]> git.openfabrics.org - ~shefty/rdma-dev.git/commitdiff
Merge branch 'x86-olpc-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
authorLinus Torvalds <torvalds@linux-foundation.org>
Thu, 21 Oct 2010 20:52:01 +0000 (13:52 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Thu, 21 Oct 2010 20:52:01 +0000 (13:52 -0700)
* 'x86-olpc-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86, olpc: XO-1 uses/depends on PCI
  x86, olpc: Register XO-1 platform devices
  x86, olpc: Add XO-1 poweroff support
  x86, olpc: Don't retry EC commands forever
  x86, olpc: Rework BIOS signature check
  x86, olpc: Only enable PCI configuration type override on XO-1

1065 files changed:
CREDITS
Documentation/DocBook/kernel-locking.tmpl
Documentation/RCU/checklist.txt
Documentation/RCU/stallwarn.txt
Documentation/RCU/trace.txt
Documentation/cputopology.txt
Documentation/kernel-parameters.txt
Documentation/kprobes.txt
Documentation/networking/e1000.txt
Documentation/networking/e1000e.txt [new file with mode: 0644]
Documentation/networking/ixgbevf.txt [changed mode: 0755->0644]
Documentation/vm/page-types.c
MAINTAINERS
Makefile
arch/Kconfig
arch/alpha/Kconfig
arch/alpha/include/asm/perf_event.h
arch/alpha/kernel/entry.S
arch/alpha/kernel/perf_event.c
arch/alpha/kernel/process.c
arch/alpha/kernel/signal.c
arch/alpha/kernel/systbls.S
arch/alpha/kernel/time.c
arch/arm/Kconfig
arch/arm/boot/compressed/Makefile
arch/arm/common/it8152.c
arch/arm/include/asm/perf_event.h
arch/arm/include/asm/pgtable.h
arch/arm/kernel/entry-common.S
arch/arm/kernel/kprobes-decode.c
arch/arm/kernel/perf_event.c
arch/arm/mach-at91/at91sam9g45_devices.c
arch/arm/mach-at91/include/mach/system.h
arch/arm/mach-bcmring/dma.c
arch/arm/mach-davinci/dm355.c
arch/arm/mach-davinci/dm365.c
arch/arm/mach-davinci/dm644x.c
arch/arm/mach-davinci/dm646x.c
arch/arm/mach-dove/include/mach/io.h
arch/arm/mach-ep93xx/dma-m2p.c
arch/arm/mach-imx/Kconfig
arch/arm/mach-imx/mach-cpuimx27.c
arch/arm/mach-ixp4xx/common-pci.c
arch/arm/mach-ixp4xx/include/mach/hardware.h
arch/arm/mach-kirkwood/include/mach/kirkwood.h
arch/arm/mach-kirkwood/pcie.c
arch/arm/mach-mmp/include/mach/system.h
arch/arm/mach-pxa/cpufreq-pxa2xx.c
arch/arm/mach-pxa/include/mach/hardware.h
arch/arm/mach-pxa/include/mach/io.h
arch/arm/mach-pxa/palm27x.c
arch/arm/mach-pxa/vpac270.c
arch/arm/mach-s5p6440/cpu.c
arch/arm/mach-s5p6442/cpu.c
arch/arm/mach-s5pc100/cpu.c
arch/arm/mach-s5pv210/clock.c
arch/arm/mach-s5pv210/cpu.c
arch/arm/mach-u300/include/mach/gpio.h
arch/arm/mach-vexpress/ct-ca9x4.c
arch/arm/mach-vexpress/v2m.c
arch/arm/mm/alignment.c
arch/arm/mm/ioremap.c
arch/arm/mm/mmu.c
arch/arm/mm/proc-v7.S
arch/arm/oprofile/Makefile
arch/arm/oprofile/common.c
arch/arm/plat-nomadik/timer.c
arch/arm/plat-omap/Kconfig
arch/arm/plat-omap/iommu.c
arch/arm/plat-omap/mcbsp.c
arch/arm/plat-omap/sram.c
arch/arm/plat-samsung/adc.c
arch/arm/plat-samsung/clock.c
arch/avr32/kernel/module.c
arch/frv/Kconfig
arch/frv/lib/Makefile
arch/frv/lib/perf_event.c [deleted file]
arch/h8300/kernel/module.c
arch/ia64/include/asm/hardirq.h
arch/ia64/include/asm/system.h
arch/m32r/include/asm/elf.h
arch/m32r/include/asm/signal.h
arch/m32r/include/asm/unistd.h
arch/m32r/kernel/.gitignore [new file with mode: 0644]
arch/m32r/kernel/entry.S
arch/m32r/kernel/ptrace.c
arch/m32r/kernel/signal.c
arch/m68k/mac/macboing.c
arch/mips/Kbuild
arch/mips/Kconfig
arch/mips/alchemy/common/prom.c
arch/mips/boot/compressed/Makefile
arch/mips/cavium-octeon/Kconfig
arch/mips/cavium-octeon/cpu.c
arch/mips/cavium-octeon/executive/Makefile
arch/mips/dec/Platform
arch/mips/include/asm/atomic.h
arch/mips/include/asm/cop2.h
arch/mips/include/asm/fcntl.h
arch/mips/include/asm/gic.h
arch/mips/include/asm/mach-tx49xx/kmalloc.h
arch/mips/include/asm/mips-boards/maltaint.h
arch/mips/include/asm/page.h
arch/mips/include/asm/siginfo.h
arch/mips/include/asm/thread_info.h
arch/mips/include/asm/unistd.h
arch/mips/jz4740/Platform
arch/mips/kernel/branch.c
arch/mips/kernel/irq-gic.c
arch/mips/kernel/kgdb.c
arch/mips/kernel/kspd.c
arch/mips/kernel/linux32.c
arch/mips/kernel/mips-mt-fpaff.c
arch/mips/kernel/ptrace.c
arch/mips/kernel/scall32-o32.S
arch/mips/kernel/scall64-64.S
arch/mips/kernel/scall64-n32.S
arch/mips/kernel/scall64-o32.S
arch/mips/kernel/signal.c
arch/mips/kernel/signal_n32.c
arch/mips/kernel/unaligned.c
arch/mips/mm/dma-default.c
arch/mips/mm/sc-rm7k.c
arch/mips/mti-malta/malta-int.c
arch/mips/pci/pci-rc32434.c
arch/mips/pnx8550/common/reset.c
arch/mips/pnx8550/common/setup.c
arch/mn10300/Kconfig
arch/mn10300/Kconfig.debug
arch/mn10300/include/asm/bitops.h
arch/mn10300/include/asm/signal.h
arch/mn10300/kernel/module.c
arch/mn10300/kernel/signal.c
arch/mn10300/mm/Makefile
arch/mn10300/mm/cache-disabled.c [new file with mode: 0644]
arch/mn10300/mm/cache.c
arch/parisc/Kconfig
arch/parisc/include/asm/perf_event.h
arch/parisc/kernel/module.c
arch/powerpc/Kconfig
arch/powerpc/include/asm/paca.h
arch/powerpc/include/asm/system.h
arch/powerpc/kernel/module.c
arch/powerpc/kernel/perf_callchain.c
arch/powerpc/kernel/perf_event.c
arch/powerpc/kernel/perf_event_fsl_emb.c
arch/powerpc/kernel/signal.c
arch/powerpc/kernel/signal_32.c
arch/powerpc/kernel/signal_64.c
arch/powerpc/kernel/time.c
arch/powerpc/platforms/512x/clock.c
arch/powerpc/platforms/52xx/efika.c
arch/powerpc/platforms/52xx/mpc52xx_common.c
arch/s390/Kconfig
arch/s390/include/asm/hardirq.h
arch/s390/include/asm/perf_event.h
arch/s390/include/asm/system.h
arch/s390/include/asm/topology.h
arch/s390/kernel/module.c
arch/s390/kernel/topology.c
arch/sh/Kconfig
arch/sh/include/asm/perf_event.h
arch/sh/kernel/module.c
arch/sh/kernel/perf_callchain.c
arch/sh/kernel/perf_event.c
arch/sh/oprofile/Makefile
arch/sh/oprofile/common.c
arch/sh/oprofile/op_impl.h [deleted file]
arch/sparc/Kconfig
arch/sparc/include/asm/jump_label.h [new file with mode: 0644]
arch/sparc/include/asm/perf_event.h
arch/sparc/kernel/Makefile
arch/sparc/kernel/jump_label.c [new file with mode: 0644]
arch/sparc/kernel/module.c
arch/sparc/kernel/pcr.c
arch/sparc/kernel/perf_event.c
arch/sparc/kernel/signal32.c
arch/sparc/kernel/signal_32.c
arch/sparc/kernel/signal_64.c
arch/tile/kernel/intvec_32.S
arch/um/drivers/hostaudio_kern.c
arch/um/drivers/net_kern.c
arch/um/drivers/ubd_kern.c
arch/um/kernel/exec.c
arch/um/kernel/internal.h
arch/um/kernel/syscall.c
arch/x86/Kconfig
arch/x86/Kconfig.debug
arch/x86/Makefile
arch/x86/boot/early_serial_console.c
arch/x86/ia32/ia32_aout.c
arch/x86/include/asm/alternative.h
arch/x86/include/asm/amd_iommu.h
arch/x86/include/asm/amd_iommu_proto.h
arch/x86/include/asm/amd_iommu_types.h
arch/x86/include/asm/amd_nb.h [new file with mode: 0644]
arch/x86/include/asm/apb_timer.h
arch/x86/include/asm/bitops.h
arch/x86/include/asm/cpu.h
arch/x86/include/asm/cpufeature.h
arch/x86/include/asm/dwarf2.h
arch/x86/include/asm/entry_arch.h
arch/x86/include/asm/fixmap.h
arch/x86/include/asm/gart.h
arch/x86/include/asm/hardirq.h
arch/x86/include/asm/hw_breakpoint.h
arch/x86/include/asm/hw_irq.h
arch/x86/include/asm/i387.h
arch/x86/include/asm/io.h
arch/x86/include/asm/irq_vectors.h
arch/x86/include/asm/jump_label.h [new file with mode: 0644]
arch/x86/include/asm/k8.h [deleted file]
arch/x86/include/asm/kvm_host.h
arch/x86/include/asm/mrst.h
arch/x86/include/asm/mwait.h [new file with mode: 0644]
arch/x86/include/asm/page_types.h
arch/x86/include/asm/perf_event_p4.h
arch/x86/include/asm/pgtable.h
arch/x86/include/asm/pgtable_64.h
arch/x86/include/asm/processor.h
arch/x86/include/asm/setup.h
arch/x86/kernel/Makefile
arch/x86/kernel/acpi/cstate.c
arch/x86/kernel/alternative.c
arch/x86/kernel/amd_iommu.c
arch/x86/kernel/amd_iommu_init.c
arch/x86/kernel/amd_nb.c [new file with mode: 0644]
arch/x86/kernel/apb_timer.c
arch/x86/kernel/aperture_64.c
arch/x86/kernel/apic/io_apic.c
arch/x86/kernel/cpu/amd.c
arch/x86/kernel/cpu/common.c
arch/x86/kernel/cpu/cpu.h
arch/x86/kernel/cpu/cpufreq/pcc-cpufreq.c
arch/x86/kernel/cpu/intel.c
arch/x86/kernel/cpu/intel_cacheinfo.c
arch/x86/kernel/cpu/mcheck/mce_amd.c
arch/x86/kernel/cpu/mcheck/therm_throt.c
arch/x86/kernel/cpu/mtrr/cleanup.c
arch/x86/kernel/cpu/mtrr/generic.c
arch/x86/kernel/cpu/perf_event.c
arch/x86/kernel/cpu/perf_event_amd.c
arch/x86/kernel/cpu/perf_event_intel.c
arch/x86/kernel/cpu/perf_event_intel_ds.c
arch/x86/kernel/cpu/perf_event_p4.c
arch/x86/kernel/cpu/perfctr-watchdog.c
arch/x86/kernel/cpu/scattered.c
arch/x86/kernel/crash_dump_64.c
arch/x86/kernel/early-quirks.c
arch/x86/kernel/early_printk.c
arch/x86/kernel/early_printk_mrst.c [new file with mode: 0644]
arch/x86/kernel/entry_32.S
arch/x86/kernel/entry_64.S
arch/x86/kernel/ftrace.c
arch/x86/kernel/hpet.c
arch/x86/kernel/hw_breakpoint.c
arch/x86/kernel/i387.c
arch/x86/kernel/irq.c
arch/x86/kernel/irq_work.c [new file with mode: 0644]
arch/x86/kernel/irqinit.c
arch/x86/kernel/jump_label.c [new file with mode: 0644]
arch/x86/kernel/k8.c [deleted file]
arch/x86/kernel/kprobes.c
arch/x86/kernel/machine_kexec_64.c
arch/x86/kernel/module.c
arch/x86/kernel/pci-gart_64.c
arch/x86/kernel/pmtimer_64.c [deleted file]
arch/x86/kernel/process_64.c
arch/x86/kernel/reboot.c
arch/x86/kernel/setup.c
arch/x86/kernel/setup_percpu.c
arch/x86/kernel/sfi.c
arch/x86/kernel/smpboot.c
arch/x86/kernel/sys_i386_32.c
arch/x86/kernel/traps.c
arch/x86/kernel/tsc.c
arch/x86/kvm/lapic.c
arch/x86/kvm/svm.c
arch/x86/kvm/vmx.c
arch/x86/kvm/x86.c
arch/x86/lguest/boot.c
arch/x86/lib/memcpy_32.c
arch/x86/lib/memcpy_64.S
arch/x86/lib/memmove_64.c
arch/x86/mm/fault.c
arch/x86/mm/init_32.c
arch/x86/mm/init_64.c
arch/x86/mm/k8topology_64.c
arch/x86/mm/kmemcheck/kmemcheck.c
arch/x86/mm/kmemcheck/opcode.c
arch/x86/mm/numa_64.c
arch/x86/mm/pgtable.c
arch/x86/mm/srat_64.c
arch/x86/mm/tlb.c
arch/x86/oprofile/backtrace.c
arch/x86/oprofile/nmi_int.c
arch/x86/xen/time.c
block/blk-map.c
block/blk-merge.c
block/bsg.c
block/cfq-iosched.c
block/elevator.c
drivers/acpi/Kconfig
drivers/acpi/acpi_pad.c
drivers/acpi/acpica/aclocal.h
drivers/acpi/acpica/exutils.c
drivers/acpi/acpica/rsutils.c
drivers/acpi/apei/Kconfig
drivers/acpi/apei/apei-base.c
drivers/acpi/apei/einj.c
drivers/acpi/apei/erst-dbg.c
drivers/acpi/apei/erst.c
drivers/acpi/apei/ghes.c
drivers/acpi/apei/hest.c
drivers/acpi/atomicio.c
drivers/acpi/battery.c
drivers/acpi/blacklist.c
drivers/acpi/bus.c
drivers/acpi/fan.c
drivers/acpi/processor_core.c
drivers/acpi/processor_driver.c
drivers/acpi/processor_perflib.c
drivers/acpi/sleep.c
