543static inline void bio_list_add(struct bio_list *bl, struct bio *bio)
544{
545 bio->bi_next = NULL;
546
547 if (bl->tail)
548 bl->tail->bi_next = bio;
549 else
550 bl->head = bio;
551
552 bl->tail = bio;
553}
554
555static inline void bio_list_add_head(struct bio_list *bl, struct bio *bio)
556{
557 bio->bi_next = bl->head;
558
559 bl->head = bio;
560
561 if (!bl->tail)
562 bl->tail = bio;
563}
564
565static inline void bio_list_merge(struct bio_list *bl, struct bio_list *bl2)
566{
567 if (!bl2->head)
568 return;
569
570 if (bl->tail)
571 bl->tail->bi_next = bl2->head;
572 else
573 bl->head = bl2->head;
574
575 bl->tail = bl2->tail;
576}
577
578static inline void bio_list_merge_head(struct bio_list *bl,
579 struct bio_list *bl2)
580{
581 if (!bl2->head)
582 return;
583
584 if (bl->head)
585 bl2->tail->bi_next = bl->head;
586 else
587 bl->tail = bl2->tail;
588
589 bl->head = bl2->head;
590}
591
592static inline struct bio *bio_list_pop(struct bio_list *bl)
593{
594 struct bio *bio = bl->head;
595
596 if (bio) {
597 bl->head = bl->head->bi_next;
598 if (!bl->head)
599 bl->tail = NULL;
600
601 bio->bi_next = NULL;
602 }
603
604 return bio;
605}
606
607static inline struct bio *bio_list_get(struct bio_list *bl)
608{
609 struct bio *bio = bl->head;
610
611 bl->head = bl->tail = NULL;
612
613 return bio;
614}
615
507#if defined(CONFIG_BLK_DEV_INTEGRITY) 616#if defined(CONFIG_BLK_DEV_INTEGRITY)
508 617
509#define bip_vec_idx(bip, idx) (&(bip->bip_vec[(idx)])) 618#define bip_vec_idx(bip, idx) (&(bip->bip_vec[(idx)]))
diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h
index 7b73bb8f1970..16ed0284d780 100644
--- a/include/linux/buffer_head.h
+++ b/include/linux/buffer_head.h
@@ -155,6 +155,7 @@ void create_empty_buffers(struct page *, unsigned long,
155 unsigned long b_state); 155 unsigned long b_state);
156void end_buffer_read_sync(struct buffer_head *bh, int uptodate); 156void end_buffer_read_sync(struct buffer_head *bh, int uptodate);
157void end_buffer_write_sync(struct buffer_head *bh, int uptodate); 157void end_buffer_write_sync(struct buffer_head *bh, int uptodate);
158void end_buffer_async_write(struct buffer_head *bh, int uptodate);
158 159
159/* Things to do with buffers at mapping->private_list */ 160/* Things to do with buffers at mapping->private_list */
160void mark_buffer_dirty_inode(struct buffer_head *bh, struct inode *inode); 161void mark_buffer_dirty_inode(struct buffer_head *bh, struct inode *inode);
@@ -197,6 +198,8 @@ extern int buffer_heads_over_limit;
197void block_invalidatepage(struct page *page, unsigned long offset); 198void block_invalidatepage(struct page *page, unsigned long offset);
198int block_write_full_page(struct page *page, get_block_t *get_block, 199int block_write_full_page(struct page *page, get_block_t *get_block,
199 struct writeback_control *wbc); 200 struct writeback_control *wbc);
201int block_write_full_page_endio(struct page *page, get_block_t *get_block,
202 struct writeback_control *wbc, bh_end_io_t *handler);
200int block_read_full_page(struct page*, get_block_t*); 203int block_read_full_page(struct page*, get_block_t*);
201int block_is_partially_uptodate(struct page *page, read_descriptor_t *desc, 204int block_is_partially_uptodate(struct page *page, read_descriptor_t *desc,
202 unsigned long from); 205 unsigned long from);
diff --git a/include/linux/debug_locks.h b/include/linux/debug_locks.h
index 493dedb7a67b..29b3ce3f2a1d 100644
--- a/include/linux/debug_locks.h
+++ b/include/linux/debug_locks.h
@@ -3,6 +3,7 @@
3 3
4#include <linux/kernel.h> 4#include <linux/kernel.h>
5#include <asm/atomic.h> 5#include <asm/atomic.h>
6#include <asm/system.h>
6 7
7struct task_struct; 8struct task_struct;
8 9
diff --git a/include/linux/fb.h b/include/linux/fb.h
index f563c5013932..330c4b1bfcaa 100644
--- a/include/linux/fb.h
+++ b/include/linux/fb.h
@@ -173,8 +173,12 @@ struct fb_fix_screeninfo {
173/* Interpretation of offset for color fields: All offsets are from the right, 173/* Interpretation of offset for color fields: All offsets are from the right,
174 * inside a "pixel" value, which is exactly 'bits_per_pixel' wide (means: you 174 * inside a "pixel" value, which is exactly 'bits_per_pixel' wide (means: you
175 * can use the offset as right argument to <<). A pixel afterwards is a bit 175 * can use the offset as right argument to <<). A pixel afterwards is a bit
176 * stream and is written to video memory as that unmodified. This implies 176 * stream and is written to video memory as that unmodified.
177 * big-endian byte order if bits_per_pixel is greater than 8. 177 *
178 * For pseudocolor: offset and length should be the same for all color
179 * components. Offset specifies the position of the least significant bit
180 * of the pallette index in a pixel value. Length indicates the number
181 * of available palette entries (i.e. # of entries = 1 << length).
178 */ 182 */
179struct fb_bitfield { 183struct fb_bitfield {
180 __u32 offset; /* beginning of bitfield */ 184 __u32 offset; /* beginning of bitfield */
diff --git a/include/linux/fiemap.h b/include/linux/fiemap.h
index 671decbd2aeb..934e22d65801 100644
--- a/include/linux/fiemap.h
+++ b/include/linux/fiemap.h
@@ -11,6 +11,8 @@
11#ifndef _LINUX_FIEMAP_H 11#ifndef _LINUX_FIEMAP_H
12#define _LINUX_FIEMAP_H 12#define _LINUX_FIEMAP_H
13 13
14#include <linux/types.h>
15
14struct fiemap_extent { 16struct fiemap_extent {
15 __u64 fe_logical; /* logical offset in bytes for the start of 17 __u64 fe_logical; /* logical offset in bytes for the start of
16 * the extent from the beginning of the file */ 18 * the extent from the beginning of the file */
diff --git a/include/linux/fs.h b/include/linux/fs.h
index 562d2855cf30..e766be0d4329 100644
--- a/include/linux/fs.h
+++ b/include/linux/fs.h
@@ -87,6 +87,60 @@ struct inodes_stat_t {
87 */ 87 */
88#define FMODE_NOCMTIME ((__force fmode_t)2048) 88#define FMODE_NOCMTIME ((__force fmode_t)2048)
89 89
90/*
91 * The below are the various read and write types that we support. Some of
92 * them include behavioral modifiers that send information down to the
93 * block layer and IO scheduler. Terminology:
94 *
95 * The block layer uses device plugging to defer IO a little bit, in
96 * the hope that we will see more IO very shortly. This increases
97 * coalescing of adjacent IO and thus reduces the number of IOs we
98 * have to send to the device. It also allows for better queuing,
99 * if the IO isn't mergeable. If the caller is going to be waiting
100 * for the IO, then he must ensure that the device is unplugged so
101 * that the IO is dispatched to the driver.
102 *
103 * All IO is handled async in Linux. This is fine for background
104 * writes, but for reads or writes that someone waits for completion
105 * on, we want to notify the block layer and IO scheduler so that they
106 * know about it. That allows them to make better scheduling
107 * decisions. So when the below references 'sync' and 'async', it
108 * is referencing this priority hint.
109 *
110 * With that in mind, the available types are:
111 *
112 * READ A normal read operation. Device will be plugged.
113 * READ_SYNC A synchronous read. Device is not plugged, caller can
114 * immediately wait on this read without caring about
115 * unplugging.
116 * READA Used for read-ahead operations. Lower priority, and the
117 * block layer could (in theory) choose to ignore this
118 * request if it runs into resource problems.
119 * WRITE A normal async write. Device will be plugged.
120 * SWRITE Like WRITE, but a special case for ll_rw_block() that
121 * tells it to lock the buffer first. Normally a buffer
122 * must be locked before doing IO.
123 * WRITE_SYNC_PLUG Synchronous write. Identical to WRITE, but passes down
124 * the hint that someone will be waiting on this IO
125 * shortly. The device must still be unplugged explicitly,
126 * WRITE_SYNC_PLUG does not do this as we could be
127 * submitting more writes before we actually wait on any
128 * of them.
129 * WRITE_SYNC Like WRITE_SYNC_PLUG, but also unplugs the device
130 * immediately after submission. The write equivalent
131 * of READ_SYNC.
132 * WRITE_ODIRECT Special case write for O_DIRECT only.
133 * SWRITE_SYNC
134 * SWRITE_SYNC_PLUG Like WRITE_SYNC/WRITE_SYNC_PLUG, but locks the buffer.
135 * See SWRITE.
136 * WRITE_BARRIER Like WRITE, but tells the block layer that all
137 * previously submitted writes must be safely on storage
138 * before this one is started. Also guarantees that when
139 * this write is complete, it itself is also safely on
140 * storage. Prevents reordering of writes on both sides
141 * of this IO.
142 *
143 */
90#define RW_MASK 1 144#define RW_MASK 1
91#define RWA_MASK 2 145#define RWA_MASK 2
92#define READ 0 146#define READ 0
@@ -102,6 +156,11 @@ struct inodes_stat_t {
102 (SWRITE | (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_NOIDLE)) 156 (SWRITE | (1 << BIO_RW_SYNCIO) | (1 << BIO_RW_NOIDLE))
103#define SWRITE_SYNC (SWRITE_SYNC_PLUG | (1 << BIO_RW_UNPLUG)) 157#define SWRITE_SYNC (SWRITE_SYNC_PLUG | (1 << BIO_RW_UNPLUG))
104#define WRITE_BARRIER (WRITE | (1 << BIO_RW_BARRIER)) 158#define WRITE_BARRIER (WRITE | (1 << BIO_RW_BARRIER))
159
160/*
161 * These aren't really reads or writes, they pass down information about
162 * parts of device that are now unused by the file system.
163 */
105#define DISCARD_NOBARRIER (1 << BIO_RW_DISCARD) 164#define DISCARD_NOBARRIER (1 << BIO_RW_DISCARD)
106#define DISCARD_BARRIER ((1 << BIO_RW_DISCARD) | (1 << BIO_RW_BARRIER)) 165#define DISCARD_BARRIER ((1 << BIO_RW_DISCARD) | (1 << BIO_RW_BARRIER))
107 166
@@ -738,9 +797,6 @@ enum inode_i_mutex_lock_class
738 I_MUTEX_QUOTA 797 I_MUTEX_QUOTA
739}; 798};
740 799
741extern void inode_double_lock(struct inode *inode1, struct inode *inode2);
742extern void inode_double_unlock(struct inode *inode1, struct inode *inode2);
743
744/* 800/*
745 * NOTE: in a 32bit arch with a preemptable kernel and 801 * NOTE: in a 32bit arch with a preemptable kernel and
746 * an UP compile the i_size_read/write must be atomic 802 * an UP compile the i_size_read/write must be atomic
@@ -2150,8 +2206,6 @@ extern ssize_t generic_file_splice_read(struct file *, loff_t *,
2150 struct pipe_inode_info *, size_t, unsigned int); 2206 struct pipe_inode_info *, size_t, unsigned int);
2151extern ssize_t generic_file_splice_write(struct pipe_inode_info *, 2207extern ssize_t generic_file_splice_write(struct pipe_inode_info *,
2152 struct file *, loff_t *, size_t, unsigned int); 2208 struct file *, loff_t *, size_t, unsigned int);
2153extern ssize_t generic_file_splice_write_nolock(struct pipe_inode_info *,
2154 struct file *, loff_t *, size_t, unsigned int);
2155extern ssize_t generic_splice_sendpage(struct pipe_inode_info *pipe, 2209extern ssize_t generic_splice_sendpage(struct pipe_inode_info *pipe,
2156 struct file *out, loff_t *, size_t len, unsigned int flags); 2210 struct file *out, loff_t *, size_t len, unsigned int flags);
2157extern long do_splice_direct(struct file *in, loff_t *ppos, struct file *out, 2211extern long do_splice_direct(struct file *in, loff_t *ppos, struct file *out,
diff --git a/include/linux/fsl_devices.h b/include/linux/fsl_devices.h
index f2a78b5e8b55..43fc95d822d5 100644
--- a/include/linux/fsl_devices.h
+++ b/include/linux/fsl_devices.h
@@ -43,10 +43,6 @@
43 * 43 *
44 */ 44 */
45 45
46/* Flags related to I2C device features */
47#define FSL_I2C_DEV_SEPARATE_DFSRR 0x00000001
48#define FSL_I2C_DEV_CLOCK_5200 0x00000002
49
50enum fsl_usb2_operating_modes { 46enum fsl_usb2_operating_modes {
51 FSL_USB2_MPH_HOST, 47 FSL_USB2_MPH_HOST,
52 FSL_USB2_DR_HOST, 48 FSL_USB2_DR_HOST,
diff --git a/include/linux/init_task.h b/include/linux/init_task.h
index dcfb93337e9a..d87247d2641f 100644
--- a/include/linux/init_task.h
+++ b/include/linux/init_task.h
@@ -15,19 +15,6 @@
15extern struct files_struct init_files; 15extern struct files_struct init_files;
16extern struct fs_struct init_fs; 16extern struct fs_struct init_fs;
17 17
18#define INIT_KIOCTX(name, which_mm) \
19{ \
20 .users = ATOMIC_INIT(1), \
21 .dead = 0, \
22 .mm = &which_mm, \
23 .user_id = 0, \
24 .next = NULL, \
25 .wait = __WAIT_QUEUE_HEAD_INITIALIZER(name.wait), \
26 .ctx_lock = __SPIN_LOCK_UNLOCKED(name.ctx_lock), \
27 .reqs_active = 0U, \
28 .max_reqs = ~0U, \
29}
30
31#define INIT_MM(name) \ 18#define INIT_MM(name) \
32{ \ 19{ \
33 .mm_rb = RB_ROOT, \ 20 .mm_rb = RB_ROOT, \
diff --git a/include/linux/libata.h b/include/linux/libata.h
index b450a2628855..3d501db36a26 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -209,6 +209,7 @@ enum {
209 209
210 /* bits 24:31 of ap->flags are reserved for LLD specific flags */ 210 /* bits 24:31 of ap->flags are reserved for LLD specific flags */
211 211
212
212 /* struct ata_port pflags */ 213 /* struct ata_port pflags */
213 ATA_PFLAG_EH_PENDING = (1 << 0), /* EH pending */ 214 ATA_PFLAG_EH_PENDING = (1 << 0), /* EH pending */
214 ATA_PFLAG_EH_IN_PROGRESS = (1 << 1), /* EH in progress */ 215 ATA_PFLAG_EH_IN_PROGRESS = (1 << 1), /* EH in progress */
@@ -225,6 +226,9 @@ enum {
225 ATA_PFLAG_PM_PENDING = (1 << 18), /* PM operation pending */ 226 ATA_PFLAG_PM_PENDING = (1 << 18), /* PM operation pending */
226 ATA_PFLAG_INIT_GTM_VALID = (1 << 19), /* initial gtm data valid */ 227 ATA_PFLAG_INIT_GTM_VALID = (1 << 19), /* initial gtm data valid */
227 228
229 ATA_PFLAG_PIO32 = (1 << 20), /* 32bit PIO */
230 ATA_PFLAG_PIO32CHANGE = (1 << 21), /* 32bit PIO can be turned on/off */
231
228 /* struct ata_queued_cmd flags */ 232 /* struct ata_queued_cmd flags */
229 ATA_QCFLAG_ACTIVE = (1 << 0), /* cmd not yet ack'd to scsi lyer */ 233 ATA_QCFLAG_ACTIVE = (1 << 0), /* cmd not yet ack'd to scsi lyer */
230 ATA_QCFLAG_DMAMAP = (1 << 1), /* SG table is DMA mapped */ 234 ATA_QCFLAG_DMAMAP = (1 << 1), /* SG table is DMA mapped */
@@ -689,7 +693,10 @@ struct ata_port {
689 struct Scsi_Host *scsi_host; /* our co-allocated scsi host */ 693 struct Scsi_Host *scsi_host; /* our co-allocated scsi host */
690 struct ata_port_operations *ops; 694 struct ata_port_operations *ops;
691 spinlock_t *lock; 695 spinlock_t *lock;
696 /* Flags owned by the EH context. Only EH should touch these once the
697 port is active */
692 unsigned long flags; /* ATA_FLAG_xxx */ 698 unsigned long flags; /* ATA_FLAG_xxx */
699 /* Flags that change dynamically, protected by ap->lock */
693 unsigned int pflags; /* ATA_PFLAG_xxx */ 700 unsigned int pflags; /* ATA_PFLAG_xxx */
694 unsigned int print_id; /* user visible unique port ID */ 701 unsigned int print_id; /* user visible unique port ID */
695 unsigned int port_no; /* 0 based port no. inside the host */ 702 unsigned int port_no; /* 0 based port no. inside the host */
@@ -1595,6 +1602,7 @@ extern void ata_sff_drain_fifo(struct ata_queued_cmd *qc);
1595extern void ata_sff_error_handler(struct ata_port *ap); 1602extern void ata_sff_error_handler(struct ata_port *ap);
1596extern void ata_sff_post_internal_cmd(struct ata_queued_cmd *qc); 1603extern void ata_sff_post_internal_cmd(struct ata_queued_cmd *qc);
1597extern int ata_sff_port_start(struct ata_port *ap); 1604extern int ata_sff_port_start(struct ata_port *ap);
1605extern int ata_sff_port_start32(struct ata_port *ap);
1598extern void ata_sff_std_ports(struct ata_ioports *ioaddr); 1606extern void ata_sff_std_ports(struct ata_ioports *ioaddr);
1599extern unsigned long ata_bmdma_mode_filter(struct ata_device *dev, 1607extern unsigned long ata_bmdma_mode_filter(struct ata_device *dev,
1600 unsigned long xfer_mask); 1608 unsigned long xfer_mask);
diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h
index ee98cd570885..06ba90c211a5 100644
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -2514,6 +2514,8 @@
2514#define PCI_DEVICE_ID_INTEL_IOAT_TBG3 0x3433 2514#define PCI_DEVICE_ID_INTEL_IOAT_TBG3 0x3433
2515#define PCI_DEVICE_ID_INTEL_82830_HB 0x3575 2515#define PCI_DEVICE_ID_INTEL_82830_HB 0x3575
2516#define PCI_DEVICE_ID_INTEL_82830_CGC 0x3577 2516#define PCI_DEVICE_ID_INTEL_82830_CGC 0x3577
2517#define PCI_DEVICE_ID_INTEL_82854_HB 0x358c
2518#define PCI_DEVICE_ID_INTEL_82854_IG 0x358e
2517#define PCI_DEVICE_ID_INTEL_82855GM_HB 0x3580 2519#define PCI_DEVICE_ID_INTEL_82855GM_HB 0x3580
2518#define PCI_DEVICE_ID_INTEL_82855GM_IG 0x3582 2520#define PCI_DEVICE_ID_INTEL_82855GM_IG 0x3582
2519#define PCI_DEVICE_ID_INTEL_E7520_MCH 0x3590 2521#define PCI_DEVICE_ID_INTEL_E7520_MCH 0x3590
diff --git a/include/linux/phy.h b/include/linux/phy.h
index 32cf14a4b034..97e40cb6b588 100644
--- a/include/linux/phy.h
+++ b/include/linux/phy.h
@@ -388,6 +388,12 @@ struct phy_driver {
388 /* Enables or disables interrupts */ 388 /* Enables or disables interrupts */
389 int (*config_intr)(struct phy_device *phydev); 389 int (*config_intr)(struct phy_device *phydev);
390 390
391 /*
392 * Checks if the PHY generated an interrupt.