drivers/acpi/sysfs.c
drivers/acpi/video_detect.c
drivers/ata/ahci.c
drivers/ata/ahci.h
drivers/ata/ahci_platform.c
drivers/ata/libahci.c
drivers/atm/iphase.c
drivers/atm/iphase.h
drivers/atm/solos-pci.c
drivers/base/topology.c
drivers/block/Kconfig
drivers/block/Makefile
drivers/block/cciss.c
drivers/block/pktcdvd.c
drivers/block/ps3disk.c
drivers/block/rbd.c [new file with mode: 0644]
drivers/block/rbd_types.h [new file with mode: 0644]
drivers/block/virtio_blk.c
drivers/char/agp/Kconfig
drivers/char/agp/amd64-agp.c
drivers/char/agp/generic.c
drivers/char/agp/intel-agp.c
drivers/char/agp/intel-agp.h
drivers/char/ipmi/ipmi_si_intf.c
drivers/char/mem.c
drivers/char/tpm/tpm.c
drivers/char/virtio_console.c
drivers/cpuidle/governors/menu.c
drivers/dma/ioat/dma_v2.c
drivers/dma/mv_xor.c
drivers/dma/shdma.c
drivers/edac/Kconfig
drivers/edac/amd64_edac.c
drivers/edac/edac_mc.c
drivers/edac/i7core_edac.c
drivers/firewire/ohci.c
drivers/firewire/ohci.h
drivers/gpu/drm/drm_buffer.c
drivers/gpu/drm/drm_gem.c
drivers/gpu/drm/drm_info.c
drivers/gpu/drm/drm_vm.c
drivers/gpu/drm/i810/i810_dma.c
drivers/gpu/drm/i830/i830_dma.c
drivers/gpu/drm/i915/i915_dma.c
drivers/gpu/drm/i915/i915_drv.c
drivers/gpu/drm/i915/i915_gem.c
drivers/gpu/drm/i915/i915_gem_evict.c
drivers/gpu/drm/i915/i915_suspend.c
drivers/gpu/drm/i915/intel_crt.c
drivers/gpu/drm/i915/intel_display.c
drivers/gpu/drm/i915/intel_dp.c
drivers/gpu/drm/i915/intel_drv.h
drivers/gpu/drm/i915/intel_fb.c
drivers/gpu/drm/i915/intel_sdvo.c
drivers/gpu/drm/nouveau/nouveau_connector.c
drivers/gpu/drm/nouveau/nouveau_gem.c
drivers/gpu/drm/radeon/atombios.h
drivers/gpu/drm/radeon/evergreen.c
drivers/gpu/drm/radeon/r100.c
drivers/gpu/drm/radeon/r600.c
drivers/gpu/drm/radeon/r600_blit_kms.c
drivers/gpu/drm/radeon/radeon.h
drivers/gpu/drm/radeon/radeon_atombios.c
drivers/gpu/drm/radeon/radeon_combios.c
drivers/gpu/drm/radeon/radeon_cursor.c
drivers/gpu/drm/radeon/radeon_display.c
drivers/gpu/drm/radeon/radeon_fb.c
drivers/gpu/drm/radeon/radeon_gem.c
drivers/gpu/drm/radeon/radeon_kms.c
drivers/gpu/drm/radeon/radeon_object.c
drivers/gpu/drm/radeon/radeon_object.h
drivers/gpu/drm/radeon/rs600.c
drivers/gpu/drm/radeon/rs690.c
drivers/gpu/drm/radeon/rv770.c
drivers/gpu/drm/ttm/ttm_bo.c
drivers/gpu/drm/ttm/ttm_bo_util.c
drivers/gpu/drm/ttm/ttm_page_alloc.c
drivers/gpu/drm/vmwgfx/vmwgfx_drv.c
drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
drivers/gpu/drm/vmwgfx/vmwgfx_fb.c
drivers/gpu/drm/vmwgfx/vmwgfx_fifo.c
drivers/gpu/drm/vmwgfx/vmwgfx_kms.c
drivers/gpu/drm/vmwgfx/vmwgfx_ldu.c
drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
drivers/gpu/vga/vgaarb.c
drivers/hid/hid-cando.c
drivers/hid/hid-core.c
drivers/hid/hid-ids.h
drivers/hid/hidraw.c
drivers/hid/usbhid/hid-quirks.c
drivers/hwmon/Kconfig
drivers/hwmon/coretemp.c
drivers/hwmon/f71882fg.c
drivers/hwmon/lis3lv02d.c
drivers/hwmon/pkgtemp.c
drivers/i2c/busses/i2c-cpm.c
drivers/i2c/busses/i2c-davinci.c
drivers/i2c/busses/i2c-ibm_iic.c
drivers/i2c/busses/i2c-imx.c
drivers/i2c/busses/i2c-mpc.c
drivers/i2c/busses/i2c-octeon.c
drivers/i2c/busses/i2c-omap.c
drivers/i2c/busses/i2c-pca-isa.c
drivers/i2c/busses/i2c-pca-platform.c
drivers/i2c/busses/i2c-s3c2410.c
drivers/i2c/i2c-core.c
drivers/idle/intel_idle.c [changed mode: 0755->0644]
drivers/infiniband/hw/cxgb3/iwch_cm.c
drivers/input/evdev.c
drivers/input/joydev.c
drivers/input/misc/hp_sdc_rtc.c
drivers/input/misc/uinput.c
drivers/input/serio/hil_mlc.c
drivers/input/serio/hp_sdc.c
drivers/input/tablet/wacom_sys.c
drivers/input/tablet/wacom_wac.c
drivers/isdn/sc/interrupt.c
drivers/leds/leds-ns2.c
drivers/macintosh/adb.c
drivers/md/bitmap.c
drivers/md/raid1.c
drivers/media/IR/ir-keytable.c
drivers/media/IR/ir-lirc-codec.c
drivers/media/IR/ir-raw-event.c
drivers/media/IR/ir-sysfs.c
drivers/media/IR/keymaps/rc-rc6-mce.c
drivers/media/IR/mceusb.c
drivers/media/dvb/dvb-usb/dib0700_core.c
drivers/media/dvb/dvb-usb/dib0700_devices.c
drivers/media/dvb/dvb-usb/opera1.c
drivers/media/dvb/frontends/dib7000p.c
drivers/media/dvb/frontends/dib7000p.h
drivers/media/dvb/siano/smscoreapi.c
drivers/media/radio/si470x/radio-si470x-i2c.c
drivers/media/video/cx231xx/Makefile
drivers/media/video/cx231xx/cx231xx-cards.c
drivers/media/video/cx25840/cx25840-core.c
drivers/media/video/cx88/Kconfig
drivers/media/video/gspca/gspca.c
drivers/media/video/gspca/sn9c20x.c
drivers/media/video/ivtv/ivtvfb.c
drivers/media/video/mem2mem_testdev.c
drivers/media/video/mt9m111.c
drivers/media/video/mt9v022.c
drivers/media/video/mx2_camera.c
drivers/media/video/pvrusb2/pvrusb2-ctrl.c
drivers/media/video/s5p-fimc/fimc-core.c
drivers/media/video/saa7134/saa7134-cards.c
drivers/media/video/saa7164/saa7164-buffer.c
drivers/media/video/uvc/uvc_driver.c
drivers/media/video/uvc/uvcvideo.h
drivers/media/video/v4l2-compat-ioctl32.c
drivers/media/video/videobuf-dma-contig.c
drivers/media/video/videobuf-dma-sg.c
drivers/mfd/max8925-core.c
drivers/mfd/wm831x-irq.c
drivers/misc/Kconfig
drivers/misc/Makefile
drivers/misc/bh1780gli.c
drivers/misc/vmw_balloon.c [new file with mode: 0644]
drivers/misc/vmware_balloon.c [deleted file]
drivers/mmc/core/core.c
drivers/mmc/host/sdhci-s3c.c
drivers/mtd/nand/mxc_nand.c
drivers/mtd/nand/omap2.c
drivers/net/3c527.c
drivers/net/3c59x.c
drivers/net/Kconfig
drivers/net/atlx/atl1.c
drivers/net/b44.c
drivers/net/bonding/bond_main.c
drivers/net/e1000e/hw.h
drivers/net/e1000e/ich8lan.c
drivers/net/e1000e/netdev.c
drivers/net/ehea/ehea_main.c
drivers/net/ehea/ehea_qmr.h
drivers/net/fec.c
drivers/net/hamradio/6pack.c
drivers/net/hamradio/mkiss.c
drivers/net/ibm_newemac/core.c
drivers/net/irda/sir_dev.c
drivers/net/netxen/netxen_nic_init.c
drivers/net/ppp_async.c
drivers/net/qlcnic/qlcnic_init.c
drivers/net/r8169.c
drivers/net/rionet.c
drivers/net/sgiseeq.c
drivers/net/skge.c
drivers/net/smsc911x.c
drivers/net/tg3.c
drivers/net/tg3.h
drivers/net/tulip/de2104x.c
drivers/net/wan/cosa.c
drivers/net/wimax/i2400m/rx.c
drivers/net/wireless/ath/ath9k/ani.c
drivers/net/wireless/iwlwifi/iwl-agn-lib.c
drivers/net/wireless/iwlwifi/iwl-core.c
drivers/net/wireless/iwlwifi/iwl3945-base.c
drivers/oprofile/oprof.c
drivers/oprofile/oprof.h
drivers/oprofile/oprofile_files.c
drivers/oprofile/oprofile_perf.c [new file with mode: 0644]
drivers/oprofile/oprofilefs.c
drivers/parport/share.c
drivers/pci/intel-iommu.c
drivers/pci/iov.c
drivers/pci/pci.h
drivers/pci/quirks.c
drivers/pcmcia/pcmcia_resource.c
drivers/pcmcia/pd6729.c
drivers/platform/x86/intel_ips.c
drivers/platform/x86/thinkpad_acpi.c
drivers/regulator/ad5398.c
drivers/regulator/core.c
drivers/regulator/isl6271a-regulator.c
drivers/regulator/max8649.c
drivers/rtc/rtc-ab3100.c
drivers/rtc/rtc-ds3232.c
drivers/rtc/rtc-s3c.c
drivers/s390/net/ctcm_main.c
drivers/scsi/scsi.c
drivers/serial/ioc3_serial.c
drivers/serial/mfd.c
drivers/serial/mrst_max3110.c
drivers/spi/spi.c
drivers/spi/spi_gpio.c
drivers/spi/spi_mpc8xxx.c
drivers/staging/ti-st/st.h
drivers/staging/ti-st/st_core.c
drivers/staging/ti-st/st_core.h
drivers/staging/ti-st/st_kim.c
drivers/staging/tm6000/Kconfig
drivers/staging/tm6000/tm6000-input.c
drivers/usb/core/Kconfig
drivers/usb/core/file.c
drivers/usb/core/message.c
drivers/usb/musb/cppi_dma.c
drivers/usb/musb/musb_gadget.c
drivers/usb/musb/musb_gadget.h
drivers/usb/musb/musb_gadget_ep0.c
drivers/usb/musb/musb_host.c
drivers/vhost/net.c
drivers/vhost/vhost.c
drivers/vhost/vhost.h
drivers/video/console/fbcon.c
drivers/video/efifb.c
drivers/video/pxa168fb.c
drivers/video/sis/sis_main.c
drivers/xen/xenbus/xenbus_probe.c
fs/affs/super.c
fs/aio.c
fs/binfmt_aout.c
fs/ceph/Kconfig
fs/ceph/Makefile
fs/ceph/README [deleted file]
fs/ceph/addr.c
fs/ceph/armor.c [deleted file]
fs/ceph/auth.c [deleted file]
fs/ceph/auth.h [deleted file]
fs/ceph/auth_none.c [deleted file]
fs/ceph/auth_none.h [deleted file]
fs/ceph/auth_x.c [deleted file]
fs/ceph/auth_x.h [deleted file]
fs/ceph/auth_x_protocol.h [deleted file]
fs/ceph/buffer.c [deleted file]
fs/ceph/buffer.h [deleted file]
fs/ceph/caps.c
fs/ceph/ceph_debug.h [deleted file]
fs/ceph/ceph_frag.c
fs/ceph/ceph_frag.h [deleted file]
fs/ceph/ceph_fs.c [deleted file]
fs/ceph/ceph_fs.h [deleted file]
fs/ceph/ceph_hash.c [deleted file]
fs/ceph/ceph_hash.h [deleted file]
fs/ceph/ceph_strings.c [deleted file]
fs/ceph/crush/crush.c [deleted file]
fs/ceph/crush/crush.h [deleted file]
fs/ceph/crush/hash.c [deleted file]
fs/ceph/crush/hash.h [deleted file]
fs/ceph/crush/mapper.c [deleted file]
fs/ceph/crush/mapper.h [deleted file]
fs/ceph/crypto.c [deleted file]
fs/ceph/crypto.h [deleted file]
fs/ceph/debugfs.c
fs/ceph/decode.h [deleted file]
fs/ceph/dir.c
fs/ceph/export.c
fs/ceph/file.c
fs/ceph/inode.c
fs/ceph/ioctl.c
fs/ceph/ioctl.h
fs/ceph/locks.c
fs/ceph/mds_client.c
fs/ceph/mds_client.h
fs/ceph/mdsmap.c
fs/ceph/mdsmap.h [deleted file]
fs/ceph/messenger.c [deleted file]
fs/ceph/messenger.h [deleted file]
fs/ceph/mon_client.c [deleted file]
fs/ceph/mon_client.h [deleted file]
fs/ceph/msgpool.c [deleted file]
fs/ceph/msgpool.h [deleted file]
fs/ceph/msgr.h [deleted file]
fs/ceph/osd_client.c [deleted file]
fs/ceph/osd_client.h [deleted file]
fs/ceph/osdmap.c [deleted file]
fs/ceph/osdmap.h [deleted file]
fs/ceph/pagelist.c [deleted file]
fs/ceph/pagelist.h [deleted file]
fs/ceph/rados.h [deleted file]
fs/ceph/snap.c
fs/ceph/strings.c [new file with mode: 0644]
fs/ceph/super.c
fs/ceph/super.h
fs/ceph/types.h [deleted file]
fs/ceph/xattr.c
fs/char_dev.c
fs/cifs/cifssmb.c
fs/cifs/inode.c
fs/compat.c
fs/exec.c
fs/exofs/inode.c
fs/fs-writeback.c
fs/fuse/dev.c
fs/gfs2/Kconfig
fs/gfs2/aops.c
fs/gfs2/bmap.c
fs/gfs2/bmap.h
fs/gfs2/dentry.c
fs/gfs2/dir.c
fs/gfs2/dir.h
fs/gfs2/export.c
fs/gfs2/file.c
fs/gfs2/glock.c
fs/gfs2/glock.h
fs/gfs2/glops.c
fs/gfs2/incore.h
fs/gfs2/inode.c
fs/gfs2/inode.h
fs/gfs2/lock_dlm.c
fs/gfs2/main.c
fs/gfs2/ops_fstype.c
fs/gfs2/ops_inode.c
fs/gfs2/quota.c
fs/gfs2/recovery.c
fs/gfs2/rgrp.c
fs/gfs2/rgrp.h
fs/gfs2/super.c
fs/gfs2/sys.c
fs/gfs2/trace_gfs2.h
fs/gfs2/trans.h
fs/gfs2/xattr.c
fs/hfs/bfind.c
fs/hfs/btree.c
fs/hfs/btree.h
fs/hfsplus/bfind.c
fs/hfsplus/bitmap.c
fs/hfsplus/brec.c
fs/hfsplus/btree.c
fs/hfsplus/catalog.c
fs/hfsplus/dir.c
fs/hfsplus/extents.c
fs/hfsplus/hfsplus_fs.h
fs/hfsplus/hfsplus_raw.h
fs/hfsplus/inode.c
fs/hfsplus/ioctl.c
fs/hfsplus/options.c
fs/hfsplus/part_tbl.c
fs/hfsplus/super.c
fs/hfsplus/unicode.c
fs/hfsplus/wrapper.c
fs/nfsd/nfsfh.h
fs/notify/Kconfig
fs/ocfs2/acl.c
fs/ocfs2/cluster/tcp.c