393 * For multi-PHY devices with shared PHY interrupt pin
394 */
395 int (*did_interrupt)(struct phy_device *phydev);
396
391 /* Clears up any memory if needed */ 397 /* Clears up any memory if needed */
392 void (*remove)(struct phy_device *phydev); 398 void (*remove)(struct phy_device *phydev);
393 399
diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h
index 8e4120285f72..c8f038554e80 100644
--- a/include/linux/pipe_fs_i.h
+++ b/include/linux/pipe_fs_i.h
@@ -134,6 +134,11 @@ struct pipe_buf_operations {
134 memory allocation, whereas PIPE_BUF makes atomicity guarantees. */ 134 memory allocation, whereas PIPE_BUF makes atomicity guarantees. */
135#define PIPE_SIZE PAGE_SIZE 135#define PIPE_SIZE PAGE_SIZE
136 136
137/* Pipe lock and unlock operations */
138void pipe_lock(struct pipe_inode_info *);
139void pipe_unlock(struct pipe_inode_info *);
140void pipe_double_lock(struct pipe_inode_info *, struct pipe_inode_info *);
141
137/* Drop the inode semaphore and wait for a pipe event, atomically */ 142/* Drop the inode semaphore and wait for a pipe event, atomically */
138void pipe_wait(struct pipe_inode_info *pipe); 143void pipe_wait(struct pipe_inode_info *pipe);
139 144
diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h
index 0cdda00f2b2a..58b2aa5312b9 100644
--- a/include/linux/rcutree.h
+++ b/include/linux/rcutree.h
@@ -161,9 +161,8 @@ struct rcu_data {
161 unsigned long offline_fqs; /* Kicked due to being offline. */ 161 unsigned long offline_fqs; /* Kicked due to being offline. */
162 unsigned long resched_ipi; /* Sent a resched IPI. */ 162 unsigned long resched_ipi; /* Sent a resched IPI. */
163 163
164 /* 5) state to allow this CPU to force_quiescent_state on others */ 164 /* 5) For future __rcu_pending statistics. */
165 long n_rcu_pending; /* rcu_pending() calls since boot. */ 165 long n_rcu_pending; /* rcu_pending() calls since boot. */
166 long n_rcu_pending_force_qs; /* when to force quiescent states. */
167 166
168 int cpu; 167 int cpu;
169}; 168};
diff --git a/include/linux/sht15.h b/include/linux/sht15.h
new file mode 100644
index 000000000000..046bce05ecab
--- /dev/null
+++ b/include/linux/sht15.h
@@ -0,0 +1,24 @@
1/*
2 * sht15.h - support for the SHT15 Temperature and Humidity Sensor
3 *
4 * Copyright (c) 2009 Jonathan Cameron
5 *
6 * Copyright (c) 2007 Wouter Horre
7 *
8 * This program is free software; you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License version 2 as
10 * published by the Free Software Foundation.
11 */
12
13/**
14 * struct sht15_platform_data - sht15 connectivity info
15 * @gpio_data: no. of gpio to which bidirectional data line is connected
16 * @gpio_sck: no. of gpio to which the data clock is connected.
17 * @supply_mv: supply voltage in mv. Overridden by regulator if available.
18 **/
19struct sht15_platform_data {
20 int gpio_data;
21 int gpio_sck;
22 int supply_mv;
23};
24
diff --git a/include/linux/splice.h b/include/linux/splice.h
index 528dcb93c2f2..5f3faa9d15ae 100644
--- a/include/linux/splice.h
+++ b/include/linux/splice.h
@@ -36,6 +36,8 @@ struct splice_desc {
36 void *data; /* cookie */ 36 void *data; /* cookie */
37 } u; 37 } u;
38 loff_t pos; /* file position */ 38 loff_t pos; /* file position */
39 size_t num_spliced; /* number of bytes already spliced */
40 bool need_wakeup; /* need to wake up writer */
39}; 41};
40 42
41struct partial_page { 43struct partial_page {
@@ -66,6 +68,16 @@ extern ssize_t splice_from_pipe(struct pipe_inode_info *, struct file *,
66 splice_actor *); 68 splice_actor *);
67extern ssize_t __splice_from_pipe(struct pipe_inode_info *, 69extern ssize_t __splice_from_pipe(struct pipe_inode_info *,
68 struct splice_desc *, splice_actor *); 70 struct splice_desc *, splice_actor *);
71extern int splice_from_pipe_feed(struct pipe_inode_info *, struct splice_desc *,
72 splice_actor *);
73extern int splice_from_pipe_next(struct pipe_inode_info *,
74 struct splice_desc *);
75extern void splice_from_pipe_begin(struct splice_desc *);
76extern void splice_from_pipe_end(struct pipe_inode_info *,
77 struct splice_desc *);
78extern int pipe_to_file(struct pipe_inode_info *, struct pipe_buffer *,
79 struct splice_desc *);
80
69extern ssize_t splice_to_pipe(struct pipe_inode_info *, 81extern ssize_t splice_to_pipe(struct pipe_inode_info *,
70 struct splice_pipe_desc *); 82 struct splice_pipe_desc *);
71extern ssize_t splice_direct_to_actor(struct file *, struct splice_desc *, 83extern ssize_t splice_direct_to_actor(struct file *, struct splice_desc *,
diff --git a/include/linux/usb/serial.h b/include/linux/usb/serial.h
index b95842542590..625e9e4639c6 100644
--- a/include/linux/usb/serial.h
+++ b/include/linux/usb/serial.h
@@ -29,7 +29,7 @@
29/** 29/**
30 * usb_serial_port: structure for the specific ports of a device. 30 * usb_serial_port: structure for the specific ports of a device.
31 * @serial: pointer back to the struct usb_serial owner of this port. 31 * @serial: pointer back to the struct usb_serial owner of this port.
32 * @tty: pointer to the corresponding tty for this port. 32 * @port: pointer to the corresponding tty_port for this port.
33 * @lock: spinlock to grab when updating portions of this structure. 33 * @lock: spinlock to grab when updating portions of this structure.
34 * @mutex: mutex used to synchronize serial_open() and serial_close() 34 * @mutex: mutex used to synchronize serial_open() and serial_close()
35 * access for this port. 35 * access for this port.
@@ -44,19 +44,22 @@
44 * @interrupt_out_endpointAddress: endpoint address for the interrupt out pipe 44 * @interrupt_out_endpointAddress: endpoint address for the interrupt out pipe
45 * for this port. 45 * for this port.
46 * @bulk_in_buffer: pointer to the bulk in buffer for this port. 46 * @bulk_in_buffer: pointer to the bulk in buffer for this port.
47 * @bulk_in_size: the size of the bulk_in_buffer, in bytes.
47 * @read_urb: pointer to the bulk in struct urb for this port. 48 * @read_urb: pointer to the bulk in struct urb for this port.
48 * @bulk_in_endpointAddress: endpoint address for the bulk in pipe for this 49 * @bulk_in_endpointAddress: endpoint address for the bulk in pipe for this
49 * port. 50 * port.
50 * @bulk_out_buffer: pointer to the bulk out buffer for this port. 51 * @bulk_out_buffer: pointer to the bulk out buffer for this port.
51 * @bulk_out_size: the size of the bulk_out_buffer, in bytes. 52 * @bulk_out_size: the size of the bulk_out_buffer, in bytes.
52 * @write_urb: pointer to the bulk out struct urb for this port. 53 * @write_urb: pointer to the bulk out struct urb for this port.
54 * @write_urb_busy: port`s writing status
53 * @bulk_out_endpointAddress: endpoint address for the bulk out pipe for this 55 * @bulk_out_endpointAddress: endpoint address for the bulk out pipe for this
54 * port. 56 * port.
55 * @write_wait: a wait_queue_head_t used by the port. 57 * @write_wait: a wait_queue_head_t used by the port.
56 * @work: work queue entry for the line discipline waking up. 58 * @work: work queue entry for the line discipline waking up.
57 * @open_count: number of times this port has been opened.
58 * @throttled: nonzero if the read urb is inactive to throttle the device 59 * @throttled: nonzero if the read urb is inactive to throttle the device
59 * @throttle_req: nonzero if the tty wants to throttle us 60 * @throttle_req: nonzero if the tty wants to throttle us
61 * @console: attached usb serial console
62 * @dev: pointer to the serial device
60 * 63 *
61 * This structure is used by the usb-serial core and drivers for the specific 64 * This structure is used by the usb-serial core and drivers for the specific
62 * ports of a device. 65 * ports of a device.
diff --git a/include/net/udp.h b/include/net/udp.h
index 93dbe294d459..90e6ce56be65 100644
--- a/include/net/udp.h
+++ b/include/net/udp.h
@@ -124,8 +124,6 @@ static inline void udp_lib_close(struct sock *sk, long timeout)
124 sk_common_release(sk); 124 sk_common_release(sk);
125} 125}
126 126
127extern int ipv4_rcv_saddr_equal(const struct sock *sk1,
128 const struct sock *sk2);
129extern int udp_lib_get_port(struct sock *sk, unsigned short snum, 127extern int udp_lib_get_port(struct sock *sk, unsigned short snum,
130 int (*)(const struct sock*,const struct sock*)); 128 int (*)(const struct sock*,const struct sock*));
131 129
diff --git a/include/sound/jack.h b/include/sound/jack.h
index 6b013c6f6a04..f236e426a706 100644
--- a/include/sound/jack.h
+++ b/include/sound/jack.h
@@ -50,6 +50,8 @@ struct snd_jack {
50 int type; 50 int type;
51 const char *id; 51 const char *id;
52 char name[100]; 52 char name[100];
53 void *private_data;
54 void (*private_free)(struct snd_jack *);
53}; 55};
54 56
55#ifdef CONFIG_SND_JACK 57#ifdef CONFIG_SND_JACK
diff --git a/include/sound/pcm.h b/include/sound/pcm.h
index 8904b1900d7f..c17296891617 100644
--- a/include/sound/pcm.h
+++ b/include/sound/pcm.h
@@ -268,7 +268,8 @@ struct snd_pcm_runtime {
268 int overrange; 268 int overrange;
269 snd_pcm_uframes_t avail_max; 269 snd_pcm_uframes_t avail_max;
270 snd_pcm_uframes_t hw_ptr_base; /* Position at buffer restart */ 270 snd_pcm_uframes_t hw_ptr_base; /* Position at buffer restart */
271 snd_pcm_uframes_t hw_ptr_interrupt; /* Position at interrupt time*/ 271 snd_pcm_uframes_t hw_ptr_interrupt; /* Position at interrupt time */
272 unsigned long hw_ptr_jiffies; /* Time when hw_ptr is updated */
272 273
273 /* -- HW params -- */ 274 /* -- HW params -- */
274 snd_pcm_access_t access; /* access mode */ 275 snd_pcm_access_t access; /* access mode */
diff --git a/include/video/cyblafb.h b/include/video/cyblafb.h
deleted file mode 100644
index d3c1d4e2c8e3..000000000000
--- a/include/video/cyblafb.h
+++ /dev/null
@@ -1,175 +0,0 @@
1
2#ifndef CYBLAFB_DEBUG
3#define CYBLAFB_DEBUG 0
4#endif
5
6#if CYBLAFB_DEBUG
7#define debug(f,a...) printk("%s:" f, __func__ , ## a);
8#else
9#define debug(f,a...)
10#endif
11
12#define output(f, a...) printk("cyblafb: " f, ## a)
13
14#define Kb (1024)
15#define Mb (Kb*Kb)
16
17/* PCI IDS of supported cards temporarily here */
18
19#define CYBERBLADEi1 0x8500
20
21/* these defines are for 'lcd' variable */
22#define LCD_STRETCH 0
23#define LCD_CENTER 1
24#define LCD_BIOS 2
25
26/* display types */
27#define DISPLAY_CRT 0
28#define DISPLAY_FP 1
29
30#define ROP_S 0xCC
31
32#define point(x,y) ((y)<<16|(x))
33
34//
35// Attribute Regs, ARxx, 3c0/3c1
36//
37#define AR00 0x00
38#define AR01 0x01
39#define AR02 0x02
40#define AR03 0x03
41#define AR04 0x04
42#define AR05 0x05
43#define AR06 0x06
44#define AR07 0x07
45#define AR08 0x08
46#define AR09 0x09
47#define AR0A 0x0A
48#define AR0B 0x0B
49#define AR0C 0x0C
50#define AR0D 0x0D
51#define AR0E 0x0E
52#define AR0F 0x0F
53#define AR10 0x10
54#define AR12 0x12
55#define AR13 0x13
56
57//
58// Sequencer Regs, SRxx, 3c4/3c5
59//
60#define SR00 0x00
61#define SR01 0x01
62#define SR02 0x02
63#define SR03 0x03
64#define SR04 0x04
65#define SR0D 0x0D
66#define SR0E 0x0E
67#define SR11 0x11
68#define SR18 0x18
69#define SR19 0x19
70
71//
72//
73//
74#define CR00 0x00
75#define CR01 0x01
76#define CR02 0x02
77#define CR03 0x03
78#define CR04 0x04
79#define CR05 0x05
80#define CR06 0x06
81#define CR07 0x07
82#define CR08 0x08
83#define CR09 0x09
84#define CR0A 0x0A
85#define CR0B 0x0B
86#define CR0C 0x0C
87#define CR0D 0x0D
88#define CR0E 0x0E
89#define CR0F 0x0F
90#define CR10 0x10
91#define CR11 0x11
92#define CR12 0x12
93#define CR13 0x13
94#define CR14 0x14
95#define CR15 0x15
96#define CR16 0x16
97#define CR17 0x17
98#define CR18 0x18
99#define CR19 0x19
100#define CR1A 0x1A
101#define CR1B 0x1B
102#define CR1C 0x1C
103#define CR1D 0x1D
104#define CR1E 0x1E
105#define CR1F 0x1F
106#define CR20 0x20
107#define CR21 0x21
108#define CR27 0x27
109#define CR29 0x29
110#define CR2A 0x2A
111#define CR2B 0x2B
112#define CR2D 0x2D
113#define CR2F 0x2F
114#define CR36 0x36
115#define CR38 0x38
116#define CR39 0x39
117#define CR3A 0x3A
118#define CR55 0x55
119#define CR56 0x56
120#define CR57 0x57
121#define CR58 0x58
122
123//
124//
125//
126
127#define GR00 0x01
128#define GR01 0x01
129#define GR02 0x02
130#define GR03 0x03
131#define GR04 0x04
132#define GR05 0x05
133#define GR06 0x06
134#define GR07 0x07
135#define GR08 0x08
136#define GR0F 0x0F
137#define GR20 0x20
138#define GR23 0x23
139#define GR2F 0x2F
140#define GR30 0x30
141#define GR31 0x31
142#define GR33 0x33
143#define GR52 0x52
144#define GR53 0x53
145#define GR5D 0x5d
146
147
148//
149// Graphics Engine
150//
151#define GEBase 0x2100 // could be mapped elsewhere if we like it
152#define GE00 (GEBase+0x00) // source 1, p 111
153#define GE04 (GEBase+0x04) // source 2, p 111
154#define GE08 (GEBase+0x08) // destination 1, p 111
155#define GE0C (GEBase+0x0C) // destination 2, p 112
156#define GE10 (GEBase+0x10) // right view base & enable, p 112
157#define GE13 (GEBase+0x13) // left view base & enable, p 112
158#define GE18 (GEBase+0x18) // block write start address, p 112
159#define GE1C (GEBase+0x1C) // block write end address, p 112
160#define GE20 (GEBase+0x20) // engine status, p 113
161#define GE24 (GEBase+0x24) // reset all GE pointers
162#define GE44 (GEBase+0x44) // command register, p 126
163#define GE48 (GEBase+0x48) // raster operation, p 127
164#define GE60 (GEBase+0x60) // foreground color, p 128
165#define GE64 (GEBase+0x64) // background color, p 128
166#define GE6C (GEBase+0x6C) // Pattern and Style, p 129, ok
167#define GE9C (GEBase+0x9C) // pixel engine data port, p 125
168#define GEB8 (GEBase+0xB8) // Destination Stride / Buffer Base 0, p 133
169#define GEBC (GEBase+0xBC) // Destination Stride / Buffer Base 1, p 133
170#define GEC0 (GEBase+0xC0) // Destination Stride / Buffer Base 2, p 133
171#define GEC4 (GEBase+0xC4) // Destination Stride / Buffer Base 3, p 133
172#define GEC8 (GEBase+0xC8) // Source Stride / Buffer Base 0, p 133
173#define GECC (GEBase+0xCC) // Source Stride / Buffer Base 1, p 133
174#define GED0 (GEBase+0xD0) // Source Stride / Buffer Base 2, p 133
175#define GED4 (GEBase+0xD4) // Source Stride / Buffer Base 3, p 133
diff --git a/init/initramfs.c b/init/initramfs.c
index 80cd713f6cc5..9ee7b7810417 100644
--- a/init/initramfs.c
+++ b/init/initramfs.c
@@ -310,7 +310,8 @@ static int __init do_name(void)
310 if (wfd >= 0) { 310 if (wfd >= 0) {
311 sys_fchown(wfd, uid, gid); 311 sys_fchown(wfd, uid, gid);
312 sys_fchmod(wfd, mode); 312 sys_fchmod(wfd, mode);
313 sys_ftruncate(wfd, body_len); 313 if (body_len)
314 sys_ftruncate(wfd, body_len);
314 vcollected = kstrdup(collected, GFP_KERNEL); 315 vcollected = kstrdup(collected, GFP_KERNEL);
315 state = CopyFile; 316 state = CopyFile;
316 } 317 }
@@ -515,6 +516,7 @@ skip:
515 initrd_end = 0; 516 initrd_end = 0;
516} 517}
517 518
519#ifdef CONFIG_BLK_DEV_RAM
518#define BUF_SIZE 1024 520#define BUF_SIZE 1024
519static void __init clean_rootfs(void) 521static void __init clean_rootfs(void)
520{ 522{
@@ -561,6 +563,7 @@ static void __init clean_rootfs(void)
561 sys_close(fd); 563 sys_close(fd);
562 kfree(buf); 564 kfree(buf);
563} 565}
566#endif
564 567
565static int __init populate_rootfs(void) 568static int __init populate_rootfs(void)
566{ 569{
diff --git a/ipc/mq_sysctl.c b/ipc/mq_sysctl.c
index 89f60ec8ee54..24ae46dfe45d 100644
--- a/ipc/mq_sysctl.c
+++ b/ipc/mq_sysctl.c
@@ -22,6 +22,7 @@
22#define MIN_MSGSIZEMAX 128 /* min value for msgsize_max */ 22#define MIN_MSGSIZEMAX 128 /* min value for msgsize_max */
23#define MAX_MSGSIZEMAX (8192*128) /* max value for msgsize_max */ 23#define MAX_MSGSIZEMAX (8192*128) /* max value for msgsize_max */
24 24
25#ifdef CONFIG_PROC_SYSCTL
25static void *get_mq(ctl_table *table) 26static void *get_mq(ctl_table *table)
26{ 27{
27 char *which = table->data; 28 char *which = table->data;
@@ -30,7 +31,6 @@ static void *get_mq(ctl_table *table)
30 return which; 31 return which;
31} 32}
32 33
33#ifdef CONFIG_PROC_SYSCTL
34static int proc_mq_dointvec(ctl_table *table, int write, struct file *filp, 34static int proc_mq_dointvec(ctl_table *table, int write, struct file *filp,
35 void __user *buffer, size_t *lenp, loff_t *ppos) 35 void __user *buffer, size_t *lenp, loff_t *ppos)
36{ 36{
diff --git a/kernel/power/swap.c b/kernel/power/swap.c
index 505f319e489c..8ba052c86d48 100644
--- a/kernel/power/swap.c
+++ b/kernel/power/swap.c
@@ -64,8 +64,6 @@ static int submit(int rw, pgoff_t page_off, struct page *page,
64 struct bio *bio; 64 struct bio *bio;
65 65
66 bio = bio_alloc(__GFP_WAIT | __GFP_HIGH, 1); 66 bio = bio_alloc(__GFP_WAIT | __GFP_HIGH, 1);
67 if (!bio)
68 return -ENOMEM;
69 bio->bi_sector = page_off * (PAGE_SIZE >> 9); 67 bio->bi_sector = page_off * (PAGE_SIZE >> 9);
70 bio->bi_bdev = resume_bdev; 68 bio->bi_bdev = resume_bdev;
71 bio->bi_end_io = end_swap_bio_read; 69 bio->bi_end_io = end_swap_bio_read;
diff --git a/kernel/ptrace.c b/kernel/ptrace.c
index 64191fa09b7e..dfcd83ceee3b 100644
--- a/kernel/ptrace.c
+++ b/kernel/ptrace.c
@@ -604,10 +604,11 @@ repeat:
604 ret = security_ptrace_traceme(current->parent); 604 ret = security_ptrace_traceme(current->parent);
605 605
606 /* 606 /*
607 * Set the ptrace bit in the process ptrace flags. 607 * Check PF_EXITING to ensure ->real_parent has not passed
608 * Then link us on our parent's ptraced list. 608 * exit_ptrace(). Otherwise we don't report the error but
609 * pretend ->real_parent untraces us right after return.
609 */ 610 */
610 if (!ret) { 611 if (!ret && !(current->real_parent->flags & PF_EXITING)) {
611 current->ptrace |= PT_PTRACED; 612 current->ptrace |= PT_PTRACED;
612 __ptrace_link(current, current->real_parent); 613 __ptrace_link(current, current->real_parent);
613 } 614 }
diff --git a/kernel/rcupdate.c b/kernel/rcupdate.c
index 2c7b8457d0d2..a967c9feb90a 100644
--- a/kernel/rcupdate.c
+++ b/kernel/rcupdate.c
@@ -58,6 +58,10 @@ static DEFINE_MUTEX(rcu_barrier_mutex);
58static struct completion rcu_barrier_completion; 58static struct completion rcu_barrier_completion;
59int rcu_scheduler_active __read_mostly; 59int rcu_scheduler_active __read_mostly;
60 60
61static atomic_t rcu_migrate_type_count = ATOMIC_INIT(0);
62static struct rcu_head rcu_migrate_head[3];
63static DECLARE_WAIT_QUEUE_HEAD(rcu_migrate_wq);
64
61/* 65/*
62 * Awaken the corresponding synchronize_rcu() instance now that a 66 * Awaken the corresponding synchronize_rcu() instance now that a
63 * grace period has elapsed. 67 * grace period has elapsed.
@@ -122,7 +126,10 @@ static void rcu_barrier_func(void *type)
122 } 126 }
123} 127}
124 128
125static inline void wait_migrated_callbacks(void); 129static inline void wait_migrated_callbacks(void)
130{
131 wait_event(rcu_migrate_wq, !atomic_read(&rcu_migrate_type_count));
132}
126 133
127/* 134/*
128 * Orchestrate the specified type of RCU barrier, waiting for all 135 * Orchestrate the specified type of RCU barrier, waiting for all
@@ -179,21 +186,12 @@ void rcu_barrier_sched(void)
179} 186}
180EXPORT_SYMBOL_GPL(rcu_barrier_sched); 187EXPORT_SYMBOL_GPL(rcu_barrier_sched);
181 188
182static atomic_t rcu_migrate_type_count = ATOMIC_INIT(0);
183static struct rcu_head rcu_migrate_head[3];
184static DECLARE_WAIT_QUEUE_HEAD(rcu_migrate_wq);
185
186static void rcu_migrate_callback(struct rcu_head *notused) 189static void rcu_migrate_callback(struct rcu_head *notused)
187{ 190{
188 if (atomic_dec_and_test(&rcu_migrate_type_count)) 191 if (atomic_dec_and_test(&rcu_migrate_type_count))
189 wake_up(&rcu_migrate_wq); 192 wake_up(&rcu_migrate_wq);
190} 193}
191 194
192static inline void wait_migrated_callbacks(void)
193{
194 wait_event(rcu_migrate_wq, !atomic_read(&rcu_migrate_type_count));
195}
196
197static int __cpuinit rcu_barrier_cpu_hotplug(struct notifier_block *self, 195static int __cpuinit rcu_barrier_cpu_hotplug(struct notifier_block *self,
198 unsigned long action, void *hcpu) 196 unsigned long action, void *hcpu)
199{ 197{
diff --git a/kernel/rcutree.c b/kernel/rcutree.c
index 7f3266922572..d2a372fb0b9b 100644
--- a/kernel/rcutree.c
+++ b/kernel/rcutree.c
@@ -530,8 +530,6 @@ static void note_new_gpnum(struct rcu_state *rsp, struct rcu_data *rdp)
530 rdp->qs_pending = 1; 530 rdp->qs_pending = 1;
531 rdp->passed_quiesc = 0; 531 rdp->passed_quiesc = 0;
532 rdp->gpnum = rsp->gpnum; 532 rdp->gpnum = rsp->gpnum;
533 rdp->n_rcu_pending_force_qs = rdp->n_rcu_pending +
534 RCU_JIFFIES_TILL_FORCE_QS;
535} 533}
536 534
537/* 535/*
@@ -578,8 +576,6 @@ rcu_start_gp(struct rcu_state *rsp, unsigned long flags)
578 rsp->gpnum++; 576 rsp->gpnum++;
579 rsp->signaled = RCU_GP_INIT; /* Hold off force_quiescent_state. */ 577 rsp->signaled = RCU_GP_INIT; /* Hold off force_quiescent_state. */
580 rsp->jiffies_force_qs = jiffies + RCU_JIFFIES_TILL_FORCE_QS; 578 rsp->jiffies_force_qs = jiffies + RCU_JIFFIES_TILL_FORCE_QS;
581 rdp->n_rcu_pending_force_qs = rdp->n_rcu_pending +
582 RCU_JIFFIES_TILL_FORCE_QS;
583 record_gp_stall_check_time(rsp); 579 record_gp_stall_check_time(rsp);
584 dyntick_record_completed(rsp, rsp->completed - 1); 580 dyntick_record_completed(rsp, rsp->completed - 1);
585 note_new_gpnum(rsp, rdp); 581 note_new_gpnum(rsp, rdp);
@@ -1055,7 +1051,6 @@ static void force_quiescent_state(struct rcu_state *rsp, int relaxed)
1055{ 1051{
1056 unsigned long flags; 1052 unsigned long flags;
1057 long lastcomp; 1053 long lastcomp;
1058 struct rcu_data *rdp = rsp->rda[smp_processor_id()];
1059 struct rcu_node *rnp = rcu_get_root(rsp); 1054 struct rcu_node *rnp = rcu_get_root(rsp);
1060 u8 signaled; 1055 u8 signaled;
1061 1056
@@ -1066,16 +1061,13 @@ static void force_quiescent_state(struct rcu_state *rsp, int relaxed)
1066 return; /* Someone else is already on the job. */ 1061 return; /* Someone else is already on the job. */
1067 } 1062 }
1068 if (relaxed && 1063 if (relaxed &&
1069 (long)(rsp->jiffies_force_qs - jiffies) >= 0 && 1064 (long)(rsp->jiffies_force_qs - jiffies) >= 0)
1070 (rdp->n_rcu_pending_force_qs - rdp->n_rcu_pending) >= 0)
1071 goto unlock_ret; /* no emergency and done recently. */ 1065 goto unlock_ret; /* no emergency and done recently. */
1072 rsp->n_force_qs++; 1066 rsp->n_force_qs++;
1073 spin_lock(&rnp->lock); 1067 spin_lock(&rnp->lock);
1074 lastcomp = rsp->completed; 1068 lastcomp = rsp->completed;
1075 signaled = rsp->signaled; 1069 signaled = rsp->signaled;
1076 rsp->jiffies_force_qs = jiffies + RCU_JIFFIES_TILL_FORCE_QS; 1070 rsp->jiffies_force_qs = jiffies + RCU_JIFFIES_TILL_FORCE_QS;
1077 rdp->n_rcu_pending_force_qs = rdp->n_rcu_pending +
1078 RCU_JIFFIES_TILL_FORCE_QS;
1079 if (lastcomp == rsp->gpnum) { 1071 if (lastcomp == rsp->gpnum) {
1080 rsp->n_force_qs_ngp++; 1072 rsp->n_force_qs_ngp++;
1081 spin_unlock(&rnp->lock); 1073 spin_unlock(&rnp->lock);
@@ -1144,8 +1136,7 @@ __rcu_process_callbacks(struct rcu_state *rsp, struct rcu_data *rdp)
1144 * If an RCU GP has gone long enough, go check for dyntick 1136 * If an RCU GP has gone long enough, go check for dyntick
1145 * idle CPUs and, if needed, send resched IPIs. 1137 * idle CPUs and, if needed, send resched IPIs.
1146 */ 1138 */
1147 if ((long)(ACCESS_ONCE(rsp->jiffies_force_qs) - jiffies) < 0 || 1139 if ((long)(ACCESS_ONCE(rsp->jiffies_force_qs) - jiffies) < 0)
1148 (rdp->n_rcu_pending_force_qs - rdp->n_rcu_pending) < 0)
1149 force_quiescent_state(rsp, 1); 1140 force_quiescent_state(rsp, 1);
1150 1141
1151 /* 1142 /*
@@ -1230,8 +1221,7 @@ __call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu),
1230 if (unlikely(++rdp->qlen > qhimark)) { 1221 if (unlikely(++rdp->qlen > qhimark)) {
1231 rdp->blimit = LONG_MAX; 1222 rdp->blimit = LONG_MAX;
1232 force_quiescent_state(rsp, 0); 1223 force_quiescent_state(rsp, 0);
1233 } else if ((long)(ACCESS_ONCE(rsp->jiffies_force_qs) - jiffies) < 0 || 1224 } else if ((long)(ACCESS_ONCE(rsp->jiffies_force_qs) - jiffies) < 0)
1234 (rdp->n_rcu_pending_force_qs - rdp->n_rcu_pending) < 0)
1235 force_quiescent_state(rsp, 1); 1225 force_quiescent_state(rsp, 1);
1236 local_irq_restore(flags); 1226 local_irq_restore(flags);
1237} 1227}
@@ -1290,8 +1280,7 @@ static int __rcu_pending(struct rcu_state *rsp, struct rcu_data *rdp)
1290 1280
1291 /* Has an RCU GP gone long enough to send resched IPIs &c? */ 1281 /* Has an RCU GP gone long enough to send resched IPIs &c? */
1292 if (ACCESS_ONCE(rsp->completed) != ACCESS_ONCE(rsp->gpnum) && 1282 if (ACCESS_ONCE(rsp->completed) != ACCESS_ONCE(rsp->gpnum) &&
1293 ((long)(ACCESS_ONCE(rsp->jiffies_force_qs) - jiffies) < 0 || 1283 ((long)(ACCESS_ONCE(rsp->jiffies_force_qs) - jiffies) < 0))
1294 (rdp->n_rcu_pending_force_qs - rdp->n_rcu_pending) < 0))
1295 return 1; 1284 return 1;
1296 1285
1297 /* nothing to do */ 1286 /* nothing to do */
diff --git a/kernel/rcutree_trace.c b/kernel/rcutree_trace.c
index 4ee954f6a8d5..4b1875ba9404 100644
--- a/kernel/rcutree_trace.c
+++ b/kernel/rcutree_trace.c
@@ -49,14 +49,12 @@ static void print_one_rcu_data(struct seq_file *m, struct rcu_data *rdp)
49{ 49{
50 if (!rdp->beenonline) 50 if (!rdp->beenonline)
51 return; 51 return;
52 seq_printf(m, "%3d%cc=%ld g=%ld pq=%d pqc=%ld qp=%d rpfq=%ld rp=%x", 52 seq_printf(m, "%3d%cc=%ld g=%ld pq=%d pqc=%ld qp=%d",
53 rdp->cpu, 53 rdp->cpu,
54 cpu_is_offline(rdp->cpu) ? '!' : ' ', 54 cpu_is_offline(rdp->cpu) ? '!' : ' ',
55 rdp->completed, rdp->gpnum, 55 rdp->completed, rdp->gpnum,
56 rdp->passed_quiesc, rdp->passed_quiesc_completed, 56 rdp->passed_quiesc, rdp->passed_quiesc_completed,
57 rdp->qs_pending, 57 rdp->qs_pending);
58 rdp->n_rcu_pending_force_qs - rdp->n_rcu_pending,
59 (int)(rdp->n_rcu_pending & 0xffff));
60#ifdef CONFIG_NO_HZ 58#ifdef CONFIG_NO_HZ
61 seq_printf(m, " dt=%d/%d dn=%d df=%lu", 59 seq_printf(m, " dt=%d/%d dn=%d df=%lu",
62 rdp->dynticks->dynticks, 60 rdp->dynticks->dynticks,
@@ -102,14 +100,12 @@ static void print_one_rcu_data_csv(struct seq_file *m, struct rcu_data *rdp)
102{ 100{
103 if (!rdp->beenonline) 101 if (!rdp->beenonline)
104 return; 102 return;
105 seq_printf(m, "%d,%s,%ld,%ld,%d,%ld,%d,%ld,%ld", 103 seq_printf(m, "%d,%s,%ld,%ld,%d,%ld,%d",
106 rdp->cpu, 104 rdp->cpu,
107 cpu_is_offline(rdp->cpu) ? "\"Y\"" : "\"N\"", 105 cpu_is_offline(rdp->cpu) ? "\"Y\"" : "\"N\"",
108 rdp->completed, rdp->gpnum, 106 rdp->completed, rdp->gpnum,
109 rdp->passed_quiesc, rdp->passed_quiesc_completed, 107 rdp->passed_quiesc, rdp->passed_quiesc_completed,
110 rdp->qs_pending, 108 rdp->qs_pending);
111 rdp->n_rcu_pending_force_qs - rdp->n_rcu_pending,
112 rdp->n_rcu_pending);
113#ifdef CONFIG_NO_HZ 109#ifdef CONFIG_NO_HZ
114 seq_printf(m, ",%d,%d,%d,%lu", 110 seq_printf(m, ",%d,%d,%d,%lu",
115 rdp->dynticks->dynticks, 111 rdp->dynticks->dynticks,
@@ -123,7 +119,7 @@ static void print_one_rcu_data_csv(struct seq_file *m, struct rcu_data *rdp)
123 119
124static int show_rcudata_csv(struct seq_file *m, void *unused) 120static int show_rcudata_csv(struct seq_file *m, void *unused)
125{ 121{
126 seq_puts(m, "\"CPU\",\"Online?\",\"c\",\"g\",\"pq\",\"pqc\",\"pq\",\"rpfq\",\"rp\","); 122 seq_puts(m, "\"CPU\",\"Online?\",\"c\",\"g\",\"pq\",\"pqc\",\"pq\",");
127#ifdef CONFIG_NO_HZ 123#ifdef CONFIG_NO_HZ
128 seq_puts(m, "\"dt\",\"dt nesting\",\"dn\",\"df\","); 124 seq_puts(m, "\"dt\",\"dt nesting\",\"dn\",\"df\",");
129#endif /* #ifdef CONFIG_NO_HZ */ 125#endif /* #ifdef CONFIG_NO_HZ */
diff --git a/kernel/sched.c b/kernel/sched.c
index 5724508c3b66..b902e587a3a0 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -4846,7 +4846,7 @@ void scheduler_tick(void)
4846#endif 4846#endif
4847} 4847}
4848 4848
4849unsigned long get_parent_ip(unsigned long addr) 4849notrace unsigned long get_parent_ip(unsigned long addr)
4850{ 4850{
4851 if (in_lock_functions(addr)) { 4851 if (in_lock_functions(addr)) {
4852 addr = CALLER_ADDR2; 4852 addr = CALLER_ADDR2;
@@ -7367,8 +7367,12 @@ static int sched_domain_debug_one(struct sched_domain *sd, int cpu, int level,
7367 cpumask_or(groupmask, groupmask, sched_group_cpus(group)); 7367 cpumask_or(groupmask, groupmask, sched_group_cpus(group));
7368 7368
7369 cpulist_scnprintf(str, sizeof(str), sched_group_cpus(group)); 7369 cpulist_scnprintf(str, sizeof(str), sched_group_cpus(group));
7370 printk(KERN_CONT " %s (__cpu_power = %d)", str, 7370
7371 group->__cpu_power); 7371 printk(KERN_CONT " %s", str);
7372 if (group->__cpu_power != SCHED_LOAD_SCALE) {
7373 printk(KERN_CONT " (__cpu_power = %d)",
7374 group->__cpu_power);
7375 }
7372 7376
7373 group = group->next; 7377 group = group->next;
7374 } while (group != sd->groups); 7378 } while (group != sd->groups);
diff --git a/kernel/softirq.c b/kernel/softirq.c
index 2fecefacdc5b..b525dd348511 100644
--- a/kernel/softirq.c
+++ b/kernel/softirq.c
@@ -472,9 +472,9 @@ void tasklet_kill(struct tasklet_struct *t)
472 printk("Attempt to kill tasklet from interrupt\n"); 472 printk("Attempt to kill tasklet from interrupt\n");
473 473
474 while (test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) { 474 while (test_and_set_bit(TASKLET_STATE_SCHED, &t->state)) {
475 do 475 do {
476 yield(); 476 yield();
477 while (test_bit(TASKLET_STATE_SCHED, &t->state)); 477 } while (test_bit(TASKLET_STATE_SCHED, &t->state));
478 } 478 }
479 tasklet_unlock_wait(t); 479 tasklet_unlock_wait(t);
480 clear_bit(TASKLET_STATE_SCHED, &t->state); 480 clear_bit(TASKLET_STATE_SCHED, &t->state);
diff --git a/kernel/sys.c b/kernel/sys.c
index 51dbb55604e8..e7998cf31498 100644
--- a/kernel/sys.c
+++ b/kernel/sys.c
@@ -360,6 +360,7 @@ SYSCALL_DEFINE4(reboot, int, magic1, int, magic2, unsigned int, cmd,
360 void __user *, arg) 360 void __user *, arg)
361{ 361{
362 char buffer[256]; 362 char buffer[256];
363 int ret = 0;
363 364
364 /* We only trust the superuser with rebooting the system. */ 365 /* We only trust the superuser with rebooting the system. */
365 if (!capable(CAP_SYS_BOOT)) 366 if (!capable(CAP_SYS_BOOT))
@@ -397,7 +398,7 @@ SYSCALL_DEFINE4(reboot, int, magic1, int, magic2, unsigned int, cmd,
397 kernel_halt(); 398 kernel_halt();
398 unlock_kernel(); 399 unlock_kernel();
399 do_exit(0); 400 do_exit(0);
400 break; 401 panic("cannot halt");
401 402
402 case LINUX_REBOOT_CMD_POWER_OFF: 403 case LINUX_REBOOT_CMD_POWER_OFF:
403 kernel_power_off(); 404 kernel_power_off();
@@ -417,29 +418,22 @@ SYSCALL_DEFINE4(reboot, int, magic1, int, magic2, unsigned int, cmd,
417 418
418#ifdef CONFIG_KEXEC 419#ifdef CONFIG_KEXEC
419 case LINUX_REBOOT_CMD_KEXEC: 420 case LINUX_REBOOT_CMD_KEXEC:
420 { 421 ret = kernel_kexec();
421 int ret; 422 break;
422 ret = kernel_kexec();
423 unlock_kernel();
424 return ret;
425 }
426#endif 423#endif
427 424
428#ifdef CONFIG_HIBERNATION 425#ifdef CONFIG_HIBERNATION
429 case LINUX_REBOOT_CMD_SW_SUSPEND: 426 case LINUX_REBOOT_CMD_SW_SUSPEND:
430 { 427 ret = hibernate();
431 int ret = hibernate(); 428 break;
432 unlock_kernel();
433 return ret;
434 }
435#endif 429#endif
436 430
437 default: 431 default:
438 unlock_kernel(); 432 ret = -EINVAL;
439 return -EINVAL; 433 break;
440 } 434 }
441 unlock_kernel(); 435 unlock_kernel();
442 return 0; 436 return ret;
443} 437}
444 438
445static void deferred_cad(struct work_struct *dummy) 439static void deferred_cad(struct work_struct *dummy)
diff --git a/kernel/sysctl.c b/kernel/sysctl.c
index 4286b62b34a0..e3d2c7dd59b9 100644
--- a/kernel/sysctl.c
+++ b/kernel/sysctl.c
@@ -902,16 +902,6 @@ static struct ctl_table kern_table[] = {
902 .proc_handler = &proc_dointvec, 902 .proc_handler = &proc_dointvec,
903 }, 903 },
904#endif 904#endif
905#ifdef CONFIG_UNEVICTABLE_LRU
906 {
907 .ctl_name = CTL_UNNUMBERED,
908 .procname = "scan_unevictable_pages",
909 .data = &scan_unevictable_pages,
910 .maxlen = sizeof(scan_unevictable_pages),
911 .mode = 0644,
912 .proc_handler = &scan_unevictable_handler,
913 },
914#endif
915#ifdef CONFIG_SLOW_WORK 905#ifdef CONFIG_SLOW_WORK
916 { 906 {
917 .ctl_name = CTL_UNNUMBERED, 907 .ctl_name = CTL_UNNUMBERED,
@@ -1302,6 +1292,16 @@ static struct ctl_table vm_table[] = {
1302 .extra2 = &one, 1292 .extra2 = &one,
1303 }, 1293 },
1304#endif 1294#endif
1295#ifdef CONFIG_UNEVICTABLE_LRU
1296 {
1297 .ctl_name = CTL_UNNUMBERED,
1298 .procname = "scan_unevictable_pages",
1299 .data = &scan_unevictable_pages,
1300 .maxlen = sizeof(scan_unevictable_pages),
1301 .mode = 0644,
1302 .proc_handler = &scan_unevictable_handler,
1303 },
1304#endif
1305/* 1305/*
1306 * NOTE: do not add new entries to this table unless you have read 1306 * NOTE: do not add new entries to this table unless you have read
1307 * Documentation/sysctl/ctl_unnumbered.txt 1307 * Documentation/sysctl/ctl_unnumbered.txt
diff --git a/lib/kobject_uevent.c b/lib/kobject_uevent.c
index 97a777ad4f59..dafeecf5b143 100644
--- a/lib/kobject_uevent.c
+++ b/lib/kobject_uevent.c
@@ -258,7 +258,7 @@ int kobject_uevent_env(struct kobject *kobj, enum kobject_action action,
258 goto exit; 258 goto exit;
259 259
260 retval = call_usermodehelper(argv[0], argv, 260 retval = call_usermodehelper(argv[0], argv,
261 env->envp, UMH_NO_WAIT); 261 env->envp, UMH_WAIT_EXEC);
262 } 262 }
263 263
264exit: 264exit:
diff --git a/mm/Kconfig b/mm/Kconfig
index b53427ad30a3..57971d2ab848 100644
--- a/mm/Kconfig
+++ b/mm/Kconfig
@@ -213,6 +213,8 @@ config UNEVICTABLE_LRU
213 will use one page flag and increase the code size a little, 213 will use one page flag and increase the code size a little,
214 say Y unless you know what you are doing. 214 say Y unless you know what you are doing.