fs/ocfs2/dir.c
fs/ocfs2/dlm/dlmcommon.h
fs/ocfs2/dlm/dlmdebug.c
fs/ocfs2/dlm/dlmdomain.c
fs/ocfs2/dlm/dlmmaster.c
fs/ocfs2/dlmglue.h
fs/ocfs2/ocfs2_fs.h
fs/ocfs2/ocfs2_ioctl.h
fs/ocfs2/refcounttree.c
fs/ocfs2/reservations.c
fs/ocfs2/suballoc.c
fs/ocfs2/symlink.c
fs/ocfs2/xattr.c
fs/proc/base.c
fs/proc/task_mmu.c
fs/proc/vmcore.c
fs/reiserfs/ioctl.c
fs/xfs/linux-2.6/xfs_sync.c
fs/xfs/xfs_log_cil.c
fs/xfs/xfs_log_priv.h
include/acpi/acpixf.h
include/asm-generic/hardirq.h
include/asm-generic/pgtable.h
include/asm-generic/vmlinux.lds.h
include/drm/drmP.h
include/drm/drm_pciids.h
include/drm/ttm/ttm_bo_api.h
include/linux/Kbuild
include/linux/acpi_pmtmr.h
include/linux/ceph/auth.h [new file with mode: 0644]
include/linux/ceph/buffer.h [new file with mode: 0644]
include/linux/ceph/ceph_debug.h [new file with mode: 0644]
include/linux/ceph/ceph_frag.h [new file with mode: 0644]
include/linux/ceph/ceph_fs.h [new file with mode: 0644]
include/linux/ceph/ceph_hash.h [new file with mode: 0644]
include/linux/ceph/debugfs.h [new file with mode: 0644]
include/linux/ceph/decode.h [new file with mode: 0644]
include/linux/ceph/libceph.h [new file with mode: 0644]
include/linux/ceph/mdsmap.h [new file with mode: 0644]
include/linux/ceph/messenger.h [new file with mode: 0644]
include/linux/ceph/mon_client.h [new file with mode: 0644]
include/linux/ceph/msgpool.h [new file with mode: 0644]
include/linux/ceph/msgr.h [new file with mode: 0644]
include/linux/ceph/osd_client.h [new file with mode: 0644]
include/linux/ceph/osdmap.h [new file with mode: 0644]
include/linux/ceph/pagelist.h [new file with mode: 0644]
include/linux/ceph/rados.h [new file with mode: 0644]
include/linux/ceph/types.h [new file with mode: 0644]
include/linux/cgroup.h
include/linux/compiler.h
include/linux/coredump.h
include/linux/cpuidle.h
include/linux/cred.h
include/linux/crush/crush.h [new file with mode: 0644]
include/linux/crush/hash.h [new file with mode: 0644]
include/linux/crush/mapper.h [new file with mode: 0644]
include/linux/debug_locks.h
include/linux/dma-mapping.h
include/linux/dmaengine.h
include/linux/dynamic_debug.h
include/linux/elevator.h
include/linux/fdtable.h
include/linux/fs.h
include/linux/ftrace_event.h
include/linux/genhd.h
include/linux/hardirq.h
include/linux/idr.h
include/linux/init_task.h
include/linux/input.h
include/linux/interrupt.h
include/linux/iocontext.h
include/linux/irq_work.h [new file with mode: 0644]
include/linux/jump_label.h [new file with mode: 0644]
include/linux/jump_label_ref.h [new file with mode: 0644]
include/linux/kernel.h
include/linux/key.h
include/linux/kvm_host.h
include/linux/lockdep.h
include/linux/mm_types.h
include/linux/module.h
include/linux/netfilter/nfnetlink_conntrack.h
include/linux/netfilter/xt_SECMARK.h
include/linux/netlink.h
include/linux/nfs_fs.h
include/linux/notifier.h
include/linux/oprofile.h
include/linux/pci_ids.h
include/linux/percpu-defs.h
include/linux/percpu.h
include/linux/perf_event.h
include/linux/radix-tree.h
include/linux/rculist.h
include/linux/rculist_nulls.h
include/linux/rcupdate.h
include/linux/rcutiny.h
include/linux/rcutree.h
include/linux/sched.h
include/linux/security.h
include/linux/selinux.h
include/linux/socket.h
include/linux/srcu.h
include/linux/stop_machine.h
include/linux/sunrpc/auth_gss.h
include/linux/topology.h
include/linux/tracepoint.h
include/linux/types.h
include/linux/wait.h
include/media/videobuf-dma-sg.h
include/net/addrconf.h
include/net/bluetooth/bluetooth.h
include/net/cls_cgroup.h
include/net/dst.h
include/net/netfilter/nf_conntrack.h
include/net/route.h
include/net/xfrm.h
include/trace/events/irq.h
include/trace/events/napi.h
include/trace/events/net.h [new file with mode: 0644]
include/trace/events/power.h
include/trace/events/sched.h
include/trace/events/skb.h
init/Kconfig
ipc/sem.c
kernel/Makefile
kernel/cgroup.c
kernel/cpuset.c
kernel/exit.c
kernel/fork.c
kernel/hrtimer.c
kernel/hung_task.c
kernel/hw_breakpoint.c
kernel/irq_work.c [new file with mode: 0644]
kernel/jump_label.c [new file with mode: 0644]
kernel/kfifo.c
kernel/kprobes.c
kernel/lockdep.c
kernel/module.c
kernel/perf_event.c
kernel/pid.c
kernel/printk.c
kernel/rcupdate.c
kernel/rcutiny.c
kernel/rcutiny_plugin.h
kernel/rcutorture.c
kernel/rcutree.c
kernel/rcutree.h
kernel/rcutree_plugin.h
kernel/rcutree_trace.c
kernel/sched.c
kernel/sched_fair.c
kernel/sched_features.h
kernel/sched_rt.c
kernel/sched_stoptask.c [new file with mode: 0644]
kernel/signal.c
kernel/smp.c
kernel/softirq.c
kernel/srcu.c
kernel/stop_machine.c
kernel/sysctl.c
kernel/sysctl_check.c
kernel/test_kprobes.c
kernel/timer.c
kernel/trace/Kconfig
kernel/trace/ftrace.c
kernel/trace/ring_buffer.c
kernel/trace/trace.c
kernel/trace/trace.h
kernel/trace/trace_event_perf.c
kernel/trace/trace_events.c
kernel/trace/trace_functions_graph.c
kernel/trace/trace_irqsoff.c
kernel/trace/trace_sched_wakeup.c
kernel/trace/trace_workqueue.c
kernel/tracepoint.c
kernel/watchdog.c
lib/Kconfig.debug
lib/bug.c
lib/dynamic_debug.c
lib/list_sort.c
lib/radix-tree.c
mm/backing-dev.c
mm/fremap.c
mm/hugetlb.c
mm/ksm.c
mm/memcontrol.c
mm/memory-failure.c
mm/memory.c
mm/mmap.c
mm/oom_kill.c
mm/page_alloc.c
mm/percpu.c
mm/rmap.c
mm/vmalloc.c
mm/vmscan.c
net/8021q/vlan_core.c
net/9p/trans_rdma.c
net/9p/trans_virtio.c
net/Kconfig
net/Makefile
net/atm/br2684.c
net/atm/mpc.c
net/bluetooth/l2cap.c
net/bluetooth/rfcomm/sock.c
net/caif/caif_socket.c
net/ceph/Kconfig [new file with mode: 0644]
net/ceph/Makefile [new file with mode: 0644]
net/ceph/armor.c [new file with mode: 0644]
net/ceph/auth.c [new file with mode: 0644]
net/ceph/auth_none.c [new file with mode: 0644]
net/ceph/auth_none.h [new file with mode: 0644]
net/ceph/auth_x.c [new file with mode: 0644]
net/ceph/auth_x.h [new file with mode: 0644]
net/ceph/auth_x_protocol.h [new file with mode: 0644]
net/ceph/buffer.c [new file with mode: 0644]
net/ceph/ceph_common.c [new file with mode: 0644]
net/ceph/ceph_fs.c [new file with mode: 0644]
net/ceph/ceph_hash.c [new file with mode: 0644]
net/ceph/ceph_strings.c [new file with mode: 0644]
net/ceph/crush/crush.c [new file with mode: 0644]
net/ceph/crush/hash.c [new file with mode: 0644]
net/ceph/crush/mapper.c [new file with mode: 0644]
net/ceph/crypto.c [new file with mode: 0644]
net/ceph/crypto.h [new file with mode: 0644]
net/ceph/debugfs.c [new file with mode: 0644]
net/ceph/messenger.c [new file with mode: 0644]
net/ceph/mon_client.c [new file with mode: 0644]
net/ceph/msgpool.c [new file with mode: 0644]
net/ceph/osd_client.c [new file with mode: 0644]
net/ceph/osdmap.c [new file with mode: 0644]
net/ceph/pagelist.c [new file with mode: 0644]
net/ceph/pagevec.c [new file with mode: 0644]
net/core/datagram.c
net/core/dev.c
net/core/ethtool.c
net/core/iovec.c
net/core/net-traces.c
net/core/skbuff.c
net/core/sock.c
net/core/stream.c
net/ipv4/Kconfig
net/ipv4/igmp.c
net/ipv4/ip_gre.c
net/ipv4/ip_output.c
net/ipv4/netfilter/ipt_REJECT.c
net/ipv4/netfilter/nf_conntrack_l3proto_ipv4_compat.c
net/ipv4/netfilter/nf_defrag_ipv4.c
net/ipv4/netfilter/nf_nat_core.c
net/ipv4/netfilter/nf_nat_snmp_basic.c
net/ipv4/route.c
net/ipv4/tcp.c
net/ipv4/tcp_input.c
net/ipv4/tcp_timer.c
net/ipv4/xfrm4_policy.c
net/ipv4/xfrm4_state.c
net/ipv6/addrconf.c
net/ipv6/addrlabel.c
net/ipv6/ip6_output.c
net/ipv6/route.c
net/ipv6/xfrm6_state.c
net/mac80211/agg-tx.c
net/mac80211/rx.c
net/mac80211/status.c
net/netfilter/core.c
net/netfilter/nf_conntrack_ecache.c
net/netfilter/nf_conntrack_extend.c
net/netfilter/nf_conntrack_netlink.c
net/netfilter/nf_conntrack_proto.c
net/netfilter/nf_conntrack_sip.c
net/netfilter/nf_conntrack_standalone.c
net/netfilter/nf_log.c
net/netfilter/nf_queue.c
net/netfilter/nf_tproxy_core.c
net/netfilter/xt_CT.c
net/netfilter/xt_SECMARK.c
net/phonet/pep.c
net/rds/page.c
net/rds/tcp_connect.c
net/rds/tcp_listen.c
net/rds/tcp_recv.c
net/rds/tcp_send.c
net/rose/af_rose.c
net/sched/cls_cgroup.c
net/sched/cls_u32.c
net/sctp/auth.c
net/sctp/socket.c
net/sunrpc/xprtsock.c
net/wireless/wext-priv.c
net/xfrm/xfrm_policy.c
net/xfrm/xfrm_state.c
samples/kfifo/dma-example.c
scripts/Makefile
scripts/Makefile.build
scripts/Makefile.lib
scripts/basic/Makefile
scripts/basic/hash.c [deleted file]
scripts/gcc-goto.sh [new file with mode: 0644]
scripts/kconfig/conf.c
scripts/kconfig/expr.h
scripts/kconfig/menu.c
scripts/kconfig/symbol.c
scripts/recordmcount.c [new file with mode: 0644]
scripts/recordmcount.h [new file with mode: 0644]
security/apparmor/.gitignore
security/apparmor/apparmorfs.c
security/capability.c
security/commoncap.c
security/security.c
security/selinux/Makefile
security/selinux/exports.c
security/selinux/hooks.c
security/selinux/include/classmap.h
security/selinux/include/security.h
security/selinux/selinuxfs.c
security/selinux/ss/Makefile [deleted file]
security/selinux/ss/avtab.c
security/selinux/ss/avtab.h
security/selinux/ss/conditional.c
security/selinux/ss/conditional.h
security/selinux/ss/ebitmap.c
security/selinux/ss/ebitmap.h
security/selinux/ss/policydb.c
security/selinux/ss/policydb.h
security/selinux/ss/services.c
security/selinux/ss/status.c [new file with mode: 0644]
security/smack/smack_lsm.c
security/tomoyo/common.c
security/tomoyo/common.h
sound/core/control.c
sound/core/rawmidi.c
sound/i2c/other/ak4xxx-adda.c
sound/oss/soundcard.c
sound/pci/hda/patch_analog.c
sound/pci/hda/patch_realtek.c
sound/pci/hda/patch_sigmatel.c
sound/pci/oxygen/oxygen.c
sound/pci/rme9652/hdsp.c
sound/pci/rme9652/hdspm.c
sound/soc/sh/migor.c
sound/soc/soc-cache.c
tools/perf/Documentation/perf-annotate.txt
tools/perf/Documentation/perf-report.txt
tools/perf/Makefile
tools/perf/builtin-annotate.c
tools/perf/builtin-report.c
tools/perf/feature-tests.mak
tools/perf/perf.h
tools/perf/scripts/python/bin/netdev-times-record [new file with mode: 0644]
tools/perf/scripts/python/bin/netdev-times-report [new file with mode: 0644]
tools/perf/scripts/python/netdev-times.py [new file with mode: 0644]
tools/perf/util/cache.h
tools/perf/util/callchain.c
tools/perf/util/callchain.h
tools/perf/util/hist.c
tools/perf/util/path.c
tools/perf/util/sort.h
tools/perf/util/symbol.c
tools/perf/util/symbol.h
tools/perf/util/trace-event-scripting.c
tools/perf/util/ui/browser.c
tools/perf/util/ui/browser.h
tools/perf/util/ui/browsers/annotate.c
tools/perf/util/ui/browsers/hists.c
tools/perf/util/ui/browsers/map.c
tools/perf/util/ui/util.c
tools/perf/util/util.h
virt/kvm/eventfd.c
virt/kvm/kvm_main.c

diff --git a/CREDITS b/CREDITS
index 72b487869788c14cd40e9535b700f4be166aa12b..41d8e63d5165b5b786db6ab7d8c14fbc49fc0107 100644 (file)
--- a/CREDITS
+++ b/CREDITS
@@ -3554,12 +3554,12 @@ E: cvance@nai.com
 D: portions of the Linux Security Module (LSM) framework and security modules
 
 N: Petr Vandrovec
-E: vandrove@vc.cvut.cz
+E: petr@vandrovec.name
 D: Small contributions to ncpfs
 D: Matrox framebuffer driver
-S: Chudenicka 8
-S: 10200 Prague 10, Hostivar
-S: Czech Republic
+S: 21513 Conradia Ct
+S: Cupertino, CA 95014
+S: USA
 
 N: Thibaut Varene
 E: T-Bone@parisc-linux.org
index a0d479d1e1dd872bd1ae7b4d17d4582df03a384c..f66f4df186908f5d6ba79171e303949683107a1e 100644 (file)
@@ -1645,7 +1645,9 @@ the amount of locking which needs to be done.
       all the readers who were traversing the list when we deleted the
       element are finished.  We use <function>call_rcu()</function> to
       register a callback which will actually destroy the object once
-      the readers are finished.
+      all pre-existing readers are finished.  Alternatively,
+      <function>synchronize_rcu()</function> may be used to block until
+      all pre-existing are finished.
     </para>
     <para>
       But how does Read Copy Update know when the readers are
@@ -1714,7 +1716,7 @@ the amount of locking which needs to be done.
 -        object_put(obj);
 +        list_del_rcu(&amp;obj-&gt;list);
          cache_num--;
-+        call_rcu(&amp;obj-&gt;rcu, cache_delete_rcu, obj);
++        call_rcu(&amp;obj-&gt;rcu, cache_delete_rcu);
  }
 
  /* Must be holding cache_lock */
@@ -1725,14 +1727,6 @@ the amount of locking which needs to be done.
          if (++cache_num > MAX_CACHE_SIZE) {
                  struct object *i, *outcast = NULL;
                  list_for_each_entry(i, &amp;cache, list) {
-@@ -85,6 +94,7 @@
-         obj-&gt;popularity = 0;
-         atomic_set(&amp;obj-&gt;refcnt, 1); /* The cache holds a reference */
-         spin_lock_init(&amp;obj-&gt;lock);
-+        INIT_RCU_HEAD(&amp;obj-&gt;rcu);
-
-         spin_lock_irqsave(&amp;cache_lock, flags);
-         __cache_add(obj);
 @@ -104,12 +114,11 @@
  struct object *cache_find(int id)
  {
index 790d1a8123760211bdcb6427b75c1b4abf2b7210..0c134f8afc6f60b1316b9551577179f1b6dc3961 100644 (file)
@@ -218,13 +218,22 @@ over a rather long period of time, but improvements are always welcome!
        include:
 
        a.      Keeping a count of the number of data-structure elements
-               used by the RCU-protected data structure, including those
-               waiting for a grace period to elapse.  Enforce a limit
-               on this number, stalling updates as needed to allow
-               previously deferred frees to complete.
-
-               Alternatively, limit only the number awaiting deferred
-               free rather than the total number of elements.
+               used by the RCU-protected data structure, including
+               those waiting for a grace period to elapse.  Enforce a
+               limit on this number, stalling updates as needed to allow
+               previously deferred frees to complete.  Alternatively,
+               limit only the number awaiting deferred free rather than
+               the total number of elements.
+
+               One way to stall the updates is to acquire the update-side
+               mutex.  (Don't try this with a spinlock -- other CPUs
+               spinning on the lock could prevent the grace period
+               from ever ending.)  Another way to stall the updates
+               is for the updates to use a wrapper function around
+               the memory allocator, so that this wrapper function
+               simulates OOM when there is too much memory awaiting an
+               RCU grace period.  There are of course many other
+               variations on this theme.
 
        b.      Limiting update rate.  For example, if updates occur only
                once per hour, then no explicit rate limiting is required,
@@ -365,3 +374,26 @@ over a rather long period of time, but improvements are always welcome!
        and the compiler to freely reorder code into and out of RCU
        read-side critical sections.  It is the responsibility of the
        RCU update-side primitives to deal with this.
+
+17.    Use CONFIG_PROVE_RCU, CONFIG_DEBUG_OBJECTS_RCU_HEAD, and
+       the __rcu sparse checks to validate your RCU code.  These
+       can help find problems as follows:
+
+       CONFIG_PROVE_RCU: check that accesses to RCU-protected data
+               structures are carried out under the proper RCU
+               read-side critical section, while holding the right
+               combination of locks, or whatever other conditions
+               are appropriate.
+
+       CONFIG_DEBUG_OBJECTS_RCU_HEAD: check that you don't pass the
+               same object to call_rcu() (or friends) before an RCU
+               grace period has elapsed since the last time that you
+               passed that same object to call_rcu() (or friends).
+
+       __rcu sparse checks: tag the pointer to the RCU-protected data
+               structure with __rcu, and sparse will warn you if you
+               access that pointer without the services of one of the
+               variants of rcu_dereference().
+
+       These debugging aids can help you find problems that are
+       otherwise extremely difficult to spot.
index 44c6dcc93d6dad8e9cee2cadb9f49fa992c5eb00..862c08ef1fde4436ddac8010ba4a1aae40328719 100644 (file)
@@ -80,6 +80,24 @@ o    A CPU looping with bottom halves disabled.  This condition can
 o      For !CONFIG_PREEMPT kernels, a CPU looping anywhere in the kernel
        without invoking schedule().
 
+o      A CPU-bound real-time task in a CONFIG_PREEMPT kernel, which might
+       happen to preempt a low-priority task in the middle of an RCU
+       read-side critical section.   This is especially damaging if
+       that low-priority task is not permitted to run on any other CPU,
+       in which case the next RCU grace period can never complete, which
+       will eventually cause the system to run out of memory and hang.
+       While the system is in the process of running itself out of
+       memory, you might see stall-warning messages.
+
+o      A CPU-bound real-time task in a CONFIG_PREEMPT_RT kernel that
+       is running at a higher priority than the RCU softirq threads.
+       This will prevent RCU callbacks from ever being invoked,
+       and in a CONFIG_TREE_PREEMPT_RCU kernel will further prevent
+       RCU grace periods from ever completing.  Either way, the
+       system will eventually run out of memory and hang.  In the
+       CONFIG_TREE_PREEMPT_RCU case, you might see stall-warning
+       messages.
+
 o      A bug in the RCU implementation.
 
 o      A hardware failure.  This is quite unlikely, but has occurred
index efd8cc95c06b1470db165a74a0d99fd33ee17dea..a851118775d84c7a1d2356ba6a6c8e6208292887 100644 (file)
@@ -125,6 +125,17 @@ o  "b" is the batch limit for this CPU.  If more than this number
        of RCU callbacks is ready to invoke, then the remainder will
        be deferred.
 
+o      "ci" is the number of RCU callbacks that have been invoked for
+       this CPU.  Note that ci+ql is the number of callbacks that have
+       been registered in absence of CPU-hotplug activity.
+
+o      "co" is the number of RCU callbacks that have been orphaned due to
+       this CPU going offline.
+
+o      "ca" is the number of RCU callbacks that have been adopted due to
+       other CPUs going offline.  Note that ci+co-ca+ql is the number of
+       RCU callbacks registered on this CPU.
+
 There is also an rcu/rcudata.csv file with the same information in
 comma-separated-variable spreadsheet format.
 
@@ -180,7 +191,7 @@ o   "s" is the "signaled" state that drives force_quiescent_state()'s
 
 o      "jfq" is the number of jiffies remaining for this grace period
        before force_quiescent_state() is invoked to help push things
-       along.  Note that CPUs in dyntick-idle mode thoughout the grace
+       along.  Note that CPUs in dyntick-idle mode throughout the grace
        period will not report on their own, but rather must be check by
        some other CPU via force_quiescent_state().
 
index f1c5c4bccd3e8ed6674903eedbc59db3cf76ff9e..902d3151f527919ab190d8f30ff9253e3da2c8d0 100644 (file)
@@ -14,25 +14,39 @@ to /proc/cpuinfo.
        identifier (rather than the kernel's).  The actual value is
        architecture and platform dependent.
 
-3) /sys/devices/system/cpu/cpuX/topology/thread_siblings:
+3) /sys/devices/system/cpu/cpuX/topology/book_id:
+
+       the book ID of cpuX. Typically it is the hardware platform's
+       identifier (rather than the kernel's).  The actual value is
+       architecture and platform dependent.
+
+4) /sys/devices/system/cpu/cpuX/topology/thread_siblings:
 
        internel kernel map of cpuX's hardware threads within the same
        core as cpuX
 
-4) /sys/devices/system/cpu/cpuX/topology/core_siblings:
+5) /sys/devices/system/cpu/cpuX/topology/core_siblings:
 
        internal kernel map of cpuX's hardware threads within the same
        physical_package_id.
 
+6) /sys/devices/system/cpu/cpuX/topology/book_siblings:
+
+       internal kernel map of cpuX's hardware threads within the same
+       book_id.
+
 To implement it in an architecture-neutral way, a new source file,
-drivers/base/topology.c, is to export the 4 attributes.
+drivers/base/topology.c, is to export the 4 or 6 attributes. The two book
+related sysfs files will only be created if CONFIG_SCHED_BOOK is selected.
 
 For an architecture to support this feature, it must define some of
 these macros in include/asm-XXX/topology.h:
 #define topology_physical_package_id(cpu)
 #define topology_core_id(cpu)
+#define topology_book_id(cpu)
 #define topology_thread_cpumask(cpu)
 #define topology_core_cpumask(cpu)
+#define topology_book_cpumask(cpu)
 
 The type of **_id is int.
 The type of siblings is (const) struct cpumask *.
@@ -45,6 +59,9 @@ not defined by include/asm-XXX/topology.h:
 3) thread_siblings: just the given CPU
 4) core_siblings: just the given CPU
 
+For architectures that don't support books (CONFIG_SCHED_BOOK) there are no
+default definitions for topology_book_id() and topology_book_cpumask().
+
 Additionally, CPU topology information is provided under
 /sys/devices/system/cpu and includes these files.  The internal
 source for the output is in brackets ("[]").
index 8dd7248508a9e12ac8908c24fb4a3649f157697a..a2ffd6be0ef79374ed6d90773f77f288fc04103e 100644 (file)
@@ -2153,6 +2153,11 @@ and is between 256 and 4096 characters. It is defined in the file
                        Reserves a hole at the top of the kernel virtual
                        address space.
 
+       reservelow=     [X86]
+                       Format: nn[K]
+                       Set the amount of memory to reserve for BIOS at
+                       the bottom of the address space.
+
        reset_devices   [KNL] Force drivers to reset the underlying device
                        during initialization.
 
@@ -2435,6 +2440,10 @@ and is between 256 and 4096 characters. It is defined in the file
                        disables clocksource verification at runtime.
                        Used to enable high-resolution timer mode on older
                        hardware, and in virtualized environment.
+                       [x86] noirqtime: Do not use TSC to do irq accounting.
+                       Used to run time disable IRQ_TIME_ACCOUNTING on any
+                       platforms where RDTSC is slow and this accounting
+                       can add overhead.
 
        turbografx.map[2|3]=    [HW,JOY]
                        TurboGraFX parallel port interface
index 1762b81fcdf2ec4a235423687865453a16196aed..741fe66d6eca4852894cac79d6ad2d3162486c7a 100644 (file)
@@ -542,9 +542,11 @@ Kprobes does not use mutexes or allocate memory except during
 registration and unregistration.
 
 Probe handlers are run with preemption disabled.  Depending on the
-architecture, handlers may also run with interrupts disabled.  In any
-case, your handler should not yield the CPU (e.g., by attempting to
-acquire a semaphore).
+architecture and optimization state, handlers may also run with
+interrupts disabled (e.g., kretprobe handlers and optimized kprobe
+handlers run without interrupt disabled on x86/x86-64).  In any case,
+your handler should not yield the CPU (e.g., by attempting to acquire
+a semaphore).
 
 Since a return probe is implemented by replacing the return
 address with the trampoline's address, stack backtraces and calls
index 2df71861e578b6ebcb2b94371d4a32fbdf7fca61..d9271e74e488a54177c548a5ad859d51ffdaff28 100644 (file)
@@ -1,82 +1,35 @@
 Linux* Base Driver for the Intel(R) PRO/1000 Family of Adapters
 ===============================================================
 
-September 26, 2006
-
+Intel Gigabit Linux driver.
+Copyright(c) 1999 - 2010 Intel Corporation.
 
 Contents
 ========
 
-- In This Release
 - Identifying Your Adapter
-- Building and Installation
 - Command Line Parameters
 - Speed and Duplex Configuration
 - Additional Configurations
-- Known Issues
 - Support
 
-
-In This Release
-===============
-
-This file describes the Linux* Base Driver for the Intel(R) PRO/1000 Family
-of Adapters.  This driver includes support for Itanium(R)2-based systems.
-
-For questions related to hardware requirements, refer to the documentation
-supplied with your Intel PRO/1000 adapter. All hardware requirements listed
-apply to use with Linux.
-
-The following features are now available in supported kernels:
- - Native VLANs
- - Channel Bonding (teaming)
- - SNMP
-
-Channel Bonding documentation can be found in the Linux kernel source:
-/Documentation/networking/bonding.txt
-
-The driver information previously displayed in the /proc filesystem is not
-supported in this release.  Alternatively, you can use ethtool (version 1.6
-or later), lspci, and ifconfig to obtain the same information.
-
-Instructions on updating ethtool can be found in the section "Additional
-Configurations" later in this document.
-
-NOTE: The Intel(R) 82562v 10/100 Network Connection only provides 10/100
-support.
-
-
 Identifying Your Adapter
 ========================
 
 For more information on how to identify your adapter, go to the Adapter &
 Driver ID Guide at:
 
-    http://support.intel.com/support/network/adapter/pro100/21397.htm
+    http://support.intel.com/support/go/network/adapter/idguide.htm
 
 For the latest Intel network drivers for Linux, refer to the following
 website.  In the search field, enter your adapter name or type, or use the
 networking link on the left to search for your adapter:
 
-    http://downloadfinder.intel.com/scripts-df/support_intel.asp
-
+    http://support.intel.com/support/go/network/adapter/home.htm
 
 Command Line Parameters
 =======================
 
-If the driver is built as a module, the  following optional parameters
-are used by entering them on the command line with the modprobe command
-using this syntax:
-
-     modprobe e1000 [<option>=<VAL1>,<VAL2>,...]
-
-For example, with two PRO/1000 PCI adapters, entering:
-
-     modprobe e1000 TxDescriptors=80,128
-
-loads the e1000 driver with 80 TX descriptors for the first adapter and
-128 TX descriptors for the second adapter.
-
 The default value for each parameter is generally the recommended setting,
 unless otherwise noted.
 