215 215
216 See Documentation/vm/unevictable-lru.txt for more information.
217
216config HAVE_MLOCK 218config HAVE_MLOCK
217 bool 219 bool
218 default y if MMU=y 220 default y if MMU=y
diff --git a/mm/filemap.c b/mm/filemap.c
index 2e2d38ebda4b..379ff0bcbf6e 100644
--- a/mm/filemap.c
+++ b/mm/filemap.c
@@ -441,6 +441,7 @@ int filemap_write_and_wait_range(struct address_space *mapping,
441 } 441 }
442 return err; 442 return err;
443} 443}
444EXPORT_SYMBOL(filemap_write_and_wait_range);
444 445
445/** 446/**
446 * add_to_page_cache_locked - add a locked page to the pagecache 447 * add_to_page_cache_locked - add a locked page to the pagecache
@@ -567,8 +568,8 @@ EXPORT_SYMBOL(wait_on_page_bit);
567 568
568/** 569/**
569 * add_page_wait_queue - Add an arbitrary waiter to a page's wait queue 570 * add_page_wait_queue - Add an arbitrary waiter to a page's wait queue
570 * @page - Page defining the wait queue of interest 571 * @page: Page defining the wait queue of interest
571 * @waiter - Waiter to add to the queue 572 * @waiter: Waiter to add to the queue
572 * 573 *
573 * Add an arbitrary @waiter to the wait queue for the nominated @page. 574 * Add an arbitrary @waiter to the wait queue for the nominated @page.
574 */ 575 */
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 2fc6d6c48238..e44fb0fbb80e 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -932,7 +932,7 @@ static int __mem_cgroup_try_charge(struct mm_struct *mm,
932 if (unlikely(!mem)) 932 if (unlikely(!mem))
933 return 0; 933 return 0;
934 934
935 VM_BUG_ON(mem_cgroup_is_obsolete(mem)); 935 VM_BUG_ON(!mem || mem_cgroup_is_obsolete(mem));
936 936
937 while (1) { 937 while (1) {
938 int ret; 938 int ret;
diff --git a/mm/mmap.c b/mm/mmap.c
index 4a3841186c11..3303d1ba8e87 100644
--- a/mm/mmap.c
+++ b/mm/mmap.c
@@ -1575,7 +1575,7 @@ static int acct_stack_growth(struct vm_area_struct *vma, unsigned long size, uns
1575 * Overcommit.. This must be the final test, as it will 1575 * Overcommit.. This must be the final test, as it will
1576 * update security statistics. 1576 * update security statistics.
1577 */ 1577 */
1578 if (security_vm_enough_memory(grow)) 1578 if (security_vm_enough_memory_mm(mm, grow))
1579 return -ENOMEM; 1579 return -ENOMEM;
1580 1580
1581 /* Ok, everything looks good - let it rip */ 1581 /* Ok, everything looks good - let it rip */
diff --git a/mm/shmem.c b/mm/shmem.c
index d94d2e9146bc..f9cb20ebb990 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -24,6 +24,7 @@
24#include <linux/init.h> 24#include <linux/init.h>
25#include <linux/vfs.h> 25#include <linux/vfs.h>
26#include <linux/mount.h> 26#include <linux/mount.h>
27#include <linux/pagemap.h>
27#include <linux/file.h> 28#include <linux/file.h>
28#include <linux/mm.h> 29#include <linux/mm.h>
29#include <linux/module.h> 30#include <linux/module.h>
@@ -43,7 +44,6 @@ static struct vfsmount *shm_mnt;
43#include <linux/exportfs.h> 44#include <linux/exportfs.h>
44#include <linux/generic_acl.h> 45#include <linux/generic_acl.h>
45#include <linux/mman.h> 46#include <linux/mman.h>
46#include <linux/pagemap.h>
47#include <linux/string.h> 47#include <linux/string.h>
48#include <linux/slab.h> 48#include <linux/slab.h>
49#include <linux/backing-dev.h> 49#include <linux/backing-dev.h>
@@ -65,13 +65,28 @@ static struct vfsmount *shm_mnt;
65#include <asm/div64.h> 65#include <asm/div64.h>
66#include <asm/pgtable.h> 66#include <asm/pgtable.h>
67 67
68/*
69 * The maximum size of a shmem/tmpfs file is limited by the maximum size of
70 * its triple-indirect swap vector - see illustration at shmem_swp_entry().
71 *
72 * With 4kB page size, maximum file size is just over 2TB on a 32-bit kernel,
73 * but one eighth of that on a 64-bit kernel. With 8kB page size, maximum
74 * file size is just over 4TB on a 64-bit kernel, but 16TB on a 32-bit kernel,
75 * MAX_LFS_FILESIZE being then more restrictive than swap vector layout.
76 *
77 * We use / and * instead of shifts in the definitions below, so that the swap
78 * vector can be tested with small even values (e.g. 20) for ENTRIES_PER_PAGE.
79 */
68#define ENTRIES_PER_PAGE (PAGE_CACHE_SIZE/sizeof(unsigned long)) 80#define ENTRIES_PER_PAGE (PAGE_CACHE_SIZE/sizeof(unsigned long))
69#define ENTRIES_PER_PAGEPAGE (ENTRIES_PER_PAGE*ENTRIES_PER_PAGE) 81#define ENTRIES_PER_PAGEPAGE ((unsigned long long)ENTRIES_PER_PAGE*ENTRIES_PER_PAGE)
70#define BLOCKS_PER_PAGE (PAGE_CACHE_SIZE/512)
71 82
72#define SHMEM_MAX_INDEX (SHMEM_NR_DIRECT + (ENTRIES_PER_PAGEPAGE/2) * (ENTRIES_PER_PAGE+1)) 83#define SHMSWP_MAX_INDEX (SHMEM_NR_DIRECT + (ENTRIES_PER_PAGEPAGE/2) * (ENTRIES_PER_PAGE+1))
73#define SHMEM_MAX_BYTES ((unsigned long long)SHMEM_MAX_INDEX << PAGE_CACHE_SHIFT) 84#define SHMSWP_MAX_BYTES (SHMSWP_MAX_INDEX << PAGE_CACHE_SHIFT)
74 85
86#define SHMEM_MAX_BYTES min_t(unsigned long long, SHMSWP_MAX_BYTES, MAX_LFS_FILESIZE)
87#define SHMEM_MAX_INDEX ((unsigned long)((SHMEM_MAX_BYTES+1) >> PAGE_CACHE_SHIFT))
88
89#define BLOCKS_PER_PAGE (PAGE_CACHE_SIZE/512)
75#define VM_ACCT(size) (PAGE_CACHE_ALIGN(size) >> PAGE_SHIFT) 90#define VM_ACCT(size) (PAGE_CACHE_ALIGN(size) >> PAGE_SHIFT)
76 91
77/* info->flags needs VM_flags to handle pagein/truncate races efficiently */ 92/* info->flags needs VM_flags to handle pagein/truncate races efficiently */
@@ -2581,7 +2596,7 @@ int shmem_unuse(swp_entry_t entry, struct page *page)
2581#define shmem_get_inode(sb, mode, dev, flags) ramfs_get_inode(sb, mode, dev) 2596#define shmem_get_inode(sb, mode, dev, flags) ramfs_get_inode(sb, mode, dev)
2582#define shmem_acct_size(flags, size) 0 2597#define shmem_acct_size(flags, size) 0
2583#define shmem_unacct_size(flags, size) do {} while (0) 2598#define shmem_unacct_size(flags, size) do {} while (0)
2584#define SHMEM_MAX_BYTES LLONG_MAX 2599#define SHMEM_MAX_BYTES MAX_LFS_FILESIZE
2585 2600
2586#endif /* CONFIG_SHMEM */ 2601#endif /* CONFIG_SHMEM */
2587 2602
diff --git a/mm/util.c b/mm/util.c
index 2599e83eea17..55bef160b9f1 100644
--- a/mm/util.c
+++ b/mm/util.c
@@ -223,6 +223,22 @@ void arch_pick_mmap_layout(struct mm_struct *mm)
223} 223}
224#endif 224#endif
225 225
226/**
227 * get_user_pages_fast() - pin user pages in memory
228 * @start: starting user address
229 * @nr_pages: number of pages from start to pin
230 * @write: whether pages will be written to
231 * @pages: array that receives pointers to the pages pinned.
232 * Should be at least nr_pages long.
233 *
234 * Attempt to pin user pages in memory without taking mm->mmap_sem.
235 * If not successful, it will fall back to taking the lock and
236 * calling get_user_pages().
237 *
238 * Returns number of pages pinned. This may be fewer than the number
239 * requested. If nr_pages is 0 or negative, returns 0. If no pages
240 * were pinned, returns -errno.
241 */
226int __attribute__((weak)) get_user_pages_fast(unsigned long start, 242int __attribute__((weak)) get_user_pages_fast(unsigned long start,
227 int nr_pages, int write, struct page **pages) 243 int nr_pages, int write, struct page **pages)
228{ 244{
diff --git a/net/802/tr.c b/net/802/tr.c
index e7eb13084d71..e874447ad144 100644
--- a/net/802/tr.c
+++ b/net/802/tr.c
@@ -561,6 +561,9 @@ static int rif_seq_show(struct seq_file *seq, void *v)
561 } 561 }
562 seq_putc(seq, '\n'); 562 seq_putc(seq, '\n');
563 } 563 }
564
565 if (dev)
566 dev_put(dev);
564 } 567 }
565 return 0; 568 return 0;
566} 569}
diff --git a/net/8021q/vlan_core.c b/net/8021q/vlan_core.c
index 654e45f5719d..c67fe6f75653 100644
--- a/net/8021q/vlan_core.c
+++ b/net/8021q/vlan_core.c
@@ -121,8 +121,10 @@ int vlan_gro_frags(struct napi_struct *napi, struct vlan_group *grp,
121 if (!skb) 121 if (!skb)
122 return NET_RX_DROP; 122 return NET_RX_DROP;
123 123
124 if (netpoll_rx_on(skb)) 124 if (netpoll_rx_on(skb)) {
125 skb->protocol = eth_type_trans(skb, skb->dev);
125 return vlan_hwaccel_receive_skb(skb, grp, vlan_tci); 126 return vlan_hwaccel_receive_skb(skb, grp, vlan_tci);
127 }
126 128
127 return napi_frags_finish(napi, skb, 129 return napi_frags_finish(napi, skb,
128 vlan_gro_common(napi, grp, vlan_tci, skb)); 130 vlan_gro_common(napi, grp, vlan_tci, skb));
diff --git a/net/core/dev.c b/net/core/dev.c
index 91d792d17e09..343883f65ea7 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -1430,7 +1430,7 @@ void netif_device_detach(struct net_device *dev)
1430{ 1430{
1431 if (test_and_clear_bit(__LINK_STATE_PRESENT, &dev->state) && 1431 if (test_and_clear_bit(__LINK_STATE_PRESENT, &dev->state) &&
1432 netif_running(dev)) { 1432 netif_running(dev)) {
1433 netif_stop_queue(dev); 1433 netif_tx_stop_all_queues(dev);
1434 } 1434 }
1435} 1435}
1436EXPORT_SYMBOL(netif_device_detach); 1436EXPORT_SYMBOL(netif_device_detach);
@@ -1445,7 +1445,7 @@ void netif_device_attach(struct net_device *dev)
1445{ 1445{
1446 if (!test_and_set_bit(__LINK_STATE_PRESENT, &dev->state) && 1446 if (!test_and_set_bit(__LINK_STATE_PRESENT, &dev->state) &&
1447 netif_running(dev)) { 1447 netif_running(dev)) {
1448 netif_wake_queue(dev); 1448 netif_tx_wake_all_queues(dev);
1449 __netdev_watchdog_up(dev); 1449 __netdev_watchdog_up(dev);
1450 } 1450 }
1451} 1451}
@@ -2328,8 +2328,10 @@ static int napi_gro_complete(struct sk_buff *skb)
2328 struct list_head *head = &ptype_base[ntohs(type) & PTYPE_HASH_MASK]; 2328 struct list_head *head = &ptype_base[ntohs(type) & PTYPE_HASH_MASK];
2329 int err = -ENOENT; 2329 int err = -ENOENT;
2330 2330
2331 if (NAPI_GRO_CB(skb)->count == 1) 2331 if (NAPI_GRO_CB(skb)->count == 1) {
2332 skb_shinfo(skb)->gso_size = 0;
2332 goto out; 2333 goto out;
2334 }
2333 2335
2334 rcu_read_lock(); 2336 rcu_read_lock();
2335 list_for_each_entry_rcu(ptype, head, list) { 2337 list_for_each_entry_rcu(ptype, head, list) {
@@ -2348,7 +2350,6 @@ static int napi_gro_complete(struct sk_buff *skb)
2348 } 2350 }
2349 2351
2350out: 2352out:
2351 skb_shinfo(skb)->gso_size = 0;
2352 return netif_receive_skb(skb); 2353 return netif_receive_skb(skb);
2353} 2354}
2354 2355
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 2bc8e27a163d..c96a6bb25430 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -928,6 +928,8 @@ static void tcp_init_metrics(struct sock *sk)
928 tcp_set_rto(sk); 928 tcp_set_rto(sk);
929 if (inet_csk(sk)->icsk_rto < TCP_TIMEOUT_INIT && !tp->rx_opt.saw_tstamp) 929 if (inet_csk(sk)->icsk_rto < TCP_TIMEOUT_INIT && !tp->rx_opt.saw_tstamp)
930 goto reset; 930 goto reset;
931
932cwnd:
931 tp->snd_cwnd = tcp_init_cwnd(tp, dst); 933 tp->snd_cwnd = tcp_init_cwnd(tp, dst);
932 tp->snd_cwnd_stamp = tcp_time_stamp; 934 tp->snd_cwnd_stamp = tcp_time_stamp;
933 return; 935 return;
@@ -942,6 +944,7 @@ reset:
942 tp->mdev = tp->mdev_max = tp->rttvar = TCP_TIMEOUT_INIT; 944 tp->mdev = tp->mdev_max = tp->rttvar = TCP_TIMEOUT_INIT;
943 inet_csk(sk)->icsk_rto = TCP_TIMEOUT_INIT; 945 inet_csk(sk)->icsk_rto = TCP_TIMEOUT_INIT;
944 } 946 }
947 goto cwnd;
945} 948}
946 949
947static void tcp_update_reordering(struct sock *sk, const int metric, 950static void tcp_update_reordering(struct sock *sk, const int metric,
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index bda08a09357d..7a1d1ce22e66 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -222,7 +222,7 @@ fail:
222 return error; 222 return error;
223} 223}
224 224
225int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2) 225static int ipv4_rcv_saddr_equal(const struct sock *sk1, const struct sock *sk2)
226{ 226{
227 struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2); 227 struct inet_sock *inet1 = inet_sk(sk1), *inet2 = inet_sk(sk2);
228 228
@@ -1823,7 +1823,6 @@ EXPORT_SYMBOL(udp_lib_getsockopt);
1823EXPORT_SYMBOL(udp_lib_setsockopt); 1823EXPORT_SYMBOL(udp_lib_setsockopt);
1824EXPORT_SYMBOL(udp_poll); 1824EXPORT_SYMBOL(udp_poll);
1825EXPORT_SYMBOL(udp_lib_get_port); 1825EXPORT_SYMBOL(udp_lib_get_port);
1826EXPORT_SYMBOL(ipv4_rcv_saddr_equal);
1827 1826
1828#ifdef CONFIG_PROC_FS 1827#ifdef CONFIG_PROC_FS
1829EXPORT_SYMBOL(udp_proc_register); 1828EXPORT_SYMBOL(udp_proc_register);
diff --git a/net/ipv6/ipv6_sockglue.c b/net/ipv6/ipv6_sockglue.c
index d31df0f4bc9a..a7fdf9a27f15 100644
--- a/net/ipv6/ipv6_sockglue.c
+++ b/net/ipv6/ipv6_sockglue.c
@@ -380,10 +380,6 @@ static int do_ipv6_setsockopt(struct sock *sk, int level, int optname,
380 default: 380 default:
381 goto sticky_done; 381 goto sticky_done;
382 } 382 }
383
384 if ((rthdr->hdrlen & 1) ||
385 (rthdr->hdrlen >> 1) != rthdr->segments_left)
386 goto sticky_done;
387 } 383 }
388 384
389 retv = 0; 385 retv = 0;
diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
index 6842dd2edd5b..8905712cfbb8 100644
--- a/net/ipv6/udp.c
+++ b/net/ipv6/udp.c
@@ -53,6 +53,8 @@ int ipv6_rcv_saddr_equal(const struct sock *sk, const struct sock *sk2)
53{ 53{
54 const struct in6_addr *sk_rcv_saddr6 = &inet6_sk(sk)->rcv_saddr; 54 const struct in6_addr *sk_rcv_saddr6 = &inet6_sk(sk)->rcv_saddr;
55 const struct in6_addr *sk2_rcv_saddr6 = inet6_rcv_saddr(sk2); 55 const struct in6_addr *sk2_rcv_saddr6 = inet6_rcv_saddr(sk2);
56 __be32 sk_rcv_saddr = inet_sk(sk)->rcv_saddr;
57 __be32 sk2_rcv_saddr = inet_rcv_saddr(sk2);
56 int sk_ipv6only = ipv6_only_sock(sk); 58 int sk_ipv6only = ipv6_only_sock(sk);
57 int sk2_ipv6only = inet_v6_ipv6only(sk2); 59 int sk2_ipv6only = inet_v6_ipv6only(sk2);
58 int addr_type = ipv6_addr_type(sk_rcv_saddr6); 60 int addr_type = ipv6_addr_type(sk_rcv_saddr6);
@@ -60,7 +62,9 @@ int ipv6_rcv_saddr_equal(const struct sock *sk, const struct sock *sk2)
60 62
61 /* if both are mapped, treat as IPv4 */ 63 /* if both are mapped, treat as IPv4 */
62 if (addr_type == IPV6_ADDR_MAPPED && addr_type2 == IPV6_ADDR_MAPPED) 64 if (addr_type == IPV6_ADDR_MAPPED && addr_type2 == IPV6_ADDR_MAPPED)
63 return ipv4_rcv_saddr_equal(sk, sk2); 65 return (!sk2_ipv6only &&
66 (!sk_rcv_saddr || !sk2_rcv_saddr ||
67 sk_rcv_saddr == sk2_rcv_saddr));
64 68
65 if (addr_type2 == IPV6_ADDR_ANY && 69 if (addr_type2 == IPV6_ADDR_ANY &&
66 !(sk2_ipv6only && addr_type == IPV6_ADDR_MAPPED)) 70 !(sk2_ipv6only && addr_type == IPV6_ADDR_MAPPED))
diff --git a/net/packet/af_packet.c b/net/packet/af_packet.c
index 74776de523ec..f546e81acc45 100644
--- a/net/packet/af_packet.c
+++ b/net/packet/af_packet.c
@@ -1758,8 +1758,9 @@ static void free_pg_vec(char **pg_vec, unsigned int order, unsigned int len)
1758 1758
1759static inline char *alloc_one_pg_vec_page(unsigned long order) 1759static inline char *alloc_one_pg_vec_page(unsigned long order)
1760{ 1760{
1761 return (char *) __get_free_pages(GFP_KERNEL | __GFP_COMP | __GFP_ZERO, 1761 gfp_t gfp_flags = GFP_KERNEL | __GFP_COMP | __GFP_ZERO | __GFP_NOWARN;
1762 order); 1762
1763 return (char *) __get_free_pages(gfp_flags, order);
1763} 1764}
1764 1765
1765static char **alloc_pg_vec(struct tpacket_req *req, int order) 1766static char **alloc_pg_vec(struct tpacket_req *req, int order)
diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c
index 0f36e8d59b29..877a7f65f707 100644
--- a/net/rose/af_rose.c
+++ b/net/rose/af_rose.c
@@ -1072,10 +1072,6 @@ static int rose_sendmsg(struct kiocb *iocb, struct socket *sock,
1072 unsigned char *asmptr; 1072 unsigned char *asmptr;
1073 int n, size, qbit = 0; 1073 int n, size, qbit = 0;
1074 1074
1075 /* ROSE empty frame has no meaning : don't send */
1076 if (len == 0)
1077 return 0;
1078
1079 if (msg->msg_flags & ~(MSG_DONTWAIT|MSG_EOR|MSG_CMSG_COMPAT)) 1075 if (msg->msg_flags & ~(MSG_DONTWAIT|MSG_EOR|MSG_CMSG_COMPAT))
1080 return -EINVAL; 1076 return -EINVAL;
1081 1077
@@ -1273,12 +1269,6 @@ static int rose_recvmsg(struct kiocb *iocb, struct socket *sock,
1273 skb_reset_transport_header(skb); 1269 skb_reset_transport_header(skb);
1274 copied = skb->len; 1270 copied = skb->len;
1275 1271
1276 /* ROSE empty frame has no meaning : ignore it */
1277 if (copied == 0) {
1278 skb_free_datagram(sk, skb);
1279 return copied;
1280 }