@@ -89,10 +42,6 @@ NOTES:  For more information about the AutoNeg, Duplex, and Speed
         parameters, see the application note at:
         http://www.intel.com/design/network/applnots/ap450.htm
 
-        A descriptor describes a data buffer and attributes related to
-        the data buffer.  This information is accessed by the hardware.
-
-
 AutoNeg
 -------
 (Supported only on adapters with copper connections)
@@ -106,7 +55,6 @@ Duplex parameters must not be specified.
 NOTE:  Refer to the Speed and Duplex section of this readme for more
        information on the AutoNeg parameter.
 
-
 Duplex
 ------
 (Supported only on adapters with copper connections)
@@ -119,7 +67,6 @@ set to auto-negotiate, the board auto-detects the correct duplex.  If the
 link partner is forced (either full or half), Duplex defaults to half-
 duplex.
 
-
 FlowControl
 -----------
 Valid Range:   0-3 (0=none, 1=Rx only, 2=Tx only, 3=Rx&Tx)
@@ -128,16 +75,16 @@ Default Value: Reads flow control settings from the EEPROM
 This parameter controls the automatic generation(Tx) and response(Rx)
 to Ethernet PAUSE frames.
 
-
 InterruptThrottleRate
 ---------------------
 (not supported on Intel(R) 82542, 82543 or 82544-based adapters)
-Valid Range:   0,1,3,100-100000 (0=off, 1=dynamic, 3=dynamic conservative)
+Valid Range:   0,1,3,4,100-100000 (0=off, 1=dynamic, 3=dynamic conservative,
+                                   4=simplified balancing)
 Default Value: 3
 
 The driver can limit the amount of interrupts per second that the adapter
-will generate for incoming packets. It does this by writing a value to the 
-adapter that is based on the maximum amount of interrupts that the adapter 
+will generate for incoming packets. It does this by writing a value to the
+adapter that is based on the maximum amount of interrupts that the adapter
 will generate per second.
 
 Setting InterruptThrottleRate to a value greater or equal to 100
@@ -146,37 +93,43 @@ per second, even if more packets have come in. This reduces interrupt
 load on the system and can lower CPU utilization under heavy load,
 but will increase latency as packets are not processed as quickly.
 
-The default behaviour of the driver previously assumed a static 
-InterruptThrottleRate value of 8000, providing a good fallback value for 
-all traffic types,but lacking in small packet performance and latency. 
-The hardware can handle many more small packets per second however, and 
+The default behaviour of the driver previously assumed a static
+InterruptThrottleRate value of 8000, providing a good fallback value for
+all traffic types,but lacking in small packet performance and latency.
+The hardware can handle many more small packets per second however, and
 for this reason an adaptive interrupt moderation algorithm was implemented.
 
 Since 7.3.x, the driver has two adaptive modes (setting 1 or 3) in which
-it dynamically adjusts the InterruptThrottleRate value based on the traffic 
+it dynamically adjusts the InterruptThrottleRate value based on the traffic
 that it receives. After determining the type of incoming traffic in the last
-timeframe, it will adjust the InterruptThrottleRate to an appropriate value 
+timeframe, it will adjust the InterruptThrottleRate to an appropriate value
 for that traffic.
 
 The algorithm classifies the incoming traffic every interval into
-classes.  Once the class is determined, the InterruptThrottleRate value is 
-adjusted to suit that traffic type the best. There are three classes defined: 
+classes.  Once the class is determined, the InterruptThrottleRate value is
+adjusted to suit that traffic type the best. There are three classes defined:
 "Bulk traffic", for large amounts of packets of normal size; "Low latency",
 for small amounts of traffic and/or a significant percentage of small
-packets; and "Lowest latency", for almost completely small packets or 
+packets; and "Lowest latency", for almost completely small packets or
 minimal traffic.
 
-In dynamic conservative mode, the InterruptThrottleRate value is set to 4000 
-for traffic that falls in class "Bulk traffic". If traffic falls in the "Low 
-latency" or "Lowest latency" class, the InterruptThrottleRate is increased 
+In dynamic conservative mode, the InterruptThrottleRate value is set to 4000
+for traffic that falls in class "Bulk traffic". If traffic falls in the "Low
+latency" or "Lowest latency" class, the InterruptThrottleRate is increased
 stepwise to 20000. This default mode is suitable for most applications.
 
 For situations where low latency is vital such as cluster or
 grid computing, the algorithm can reduce latency even more when
 InterruptThrottleRate is set to mode 1. In this mode, which operates
-the same as mode 3, the InterruptThrottleRate will be increased stepwise to 
+the same as mode 3, the InterruptThrottleRate will be increased stepwise to
 70000 for traffic in class "Lowest latency".
 
+In simplified mode the interrupt rate is based on the ratio of Tx and
+Rx traffic.  If the bytes per second rate is approximately equal, the
+interrupt rate will drop as low as 2000 interrupts per second.  If the
+traffic is mostly transmit or mostly receive, the interrupt rate could
+be as high as 8000.
+
 Setting InterruptThrottleRate to 0 turns off any interrupt moderation
 and may improve small packet latency, but is generally not suitable
 for bulk throughput traffic.
@@ -212,8 +165,6 @@ NOTE:  When e1000 is loaded with default settings and multiple adapters
        be platform-specific.  If CPU utilization is not a concern, use
        RX_POLLING (NAPI) and default driver settings.
 
-
-
 RxDescriptors
 -------------
 Valid Range:   80-256 for 82542 and 82543-based adapters
@@ -225,15 +176,14 @@ by the driver.  Increasing this value allows the driver to buffer more
 incoming packets, at the expense of increased system memory utilization.
 
 Each descriptor is 16 bytes.  A receive buffer is also allocated for each
-descriptor and can be either 2048, 4096, 8192, or 16384 bytes, depending 
+descriptor and can be either 2048, 4096, 8192, or 16384 bytes, depending
 on the MTU setting. The maximum MTU size is 16110.
 
-NOTE:  MTU designates the frame size.  It only needs to be set for Jumbo 
-       Frames.  Depending on the available system resources, the request 
-       for a higher number of receive descriptors may be denied.  In this 
+NOTE:  MTU designates the frame size.  It only needs to be set for Jumbo
+       Frames.  Depending on the available system resources, the request
+       for a higher number of receive descriptors may be denied.  In this
        case, use a lower number.
 
-
 RxIntDelay
 ----------
 Valid Range:   0-65535 (0=off)
@@ -254,7 +204,6 @@ CAUTION:  When setting RxIntDelay to a value other than 0, adapters may
           restoring the network connection.  To eliminate the potential
           for the hang ensure that RxIntDelay is set to 0.
 
-
 RxAbsIntDelay
 -------------
 (This parameter is supported only on 82540, 82545 and later adapters.)
@@ -268,7 +217,6 @@ packet is received within the set amount of time.  Proper tuning,
 along with RxIntDelay, may improve traffic throughput in specific network
 conditions.
 
-
 Speed
 -----
 (This parameter is supported only on adapters with copper connections.)
@@ -280,7 +228,6 @@ Speed forces the line speed to the specified value in megabits per second
 partner is set to auto-negotiate, the board will auto-detect the correct
 speed.  Duplex should also be set when Speed is set to either 10 or 100.
 
-
 TxDescriptors
 -------------
 Valid Range:   80-256 for 82542 and 82543-based adapters
@@ -295,6 +242,36 @@ NOTE:  Depending on the available system resources, the request for a
        higher number of transmit descriptors may be denied.  In this case,
        use a lower number.
 
+TxDescriptorStep
+----------------
+Valid Range:    1 (use every Tx Descriptor)
+               4 (use every 4th Tx Descriptor)
+
+Default Value:  1 (use every Tx Descriptor)
+
+On certain non-Intel architectures, it has been observed that intense TX
+traffic bursts of short packets may result in an improper descriptor
+writeback. If this occurs, the driver will report a "TX Timeout" and reset
+the adapter, after which the transmit flow will restart, though data may
+have stalled for as much as 10 seconds before it resumes.
+
+The improper writeback does not occur on the first descriptor in a system
+memory cache-line, which is typically 32 bytes, or 4 descriptors long.
+
+Setting TxDescriptorStep to a value of 4 will ensure that all TX descriptors
+are aligned to the start of a system memory cache line, and so this problem
+will not occur.
+
+NOTES: Setting TxDescriptorStep to 4 effectively reduces the number of
+       TxDescriptors available for transmits to 1/4 of the normal allocation.
+       This has a possible negative performance impact, which may be
+       compensated for by allocating more descriptors using the TxDescriptors
+       module parameter.
+
+       There are other conditions which may result in "TX Timeout", which will
+       not be resolved by the use of the TxDescriptorStep parameter. As the
+       issue addressed by this parameter has never been observed on Intel
+       Architecture platforms, it should not be used on Intel platforms.
 
 TxIntDelay
 ----------
@@ -307,7 +284,6 @@ efficiency if properly tuned for specific network traffic.  If the
 system is reporting dropped transmits, this value may be set too high
 causing the driver to run out of available transmit descriptors.
 
-
 TxAbsIntDelay
 -------------
 (This parameter is supported only on 82540, 82545 and later adapters.)
@@ -330,6 +306,35 @@ Default Value: 1
 A value of '1' indicates that the driver should enable IP checksum
 offload for received packets (both UDP and TCP) to the adapter hardware.
 
+Copybreak
+---------
+Valid Range:   0-xxxxxxx (0=off)
+Default Value: 256
+Usage: insmod e1000.ko copybreak=128
+
+Driver copies all packets below or equaling this size to a fresh Rx
+buffer before handing it up the stack.
+
+This parameter is different than other parameters, in that it is a
+single (not 1,1,1 etc.) parameter applied to all driver instances and
+it is also available during runtime at
+/sys/module/e1000/parameters/copybreak
+
+SmartPowerDownEnable
+--------------------
+Valid Range: 0-1
+Default Value:  0 (disabled)
+
+Allows PHY to turn off in lower power states. The user can turn off
+this parameter in supported chipsets.
+
+KumeranLockLoss
+---------------
+Valid Range: 0-1
+Default Value: 1 (enabled)
+
+This workaround skips resetting the PHY at shutdown for the initial
+silicon releases of ICH8 systems.
 
 Speed and Duplex Configuration
 ==============================
@@ -385,40 +390,9 @@ If the link partner is forced to a specific speed and duplex, then this
 parameter should not be used.  Instead, use the Speed and Duplex parameters
 previously mentioned to force the adapter to the same speed and duplex.
 
-
 Additional Configurations
 =========================
 
-  Configuring the Driver on Different Distributions
-  -------------------------------------------------
-  Configuring a network driver to load properly when the system is started
-  is distribution dependent.  Typically, the configuration process involves
-  adding an alias line to /etc/modules.conf or /etc/modprobe.conf as well
-  as editing other system startup scripts and/or configuration files.  Many
-  popular Linux distributions ship with tools to make these changes for you.
-  To learn the proper way to configure a network device for your system,
-  refer to your distribution documentation.  If during this process you are
-  asked for the driver or module name, the name for the Linux Base Driver
-  for the Intel(R) PRO/1000 Family of Adapters is e1000.
-
-  As an example, if you install the e1000 driver for two PRO/1000 adapters
-  (eth0 and eth1) and set the speed and duplex to 10full and 100half, add
-  the following to modules.conf or or modprobe.conf:
-
-       alias eth0 e1000
-       alias eth1 e1000
-       options e1000 Speed=10,100 Duplex=2,1
-
-  Viewing Link Messages
-  ---------------------
-  Link messages will not be displayed to the console if the distribution is
-  restricting system messages.  In order to see network driver link messages
-  on your console, set dmesg to eight by entering the following:
-
-       dmesg -n 8
-
-  NOTE: This setting is not saved across reboots.
-
   Jumbo Frames
   ------------
   Jumbo Frames support is enabled by changing the MTU to a value larger than
@@ -437,9 +411,11 @@ Additional Configurations
    setting in a different location.
 
   Notes:
-
-  - To enable Jumbo Frames, increase the MTU size on the interface beyond
-    1500.
+  Degradation in throughput performance may be observed in some Jumbo frames
+  environments. If this is observed, increasing the application's socket buffer
+  size and/or increasing the /proc/sys/net/ipv4/tcp_*mem entry values may help.
+  See the specific application manual and /usr/src/linux*/Documentation/
+  networking/ip-sysctl.txt for more details.
 
   - The maximum MTU setting for Jumbo Frames is 16110.  This value coincides
     with the maximum Jumbo Frames size of 16128.
@@ -447,40 +423,11 @@ Additional Configurations
   - Using Jumbo Frames at 10 or 100 Mbps may result in poor performance or
     loss of link.
 
-  - Some Intel gigabit adapters that support Jumbo Frames have a frame size
-    limit of 9238 bytes, with a corresponding MTU size limit of 9216 bytes.
-    The adapters with this limitation are based on the Intel(R) 82571EB,
-    82572EI, 82573L and 80003ES2LAN controller.  These correspond to the
-    following product names:
-     Intel(R) PRO/1000 PT Server Adapter
-     Intel(R) PRO/1000 PT Desktop Adapter
-     Intel(R) PRO/1000 PT Network Connection
-     Intel(R) PRO/1000 PT Dual Port Server Adapter
-     Intel(R) PRO/1000 PT Dual Port Network Connection
-     Intel(R) PRO/1000 PF Server Adapter
-     Intel(R) PRO/1000 PF Network Connection
-     Intel(R) PRO/1000 PF Dual Port Server Adapter
-     Intel(R) PRO/1000 PB Server Connection
-     Intel(R) PRO/1000 PL Network Connection
-     Intel(R) PRO/1000 EB Network Connection with I/O Acceleration
-     Intel(R) PRO/1000 EB Backplane Connection with I/O Acceleration
-     Intel(R) PRO/1000 PT Quad Port Server Adapter
-
   - Adapters based on the Intel(R) 82542 and 82573V/E controller do not
     support Jumbo Frames. These correspond to the following product names:
      Intel(R) PRO/1000 Gigabit Server Adapter
      Intel(R) PRO/1000 PM Network Connection
 
-  - The following adapters do not support Jumbo Frames:
-     Intel(R) 82562V 10/100 Network Connection
-     Intel(R) 82566DM Gigabit Network Connection
-     Intel(R) 82566DC Gigabit Network Connection
-     Intel(R) 82566MM Gigabit Network Connection
-     Intel(R) 82566MC Gigabit Network Connection
-     Intel(R) 82562GT 10/100 Network Connection
-     Intel(R) 82562G 10/100 Network Connection
-
-
   Ethtool
   -------
   The driver utilizes the ethtool interface for driver configuration and
@@ -490,142 +437,14 @@ Additional Configurations
   The latest release of ethtool can be found from
   http://sourceforge.net/projects/gkernel.
 
-  NOTE: Ethtool 1.6 only supports a limited set of ethtool options.  Support
-  for a more complete ethtool feature set can be enabled by upgrading
-  ethtool to ethtool-1.8.1.
-
   Enabling Wake on LAN* (WoL)
   ---------------------------
-  WoL is configured through the Ethtool* utility.  Ethtool is included with
-  all versions of Red Hat after Red Hat 7.2.  For other Linux distributions,
-  download and install Ethtool from the following website:
-  http://sourceforge.net/projects/gkernel.
-
-  For instructions on enabling WoL with Ethtool, refer to the website listed
-  above.
+  WoL is configured through the Ethtool* utility.
 
   WoL will be enabled on the system during the next shut down or reboot.
   For this driver version, in order to enable WoL, the e1000 driver must be
   loaded when shutting down or rebooting the system.
 
-  Wake On LAN is only supported on port A for the following devices:
-  Intel(R) PRO/1000 PT Dual Port Network Connection
-  Intel(R) PRO/1000 PT Dual Port Server Connection
-  Intel(R) PRO/1000 PT Dual Port Server Adapter
-  Intel(R) PRO/1000 PF Dual Port Server Adapter
-  Intel(R) PRO/1000 PT Quad Port Server Adapter
-
-  NAPI
-  ----
-  NAPI (Rx polling mode) is enabled in the e1000 driver.
-
-  See www.cyberus.ca/~hadi/usenix-paper.tgz for more information on NAPI.
-
-
-Known Issues
-============
-
-Dropped Receive Packets on Half-duplex 10/100 Networks
-------------------------------------------------------
-If you have an Intel PCI Express adapter running at 10mbps or 100mbps, half-
-duplex, you may observe occasional dropped receive packets.  There are no
-workarounds for this problem in this network configuration.  The network must
-be updated to operate in full-duplex, and/or 1000mbps only.
-
-Jumbo Frames System Requirement
--------------------------------
-Memory allocation failures have been observed on Linux systems with 64 MB
-of RAM or less that are running Jumbo Frames.  If you are using Jumbo
-Frames, your system may require more than the advertised minimum
-requirement of 64 MB of system memory.
-
-Performance Degradation with Jumbo Frames
------------------------------------------
-Degradation in throughput performance may be observed in some Jumbo frames
-environments.  If this is observed, increasing the application's socket
-buffer size and/or increasing the /proc/sys/net/ipv4/tcp_*mem entry values
-may help.  See the specific application manual and
-/usr/src/linux*/Documentation/
-networking/ip-sysctl.txt for more details.
-
-Jumbo Frames on Foundry BigIron 8000 switch
--------------------------------------------
-There is a known issue using Jumbo frames when connected to a Foundry
-BigIron 8000 switch.  This is a 3rd party limitation.  If you experience
-loss of packets, lower the MTU size.
-
-Allocating Rx Buffers when Using Jumbo Frames 
----------------------------------------------
-Allocating Rx buffers when using Jumbo Frames on 2.6.x kernels may fail if 
-the available memory is heavily fragmented. This issue may be seen with PCI-X 
-adapters or with packet split disabled. This can be reduced or eliminated 
-by changing the amount of available memory for receive buffer allocation, by
-increasing /proc/sys/vm/min_free_kbytes. 
-
-Multiple Interfaces on Same Ethernet Broadcast Network
-------------------------------------------------------
-Due to the default ARP behavior on Linux, it is not possible to have
-one system on two IP networks in the same Ethernet broadcast domain
-(non-partitioned switch) behave as expected.  All Ethernet interfaces
-will respond to IP traffic for any IP address assigned to the system.
-This results in unbalanced receive traffic.
-
-If you have multiple interfaces in a server, either turn on ARP
-filtering by entering:
-
-    echo 1 > /proc/sys/net/ipv4/conf/all/arp_filter
-(this only works if your kernel's version is higher than 2.4.5),
-
-NOTE: This setting is not saved across reboots.  The configuration
-change can be made permanent by adding the line:
-    net.ipv4.conf.all.arp_filter = 1
-to the file /etc/sysctl.conf
-
-      or,
-
-install the interfaces in separate broadcast domains (either in
-different switches or in a switch partitioned to VLANs).
-
-82541/82547 can't link or are slow to link with some link partners
------------------------------------------------------------------
-There is a known compatibility issue with 82541/82547 and some
-low-end switches where the link will not be established, or will
-be slow to establish.  In particular, these switches are known to
-be incompatible with 82541/82547:
-
-    Planex FXG-08TE
-    I-O Data ETG-SH8
-
-To workaround this issue, the driver can be compiled with an override
-of the PHY's master/slave setting.  Forcing master or forcing slave
-mode will improve time-to-link.
-
-    # make CFLAGS_EXTRA=-DE1000_MASTER_SLAVE=<n>
-
-Where <n> is:
-
-    0 = Hardware default
-    1 = Master mode
-    2 = Slave mode
-    3 = Auto master/slave
-
-Disable rx flow control with ethtool
-------------------------------------
-In order to disable receive flow control using ethtool, you must turn
-off auto-negotiation on the same command line.
-
-For example:
-
-   ethtool -A eth? autoneg off rx off
-
-Unplugging network cable while ethtool -p is running
-----------------------------------------------------
-In kernel versions 2.5.50 and later (including 2.6 kernel), unplugging
-the network cable while ethtool -p is running will cause the system to
-become unresponsive to keyboard commands, except for control-alt-delete.
-Restarting the system appears to be the only remedy.
-
-
 Support
 =======
 
diff --git a/Documentation/networking/e1000e.txt b/Documentation/networking/e1000e.txt
new file mode 100644 (file)
index 0000000..6aa048b
--- /dev/null
@@ -0,0 +1,302 @@
+Linux* Driver for Intel(R) Network Connection
+===============================================================
+
+Intel Gigabit Linux driver.
+Copyright(c) 1999 - 2010 Intel Corporation.
+
+Contents
+========
+
+- Identifying Your Adapter
+- Command Line Parameters
+- Additional Configurations
+- Support
+
+Identifying Your Adapter
+========================
+
+The e1000e driver supports all PCI Express Intel(R) Gigabit Network
+Connections, except those that are 82575, 82576 and 82580-based*.
+
+* NOTE: The Intel(R) PRO/1000 P Dual Port Server Adapter is supported by
+  the e1000 driver, not the e1000e driver due to the 82546 part being used
+  behind a PCI Express bridge.
+
+For more information on how to identify your adapter, go to the Adapter &
+Driver ID Guide at:
+
+    http://support.intel.com/support/go/network/adapter/idguide.htm
+
+For the latest Intel network drivers for Linux, refer to the following
+website.  In the search field, enter your adapter name or type, or use the
+networking link on the left to search for your adapter:
+
+    http://support.intel.com/support/go/network/adapter/home.htm
+
+Command Line Parameters
+=======================
+
+The default value for each parameter is generally the recommended setting,
+unless otherwise noted.
+
+NOTES:  For more information about the InterruptThrottleRate,
+        RxIntDelay, TxIntDelay, RxAbsIntDelay, and TxAbsIntDelay
+        parameters, see the application note at:
+        http://www.intel.com/design/network/applnots/ap450.htm
+
+InterruptThrottleRate
+---------------------
+Valid Range:   0,1,3,4,100-100000 (0=off, 1=dynamic, 3=dynamic conservative,
+                                   4=simplified balancing)
+Default Value: 3
+
+The driver can limit the amount of interrupts per second that the adapter
+will generate for incoming packets. It does this by writing a value to the
+adapter that is based on the maximum amount of interrupts that the adapter
+will generate per second.
+
+Setting InterruptThrottleRate to a value greater or equal to 100
+will program the adapter to send out a maximum of that many interrupts
+per second, even if more packets have come in. This reduces interrupt
+load on the system and can lower CPU utilization under heavy load,
+but will increase latency as packets are not processed as quickly.
+
+The driver has two adaptive modes (setting 1 or 3) in which
+it dynamically adjusts the InterruptThrottleRate value based on the traffic
+that it receives. After determining the type of incoming traffic in the last
+timeframe, it will adjust the InterruptThrottleRate to an appropriate value
+for that traffic.
+
+The algorithm classifies the incoming traffic every interval into
+classes.  Once the class is determined, the InterruptThrottleRate value is
+adjusted to suit that traffic type the best. There are three classes defined:
+"Bulk traffic", for large amounts of packets of normal size; "Low latency",
+for small amounts of traffic and/or a significant percentage of small
+packets; and "Lowest latency", for almost completely small packets or
+minimal traffic.
+
+In dynamic conservative mode, the InterruptThrottleRate value is set to 4000
+for traffic that falls in class "Bulk traffic". If traffic falls in the "Low
+latency" or "Lowest latency" class, the InterruptThrottleRate is increased
+stepwise to 20000. This default mode is suitable for most applications.
+
+For situations where low latency is vital such as cluster or
+grid computing, the algorithm can reduce latency even more when
+InterruptThrottleRate is set to mode 1. In this mode, which operates
+the same as mode 3, the InterruptThrottleRate will be increased stepwise to
+70000 for traffic in class "Lowest latency".
+
+In simplified mode the interrupt rate is based on the ratio of Tx and
+Rx traffic.  If the bytes per second rate is approximately equal the
+interrupt rate will drop as low as 2000 interrupts per second.  If the
+traffic is mostly transmit or mostly receive, the interrupt rate could
+be as high as 8000.
+
+Setting InterruptThrottleRate to 0 turns off any interrupt moderation
+and may improve small packet latency, but is generally not suitable
+for bulk throughput traffic.
+
+NOTE:  InterruptThrottleRate takes precedence over the TxAbsIntDelay and
+       RxAbsIntDelay parameters.  In other words, minimizing the receive
+       and/or transmit absolute delays does not force the controller to
+       generate more interrupts than what the Interrupt Throttle Rate
+       allows.
+
+NOTE:  When e1000e is loaded with default settings and multiple adapters
+       are in use simultaneously, the CPU utilization may increase non-
+       linearly.  In order to limit the CPU utilization without impacting
+       the overall throughput, we recommend that you load the driver as
+       follows:
+
+           modprobe e1000e InterruptThrottleRate=3000,3000,3000
+
+       This sets the InterruptThrottleRate to 3000 interrupts/sec for
+       the first, second, and third instances of the driver.  The range
+       of 2000 to 3000 interrupts per second works on a majority of
+       systems and is a good starting point, but the optimal value will
+       be platform-specific.  If CPU utilization is not a concern, use
+       RX_POLLING (NAPI) and default driver settings.
+
+RxIntDelay
+----------
+Valid Range:   0-65535 (0=off)
+Default Value: 0
+
+This value delays the generation of receive interrupts in units of 1.024
+microseconds.  Receive interrupt reduction can improve CPU efficiency if
+properly tuned for specific network traffic.  Increasing this value adds
+extra latency to frame reception and can end up decreasing the throughput
+of TCP traffic.  If the system is reporting dropped receives, this value
+may be set too high, causing the driver to run out of available receive
+descriptors.
+
+CAUTION:  When setting RxIntDelay to a value other than 0, adapters may
+          hang (stop transmitting) under certain network conditions.  If
+          this occurs a NETDEV WATCHDOG message is logged in the system
+          event log.  In addition, the controller is automatically reset,
+          restoring the network connection.  To eliminate the potential
+          for the hang ensure that RxIntDelay is set to 0.
+
+RxAbsIntDelay
+-------------
+Valid Range:   0-65535 (0=off)
+Default Value: 8
+
+This value, in units of 1.024 microseconds, limits the delay in which a
+receive interrupt is generated.  Useful only if RxIntDelay is non-zero,
+this value ensures that an interrupt is generated after the initial
+packet is received within the set amount of time.  Proper tuning,
+along with RxIntDelay, may improve traffic throughput in specific network
+conditions.
+
+TxIntDelay
+----------
+Valid Range:   0-65535 (0=off)
+Default Value: 8
+
+This value delays the generation of transmit interrupts in units of
+1.024 microseconds.  Transmit interrupt reduction can improve CPU
+efficiency if properly tuned for specific network traffic.  If the
+system is reporting dropped transmits, this value may be set too high
+causing the driver to run out of available transmit descriptors.
+
+TxAbsIntDelay
+-------------
+Valid Range:   0-65535 (0=off)
+Default Value: 32
+
+This value, in units of 1.024 microseconds, limits the delay in which a
+transmit interrupt is generated.  Useful only if TxIntDelay is non-zero,
+this value ensures that an interrupt is generated after the initial
+packet is sent on the wire within the set amount of time.  Proper tuning,
+along with TxIntDelay, may improve traffic throughput in specific
+network conditions.
+
+Copybreak
+---------
+Valid Range:   0-xxxxxxx (0=off)
+Default Value: 256
+
+Driver copies all packets below or equaling this size to a fresh Rx
+buffer before handing it up the stack.
+
+This parameter is different than other parameters, in that it is a
+single (not 1,1,1 etc.) parameter applied to all driver instances and
+it is also available during runtime at
+/sys/module/e1000e/parameters/copybreak
+
+SmartPowerDownEnable
+--------------------
+Valid Range: 0-1
+Default Value:  0 (disabled)
+
+Allows PHY to turn off in lower power states. The user can set this parameter
+in supported chipsets.
+
+KumeranLockLoss
+---------------
+Valid Range: 0-1
+Default Value: 1 (enabled)
+
+This workaround skips resetting the PHY at shutdown for the initial
+silicon releases of ICH8 systems.
+
+IntMode
+-------
+Valid Range: 0-2 (0=legacy, 1=MSI, 2=MSI-X)
+Default Value: 2
+
+Allows changing the interrupt mode at module load time, without requiring a
+recompile. If the driver load fails to enable a specific interrupt mode, the
+driver will try other interrupt modes, from least to most compatible.  The
+interrupt order is MSI-X, MSI, Legacy.  If specifying MSI (IntMode=1)
+interrupts, only MSI and Legacy will be attempted.
+
+CrcStripping
+------------
+Valid Range: 0-1
+Default Value: 1 (enabled)
+
+Strip the CRC from received packets before sending up the network stack.  If
+you have a machine with a BMC enabled but cannot receive IPMI traffic after
+loading or enabling the driver, try disabling this feature.
+
+WriteProtectNVM
+---------------
+Valid Range: 0-1
+Default Value: 1 (enabled)
+
+Set the hardware to ignore all write/erase cycles to the GbE region in the
+ICHx NVM (non-volatile memory).  This feature can be disabled by the
+WriteProtectNVM module parameter (enabled by default) only after a hardware
+reset, but the machine must be power cycled before trying to enable writes.
+
+Note: the kernel boot option iomem=relaxed may need to be set if the kernel
+config option CONFIG_STRICT_DEVMEM=y, if the root user wants to write the
+NVM from user space via ethtool.
+
+Additional Configurations
+=========================
+
+  Jumbo Frames
+  ------------
+  Jumbo Frames support is enabled by changing the MTU to a value larger than
+  the default of 1500.  Use the ifconfig command to increase the MTU size.
+  For example:
+
+       ifconfig eth<x> mtu 9000 up
+
+  This setting is not saved across reboots.
+
+  Notes:
+
+  - The maximum MTU setting for Jumbo Frames is 9216.  This value coincides
+    with the maximum Jumbo Frames size of 9234 bytes.
+
+  - Using Jumbo Frames at 10 or 100 Mbps is not supported and may result in
+    poor performance or loss of link.
+
+  - Some adapters limit Jumbo Frames sized packets to a maximum of
+    4096 bytes and some adapters do not support Jumbo Frames.
+
+
+  Ethtool
+  -------
+  The driver utilizes the ethtool interface for driver configuration and
+  diagnostics, as well as displaying statistical information.  We
+  strongly recommend downloading the latest version of Ethtool at:
+
+  http://sourceforge.net/projects/gkernel.
+
+  Speed and Duplex
+  ----------------
+  Speed and Duplex are configured through the Ethtool* utility. For
+  instructions,  refer to the Ethtool man page.
+
+  Enabling Wake on LAN* (WoL)
+  ---------------------------
+  WoL is configured through the Ethtool* utility. For instructions on
+  enabling WoL with Ethtool, refer to the Ethtool man page.
+
+  WoL will be enabled on the system during the next shut down or reboot.
+  For this driver version, in order to enable WoL, the e1000e driver must be
+  loaded when shutting down or rebooting the system.
+
+  In most cases Wake On LAN is only supported on port A for multiple port
+  adapters. To verify if a port supports Wake on LAN run ethtool eth<X>.
+
+
+Support
+=======
+
+For general information, go to the Intel support website at:
+
+    www.intel.com/support/
+
+or the Intel Wired Networking project hosted by Sourceforge at:
+
+    http://sourceforge.net/projects/e1000
+
+If an issue is identified with the released source code on the supported
+kernel with a supported adapter, email the specific information related
+to the issue to e1000-devel@lists.sf.net
old mode 100755 (executable)
new mode 100644 (file)
index 19015de..21dd5d1
@@ -1,19 +1,16 @@
 Linux* Base Driver for Intel(R) Network Connection
 ==================================================
 