1281
1282 if (copied > size) { 1272 if (copied > size) {
1283 copied = size; 1273 copied = size;
1284 msg->msg_flags |= MSG_TRUNC; 1274 msg->msg_flags |= MSG_TRUNC;
diff --git a/net/sched/em_meta.c b/net/sched/em_meta.c
index 72cf86e3c090..fad596bf32d7 100644
--- a/net/sched/em_meta.c
+++ b/net/sched/em_meta.c
@@ -176,8 +176,10 @@ META_COLLECTOR(var_dev)
176 176
177META_COLLECTOR(int_vlan_tag) 177META_COLLECTOR(int_vlan_tag)
178{ 178{
179 unsigned short uninitialized_var(tag); 179 unsigned short tag;
180 if (vlan_get_tag(skb, &tag) < 0) 180
181 tag = vlan_tx_tag_get(skb);
182 if (!tag && __vlan_get_tag(skb, &tag))
181 *err = -1; 183 *err = -1;
182 else 184 else
183 dst->value = tag; 185 dst->value = tag;
diff --git a/security/tomoyo/common.c b/security/tomoyo/common.c
index a0affd9cfca8..d4d41b3efc7c 100644
--- a/security/tomoyo/common.c
+++ b/security/tomoyo/common.c
@@ -5,7 +5,7 @@
5 * 5 *
6 * Copyright (C) 2005-2009 NTT DATA CORPORATION 6 * Copyright (C) 2005-2009 NTT DATA CORPORATION
7 * 7 *
8 * Version: 2.2.0-pre 2009/02/01 8 * Version: 2.2.0 2009/04/01
9 * 9 *
10 */ 10 */
11 11
@@ -1773,7 +1773,7 @@ void tomoyo_load_policy(const char *filename)
1773 envp[2] = NULL; 1773 envp[2] = NULL;
1774 call_usermodehelper(argv[0], argv, envp, 1); 1774 call_usermodehelper(argv[0], argv, envp, 1);
1775 1775
1776 printk(KERN_INFO "TOMOYO: 2.2.0-pre 2009/02/01\n"); 1776 printk(KERN_INFO "TOMOYO: 2.2.0 2009/04/01\n");
1777 printk(KERN_INFO "Mandatory Access Control activated.\n"); 1777 printk(KERN_INFO "Mandatory Access Control activated.\n");
1778 tomoyo_policy_loaded = true; 1778 tomoyo_policy_loaded = true;
1779 { /* Check all profiles currently assigned to domains are defined. */ 1779 { /* Check all profiles currently assigned to domains are defined. */
@@ -1800,7 +1800,7 @@ void tomoyo_load_policy(const char *filename)
1800static int tomoyo_read_version(struct tomoyo_io_buffer *head) 1800static int tomoyo_read_version(struct tomoyo_io_buffer *head)
1801{ 1801{
1802 if (!head->read_eof) { 1802 if (!head->read_eof) {
1803 tomoyo_io_printf(head, "2.2.0-pre"); 1803 tomoyo_io_printf(head, "2.2.0");
1804 head->read_eof = true; 1804 head->read_eof = true;
1805 } 1805 }
1806 return 0; 1806 return 0;
diff --git a/security/tomoyo/common.h b/security/tomoyo/common.h
index e77e6a6de0f2..678f4ff16aa4 100644
--- a/security/tomoyo/common.h
+++ b/security/tomoyo/common.h
@@ -5,7 +5,7 @@
5 * 5 *
6 * Copyright (C) 2005-2009 NTT DATA CORPORATION 6 * Copyright (C) 2005-2009 NTT DATA CORPORATION
7 * 7 *
8 * Version: 2.2.0-pre 2009/02/01 8 * Version: 2.2.0 2009/04/01
9 * 9 *
10 */ 10 */
11 11
diff --git a/security/tomoyo/domain.c b/security/tomoyo/domain.c
index 2f2b449ffd2d..2d6748741a26 100644
--- a/security/tomoyo/domain.c
+++ b/security/tomoyo/domain.c
@@ -5,7 +5,7 @@
5 * 5 *
6 * Copyright (C) 2005-2009 NTT DATA CORPORATION 6 * Copyright (C) 2005-2009 NTT DATA CORPORATION
7 * 7 *
8 * Version: 2.2.0-pre 2009/02/01 8 * Version: 2.2.0 2009/04/01
9 * 9 *
10 */ 10 */
11 11
diff --git a/security/tomoyo/file.c b/security/tomoyo/file.c
index 65f50c1c5ee9..2316da8ec5bc 100644
--- a/security/tomoyo/file.c
+++ b/security/tomoyo/file.c
@@ -5,7 +5,7 @@
5 * 5 *
6 * Copyright (C) 2005-2009 NTT DATA CORPORATION 6 * Copyright (C) 2005-2009 NTT DATA CORPORATION
7 * 7 *
8 * Version: 2.2.0-pre 2009/02/01 8 * Version: 2.2.0 2009/04/01
9 * 9 *
10 */ 10 */
11 11
diff --git a/security/tomoyo/realpath.c b/security/tomoyo/realpath.c
index 3bbe01a7a4b5..bf8e2b451687 100644
--- a/security/tomoyo/realpath.c
+++ b/security/tomoyo/realpath.c
@@ -5,7 +5,7 @@
5 * 5 *
6 * Copyright (C) 2005-2009 NTT DATA CORPORATION 6 * Copyright (C) 2005-2009 NTT DATA CORPORATION
7 * 7 *
8 * Version: 2.2.0-pre 2009/02/01 8 * Version: 2.2.0 2009/04/01
9 * 9 *
10 */ 10 */
11 11
diff --git a/security/tomoyo/realpath.h b/security/tomoyo/realpath.h
index 7ec9fc9cbc07..78217a37960b 100644
--- a/security/tomoyo/realpath.h
+++ b/security/tomoyo/realpath.h
@@ -5,7 +5,7 @@
5 * 5 *
6 * Copyright (C) 2005-2009 NTT DATA CORPORATION 6 * Copyright (C) 2005-2009 NTT DATA CORPORATION
7 * 7 *
8 * Version: 2.2.0-pre 2009/02/01 8 * Version: 2.2.0 2009/04/01
9 * 9 *
10 */ 10 */
11 11
diff --git a/security/tomoyo/tomoyo.c b/security/tomoyo/tomoyo.c
index 3eeeae12c4dc..5b481912752a 100644
--- a/security/tomoyo/tomoyo.c
+++ b/security/tomoyo/tomoyo.c
@@ -5,7 +5,7 @@
5 * 5 *
6 * Copyright (C) 2005-2009 NTT DATA CORPORATION 6 * Copyright (C) 2005-2009 NTT DATA CORPORATION
7 * 7 *
8 * Version: 2.2.0-pre 2009/02/01 8 * Version: 2.2.0 2009/04/01
9 * 9 *
10 */ 10 */
11 11
diff --git a/security/tomoyo/tomoyo.h b/security/tomoyo/tomoyo.h
index a0c8f6e0bea4..41c6ebafb9c5 100644
--- a/security/tomoyo/tomoyo.h
+++ b/security/tomoyo/tomoyo.h
@@ -5,7 +5,7 @@
5 * 5 *
6 * Copyright (C) 2005-2009 NTT DATA CORPORATION 6 * Copyright (C) 2005-2009 NTT DATA CORPORATION
7 * 7 *
8 * Version: 2.2.0-pre 2009/02/01 8 * Version: 2.2.0 2009/04/01
9 * 9 *
10 */ 10 */
11 11
diff --git a/sound/core/control.c b/sound/core/control.c
index 4b20fa2b7e6d..17b8d47a5cd0 100644
--- a/sound/core/control.c
+++ b/sound/core/control.c
@@ -723,14 +723,11 @@ static int snd_ctl_elem_read_user(struct snd_card *card,
723{ 723{
724 struct snd_ctl_elem_value *control; 724 struct snd_ctl_elem_value *control;
725 int result; 725 int result;
726 726
727 control = kmalloc(sizeof(*control), GFP_KERNEL); 727 control = memdup_user(_control, sizeof(*control));
728 if (control == NULL) 728 if (IS_ERR(control))
729 return -ENOMEM; 729 return PTR_ERR(control);
730 if (copy_from_user(control, _control, sizeof(*control))) { 730
731 kfree(control);
732 return -EFAULT;
733 }
734 snd_power_lock(card); 731 snd_power_lock(card);
735 result = snd_power_wait(card, SNDRV_CTL_POWER_D0); 732 result = snd_power_wait(card, SNDRV_CTL_POWER_D0);
736 if (result >= 0) 733 if (result >= 0)
@@ -784,13 +781,10 @@ static int snd_ctl_elem_write_user(struct snd_ctl_file *file,
784 struct snd_card *card; 781 struct snd_card *card;
785 int result; 782 int result;
786 783
787 control = kmalloc(sizeof(*control), GFP_KERNEL); 784 control = memdup_user(_control, sizeof(*control));
788 if (control == NULL) 785 if (IS_ERR(control))
789 return -ENOMEM; 786 return PTR_ERR(control);
790 if (copy_from_user(control, _control, sizeof(*control))) { 787
791 kfree(control);
792 return -EFAULT;
793 }
794 card = file->card; 788 card = file->card;
795 snd_power_lock(card); 789 snd_power_lock(card);
796 result = snd_power_wait(card, SNDRV_CTL_POWER_D0); 790 result = snd_power_wait(card, SNDRV_CTL_POWER_D0);
@@ -916,13 +910,10 @@ static int snd_ctl_elem_user_tlv(struct snd_kcontrol *kcontrol,
916 if (op_flag > 0) { 910 if (op_flag > 0) {
917 if (size > 1024 * 128) /* sane value */ 911 if (size > 1024 * 128) /* sane value */
918 return -EINVAL; 912 return -EINVAL;
919 new_data = kmalloc(size, GFP_KERNEL); 913
920 if (new_data == NULL) 914 new_data = memdup_user(tlv, size);
921 return -ENOMEM; 915 if (IS_ERR(new_data))
922 if (copy_from_user(new_data, tlv, size)) { 916 return PTR_ERR(new_data);
923 kfree(new_data);
924 return -EFAULT;
925 }
926 change = ue->tlv_data_size != size; 917 change = ue->tlv_data_size != size;
927 if (!change) 918 if (!change)
928 change = memcmp(ue->tlv_data, new_data, size); 919 change = memcmp(ue->tlv_data, new_data, size);
diff --git a/sound/core/jack.c b/sound/core/jack.c
index c8254c667c62..d54d1a05fe65 100644
--- a/sound/core/jack.c
+++ b/sound/core/jack.c
@@ -35,6 +35,9 @@ static int snd_jack_dev_free(struct snd_device *device)
35{ 35{
36 struct snd_jack *jack = device->device_data; 36 struct snd_jack *jack = device->device_data;
37 37
38 if (jack->private_free)
39 jack->private_free(jack);
40
38 /* If the input device is registered with the input subsystem 41 /* If the input device is registered with the input subsystem
39 * then we need to use a different deallocator. */ 42 * then we need to use a different deallocator. */
40 if (jack->registered) 43 if (jack->registered)
diff --git a/sound/core/pcm_compat.c b/sound/core/pcm_compat.c
index 36d7a5998234..08bfed594a83 100644
--- a/sound/core/pcm_compat.c
+++ b/sound/core/pcm_compat.c
@@ -232,14 +232,11 @@ static int snd_pcm_ioctl_hw_params_compat(struct snd_pcm_substream *substream,
232 if (! (runtime = substream->runtime)) 232 if (! (runtime = substream->runtime))
233 return -ENOTTY; 233 return -ENOTTY;
234 234
235 data = kmalloc(sizeof(*data), GFP_KERNEL);
236 if (data == NULL)
237 return -ENOMEM;
238 /* only fifo_size is different, so just copy all */ 235 /* only fifo_size is different, so just copy all */
239 if (copy_from_user(data, data32, sizeof(*data32))) { 236 data = memdup_user(data32, sizeof(*data32));
240 err = -EFAULT; 237 if (IS_ERR(data))
241 goto error; 238 return PTR_ERR(data);
242 } 239
243 if (refine) 240 if (refine)
244 err = snd_pcm_hw_refine(substream, data); 241 err = snd_pcm_hw_refine(substream, data);
245 else 242 else
diff --git a/sound/core/pcm_lib.c b/sound/core/pcm_lib.c
index fbb2e391591e..63d088f2265f 100644
--- a/sound/core/pcm_lib.c
+++ b/sound/core/pcm_lib.c
@@ -209,9 +209,11 @@ static int snd_pcm_update_hw_ptr_interrupt(struct snd_pcm_substream *substream)
209{ 209{
210 struct snd_pcm_runtime *runtime = substream->runtime; 210 struct snd_pcm_runtime *runtime = substream->runtime;
211 snd_pcm_uframes_t pos; 211 snd_pcm_uframes_t pos;
212 snd_pcm_uframes_t new_hw_ptr, hw_ptr_interrupt, hw_base; 212 snd_pcm_uframes_t old_hw_ptr, new_hw_ptr, hw_ptr_interrupt, hw_base;
213 snd_pcm_sframes_t delta; 213 snd_pcm_sframes_t hdelta, delta;
214 unsigned long jdelta;
214 215
216 old_hw_ptr = runtime->status->hw_ptr;
215 pos = snd_pcm_update_hw_ptr_pos(substream, runtime); 217 pos = snd_pcm_update_hw_ptr_pos(substream, runtime);
216 if (pos == SNDRV_PCM_POS_XRUN) { 218 if (pos == SNDRV_PCM_POS_XRUN) {
217 xrun(substream); 219 xrun(substream);
@@ -247,7 +249,30 @@ static int snd_pcm_update_hw_ptr_interrupt(struct snd_pcm_substream *substream)
247 new_hw_ptr = hw_base + pos; 249 new_hw_ptr = hw_base + pos;
248 } 250 }
249 } 251 }
250 if (delta > runtime->period_size) { 252 hdelta = new_hw_ptr - old_hw_ptr;
253 jdelta = jiffies - runtime->hw_ptr_jiffies;
254 if (((hdelta * HZ) / runtime->rate) > jdelta + HZ/100) {
255 delta = jdelta /
256 (((runtime->period_size * HZ) / runtime->rate)
257 + HZ/100);
258 hw_ptr_error(substream,
259 "hw_ptr skipping! [Q] "
260 "(pos=%ld, delta=%ld, period=%ld, "
261 "jdelta=%lu/%lu/%lu)\n",
262 (long)pos, (long)hdelta,
263 (long)runtime->period_size, jdelta,
264 ((hdelta * HZ) / runtime->rate), delta);
265 hw_ptr_interrupt = runtime->hw_ptr_interrupt +
266 runtime->period_size * delta;
267 if (hw_ptr_interrupt >= runtime->boundary)
268 hw_ptr_interrupt -= runtime->boundary;
269 /* rebase to interrupt position */
270 hw_base = new_hw_ptr = hw_ptr_interrupt;
271 /* align hw_base to buffer_size */
272 hw_base -= hw_base % runtime->buffer_size;
273 delta = 0;
274 }
275 if (delta > runtime->period_size + runtime->period_size / 2) {
251 hw_ptr_error(substream, 276 hw_ptr_error(substream,
252 "Lost interrupts? " 277 "Lost interrupts? "
253 "(stream=%i, delta=%ld, intr_ptr=%ld)\n", 278 "(stream=%i, delta=%ld, intr_ptr=%ld)\n",
@@ -263,6 +288,7 @@ static int snd_pcm_update_hw_ptr_interrupt(struct snd_pcm_substream *substream)
263 288
264 runtime->hw_ptr_base = hw_base; 289 runtime->hw_ptr_base = hw_base;
265 runtime->status->hw_ptr = new_hw_ptr; 290 runtime->status->hw_ptr = new_hw_ptr;
291 runtime->hw_ptr_jiffies = jiffies;
266 runtime->hw_ptr_interrupt = hw_ptr_interrupt; 292 runtime->hw_ptr_interrupt = hw_ptr_interrupt;
267 293
268 return snd_pcm_update_hw_ptr_post(substream, runtime); 294 return snd_pcm_update_hw_ptr_post(substream, runtime);
@@ -275,6 +301,7 @@ int snd_pcm_update_hw_ptr(struct snd_pcm_substream *substream)
275 snd_pcm_uframes_t pos; 301 snd_pcm_uframes_t pos;
276 snd_pcm_uframes_t old_hw_ptr, new_hw_ptr, hw_base; 302 snd_pcm_uframes_t old_hw_ptr, new_hw_ptr, hw_base;
277 snd_pcm_sframes_t delta; 303 snd_pcm_sframes_t delta;
304 unsigned long jdelta;
278 305
279 old_hw_ptr = runtime->status->hw_ptr; 306 old_hw_ptr = runtime->status->hw_ptr;
280 pos = snd_pcm_update_hw_ptr_pos(substream, runtime); 307 pos = snd_pcm_update_hw_ptr_pos(substream, runtime);
@@ -286,14 +313,15 @@ int snd_pcm_update_hw_ptr(struct snd_pcm_substream *substream)
286 new_hw_ptr = hw_base + pos; 313 new_hw_ptr = hw_base + pos;
287 314
288 delta = new_hw_ptr - old_hw_ptr; 315 delta = new_hw_ptr - old_hw_ptr;
316 jdelta = jiffies - runtime->hw_ptr_jiffies;
289 if (delta < 0) { 317 if (delta < 0) {
290 delta += runtime->buffer_size; 318 delta += runtime->buffer_size;
291 if (delta < 0) { 319 if (delta < 0) {
292 hw_ptr_error(substream, 320 hw_ptr_error(substream,
293 "Unexpected hw_pointer value [2] " 321 "Unexpected hw_pointer value [2] "
294 "(stream=%i, pos=%ld, old_ptr=%ld)\n", 322 "(stream=%i, pos=%ld, old_ptr=%ld, jdelta=%li)\n",
295 substream->stream, (long)pos, 323 substream->stream, (long)pos,
296 (long)old_hw_ptr); 324 (long)old_hw_ptr, jdelta);
297 return 0; 325 return 0;
298 } 326 }
299 hw_base += runtime->buffer_size; 327 hw_base += runtime->buffer_size;
@@ -301,12 +329,13 @@ int snd_pcm_update_hw_ptr(struct snd_pcm_substream *substream)
301 hw_base = 0; 329 hw_base = 0;
302 new_hw_ptr = hw_base + pos; 330 new_hw_ptr = hw_base + pos;
303 } 331 }
304 if (delta > runtime->period_size && runtime->periods > 1) { 332 if (((delta * HZ) / runtime->rate) > jdelta + HZ/100) {
305 hw_ptr_error(substream, 333 hw_ptr_error(substream,
306 "hw_ptr skipping! " 334 "hw_ptr skipping! "
307 "(pos=%ld, delta=%ld, period=%ld)\n", 335 "(pos=%ld, delta=%ld, period=%ld, jdelta=%lu/%lu)\n",
308 (long)pos, (long)delta, 336 (long)pos, (long)delta,
309 (long)runtime->period_size); 337 (long)runtime->period_size, jdelta,
338 ((delta * HZ) / runtime->rate));
310 return 0; 339 return 0;
311 } 340 }
312 if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK && 341 if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK &&
@@ -315,6 +344,7 @@ int snd_pcm_update_hw_ptr(struct snd_pcm_substream *substream)
315 344
316 runtime->hw_ptr_base = hw_base; 345 runtime->hw_ptr_base = hw_base;
317 runtime->status->hw_ptr = new_hw_ptr; 346 runtime->status->hw_ptr = new_hw_ptr;
347 runtime->hw_ptr_jiffies = jiffies;
318 348
319 return snd_pcm_update_hw_ptr_post(substream, runtime); 349 return snd_pcm_update_hw_ptr_post(substream, runtime);
320} 350}
@@ -1441,6 +1471,7 @@ static int snd_pcm_lib_ioctl_reset(struct snd_pcm_substream *substream,
1441 runtime->status->hw_ptr %= runtime->buffer_size; 1471 runtime->status->hw_ptr %= runtime->buffer_size;
1442 else 1472 else
1443 runtime->status->hw_ptr = 0; 1473 runtime->status->hw_ptr = 0;
1474 runtime->hw_ptr_jiffies = jiffies;
1444 snd_pcm_stream_unlock_irqrestore(substream, flags); 1475 snd_pcm_stream_unlock_irqrestore(substream, flags);
1445 return 0; 1476 return 0;
1446} 1477}
diff --git a/sound/core/pcm_native.c b/sound/core/pcm_native.c
index a151fb01ba82..fc6f98e257df 100644
--- a/sound/core/pcm_native.c
+++ b/sound/core/pcm_native.c
@@ -327,21 +327,16 @@ static int snd_pcm_hw_refine_user(struct snd_pcm_substream *substream,
327 struct snd_pcm_hw_params *params; 327 struct snd_pcm_hw_params *params;
328 int err; 328 int err;
329 329
330 params = kmalloc(sizeof(*params), GFP_KERNEL); 330 params = memdup_user(_params, sizeof(*params));
331 if (!params) { 331 if (IS_ERR(params))
332 err = -ENOMEM; 332 return PTR_ERR(params);
333 goto out; 333
334 }
335 if (copy_from_user(params, _params, sizeof(*params))) {
336 err = -EFAULT;
337 goto out;
338 }
339 err = snd_pcm_hw_refine(substream, params); 334 err = snd_pcm_hw_refine(substream, params);
340 if (copy_to_user(_params, params, sizeof(*params))) { 335 if (copy_to_user(_params, params, sizeof(*params))) {