-November 24, 2009
+Intel Gigabit Linux driver.
+Copyright(c) 1999 - 2010 Intel Corporation.
 
 Contents
 ========
 
-- In This Release
 - Identifying Your Adapter
 - Known Issues/Troubleshooting
 - Support
 
-In This Release
-===============
-
 This file describes the ixgbevf Linux* Base Driver for Intel Network
 Connection.
 
@@ -33,7 +30,7 @@ Identifying Your Adapter
 For more information on how to identify your adapter, go to the Adapter &
 Driver ID Guide at:
 
-    http://support.intel.com/support/network/sb/CS-008441.htm
+    http://support.intel.com/support/go/network/adapter/idguide.htm
 
 Known Issues/Troubleshooting
 ============================
@@ -57,34 +54,3 @@ or the Intel Wired Networking project hosted by Sourceforge at:
 If an issue is identified with the released source code on the supported
 kernel with a supported adapter, email the specific information related
 to the issue to e1000-devel@lists.sf.net
-
-License
-=======
-
-Intel 10 Gigabit Linux driver.
-Copyright(c) 1999 - 2009 Intel Corporation.
-
-This program is free software; you can redistribute it and/or modify it
-under the terms and conditions of the GNU General Public License,
-version 2, as published by the Free Software Foundation.
-
-This program is distributed in the hope it will be useful, but WITHOUT
-ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
-FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
-more details.
-
-You should have received a copy of the GNU General Public License along with
-this program; if not, write to the Free Software Foundation, Inc.,
-51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
-
-The full GNU General Public License is included in this distribution in
-the file called "COPYING".
-
-Trademarks
-==========
-
-Intel, Itanium, and Pentium are trademarks or registered trademarks of
-Intel Corporation or its subsidiaries in the United States and other
-countries.
-
-* Other names and brands may be claimed as the property of others.
index ccd951fa94eeda42a5a50c854b6b92e512fe6724..cc96ee2666f2e5f10f13b83b1fab72bcf8e18924 100644 (file)
@@ -478,7 +478,7 @@ static void prepare_hwpoison_fd(void)
        }
 
        if (opt_unpoison && !hwpoison_forget_fd) {
-               sprintf(buf, "%s/renew-pfn", hwpoison_debug_fs);
+               sprintf(buf, "%s/unpoison-pfn", hwpoison_debug_fs);
                hwpoison_forget_fd = checked_open(buf, O_WRONLY);
        }
 }
index 50b8148448fdfe47b2abe1c838636e4b5ed6391c..3d4179fbc5263ff4b624753c9ce06004286ce246 100644 (file)
@@ -962,6 +962,23 @@ W: http://www.fluff.org/ben/linux/
 S:     Maintained
 F:     arch/arm/mach-s3c6410/
 
+ARM/S5P ARM ARCHITECTURES
+M:     Kukjin Kim <kgene.kim@samsung.com>
+L:     linux-arm-kernel@lists.infradead.org (moderated for non-subscribers)
+L:     linux-samsung-soc@vger.kernel.org (moderated for non-subscribers)
+S:     Maintained
+F:     arch/arm/mach-s5p*/
+
+ARM/SAMSUNG S5P SERIES FIMC SUPPORT
+M:     Kyungmin Park <kyungmin.park@samsung.com>
+M:     Sylwester Nawrocki <s.nawrocki@samsung.com>
+L:     linux-arm-kernel@lists.infradead.org
+L:     linux-media@vger.kernel.org
+S:     Maintained
+F:     arch/arm/plat-s5p/dev-fimc*
+F:     arch/arm/plat-samsung/include/plat/*fimc*
+F:     drivers/media/video/s5p-fimc/
+
 ARM/SHMOBILE ARM ARCHITECTURE
 M:     Paul Mundt <lethal@linux-sh.org>
 M:     Magnus Damm <magnus.damm@gmail.com>
@@ -1220,7 +1237,7 @@ F:        drivers/auxdisplay/
 F:     include/linux/cfag12864b.h
 
 AVR32 ARCHITECTURE
-M:     Haavard Skinnemoen <hskinnemoen@atmel.com>
+M:     Hans-Christian Egtvedt <hans-christian.egtvedt@atmel.com>
 W:     http://www.atmel.com/products/AVR32/
 W:     http://avr32linux.org/
 W:     http://avrfreaks.net/
@@ -1228,7 +1245,7 @@ S:        Supported
 F:     arch/avr32/
 
 AVR32/AT32AP MACHINE SUPPORT
-M:     Haavard Skinnemoen <hskinnemoen@atmel.com>
+M:     Hans-Christian Egtvedt <hans-christian.egtvedt@atmel.com>
 S:     Supported
 F:     arch/avr32/mach-at32ap/
 
@@ -1510,6 +1527,8 @@ T:        git git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client.git
 S:     Supported
 F:     Documentation/filesystems/ceph.txt
 F:     fs/ceph
+F:     net/ceph
+F:     include/linux/ceph
 
 CERTIFIED WIRELESS USB (WUSB) SUBSYSTEM:
 M:     David Vrabel <david.vrabel@csr.com>
@@ -2199,6 +2218,12 @@ W:       http://acpi4asus.sf.net
 S:     Maintained
 F:     drivers/platform/x86/eeepc-laptop.c
 
+EFIFB FRAMEBUFFER DRIVER
+L:     linux-fbdev@vger.kernel.org
+M:     Peter Jones <pjones@redhat.com>
+S:     Maintained
+F:     drivers/video/efifb.c
+
 EFS FILESYSTEM
 W:     http://aeschi.ch.eu.org/efs/
 S:     Orphan
@@ -2522,7 +2547,7 @@ S:        Supported
 F:     drivers/scsi/gdt*
 
 GENERIC GPIO I2C DRIVER
-M:     Haavard Skinnemoen <hskinnemoen@atmel.com>
+M:     Haavard Skinnemoen <hskinnemoen@gmail.com>
 S:     Supported
 F:     drivers/i2c/busses/i2c-gpio.c
 F:     include/linux/i2c-gpio.h
@@ -2662,6 +2687,8 @@ M:        Guenter Roeck <guenter.roeck@ericsson.com>
 L:     lm-sensors@lm-sensors.org
 W:     http://www.lm-sensors.org/
 T:     quilt kernel.org/pub/linux/kernel/people/jdelvare/linux-2.6/jdelvare-hwmon/
+T:     quilt kernel.org/pub/linux/kernel/people/groeck/linux-staging/
+T:     git git://git.kernel.org/pub/scm/linux/kernel/git/groeck/linux-staging.git
 S:     Maintained
 F:     Documentation/hwmon/
 F:     drivers/hwmon/
@@ -3048,16 +3075,27 @@ L:      netdev@vger.kernel.org
 S:     Maintained
 F:     drivers/net/ixp2000/
 
-INTEL ETHERNET DRIVERS (e100/e1000/e1000e/igb/igbvf/ixgb/ixgbe)
+INTEL ETHERNET DRIVERS (e100/e1000/e1000e/igb/igbvf/ixgb/ixgbe/ixgbevf)
 M:     Jeff Kirsher <jeffrey.t.kirsher@intel.com>
 M:     Jesse Brandeburg <jesse.brandeburg@intel.com>
 M:     Bruce Allan <bruce.w.allan@intel.com>
-M:     Alex Duyck <alexander.h.duyck@intel.com>
+M:     Carolyn Wyborny <carolyn.wyborny@intel.com>
+M:     Don Skidmore <donald.c.skidmore@intel.com>
+M:     Greg Rose <gregory.v.rose@intel.com>
 M:     PJ Waskiewicz <peter.p.waskiewicz.jr@intel.com>
+M:     Alex Duyck <alexander.h.duyck@intel.com>
 M:     John Ronciak <john.ronciak@intel.com>
 L:     e1000-devel@lists.sourceforge.net
 W:     http://e1000.sourceforge.net/
 S:     Supported
+F:     Documentation/networking/e100.txt
+F:     Documentation/networking/e1000.txt
+F:     Documentation/networking/e1000e.txt
+F:     Documentation/networking/igb.txt
+F:     Documentation/networking/igbvf.txt
+F:     Documentation/networking/ixgb.txt
+F:     Documentation/networking/ixgbe.txt
+F:     Documentation/networking/ixgbevf.txt
 F:     drivers/net/e100.c
 F:     drivers/net/e1000/
 F:     drivers/net/e1000e/
@@ -3065,6 +3103,7 @@ F:        drivers/net/igb/
 F:     drivers/net/igbvf/
 F:     drivers/net/ixgb/
 F:     drivers/net/ixgbe/
+F:     drivers/net/ixgbevf/
 
 INTEL PRO/WIRELESS 2100 NETWORK CONNECTION SUPPORT
 L:     linux-wireless@vger.kernel.org
@@ -3125,7 +3164,7 @@ F:        drivers/net/ioc3-eth.c
 
 IOC3 SERIAL DRIVER
 M:     Pat Gefre <pfg@sgi.com>
-L:     linux-mips@linux-mips.org
+L:     linux-serial@vger.kernel.org
 S:     Maintained
 F:     drivers/serial/ioc3_serial.c
 
@@ -3773,9 +3812,8 @@ W:        http://www.syskonnect.com
 S:     Supported
 
 MATROX FRAMEBUFFER DRIVER
-M:     Petr Vandrovec <vandrove@vc.cvut.cz>
 L:     linux-fbdev@vger.kernel.org
-S:     Maintained
+S:     Orphan
 F:     drivers/video/matrox/matroxfb_*
 F:     include/linux/matroxfb.h
 
@@ -3899,10 +3937,8 @@ F:       Documentation/serial/moxa-smartio
 F:     drivers/char/mxser.*
 
 MSI LAPTOP SUPPORT
-M:     Lennart Poettering <mzxreary@0pointer.de>
+M:     Lee, Chun-Yi <jlee@novell.com>
 L:     platform-driver-x86@vger.kernel.org
-W:     https://tango.0pointer.de/mailman/listinfo/s270-linux
-W:     http://0pointer.de/lennart/tchibo.html
 S:     Maintained
 F:     drivers/platform/x86/msi-laptop.c
 
@@ -3919,8 +3955,10 @@ S:       Supported
 F:     drivers/mfd/
 
 MULTIMEDIA CARD (MMC), SECURE DIGITAL (SD) AND SDIO SUBSYSTEM
-S:     Orphan
+M:     Chris Ball <cjb@laptop.org>
 L:     linux-mmc@vger.kernel.org
+T:     git git://git.kernel.org/pub/scm/linux/kernel/git/cjb/mmc.git
+S:     Maintained
 F:     drivers/mmc/
 F:     include/linux/mmc/
 
@@ -3962,8 +4000,8 @@ S:        Maintained
 F:     drivers/net/natsemi.c
 
 NCP FILESYSTEM
-M:     Petr Vandrovec <vandrove@vc.cvut.cz>
-S:     Maintained
+M:     Petr Vandrovec <petr@vandrovec.name>
+S:     Odd Fixes
 F:     fs/ncpfs/
 
 NCR DUAL 700 SCSI DRIVER (MICROCHANNEL)
@@ -4769,6 +4807,15 @@ F:       fs/qnx4/
 F:     include/linux/qnx4_fs.h
 F:     include/linux/qnxtypes.h
 
+RADOS BLOCK DEVICE (RBD)
+F:     include/linux/qnxtypes.h
+M:     Yehuda Sadeh <yehuda@hq.newdream.net>
+M:     Sage Weil <sage@newdream.net>
+M:     ceph-devel@vger.kernel.org
+S:     Supported
+F:     drivers/block/rbd.c
+F:     drivers/block/rbd_types.h
+
 RADEON FRAMEBUFFER DISPLAY DRIVER
 M:     Benjamin Herrenschmidt <benh@kernel.crashing.org>
 L:     linux-fbdev@vger.kernel.org
@@ -4994,6 +5041,12 @@ F:       drivers/media/common/saa7146*
 F:     drivers/media/video/*7146*
 F:     include/media/*7146*
 
+SAMSUNG AUDIO (ASoC) DRIVERS
+M:     Jassi Brar <jassi.brar@samsung.com>
+L:     alsa-devel@alsa-project.org (moderated for non-subscribers)
+S:     Supported
+F:     sound/soc/s3c24xx
+
 TLG2300 VIDEO4LINUX-2 DRIVER
 M:     Huang Shijie <shijie8@gmail.com>
 M:     Kang Yong <kangyong@telegent.com>
@@ -5091,8 +5144,10 @@ S:       Maintained
 F:     drivers/mmc/host/sdricoh_cs.c
 