341 if (!err) 336 if (!err)
342 err = -EFAULT; 337 err = -EFAULT;
343 } 338 }
344out: 339
345 kfree(params); 340 kfree(params);
346 return err; 341 return err;
347} 342}
@@ -465,21 +460,16 @@ static int snd_pcm_hw_params_user(struct snd_pcm_substream *substream,
465 struct snd_pcm_hw_params *params; 460 struct snd_pcm_hw_params *params;
466 int err; 461 int err;
467 462
468 params = kmalloc(sizeof(*params), GFP_KERNEL); 463 params = memdup_user(_params, sizeof(*params));
469 if (!params) { 464 if (IS_ERR(params))
470 err = -ENOMEM; 465 return PTR_ERR(params);
471 goto out; 466
472 }
473 if (copy_from_user(params, _params, sizeof(*params))) {
474 err = -EFAULT;
475 goto out;
476 }
477 err = snd_pcm_hw_params(substream, params); 467 err = snd_pcm_hw_params(substream, params);
478 if (copy_to_user(_params, params, sizeof(*params))) { 468 if (copy_to_user(_params, params, sizeof(*params))) {
479 if (!err) 469 if (!err)
480 err = -EFAULT; 470 err = -EFAULT;
481 } 471 }
482out: 472
483 kfree(params); 473 kfree(params);
484 return err; 474 return err;
485} 475}
@@ -2593,13 +2583,11 @@ static int snd_pcm_playback_ioctl1(struct file *file,
2593 return -EFAULT; 2583 return -EFAULT;
2594 if (copy_from_user(&xfern, _xfern, sizeof(xfern))) 2584 if (copy_from_user(&xfern, _xfern, sizeof(xfern)))
2595 return -EFAULT; 2585 return -EFAULT;
2596 bufs = kmalloc(sizeof(void *) * runtime->channels, GFP_KERNEL); 2586
2597 if (bufs == NULL) 2587 bufs = memdup_user(xfern.bufs,
2598 return -ENOMEM; 2588 sizeof(void *) * runtime->channels);
2599 if (copy_from_user(bufs, xfern.bufs, sizeof(void *) * runtime->channels)) { 2589 if (IS_ERR(bufs))
2600 kfree(bufs); 2590 return PTR_ERR(bufs);
2601 return -EFAULT;
2602 }
2603 result = snd_pcm_lib_writev(substream, bufs, xfern.frames); 2591 result = snd_pcm_lib_writev(substream, bufs, xfern.frames);
2604 kfree(bufs); 2592 kfree(bufs);
2605 __put_user(result, &_xfern->result); 2593 __put_user(result, &_xfern->result);
@@ -2675,13 +2663,11 @@ static int snd_pcm_capture_ioctl1(struct file *file,
2675 return -EFAULT; 2663 return -EFAULT;
2676 if (copy_from_user(&xfern, _xfern, sizeof(xfern))) 2664 if (copy_from_user(&xfern, _xfern, sizeof(xfern)))
2677 return -EFAULT; 2665 return -EFAULT;
2678 bufs = kmalloc(sizeof(void *) * runtime->channels, GFP_KERNEL); 2666
2679 if (bufs == NULL) 2667 bufs = memdup_user(xfern.bufs,
2680 return -ENOMEM; 2668 sizeof(void *) * runtime->channels);
2681 if (copy_from_user(bufs, xfern.bufs, sizeof(void *) * runtime->channels)) { 2669 if (IS_ERR(bufs))
2682 kfree(bufs); 2670 return PTR_ERR(bufs);
2683 return -EFAULT;
2684 }
2685 result = snd_pcm_lib_readv(substream, bufs, xfern.frames); 2671 result = snd_pcm_lib_readv(substream, bufs, xfern.frames);
2686 kfree(bufs); 2672 kfree(bufs);
2687 __put_user(result, &_xfern->result); 2673 __put_user(result, &_xfern->result);
@@ -3312,18 +3298,12 @@ static int snd_pcm_hw_refine_old_user(struct snd_pcm_substream *substream,
3312 int err; 3298 int err;
3313 3299
3314 params = kmalloc(sizeof(*params), GFP_KERNEL); 3300 params = kmalloc(sizeof(*params), GFP_KERNEL);
3315 if (!params) { 3301 if (!params)
3316 err = -ENOMEM; 3302 return -ENOMEM;
3317 goto out;
3318 }
3319 oparams = kmalloc(sizeof(*oparams), GFP_KERNEL);
3320 if (!oparams) {
3321 err = -ENOMEM;
3322 goto out;
3323 }
3324 3303
3325 if (copy_from_user(oparams, _oparams, sizeof(*oparams))) { 3304 oparams = memdup_user(_oparams, sizeof(*oparams));
3326 err = -EFAULT; 3305 if (IS_ERR(oparams)) {
3306 err = PTR_ERR(oparams);
3327 goto out; 3307 goto out;
3328 } 3308 }
3329 snd_pcm_hw_convert_from_old_params(params, oparams); 3309 snd_pcm_hw_convert_from_old_params(params, oparams);
@@ -3333,9 +3313,10 @@ static int snd_pcm_hw_refine_old_user(struct snd_pcm_substream *substream,
3333 if (!err) 3313 if (!err)
3334 err = -EFAULT; 3314 err = -EFAULT;
3335 } 3315 }
3316
3317 kfree(oparams);
3336out: 3318out:
3337 kfree(params); 3319 kfree(params);
3338 kfree(oparams);
3339 return err; 3320 return err;
3340} 3321}
3341 3322
@@ -3347,17 +3328,12 @@ static int snd_pcm_hw_params_old_user(struct snd_pcm_substream *substream,
3347 int err; 3328 int err;
3348 3329
3349 params = kmalloc(sizeof(*params), GFP_KERNEL); 3330 params = kmalloc(sizeof(*params), GFP_KERNEL);
3350 if (!params) { 3331 if (!params)
3351 err = -ENOMEM; 3332 return -ENOMEM;
3352 goto out; 3333
3353 } 3334 oparams = memdup_user(_oparams, sizeof(*oparams));
3354 oparams = kmalloc(sizeof(*oparams), GFP_KERNEL); 3335 if (IS_ERR(oparams)) {
3355 if (!oparams) { 3336 err = PTR_ERR(oparams);
3356 err = -ENOMEM;
3357 goto out;
3358 }
3359 if (copy_from_user(oparams, _oparams, sizeof(*oparams))) {
3360 err = -EFAULT;
3361 goto out; 3337 goto out;
3362 } 3338 }
3363 snd_pcm_hw_convert_from_old_params(params, oparams); 3339 snd_pcm_hw_convert_from_old_params(params, oparams);
@@ -3367,9 +3343,10 @@ static int snd_pcm_hw_params_old_user(struct snd_pcm_substream *substream,
3367 if (!err) 3343 if (!err)
3368 err = -EFAULT; 3344 err = -EFAULT;
3369 } 3345 }
3346
3347 kfree(oparams);
3370out: 3348out:
3371 kfree(params); 3349 kfree(params);
3372 kfree(oparams);
3373 return err; 3350 return err;
3374} 3351}
3375#endif /* CONFIG_SND_SUPPORT_OLD_API */ 3352#endif /* CONFIG_SND_SUPPORT_OLD_API */
diff --git a/sound/core/seq/seq_compat.c b/sound/core/seq/seq_compat.c
index 38693f47c262..c956fe462569 100644
--- a/sound/core/seq/seq_compat.c
+++ b/sound/core/seq/seq_compat.c
@@ -48,12 +48,11 @@ static int snd_seq_call_port_info_ioctl(struct snd_seq_client *client, unsigned
48 struct snd_seq_port_info *data; 48 struct snd_seq_port_info *data;
49 mm_segment_t fs; 49 mm_segment_t fs;
50 50
51 data = kmalloc(sizeof(*data), GFP_KERNEL); 51 data = memdup_user(data32, sizeof(*data32));
52 if (! data) 52 if (IS_ERR(data))
53 return -ENOMEM; 53 return PTR_ERR(data);
54 54
55 if (copy_from_user(data, data32, sizeof(*data32)) || 55 if (get_user(data->flags, &data32->flags) ||
56 get_user(data->flags, &data32->flags) ||
57 get_user(data->time_queue, &data32->time_queue)) 56 get_user(data->time_queue, &data32->time_queue))
58 goto error; 57 goto error;
59 data->kernel = NULL; 58 data->kernel = NULL;
diff --git a/sound/core/timer.c b/sound/core/timer.c
index 3f0050d0b71e..8f8b17ac074d 100644
--- a/sound/core/timer.c
+++ b/sound/core/timer.c
@@ -1395,13 +1395,10 @@ static int snd_timer_user_ginfo(struct file *file,
1395 struct list_head *p; 1395 struct list_head *p;
1396 int err = 0; 1396 int err = 0;
1397 1397
1398 ginfo = kmalloc(sizeof(*ginfo), GFP_KERNEL); 1398 ginfo = memdup_user(_ginfo, sizeof(*ginfo));
1399 if (! ginfo) 1399 if (IS_ERR(ginfo))
1400 return -ENOMEM; 1400 return PTR_ERR(ginfo);
1401 if (copy_from_user(ginfo, _ginfo, sizeof(*ginfo))) { 1401
1402 kfree(ginfo);
1403 return -EFAULT;
1404 }
1405 tid = ginfo->tid; 1402 tid = ginfo->tid;
1406 memset(ginfo, 0, sizeof(*ginfo)); 1403 memset(ginfo, 0, sizeof(*ginfo));
1407 ginfo->tid = tid; 1404 ginfo->tid = tid;
diff --git a/sound/isa/sb/sb16_csp.c b/sound/isa/sb/sb16_csp.c
index 49037d074c71..bdc8dde4e4a2 100644
--- a/sound/isa/sb/sb16_csp.c
+++ b/sound/isa/sb/sb16_csp.c
@@ -684,15 +684,16 @@ static int snd_sb_csp_load(struct snd_sb_csp * p, const unsigned char *buf, int
684 684
685static int snd_sb_csp_load_user(struct snd_sb_csp * p, const unsigned char __user *buf, int size, int load_flags) 685static int snd_sb_csp_load_user(struct snd_sb_csp * p, const unsigned char __user *buf, int size, int load_flags)
686{ 686{
687 int err = -ENOMEM; 687 int err;
688 unsigned char *kbuf = kmalloc(size, GFP_KERNEL); 688 unsigned char *kbuf;
689 if (kbuf) { 689
690 if (copy_from_user(kbuf, buf, size)) 690 kbuf = memdup_user(buf, size);
691 err = -EFAULT; 691 if (IS_ERR(kbuf))
692 else 692 return PTR_ERR(kbuf);
693 err = snd_sb_csp_load(p, kbuf, size, load_flags); 693
694 kfree(kbuf); 694 err = snd_sb_csp_load(p, kbuf, size, load_flags);
695 } 695
696 kfree(kbuf);
696 return err; 697 return err;
697} 698}
698 699
diff --git a/sound/isa/wavefront/wavefront_fx.c b/sound/isa/wavefront/wavefront_fx.c
index a4345fc07561..2bb1cee09255 100644
--- a/sound/isa/wavefront/wavefront_fx.c
+++ b/sound/isa/wavefront/wavefront_fx.c
@@ -202,15 +202,11 @@ snd_wavefront_fx_ioctl (struct snd_hwdep *sdev, struct file *file,
202 "> 512 bytes to FX\n"); 202 "> 512 bytes to FX\n");
203 return -EIO; 203 return -EIO;
204 } 204 }
205 page_data = kmalloc(r.data[2] * sizeof(short), GFP_KERNEL); 205 page_data = memdup_user((unsigned char __user *)
206 if (!page_data) 206 r.data[3],
207 return -ENOMEM; 207 r.data[2] * sizeof(short));
208 if (copy_from_user (page_data, 208 if (IS_ERR(page_data))
209 (unsigned char __user *) r.data[3], 209 return PTR_ERR(page_data);
210 r.data[2] * sizeof(short))) {
211 kfree(page_data);
212 return -EFAULT;
213 }
214 pd = page_data; 210 pd = page_data;
215 } 211 }
216 212
diff --git a/sound/isa/wavefront/wavefront_synth.c b/sound/isa/wavefront/wavefront_synth.c
index beb312cca75b..5d4ff48c4345 100644
--- a/sound/isa/wavefront/wavefront_synth.c
+++ b/sound/isa/wavefront/wavefront_synth.c
@@ -1664,12 +1664,11 @@ snd_wavefront_synth_ioctl (struct snd_hwdep *hw, struct file *file,
1664 break; 1664 break;
1665 1665
1666 case WFCTL_WFCMD: 1666 case WFCTL_WFCMD:
1667 wc = kmalloc(sizeof(*wc), GFP_KERNEL); 1667 wc = memdup_user(argp, sizeof(*wc));
1668 if (! wc) 1668 if (IS_ERR(wc))
1669 return -ENOMEM; 1669 return PTR_ERR(wc);
1670 if (copy_from_user (wc, argp, sizeof (*wc))) 1670
1671 err = -EFAULT; 1671 if (wavefront_synth_control (acard, wc) < 0)
1672 else if (wavefront_synth_control (acard, wc) < 0)
1673 err = -EIO; 1672 err = -EIO;
1674 else if (copy_to_user (argp, wc, sizeof (*wc))) 1673 else if (copy_to_user (argp, wc, sizeof (*wc)))
1675 err = -EFAULT; 1674 err = -EFAULT;
diff --git a/sound/pci/emu10k1/emufx.c b/sound/pci/emu10k1/emufx.c
index 191e1cd9997d..4b302d86f5f2 100644
--- a/sound/pci/emu10k1/emufx.c
+++ b/sound/pci/emu10k1/emufx.c
@@ -2493,24 +2493,17 @@ static int snd_emu10k1_fx8010_ioctl(struct snd_hwdep * hw, struct file *file, un
2493 case SNDRV_EMU10K1_IOCTL_CODE_POKE: 2493 case SNDRV_EMU10K1_IOCTL_CODE_POKE:
2494 if (!capable(CAP_SYS_ADMIN)) 2494 if (!capable(CAP_SYS_ADMIN))
2495 return -EPERM; 2495 return -EPERM;
2496 icode = kmalloc(sizeof(*icode), GFP_KERNEL); 2496
2497 if (icode == NULL) 2497 icode = memdup_user(argp, sizeof(*icode));
2498 return -ENOMEM; 2498 if (IS_ERR(icode))
2499 if (copy_from_user(icode, argp, sizeof(*icode))) { 2499 return PTR_ERR(icode);
2500 kfree(icode);
2501 return -EFAULT;
2502 }
2503 res = snd_emu10k1_icode_poke(emu, icode); 2500 res = snd_emu10k1_icode_poke(emu, icode);
2504 kfree(icode); 2501 kfree(icode);
2505 return res; 2502 return res;
2506 case SNDRV_EMU10K1_IOCTL_CODE_PEEK: 2503 case SNDRV_EMU10K1_IOCTL_CODE_PEEK:
2507 icode = kmalloc(sizeof(*icode), GFP_KERNEL); 2504 icode = memdup_user(argp, sizeof(*icode));
2508 if (icode == NULL) 2505 if (IS_ERR(icode))
2509 return -ENOMEM; 2506 return PTR_ERR(icode);
2510 if (copy_from_user(icode, argp, sizeof(*icode))) {
2511 kfree(icode);
2512 return -EFAULT;
2513 }
2514 res = snd_emu10k1_icode_peek(emu, icode); 2507 res = snd_emu10k1_icode_peek(emu, icode);
2515 if (res == 0 && copy_to_user(argp, icode, sizeof(*icode))) { 2508 if (res == 0 && copy_to_user(argp, icode, sizeof(*icode))) {
2516 kfree(icode); 2509 kfree(icode);
@@ -2519,24 +2512,16 @@ static int snd_emu10k1_fx8010_ioctl(struct snd_hwdep * hw, struct file *file, un
2519 kfree(icode); 2512 kfree(icode);
2520 return res; 2513 return res;
2521 case SNDRV_EMU10K1_IOCTL_PCM_POKE: 2514 case SNDRV_EMU10K1_IOCTL_PCM_POKE:
2522 ipcm = kmalloc(sizeof(*ipcm), GFP_KERNEL); 2515 ipcm = memdup_user(argp, sizeof(*ipcm));
2523 if (ipcm == NULL) 2516 if (IS_ERR(ipcm))
2524 return -ENOMEM; 2517 return PTR_ERR(ipcm);
2525 if (copy_from_user(ipcm, argp, sizeof(*ipcm))) {
2526 kfree(ipcm);
2527 return -EFAULT;
2528 }
2529 res = snd_emu10k1_ipcm_poke(emu, ipcm); 2518 res = snd_emu10k1_ipcm_poke(emu, ipcm);
2530 kfree(ipcm); 2519 kfree(ipcm);
2531 return res; 2520 return res;
2532 case SNDRV_EMU10K1_IOCTL_PCM_PEEK: 2521 case SNDRV_EMU10K1_IOCTL_PCM_PEEK:
2533 ipcm = kzalloc(sizeof(*ipcm), GFP_KERNEL); 2522 ipcm = memdup_user(argp, sizeof(*ipcm));
2534 if (ipcm == NULL) 2523 if (IS_ERR(ipcm))
2535 return -ENOMEM; 2524 return PTR_ERR(ipcm);
2536 if (copy_from_user(ipcm, argp, sizeof(*ipcm))) {
2537 kfree(ipcm);
2538 return -EFAULT;
2539 }
2540 res = snd_emu10k1_ipcm_peek(emu, ipcm); 2525 res = snd_emu10k1_ipcm_peek(emu, ipcm);
2541 if (res == 0 && copy_to_user(argp, ipcm, sizeof(*ipcm))) { 2526 if (res == 0 && copy_to_user(argp, ipcm, sizeof(*ipcm))) {
2542 kfree(ipcm); 2527 kfree(ipcm);
diff --git a/sound/pci/hda/hda_codec.c b/sound/pci/hda/hda_codec.c
index a4e5e5952115..fd6e6f337d10 100644
--- a/sound/pci/hda/hda_codec.c
+++ b/sound/pci/hda/hda_codec.c
@@ -2250,7 +2250,11 @@ int snd_hda_codec_write_cache(struct hda_codec *codec, hda_nid_t nid,
2250 err = bus->ops.command(bus, res); 2250 err = bus->ops.command(bus, res);
2251 if (!err) { 2251 if (!err) {
2252 struct hda_cache_head *c; 2252 struct hda_cache_head *c;
2253 u32 key = build_cmd_cache_key(nid, verb); 2253 u32 key;
2254 /* parm may contain the verb stuff for get/set amp */
2255 verb = verb | (parm >> 8);
2256 parm &= 0xff;
2257 key = build_cmd_cache_key(nid, verb);
2254 c = get_alloc_hash(&codec->cmd_cache, key); 2258 c = get_alloc_hash(&codec->cmd_cache, key);
2255 if (c) 2259 if (c)
2256 c->val = parm; 2260 c->val = parm;
diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c
index 30829ee920c3..bc882f8f163c 100644
--- a/sound/pci/hda/hda_intel.c
+++ b/sound/pci/hda/hda_intel.c
@@ -312,6 +312,9 @@ struct azx_dev {
312 unsigned int period_bytes; /* size of the period in bytes */ 312 unsigned int period_bytes; /* size of the period in bytes */
313 unsigned int frags; /* number for period in the play buffer */ 313 unsigned int frags; /* number for period in the play buffer */
314 unsigned int fifo_size; /* FIFO size */ 314 unsigned int fifo_size; /* FIFO size */
315 unsigned int start_flag: 1; /* stream full start flag */
316 unsigned long start_jiffies; /* start + minimum jiffies */
317 unsigned long min_jiffies; /* minimum jiffies before position is valid */
315 318
316 void __iomem *sd_addr; /* stream descriptor pointer */ 319 void __iomem *sd_addr; /* stream descriptor pointer */
317 320
@@ -330,7 +333,6 @@ struct azx_dev {
330 unsigned int opened :1; 333 unsigned int opened :1;
331 unsigned int running :1; 334 unsigned int running :1;
332 unsigned int irq_pending :1; 335 unsigned int irq_pending :1;
333 unsigned int irq_ignore :1;
334 /* 336 /*
335 * For VIA: 337 * For VIA:
336 * A flag to ensure DMA position is 0 338 * A flag to ensure DMA position is 0
@@ -975,7 +977,7 @@ static irqreturn_t azx_interrupt(int irq, void *dev_id)
975 struct azx *chip = dev_id; 977 struct azx *chip = dev_id;
976 struct azx_dev *azx_dev; 978 struct azx_dev *azx_dev;
977 u32 status; 979 u32 status;
978 int i; 980 int i, ok;
979 981
980 spin_lock(&chip->reg_lock); 982 spin_lock(&chip->reg_lock);
981 983
@@ -991,18 +993,14 @@ static irqreturn_t azx_interrupt(int irq, void *dev_id)
991 azx_sd_writeb(azx_dev, SD_STS, SD_INT_MASK); 993 azx_sd_writeb(azx_dev, SD_STS, SD_INT_MASK);
992 if (!azx_dev->substream || !azx_dev->running) 994 if (!azx_dev->substream || !azx_dev->running)
993 continue; 995 continue;
994 /* ignore the first dummy IRQ (due to pos_adj) */
995 if (azx_dev->irq_ignore) {
996 azx_dev->irq_ignore = 0;
997 continue;
998 }
999 /* check whether this IRQ is really acceptable */ 996 /* check whether this IRQ is really acceptable */
1000 if (azx_position_ok(chip, azx_dev)) { 997 ok = azx_position_ok(chip, azx_dev);
998 if (ok == 1) {
1001 azx_dev->irq_pending = 0; 999 azx_dev->irq_pending = 0;
1002 spin_unlock(&chip->reg_lock); 1000 spin_unlock(&chip->reg_lock);
1003 snd_pcm_period_elapsed(azx_dev->substream); 1001 snd_pcm_period_elapsed(azx_dev->substream);
1004 spin_lock(&chip->reg_lock); 1002 spin_lock(&chip->reg_lock);
1005 } else if (chip->bus && chip->bus->workq) { 1003 } else if (ok == 0 && chip->bus && chip->bus->workq) {
1006 /* bogus IRQ, process it later */ 1004 /* bogus IRQ, process it later */
1007 azx_dev->irq_pending = 1; 1005 azx_dev->irq_pending = 1;
1008 queue_work(chip->bus->workq, 1006 queue_work(chip->bus->workq,
@@ -1088,7 +1086,6 @@ static int azx_setup_periods(struct azx *chip,
1088 bdl = (u32 *)azx_dev->bdl.area; 1086 bdl = (u32 *)azx_dev->bdl.area;
1089 ofs = 0; 1087 ofs = 0;
1090 azx_dev->frags = 0; 1088 azx_dev->frags = 0;
1091 azx_dev->irq_ignore = 0;
1092 pos_adj = bdl_pos_adj[chip->dev_index]; 1089 pos_adj = bdl_pos_adj[chip->dev_index];
1093 if (pos_adj > 0) { 1090 if (pos_adj > 0) {
1094 struct snd_pcm_runtime *runtime = substream->runtime; 1091 struct snd_pcm_runtime *runtime = substream->runtime;
@@ -1109,7 +1106,6 @@ static int azx_setup_periods(struct azx *chip,
1109 &bdl, ofs, pos_adj, 1); 1106 &bdl, ofs, pos_adj, 1);
1110 if (ofs < 0) 1107 if (ofs < 0)
1111 goto error; 1108 goto error;
1112 azx_dev->irq_ignore = 1;
1113 } 1109 }
1114 } else 1110 } else
1115 pos_adj = 0; 1111 pos_adj = 0;
@@ -1155,6 +1151,9 @@ static void azx_stream_reset(struct azx *chip, struct azx_dev *azx_dev)
1155 while (((val = azx_sd_readb(azx_dev, SD_CTL)) & SD_CTL_STREAM_RESET) && 1151 while (((val = azx_sd_readb(azx_dev, SD_CTL)) & SD_CTL_STREAM_RESET) &&
1156 --timeout) 1152 --timeout)
1157 ; 1153 ;
1154
1155 /* reset first position - may not be synced with hw at this time */
1156 *azx_dev->posbuf = 0;
1158} 1157}
1159 1158
1160/* 1159/*
@@ -1409,7 +1408,6 @@ static int azx_pcm_open(struct snd_pcm_substream *substream)
1409 snd_pcm_set_sync(substream); 1408 snd_pcm_set_sync(substream);
1410 mutex_unlock(&chip->open_mutex); 1409 mutex_unlock(&chip->open_mutex);
1411 1410
1412 azx_stream_reset(chip, azx_dev);
1413 return 0; 1411 return 0;
1414} 1412}
1415 1413
@@ -1474,6 +1472,7 @@ static int azx_pcm_prepare(struct snd_pcm_substream *substream)
1474 unsigned int bufsize, period_bytes, format_val; 1472 unsigned int bufsize, period_bytes, format_val;
1475 int err; 1473 int err;
1476 1474
1475 azx_stream_reset(chip, azx_dev);
1477 format_val = snd_hda_calc_stream_format(runtime->rate, 1476 format_val = snd_hda_calc_stream_format(runtime->rate,
1478 runtime->channels, 1477 runtime->channels,
1479 runtime->format, 1478 runtime->format,
@@ -1502,6 +1501,8 @@ static int azx_pcm_prepare(struct snd_pcm_substream *substream)
1502 return err; 1501 return err;
1503 } 1502 }
1504 1503
1504 azx_dev->min_jiffies = (runtime->period_size * HZ) /
1505 (runtime->rate * 2);
1505 azx_setup_controller(chip, azx_dev); 1506 azx_setup_controller(chip, azx_dev);
1506 if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) 1507 if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK)
1507 azx_dev->fifo_size = azx_sd_readw(azx_dev, SD_FIFOSIZE) + 1; 1508 azx_dev->fifo_size = azx_sd_readw(azx_dev, SD_FIFOSIZE) + 1;
@@ -1518,13 +1519,14 @@ static int azx_pcm_trigger(struct snd_pcm_substream *substream, int cmd)
1518 struct azx *chip = apcm->chip; 1519 struct azx *chip = apcm->chip;
1519 struct azx_dev *azx_dev; 1520 struct azx_dev *azx_dev;
1520 struct snd_pcm_substream *s; 1521 struct snd_pcm_substream *s;
1521 int start, nsync = 0, sbits = 0; 1522 int rstart = 0, start, nsync = 0, sbits = 0;
1522 int nwait, timeout; 1523 int nwait, timeout;
1523 1524
1524 switch (cmd) { 1525 switch (cmd) {
1526 case SNDRV_PCM_TRIGGER_START:
1527 rstart = 1;
1525 case SNDRV_PCM_TRIGGER_PAUSE_RELEASE: 1528 case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:
1526 case SNDRV_PCM_TRIGGER_RESUME: 1529 case SNDRV_PCM_TRIGGER_RESUME:
1527 case SNDRV_PCM_TRIGGER_START:
1528 start = 1; 1530 start = 1;
1529 break; 1531 break;
1530 case SNDRV_PCM_TRIGGER_PAUSE_PUSH: 1532 case SNDRV_PCM_TRIGGER_PAUSE_PUSH:
@@ -1554,6 +1556,10 @@ static int azx_pcm_trigger(struct snd_pcm_substream *substream, int cmd)
1554 if (s->pcm->card != substream->pcm->card) 1556 if (s->pcm->card != substream->pcm->card)
1555 continue; 1557 continue;
1556 azx_dev = get_azx_dev(s); 1558 azx_dev = get_azx_dev(s);
1559 if (rstart) {
1560 azx_dev->start_flag = 1;
1561 azx_dev->start_jiffies = jiffies + azx_dev->min_jiffies;
1562 }
1557 if (start) 1563 if (start)
1558 azx_stream_start(chip, azx_dev); 1564 azx_stream_start(chip, azx_dev);
1559 else 1565 else
@@ -1703,6 +1709,11 @@ static int azx_position_ok(struct azx *chip, struct azx_dev *azx_dev)
1703{ 1709{
1704 unsigned int pos; 1710 unsigned int pos;
1705 1711
1712 if (azx_dev->start_flag &&
1713 time_before_eq(jiffies, azx_dev->start_jiffies))
1714 return -1; /* bogus (too early) interrupt */
1715 azx_dev->start_flag = 0;
1716
1706 pos = azx_get_position(chip, azx_dev); 1717 pos = azx_get_position(chip, azx_dev);
1707 if (chip->position_fix == POS_FIX_AUTO) { 1718 if (chip->position_fix == POS_FIX_AUTO) {
1708 if (!pos) { 1719 if (!pos) {
@@ -2260,11 +2271,11 @@ static int __devinit azx_create(struct snd_card *card, struct pci_dev *pci,
2260 gcap &= ~0x01; 2271 gcap &= ~0x01;
2261 2272
2262 /* allow 64bit DMA address if supported by H/W */ 2273 /* allow 64bit DMA address if supported by H/W */
2263 if ((gcap & 0x01) && !pci_set_dma_mask(pci, DMA_64BIT_MASK)) 2274 if ((gcap & 0x01) && !pci_set_dma_mask(pci, DMA_BIT_MASK(64)))
2264 pci_set_consistent_dma_mask(pci, DMA_64BIT_MASK); 2275 pci_set_consistent_dma_mask(pci, DMA_BIT_MASK(64));
2265 else { 2276 else {
2266 pci_set_dma_mask(pci, DMA_32BIT_MASK); 2277 pci_set_dma_mask(pci, DMA_BIT_MASK(32));
2267 pci_set_consistent_dma_mask(pci, DMA_32BIT_MASK); 2278 pci_set_consistent_dma_mask(pci, DMA_BIT_MASK(32));
2268 } 2279 }
2269 2280
2270 /* read number of streams from GCAP register instead of using 2281 /* read number of streams from GCAP register instead of using
diff --git a/sound/pci/hda/patch_conexant.c b/sound/pci/hda/patch_conexant.c
index 1f2ad76ca94b..56ce19e68cb5 100644
--- a/sound/pci/hda/patch_conexant.c
+++ b/sound/pci/hda/patch_conexant.c
@@ -350,12 +350,20 @@ static int conexant_mux_enum_put(struct snd_kcontrol *kcontrol,
350} 350}
351 351
352#ifdef CONFIG_SND_JACK 352#ifdef CONFIG_SND_JACK
353static void conexant_free_jack_priv(struct snd_jack *jack)
354{
355 struct conexant_jack *jacks = jack->private_data;
356 jacks->nid = 0;
357 jacks->jack = NULL;
358}
359
353static int conexant_add_jack(struct hda_codec *codec, 360static int conexant_add_jack(struct hda_codec *codec,
354 hda_nid_t nid, int type) 361 hda_nid_t nid, int type)
355{ 362{
356 struct conexant_spec *spec; 363 struct conexant_spec *spec;
357 struct conexant_jack *jack; 364 struct conexant_jack *jack;
358 const char *name; 365 const char *name;
366 int err;
359 367
360 spec = codec->spec; 368 spec = codec->spec;
361 snd_array_init(&spec->jacks, sizeof(*jack), 32); 369 snd_array_init(&spec->jacks, sizeof(*jack), 32);
@@ -368,7 +376,12 @@ static int conexant_add_jack(struct hda_codec *codec,
368 jack->nid = nid; 376 jack->nid = nid;
369 jack->type = type; 377 jack->type = type;
370 378
371 return snd_jack_new(codec->bus->card, name, type, &jack->jack); 379 err = snd_jack_new(codec->bus->card, name, type, &jack->jack);
380 if (err < 0)
381 return err;
382 jack->jack->private_data = jack;
383 jack->jack->private_free = conexant_free_jack_priv;
384 return 0;
372} 385}
373 386
374static void conexant_report_jack(struct hda_codec *codec, hda_nid_t nid) 387static void conexant_report_jack(struct hda_codec *codec, hda_nid_t nid)
@@ -455,8 +468,10 @@ static void conexant_free(struct hda_codec *codec)
455 if (spec->jacks.list) { 468 if (spec->jacks.list) {
456 struct conexant_jack *jacks = spec->jacks.list; 469 struct conexant_jack *jacks = spec->jacks.list;
457 int i; 470 int i;
458 for (i = 0; i < spec->jacks.used; i++) 471 for (i = 0; i < spec->jacks.used; i++, jacks++) {
459 snd_device_free(codec->bus->card, &jacks[i].jack); 472 if (jacks->jack)
473 snd_device_free(codec->bus->card, jacks->jack);
474 }
460 snd_array_free(&spec->jacks); 475 snd_array_free(&spec->jacks);
461 } 476 }
462#endif 477#endif
diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c
index f35e58a2d921..6ed787eedd06 100644
--- a/sound/pci/hda/patch_realtek.c
+++ b/sound/pci/hda/patch_realtek.c
@@ -8742,10 +8742,9 @@ static struct snd_pci_quirk alc883_cfg_tbl[] = {
8742 SND_PCI_QUIRK_VENDOR(0x1558, "Clevo laptop", ALC883_LAPTOP_EAPD), 8742 SND_PCI_QUIRK_VENDOR(0x1558, "Clevo laptop", ALC883_LAPTOP_EAPD),
8743 SND_PCI_QUIRK(0x15d9, 0x8780, "Supermicro PDSBA", ALC883_3ST_6ch), 8743 SND_PCI_QUIRK(0x15d9, 0x8780, "Supermicro PDSBA", ALC883_3ST_6ch),
8744 SND_PCI_QUIRK(0x161f, 0x2054, "Medion laptop", ALC883_MEDION), 8744 SND_PCI_QUIRK(0x161f, 0x2054, "Medion laptop", ALC883_MEDION),
8745 SND_PCI_QUIRK(0x1734, 0x1107, "FSC AMILO Xi2550", 8745 SND_PCI_QUIRK_MASK(0x1734, 0xfff0, 0x1100, "FSC AMILO Xi/Pi25xx",
8746 ALC883_FUJITSU_PI2515), 8746 ALC883_FUJITSU_PI2515),
8747 SND_PCI_QUIRK(0x1734, 0x1108, "Fujitsu AMILO Pi2515", ALC883_FUJITSU_PI2515), 8747 SND_PCI_QUIRK_MASK(0x1734, 0xfff0, 0x1130, "Fujitsu AMILO Xa35xx",
8748 SND_PCI_QUIRK(0x1734, 0x113d, "Fujitsu AMILO Xa3530",
8749 ALC888_FUJITSU_XA3530), 8748 ALC888_FUJITSU_XA3530),
8750 SND_PCI_QUIRK(0x17aa, 0x101e, "Lenovo 101e", ALC883_LENOVO_101E_2ch), 8749 SND_PCI_QUIRK(0x17aa, 0x101e, "Lenovo 101e", ALC883_LENOVO_101E_2ch),
8751 SND_PCI_QUIRK(0x17aa, 0x2085, "Lenovo NB0763", ALC883_LENOVO_NB0763), 8750 SND_PCI_QUIRK(0x17aa, 0x2085, "Lenovo NB0763", ALC883_LENOVO_NB0763),
diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c
index 61996a2f45df..ce30b459aee6 100644
--- a/sound/pci/hda/patch_sigmatel.c
+++ b/sound/pci/hda/patch_sigmatel.c
@@ -3851,6 +3851,15 @@ static void stac_gpio_set(struct hda_codec *codec, unsigned int mask,
3851 AC_VERB_SET_GPIO_DATA, gpiostate); /* sync */ 3851 AC_VERB_SET_GPIO_DATA, gpiostate); /* sync */
3852} 3852}
3853 3853
3854#ifdef CONFIG_SND_JACK
3855static void stac92xx_free_jack_priv(struct snd_jack *jack)
3856{
3857 struct sigmatel_jack *jacks = jack->private_data;
3858 jacks->nid = 0;
3859 jacks->jack = NULL;
3860}
3861#endif
3862
3854static int stac92xx_add_jack(struct hda_codec *codec, 3863static int stac92xx_add_jack(struct hda_codec *codec,
3855 hda_nid_t nid, int type) 3864 hda_nid_t nid, int type)
3856{ 3865{
@@ -3860,6 +3869,7 @@ static int stac92xx_add_jack(struct hda_codec *codec,
3860 int def_conf = snd_hda_codec_get_pincfg(codec, nid); 3869 int def_conf = snd_hda_codec_get_pincfg(codec, nid);
3861 int connectivity = get_defcfg_connect(def_conf); 3870 int connectivity = get_defcfg_connect(def_conf);
3862 char name[32]; 3871 char name[32];
3872 int err;
3863 3873
3864 if (connectivity && connectivity != AC_JACK_PORT_FIXED) 3874 if (connectivity && connectivity != AC_JACK_PORT_FIXED)
3865 return 0; 3875 return 0;
@@ -3876,10 +3886,15 @@ static int stac92xx_add_jack(struct hda_codec *codec,
3876 snd_hda_get_jack_connectivity(def_conf), 3886 snd_hda_get_jack_connectivity(def_conf),
3877 snd_hda_get_jack_location(def_conf)); 3887 snd_hda_get_jack_location(def_conf));
3878 3888
3879 return snd_jack_new(codec->bus->card, name, type, &jack->jack); 3889 err = snd_jack_new(codec->bus->card, name, type, &jack->jack);
3880#else 3890 if (err < 0) {
3881 return 0; 3891 jack->nid = 0;
3892 return err;
3893 }
3894 jack->jack->private_data = jack;
3895 jack->jack->private_free = stac92xx_free_jack_priv;
3882#endif 3896#endif
3897 return 0;
3883} 3898}
3884 3899
3885static int stac_add_event(struct sigmatel_spec *spec, hda_nid_t nid, 3900static int stac_add_event(struct sigmatel_spec *spec, hda_nid_t nid,
@@ -4138,8 +4153,10 @@ static void stac92xx_free_jacks(struct hda_codec *codec)
4138 if (!codec->bus->shutdown && spec->jacks.list) { 4153 if (!codec->bus->shutdown && spec->jacks.list) {
4139 struct sigmatel_jack *jacks = spec->jacks.list; 4154 struct sigmatel_jack *jacks = spec->jacks.list;
4140 int i; 4155 int i;
4141 for (i = 0; i < spec->jacks.used; i++) 4156 for (i = 0; i < spec->jacks.used; i++, jacks++) {
4142 snd_device_free(codec->bus->card, &jacks[i].jack); 4157 if (jacks->jack)
4158 snd_device_free(codec->bus->card, jacks->jack);
4159 }
4143 } 4160 }
4144 snd_array_free(&spec->jacks); 4161 snd_array_free(&spec->jacks);
4145#endif 4162#endif
diff --git a/sound/pci/intel8x0.c b/sound/pci/intel8x0.c
index 57648810eaf1..5dced5b79387 100644
--- a/sound/pci/intel8x0.c
+++ b/sound/pci/intel8x0.c
@@ -355,6 +355,9 @@ struct ichdev {
355 unsigned int fragsize1; 355 unsigned int fragsize1;
356 unsigned int position; 356 unsigned int position;
357 unsigned int pos_shift; 357 unsigned int pos_shift;
358 unsigned int last_pos;
359 unsigned long last_pos_jiffies;
360 unsigned int jiffy_to_bytes;
358 int frags; 361 int frags;
359 int lvi; 362 int lvi;
360 int lvi_frag; 363 int lvi_frag;
@@ -838,7 +841,10 @@ static int snd_intel8x0_pcm_trigger(struct snd_pcm_substream *substream, int cmd
838 ichdev->suspended = 0; 841 ichdev->suspended = 0;
839 /* fallthru */ 842 /* fallthru */
840 case SNDRV_PCM_TRIGGER_START: 843 case SNDRV_PCM_TRIGGER_START:
844 case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:
841 val = ICH_IOCE | ICH_STARTBM; 845 val = ICH_IOCE | ICH_STARTBM;
846 ichdev->last_pos = ichdev->position;
847 ichdev->last_pos_jiffies = jiffies;
842 break; 848 break;
843 case SNDRV_PCM_TRIGGER_SUSPEND: 849 case SNDRV_PCM_TRIGGER_SUSPEND:
844 ichdev->suspended = 1; 850 ichdev->suspended = 1;
@@ -849,9 +855,6 @@ static int snd_intel8x0_pcm_trigger(struct snd_pcm_substream *substream, int cmd
849 case SNDRV_PCM_TRIGGER_PAUSE_PUSH: 855 case SNDRV_PCM_TRIGGER_PAUSE_PUSH:
850 val = ICH_IOCE; 856 val = ICH_IOCE;
851 break; 857 break;
852 case SNDRV_PCM_TRIGGER_PAUSE_RELEASE:
853 val = ICH_IOCE | ICH_STARTBM;
854 break;
855 default: 858 default:
856 return -EINVAL; 859 return -EINVAL;
857 } 860 }
@@ -1045,6 +1048,7 @@ static int snd_intel8x0_pcm_prepare(struct snd_pcm_substream *substream)
1045 ichdev->pos_shift = (runtime->sample_bits > 16) ? 2 : 1; 1048 ichdev->pos_shift = (runtime->sample_bits > 16) ? 2 : 1;
1046 } 1049 }
1047 snd_intel8x0_setup_periods(chip, ichdev); 1050 snd_intel8x0_setup_periods(chip, ichdev);
1051 ichdev->jiffy_to_bytes = (runtime->rate * 4 * ichdev->pos_shift) / HZ;
1048 return 0; 1052 return 0;
1049} 1053}
1050 1054
@@ -1053,7 +1057,7 @@ static snd_pcm_uframes_t snd_intel8x0_pcm_pointer(struct snd_pcm_substream *subs
1053 struct intel8x0 *chip = snd_pcm_substream_chip(substream); 1057 struct intel8x0 *chip = snd_pcm_substream_chip(substream);
1054 struct ichdev *ichdev = get_ichdev(substream); 1058 struct ichdev *ichdev = get_ichdev(substream);
1055 size_t ptr1, ptr; 1059 size_t ptr1, ptr;
1056 int civ, timeout = 100; 1060 int civ, timeout = 10;
1057 unsigned int position; 1061 unsigned int position;
1058 1062
1059 spin_lock(&chip->reg_lock); 1063 spin_lock(&chip->reg_lock);
@@ -1069,9 +1073,19 @@ static snd_pcm_uframes_t snd_intel8x0_pcm_pointer(struct snd_pcm_substream *subs
1069 ptr1 == igetword(chip, ichdev->reg_offset + ichdev->roff_picb)) 1073 ptr1 == igetword(chip, ichdev->reg_offset + ichdev->roff_picb))
1070 break; 1074 break;
1071 } while (timeout--); 1075 } while (timeout--);
1072 ptr1 <<= ichdev->pos_shift; 1076 if (ptr1 != 0) {
1073 ptr = ichdev->fragsize1 - ptr1; 1077 ptr1 <<= ichdev->pos_shift;
1074 ptr += position; 1078 ptr = ichdev->fragsize1 - ptr1;
1079 ptr += position;
1080 ichdev->last_pos = ptr;
1081 ichdev->last_pos_jiffies = jiffies;
1082 } else {
1083 ptr1 = jiffies - ichdev->last_pos_jiffies;
1084 if (ptr1)
1085 ptr1 -= 1;
1086 ptr = ichdev->last_pos + ptr1 * ichdev->jiffy_to_bytes;
1087 ptr %= ichdev->size;
1088 }
1075 spin_unlock(&chip->reg_lock); 1089 spin_unlock(&chip->reg_lock);
1076 if (ptr >= ichdev->size) 1090 if (ptr >= ichdev->size)
1077 return 0; 1091 return 0;
@@ -2661,12 +2675,14 @@ static void __devinit intel8x0_measure_ac97_clock(struct intel8x0 *chip)
2661 struct snd_pcm_substream *subs; 2675 struct snd_pcm_substream *subs;
2662 struct ichdev *ichdev; 2676 struct ichdev *ichdev;
2663 unsigned long port; 2677 unsigned long port;
2664 unsigned long pos, t; 2678 unsigned long pos, pos1, t;
2665 struct timeval start_time, stop_time; 2679 int civ, timeout = 1000, attempt = 1;
2680 struct timespec start_time, stop_time;
2666 2681
2667 if (chip->ac97_bus->clock != 48000) 2682 if (chip->ac97_bus->clock != 48000)
2668 return; /* specified in module option */ 2683 return; /* specified in module option */
2669 2684
2685 __again:
2670 subs = chip->pcm[0]->streams[0].substream; 2686 subs = chip->pcm[0]->streams[0].substream;
2671 if (! subs || subs->dma_buffer.bytes < INTEL8X0_TESTBUF_SIZE) { 2687 if (! subs || subs->dma_buffer.bytes < INTEL8X0_TESTBUF_SIZE) {
2672 snd_printk(KERN_WARNING "no playback buffer allocated - aborting measure ac97 clock\n"); 2688 snd_printk(KERN_WARNING "no playback buffer allocated - aborting measure ac97 clock\n");
@@ -2674,7 +2690,7 @@ static void __devinit intel8x0_measure_ac97_clock(struct intel8x0 *chip)
2674 } 2690 }
2675 ichdev = &chip->ichd[ICHD_PCMOUT]; 2691 ichdev = &chip->ichd[ICHD_PCMOUT];
2676 ichdev->physbuf = subs->dma_buffer.addr; 2692 ichdev->physbuf = subs->dma_buffer.addr;
2677 ichdev->size = chip->ichd[ICHD_PCMOUT].fragsize = INTEL8X0_TESTBUF_SIZE; 2693 ichdev->size = ichdev->fragsize = INTEL8X0_TESTBUF_SIZE;
2678 ichdev->substream = NULL; /* don't process interrupts */ 2694 ichdev->substream = NULL; /* don't process interrupts */
2679 2695
2680 /* set rate */ 2696 /* set rate */
@@ -2693,16 +2709,31 @@ static void __devinit intel8x0_measure_ac97_clock(struct intel8x0 *chip)
2693 iputbyte(chip, port + ICH_REG_OFF_CR, ICH_IOCE); 2709 iputbyte(chip, port + ICH_REG_OFF_CR, ICH_IOCE);
2694 iputdword(chip, ICHREG(ALI_DMACR), 1 << ichdev->ali_slot); 2710 iputdword(chip, ICHREG(ALI_DMACR), 1 << ichdev->ali_slot);
2695 } 2711 }
2696 do_gettimeofday(&start_time); 2712 do_posix_clock_monotonic_gettime(&start_time);
2697 spin_unlock_irq(&chip->reg_lock); 2713 spin_unlock_irq(&chip->reg_lock);
2698 msleep(50); 2714 msleep(50);
2699 spin_lock_irq(&chip->reg_lock); 2715 spin_lock_irq(&chip->reg_lock);
2700 /* check the position */ 2716 /* check the position */
2701 pos = ichdev->fragsize1; 2717 do {
2702 pos -= igetword(chip, ichdev->reg_offset + ichdev->roff_picb) << ichdev->pos_shift; 2718 civ = igetbyte(chip, ichdev->reg_offset + ICH_REG_OFF_CIV);
2703 pos += ichdev->position; 2719 pos1 = igetword(chip, ichdev->reg_offset + ichdev->roff_picb);
2720 if (pos1 == 0) {
2721 udelay(10);
2722 continue;
2723 }
2724 if (civ == igetbyte(chip, ichdev->reg_offset + ICH_REG_OFF_CIV) &&
2725 pos1 == igetword(chip, ichdev->reg_offset + ichdev->roff_picb))
2726 break;
2727 } while (timeout--);
2728 if (pos1 == 0) { /* oops, this value is not reliable */
2729 pos = 0;
2730 } else {
2731 pos = ichdev->fragsize1;
2732 pos -= pos1 << ichdev->pos_shift;
2733 pos += ichdev->position;
2734 }
2704 chip->in_measurement = 0; 2735 chip->in_measurement = 0;
2705 do_gettimeofday(&stop_time); 2736 do_posix_clock_monotonic_gettime(&stop_time);
2706 /* stop */ 2737 /* stop */
2707 if (chip->device_type == DEVICE_ALI) { 2738 if (chip->device_type == DEVICE_ALI) {
2708 iputdword(chip, ICHREG(ALI_DMACR), 1 << (ichdev->ali_slot + 16)); 2739 iputdword(chip, ICHREG(ALI_DMACR), 1 << (ichdev->ali_slot + 16));
@@ -2717,19 +2748,37 @@ static void __devinit intel8x0_measure_ac97_clock(struct intel8x0 *chip)
2717 iputbyte(chip, port + ICH_REG_OFF_CR, ICH_RESETREGS); 2748 iputbyte(chip, port + ICH_REG_OFF_CR, ICH_RESETREGS);
2718 spin_unlock_irq(&chip->reg_lock); 2749 spin_unlock_irq(&chip->reg_lock);
2719 2750
2751 if (pos == 0) {
2752 snd_printk(KERN_ERR "intel8x0: measure - unreliable DMA position..\n");
2753 __retry:
2754 if (attempt < 2) {
2755 attempt++;
2756 goto __again;
2757 }
2758 return;
2759 }
2760
2761 pos /= 4;
2720 t = stop_time.tv_sec - start_time.tv_sec; 2762 t = stop_time.tv_sec - start_time.tv_sec;
2721 t *= 1000000; 2763 t *= 1000000;
2722 t += stop_time.tv_usec - start_time.tv_usec; 2764 t += (stop_time.tv_nsec - start_time.tv_nsec) / 1000;
2723 printk(KERN_INFO "%s: measured %lu usecs\n", __func__, t); 2765 printk(KERN_INFO "%s: measured %lu usecs (%lu samples)\n", __func__, t, pos);
2724 if (t == 0) { 2766 if (t == 0) {
2725 snd_printk(KERN_ERR "?? calculation error..\n"); 2767 snd_printk(KERN_ERR "intel8x0: ?? calculation error..\n");
2726 return; 2768 goto __retry;
2727 } 2769 }
2728 pos = (pos / 4) * 1000; 2770 pos *= 1000;
2729 pos = (pos / t) * 1000 + ((pos % t) * 1000) / t; 2771 pos = (pos / t) * 1000 + ((pos % t) * 1000) / t;
2730 if (pos < 40000 || pos >= 60000) 2772 if (pos < 40000 || pos >= 60000) {
2731 /* abnormal value. hw problem? */ 2773 /* abnormal value. hw problem? */
2732 printk(KERN_INFO "intel8x0: measured clock %ld rejected\n", pos); 2774 printk(KERN_INFO "intel8x0: measured clock %ld rejected\n", pos);
2775 goto __retry;
2776 } else if (pos > 40500 && pos < 41500)
2777 /* first exception - 41000Hz reference clock */
2778 chip->ac97_bus->clock = 41000;
2779 else if (pos > 43600 && pos < 44600)
2780 /* second exception - 44100HZ reference clock */
2781 chip->ac97_bus->clock = 44100;
2733 else if (pos < 47500 || pos > 48500) 2782 else if (pos < 47500 || pos > 48500)
2734 /* not 48000Hz, tuning the clock.. */ 2783 /* not 48000Hz, tuning the clock.. */
2735 chip->ac97_bus->clock = (chip->ac97_bus->clock * 48000) / pos; 2784 chip->ac97_bus->clock = (chip->ac97_bus->clock * 48000) / pos;
diff --git a/sound/soc/pxa/magician.c b/sound/soc/pxa/magician.c
index f7c4544f7859..0625c342a1c9 100644
--- a/sound/soc/pxa/magician.c
+++ b/sound/soc/pxa/magician.c
@@ -27,8 +27,6 @@
27#include <sound/soc.h> 27#include <sound/soc.h>
28#include <sound/soc-dapm.h> 28#include <sound/soc-dapm.h>
29 29
30#include <mach/pxa-regs.h>
31#include <mach/hardware.h>
32#include <mach/magician.h> 30#include <mach/magician.h>
33#include <asm/mach-types.h> 31#include <asm/mach-types.h>
34#include "../codecs/uda1380.h" 32#include "../codecs/uda1380.h"
diff --git a/sound/soc/s3c24xx/Kconfig b/sound/soc/s3c24xx/Kconfig
index 2f3a21eee051..df494d1e346f 100644
--- a/sound/soc/s3c24xx/Kconfig
+++ b/sound/soc/s3c24xx/Kconfig
@@ -1,10 +1,10 @@
1config SND_S3C24XX_SOC 1config SND_S3C24XX_SOC
2 tristate "SoC Audio for the Samsung S3CXXXX chips" 2 tristate "SoC Audio for the Samsung S3CXXXX chips"
3 depends on ARCH_S3C2410 || ARCH_S3C64XX 3 depends on ARCH_S3C2410
4 help 4 help
5 Say Y or M if you want to add support for codecs attached to 5 Say Y or M if you want to add support for codecs attached to
6 the S3C24XX and S3C64XX AC97, I2S or SSP interface. You will 6 the S3C24XX AC97 or I2S interfaces. You will also need to
7 also need to select the audio interfaces to support below. 7 select the audio interfaces to support below.
8 8
9config SND_S3C24XX_SOC_I2S 9config SND_S3C24XX_SOC_I2S
10 tristate 10 tristate
diff --git a/sound/usb/caiaq/Makefile b/sound/usb/caiaq/Makefile
index 23dadd5a11cd..388999653aaa 100644
--- a/sound/usb/caiaq/Makefile
+++ b/sound/usb/caiaq/Makefile
@@ -1,4 +1,4 @@
1snd-usb-caiaq-y := caiaq-device.o caiaq-audio.o caiaq-midi.o caiaq-control.o 1snd-usb-caiaq-y := device.o audio.o midi.o control.o
2snd-usb-caiaq-$(CONFIG_SND_USB_CAIAQ_INPUT) += caiaq-input.o 2snd-usb-caiaq-$(CONFIG_SND_USB_CAIAQ_INPUT) += input.o
3 3
4obj-$(CONFIG_SND_USB_CAIAQ) += snd-usb-caiaq.o 4obj-$(CONFIG_SND_USB_CAIAQ) += snd-usb-caiaq.o
diff --git a/sound/usb/caiaq/caiaq-audio.c b/sound/usb/caiaq/audio.c
index 08d51e0c9fea..3f45c0fe61ab 100644
--- a/sound/usb/caiaq/caiaq-audio.c
+++ b/sound/usb/caiaq/audio.c
@@ -16,20 +16,14 @@
16 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA 16 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17*/ 17*/
18 18
19#include <linux/spinlock.h>
19#include <linux/init.h> 20#include <linux/init.h>
20#include <linux/module.h>
21#include <linux/moduleparam.h>
22#include <linux/interrupt.h>
23#include <linux/usb.h> 21#include <linux/usb.h>
24#include <linux/spinlock.h>
25#include <sound/core.h> 22#include <sound/core.h>
26#include <sound/initval.h>
27#include <sound/pcm.h> 23#include <sound/pcm.h>
28#include <sound/rawmidi.h>
29#include <linux/input.h>
30 24
31#include "caiaq-device.h" 25#include "device.h"
32#include "caiaq-audio.h" 26#include "audio.h"
33 27
34#define N_URBS 32 28#define N_URBS 32
35#define CLOCK_DRIFT_TOLERANCE 5 29#define CLOCK_DRIFT_TOLERANCE 5
diff --git a/sound/usb/caiaq/caiaq-audio.h b/sound/usb/caiaq/audio.h
index 8ab1f8d9529e..8ab1f8d9529e 100644
--- a/sound/usb/caiaq/caiaq-audio.h
+++ b/sound/usb/caiaq/audio.h
diff --git a/sound/usb/caiaq/caiaq-control.c b/sound/usb/caiaq/control.c
index e92c2bbf4fe9..537102ba6b9d 100644
--- a/sound/usb/caiaq/caiaq-control.c
+++ b/sound/usb/caiaq/control.c
@@ -18,17 +18,13 @@
18 */ 18 */
19 19
20#include <linux/init.h> 20#include <linux/init.h>
21#include <linux/interrupt.h>
22#include <linux/usb.h> 21#include <linux/usb.h>
22#include <sound/control.h>
23#include <sound/core.h> 23#include <sound/core.h>
24#include <sound/initval.h>
25#include <sound/pcm.h> 24#include <sound/pcm.h>
26#include <sound/rawmidi.h>
27#include <sound/control.h>
28#include <linux/input.h>
29 25
30#include "caiaq-device.h" 26#include "device.h"
31#include "caiaq-control.h" 27#include "control.h"
32 28
33#define CNT_INTVAL 0x10000 29#define CNT_INTVAL 0x10000
34 30
diff --git a/sound/usb/caiaq/caiaq-control.h b/sound/usb/caiaq/control.h
index 2e7ab1aa4fb3..2e7ab1aa4fb3 100644
--- a/sound/usb/caiaq/caiaq-control.h
+++ b/sound/usb/caiaq/control.h
diff --git a/sound/usb/caiaq/caiaq-device.c b/sound/usb/caiaq/device.c
index cf573a982fdc..6d517705da0e 100644
--- a/sound/usb/caiaq/caiaq-device.c
+++ b/sound/usb/caiaq/device.c
@@ -19,27 +19,20 @@
19 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA 19 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
20*/ 20*/
21 21
22#include <linux/init.h>
23#include <linux/module.h>
24#include <linux/moduleparam.h> 22#include <linux/moduleparam.h>
25#include <linux/interrupt.h> 23#include <linux/interrupt.h>
24#include <linux/module.h>
25#include <linux/init.h>
26#include <linux/usb.h> 26#include <linux/usb.h>
27#include <linux/input.h>
28#include <linux/spinlock.h>
29#include <sound/core.h>
30#include <sound/initval.h> 27#include <sound/initval.h>
28#include <sound/core.h>
31#include <sound/pcm.h> 29#include <sound/pcm.h>
32#include <sound/rawmidi.h>
33#include <sound/control.h>
34
35#include "caiaq-device.h"
36#include "caiaq-audio.h"
37#include "caiaq-midi.h"
38#include "caiaq-control.h"
39 30
40#ifdef CONFIG_SND_USB_CAIAQ_INPUT 31#include "device.h"
41#include "caiaq-input.h" 32#include "audio.h"
42#endif 33#include "midi.h"
34#include "control.h"
35#include "input.h"
43 36
44MODULE_AUTHOR("Daniel Mack <daniel@caiaq.de>"); 37MODULE_AUTHOR("Daniel Mack <daniel@caiaq.de>");
45MODULE_DESCRIPTION("caiaq USB audio, version 1.3.13"); 38MODULE_DESCRIPTION("caiaq USB audio, version 1.3.13");
diff --git a/sound/usb/caiaq/caiaq-device.h b/sound/usb/caiaq/device.h
index 4cce1ad7493d..4cce1ad7493d 100644
--- a/sound/usb/caiaq/caiaq-device.h
+++ b/sound/usb/caiaq/device.h
diff --git a/sound/usb/caiaq/caiaq-input.c b/sound/usb/caiaq/input.c
index f743847a5e5a..a48d309bd94c 100644
--- a/sound/usb/caiaq/caiaq-input.c
+++ b/sound/usb/caiaq/input.c
@@ -17,17 +17,12 @@
17*/ 17*/
18 18
19#include <linux/init.h> 19#include <linux/init.h>
20#include <linux/module.h>
21#include <linux/moduleparam.h>
22#include <linux/input.h>
23#include <linux/usb.h> 20#include <linux/usb.h>
24#include <linux/usb/input.h> 21#include <linux/usb/input.h>
25#include <linux/spinlock.h>
26#include <sound/core.h>
27#include <sound/rawmidi.h>
28#include <sound/pcm.h> 22#include <sound/pcm.h>
29#include "caiaq-device.h" 23
30#include "caiaq-input.h" 24#include "device.h"
25#include "input.h"
31 26
32static unsigned short keycode_ak1[] = { KEY_C, KEY_B, KEY_A }; 27static unsigned short keycode_ak1[] = { KEY_C, KEY_B, KEY_A };
33static unsigned short keycode_rk2[] = { KEY_1, KEY_2, KEY_3, KEY_4, 28static unsigned short keycode_rk2[] = { KEY_1, KEY_2, KEY_3, KEY_4,
diff --git a/sound/usb/caiaq/caiaq-input.h b/sound/usb/caiaq/input.h
index ced535577864..ced535577864 100644
--- a/sound/usb/caiaq/caiaq-input.h
+++ b/sound/usb/caiaq/input.h
diff --git a/sound/usb/caiaq/caiaq-midi.c b/sound/usb/caiaq/midi.c
index f19fd360c936..8fa8cd88d763 100644
--- a/sound/usb/caiaq/caiaq-midi.c
+++ b/sound/usb/caiaq/midi.c
@@ -16,20 +16,13 @@
16 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA 16 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
17*/ 17*/
18 18
19#include <linux/init.h>
20#include <linux/module.h>
21#include <linux/moduleparam.h>
22#include <linux/interrupt.h>
23#include <linux/usb.h> 19#include <linux/usb.h>
24#include <linux/input.h>
25#include <linux/spinlock.h>
26#include <sound/core.h>
27#include <sound/rawmidi.h> 20#include <sound/rawmidi.h>
21#include <sound/core.h>
28#include <sound/pcm.h> 22#include <sound/pcm.h>
29 23
30#include "caiaq-device.h" 24#include "device.h"
31#include "caiaq-midi.h" 25#include "midi.h"
32
33 26
34static int snd_usb_caiaq_midi_input_open(struct snd_rawmidi_substream *substream) 27static int snd_usb_caiaq_midi_input_open(struct snd_rawmidi_substream *substream)
35{ 28{
diff --git a/sound/usb/caiaq/caiaq-midi.h b/sound/usb/caiaq/midi.h
index 9d16db027fc3..9d16db027fc3 100644
--- a/sound/usb/caiaq/caiaq-midi.h
+++ b/sound/usb/caiaq/midi.h
diff --git a/sound/usb/usx2y/us122l.c b/sound/usb/usx2y/us122l.c
index 98276aafefe6..012ff1f6f8af 100644
--- a/sound/usb/usx2y/us122l.c
+++ b/sound/usb/usx2y/us122l.c
@@ -349,14 +349,10 @@ static int usb_stream_hwdep_ioctl(struct snd_hwdep *hw, struct file *file,
349 if (cmd != SNDRV_USB_STREAM_IOCTL_SET_PARAMS) 349 if (cmd != SNDRV_USB_STREAM_IOCTL_SET_PARAMS)
350 return -ENOTTY; 350 return -ENOTTY;
351 351
352 cfg = kmalloc(sizeof(*cfg), GFP_KERNEL); 352 cfg = memdup_user((void *)arg, sizeof(*cfg));
353 if (!cfg) 353 if (IS_ERR(cfg))
354 return -ENOMEM; 354 return PTR_ERR(cfg);
355 355
356 if (copy_from_user(cfg, (void *)arg, sizeof(*cfg))) {
357 err = -EFAULT;
358 goto free;
359 }
360 if (cfg->version != USB_STREAM_INTERFACE_VERSION) { 356 if (cfg->version != USB_STREAM_INTERFACE_VERSION) {
361 err = -ENXIO; 357 err = -ENXIO;
362 goto free; 358 goto free;
diff --git a/sound/usb/usx2y/usX2Yhwdep.c b/sound/usb/usx2y/usX2Yhwdep.c
index 4af8740db717..f3d8f71265dd 100644
--- a/sound/usb/usx2y/usX2Yhwdep.c
+++ b/sound/usb/usx2y/usX2Yhwdep.c
@@ -203,13 +203,12 @@ static int snd_usX2Y_hwdep_dsp_load(struct snd_hwdep *hw,
203 203
204 if (access_ok(VERIFY_READ, dsp->image, dsp->length)) { 204 if (access_ok(VERIFY_READ, dsp->image, dsp->length)) {
205 struct usb_device* dev = priv->chip.dev; 205 struct usb_device* dev = priv->chip.dev;
206 char *buf = kmalloc(dsp->length, GFP_KERNEL); 206 char *buf;
207 if (!buf) 207
208 return -ENOMEM; 208 buf = memdup_user(dsp->image, dsp->length);
209 if (copy_from_user(buf, dsp->image, dsp->length)) { 209 if (IS_ERR(buf))
210 kfree(buf); 210 return PTR_ERR(buf);
211 return -EFAULT; 211
212 }
213 err = usb_set_interface(dev, 0, 1); 212 err = usb_set_interface(dev, 0, 1);
214 if (err) 213 if (err)
215 snd_printk(KERN_ERR "usb_set_interface error \n"); 214 snd_printk(KERN_ERR "usb_set_interface error \n");