 SECURE DIGITAL HOST CONTROLLER INTERFACE (SDHCI) DRIVER
-S:     Orphan
+M:     Chris Ball <cjb@laptop.org>
 L:     linux-mmc@vger.kernel.org
+T:     git git://git.kernel.org/pub/scm/linux/kernel/git/cjb/mmc.git
+S:     Maintained
 F:     drivers/mmc/host/sdhci.*
 
 SECURE DIGITAL HOST CONTROLLER INTERFACE, OPEN FIRMWARE BINDINGS (SDHCI-OF)
@@ -6434,8 +6489,10 @@ F:       include/linux/wm97xx.h
 WOLFSON MICROELECTRONICS DRIVERS
 M:     Mark Brown <broonie@opensource.wolfsonmicro.com>
 M:     Ian Lartey <ian@opensource.wolfsonmicro.com>
+M:     Dimitris Papastamos <dp@opensource.wolfsonmicro.com>
+T:     git git://opensource.wolfsonmicro.com/linux-2.6-asoc
 T:     git git://opensource.wolfsonmicro.com/linux-2.6-audioplus
-W:     http://opensource.wolfsonmicro.com/node/8
+W:     http://opensource.wolfsonmicro.com/content/linux-drivers-wolfson-devices
 S:     Supported
 F:     Documentation/hwmon/wm83??
 F:     drivers/leds/leds-wm83*.c
index 3133a5772eeb13e85f4eedf994ac662213b3e48c..d3c10719bbbd5b311c51f9a407184fd30847e6ea 100644 (file)
--- a/Makefile
+++ b/Makefile
@@ -1,8 +1,8 @@
 VERSION = 2
 PATCHLEVEL = 6
 SUBLEVEL = 36
-EXTRAVERSION = -rc5
-NAME = Sheep on Meth
+EXTRAVERSION =
+NAME = Flesh-Eating Bats with Fangs
 
 # *DOCUMENTATION*
 # To see a list of typical targets execute "make help"
@@ -568,6 +568,12 @@ endif
 
 ifdef CONFIG_FUNCTION_TRACER
 KBUILD_CFLAGS  += -pg
+ifdef CONFIG_DYNAMIC_FTRACE
+       ifdef CONFIG_HAVE_C_RECORDMCOUNT
+               BUILD_C_RECORDMCOUNT := y
+               export BUILD_C_RECORDMCOUNT
+       endif
+endif
 endif
 
 # We trigger additional mismatches with less inlining
@@ -591,6 +597,11 @@ KBUILD_CFLAGS      += $(call cc-option,-fno-strict-overflow)
 # conserve stack if available
 KBUILD_CFLAGS   += $(call cc-option,-fconserve-stack)
 
+# check for 'asm goto'
+ifeq ($(shell $(CONFIG_SHELL) $(srctree)/scripts/gcc-goto.sh $(CC)), y)
+       KBUILD_CFLAGS += -DCC_HAVE_ASM_GOTO
+endif
+
 # Add user supplied CPPFLAGS, AFLAGS and CFLAGS as the last assignments
 # But warn user when we do so
 warn-assign = \
index 4877a8c8ee1697599289f35107824f95ba8daf84..53d7f619a1b9b46643c59cd3c7d4ace354148c20 100644 (file)
@@ -32,8 +32,9 @@ config HAVE_OPROFILE
 
 config KPROBES
        bool "Kprobes"
-       depends on KALLSYMS && MODULES
+       depends on MODULES
        depends on HAVE_KPROBES
+       select KALLSYMS
        help
          Kprobes allows you to trap at almost any kernel address and
          execute a callback function.  register_kprobe() establishes
@@ -45,7 +46,6 @@ config OPTPROBES
        def_bool y
        depends on KPROBES && HAVE_OPTPROBES
        depends on !PREEMPT
-       select KALLSYMS_ALL
 
 config HAVE_EFFICIENT_UNALIGNED_ACCESS
        bool
@@ -158,4 +158,7 @@ config HAVE_PERF_EVENTS_NMI
          subsystem.  Also has support for calculating CPU cycle events
          to determine how many clock cycles in a given period.
 
+config HAVE_ARCH_JUMP_LABEL
+       bool
+
 source "kernel/gcov/Kconfig"
index b9647bb66d1388d9c13503748252c2a1df4ac830..d04ccd73af45f6487f442a0dba26921db884e218 100644 (file)
@@ -9,6 +9,7 @@ config ALPHA
        select HAVE_IDE
        select HAVE_OPROFILE
        select HAVE_SYSCALL_WRAPPERS
+       select HAVE_IRQ_WORK
        select HAVE_PERF_EVENTS
        select HAVE_DMA_ATTRS
        help
index 4157cd3c44a96d2811410b334a35f11923ba6fbf..fe792ca818f64c4d9954e8b62e5f8064a5d5777e 100644 (file)
@@ -1,11 +1,6 @@
 #ifndef __ASM_ALPHA_PERF_EVENT_H
 #define __ASM_ALPHA_PERF_EVENT_H
 
-/* Alpha only supports software events through this interface. */
-extern void set_perf_event_pending(void);
-
-#define PERF_EVENT_INDEX_OFFSET 0
-
 #ifdef CONFIG_PERF_EVENTS
 extern void init_hw_perf_events(void);
 #else
index ab1ee0ab082b996fd0d656b62baa2fcccab7b0b8..6d159cee5f2f43a48bc9aac8355069a3e588ef7f 100644 (file)
@@ -73,8 +73,6 @@
        ldq     $20, HAE_REG($19);      \
        stq     $21, HAE_CACHE($19);    \
        stq     $21, 0($20);            \
-       ldq     $0, 0($sp);             \
-       ldq     $1, 8($sp);             \
 99:;                                   \
        ldq     $19, 72($sp);           \
        ldq     $20, 80($sp);           \
@@ -316,7 +314,7 @@ ret_from_sys_call:
        cmovne  $26, 0, $19             /* $19 = 0 => non-restartable */
        ldq     $0, SP_OFF($sp)
        and     $0, 8, $0
-       beq     $0, restore_all
+       beq     $0, ret_to_kernel
 ret_to_user:
        /* Make sure need_resched and sigpending don't change between
                sampling and the rti.  */
@@ -329,6 +327,11 @@ restore_all:
        RESTORE_ALL
        call_pal PAL_rti
 
+ret_to_kernel:
+       lda     $16, 7
+       call_pal PAL_swpipl
+       br restore_all
+
        .align 3
 $syscall_error:
        /*
@@ -657,7 +660,7 @@ kernel_thread:
        /* We don't actually care for a3 success widgetry in the kernel.
           Not for positive errno values.  */
        stq     $0, 0($sp)              /* $0 */
-       br      restore_all
+       br      ret_to_kernel
 .end kernel_thread
 
 /*
@@ -911,15 +914,6 @@ sys_execve:
        jmp     $31, do_sys_execve
 .end sys_execve
 
-       .align  4
-       .globl  osf_sigprocmask
-       .ent    osf_sigprocmask
-osf_sigprocmask:
-       .prologue 0
-       mov     $sp, $18
-       jmp     $31, sys_osf_sigprocmask
-.end osf_sigprocmask
-
        .align  4
        .globl  alpha_ni_syscall
        .ent    alpha_ni_syscall
index 85d8e4f58c83ce612269162b9635bd49059c39dd..1cc49683fb69b2a5f96639e71a2f1af821479e77 100644 (file)
@@ -307,7 +307,7 @@ again:
                             new_raw_count) != prev_raw_count)
                goto again;
 
-       delta = (new_raw_count  - (prev_raw_count & alpha_pmu->pmc_count_mask[idx])) + ovf;
+       delta = (new_raw_count - (prev_raw_count & alpha_pmu->pmc_count_mask[idx])) + ovf;
 
        /* It is possible on very rare occasions that the PMC has overflowed
         * but the interrupt is yet to come.  Detect and fix this situation.
@@ -402,14 +402,13 @@ static void maybe_change_configuration(struct cpu_hw_events *cpuc)
                struct hw_perf_event *hwc = &pe->hw;
                int idx = hwc->idx;
 
-               if (cpuc->current_idx[j] != PMC_NO_INDEX) {
-                       cpuc->idx_mask |= (1<<cpuc->current_idx[j]);
-                       continue;
+               if (cpuc->current_idx[j] == PMC_NO_INDEX) {
+                       alpha_perf_event_set_period(pe, hwc, idx);
+                       cpuc->current_idx[j] = idx;
                }
 
-               alpha_perf_event_set_period(pe, hwc, idx);
-               cpuc->current_idx[j] = idx;
-               cpuc->idx_mask |= (1<<cpuc->current_idx[j]);
+               if (!(hwc->state & PERF_HES_STOPPED))
+                       cpuc->idx_mask |= (1<<cpuc->current_idx[j]);
        }
        cpuc->config = cpuc->event[0]->hw.config_base;
 }
@@ -420,12 +419,13 @@ static void maybe_change_configuration(struct cpu_hw_events *cpuc)
  *  - this function is called from outside this module via the pmu struct
  *    returned from perf event initialisation.
  */
-static int alpha_pmu_enable(struct perf_event *event)
+static int alpha_pmu_add(struct perf_event *event, int flags)
 {
        struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
+       struct hw_perf_event *hwc = &event->hw;
        int n0;
        int ret;
-       unsigned long flags;
+       unsigned long irq_flags;
 
        /*
         * The Sparc code has the IRQ disable first followed by the perf
@@ -435,8 +435,8 @@ static int alpha_pmu_enable(struct perf_event *event)
         * nevertheless we disable the PMCs first to enable a potential
         * final PMI to occur before we disable interrupts.
         */
-       perf_disable();
-       local_irq_save(flags);
+       perf_pmu_disable(event->pmu);
+       local_irq_save(irq_flags);
 
        /* Default to error to be returned */
        ret = -EAGAIN;
@@ -455,8 +455,12 @@ static int alpha_pmu_enable(struct perf_event *event)
                }
        }
 
-       local_irq_restore(flags);
-       perf_enable();
+       hwc->state = PERF_HES_UPTODATE;
+       if (!(flags & PERF_EF_START))
+               hwc->state |= PERF_HES_STOPPED;
+
+       local_irq_restore(irq_flags);
+       perf_pmu_enable(event->pmu);
 
        return ret;
 }
@@ -467,15 +471,15 @@ static int alpha_pmu_enable(struct perf_event *event)
  *  - this function is called from outside this module via the pmu struct
  *    returned from perf event initialisation.
  */
-static void alpha_pmu_disable(struct perf_event *event)
+static void alpha_pmu_del(struct perf_event *event, int flags)
 {
        struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
        struct hw_perf_event *hwc = &event->hw;
-       unsigned long flags;
+       unsigned long irq_flags;
        int j;
 
-       perf_disable();
-       local_irq_save(flags);
+       perf_pmu_disable(event->pmu);
+       local_irq_save(irq_flags);
 
        for (j = 0; j < cpuc->n_events; j++) {
                if (event == cpuc->event[j]) {
@@ -501,8 +505,8 @@ static void alpha_pmu_disable(struct perf_event *event)
                }
        }
 
-       local_irq_restore(flags);
-       perf_enable();
+       local_irq_restore(irq_flags);
+       perf_pmu_enable(event->pmu);
 }
 
 
@@ -514,13 +518,44 @@ static void alpha_pmu_read(struct perf_event *event)
 }
 
 
-static void alpha_pmu_unthrottle(struct perf_event *event)
+static void alpha_pmu_stop(struct perf_event *event, int flags)
+{
+       struct hw_perf_event *hwc = &event->hw;
+       struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
+
+       if (!(hwc->state & PERF_HES_STOPPED)) {
+               cpuc->idx_mask &= ~(1UL<<hwc->idx);
+               hwc->state |= PERF_HES_STOPPED;
+       }
+
+       if ((flags & PERF_EF_UPDATE) && !(hwc->state & PERF_HES_UPTODATE)) {
+               alpha_perf_event_update(event, hwc, hwc->idx, 0);
+               hwc->state |= PERF_HES_UPTODATE;
+       }
+
+       if (cpuc->enabled)
+               wrperfmon(PERFMON_CMD_DISABLE, (1UL<<hwc->idx));
+}
+
+
+static void alpha_pmu_start(struct perf_event *event, int flags)
 {
        struct hw_perf_event *hwc = &event->hw;
        struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
 
+       if (WARN_ON_ONCE(!(hwc->state & PERF_HES_STOPPED)))
+               return;
+
+       if (flags & PERF_EF_RELOAD) {
+               WARN_ON_ONCE(!(hwc->state & PERF_HES_UPTODATE));
+               alpha_perf_event_set_period(event, hwc, hwc->idx);
+       }
+
+       hwc->state = 0;
+
        cpuc->idx_mask |= 1UL<<hwc->idx;
-       wrperfmon(PERFMON_CMD_ENABLE, (1UL<<hwc->idx));
+       if (cpuc->enabled)
+               wrperfmon(PERFMON_CMD_ENABLE, (1UL<<hwc->idx));
 }
 
 
@@ -642,39 +677,36 @@ static int __hw_perf_event_init(struct perf_event *event)
        return 0;
 }
 
-static const struct pmu pmu = {
-       .enable         = alpha_pmu_enable,
-       .disable        = alpha_pmu_disable,
-       .read           = alpha_pmu_read,
-       .unthrottle     = alpha_pmu_unthrottle,
-};
-
-
 /*
  * Main entry point to initialise a HW performance event.
  */
-const struct pmu *hw_perf_event_init(struct perf_event *event)
+static int alpha_pmu_event_init(struct perf_event *event)
 {
        int err;
 
+       switch (event->attr.type) {
+       case PERF_TYPE_RAW:
+       case PERF_TYPE_HARDWARE:
+       case PERF_TYPE_HW_CACHE:
+               break;
+
+       default:
+               return -ENOENT;
+       }
+
        if (!alpha_pmu)
-               return ERR_PTR(-ENODEV);
+               return -ENODEV;
 
        /* Do the real initialisation work. */
        err = __hw_perf_event_init(event);
 
-       if (err)
-               return ERR_PTR(err);
-
-       return &pmu;
+       return err;
 }
 
-
-
 /*
  * Main entry point - enable HW performance counters.
  */
-void hw_perf_enable(void)
+static void alpha_pmu_enable(struct pmu *pmu)
 {
        struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
 
@@ -700,7 +732,7 @@ void hw_perf_enable(void)
  * Main entry point - disable HW performance counters.
  */
 
-void hw_perf_disable(void)
+static void alpha_pmu_disable(struct pmu *pmu)
 {
        struct cpu_hw_events *cpuc = &__get_cpu_var(cpu_hw_events);
 
@@ -713,6 +745,17 @@ void hw_perf_disable(void)
        wrperfmon(PERFMON_CMD_DISABLE, cpuc->idx_mask);
 }
 
+static struct pmu pmu = {
+       .pmu_enable     = alpha_pmu_enable,
+       .pmu_disable    = alpha_pmu_disable,
+       .event_init     = alpha_pmu_event_init,
+       .add            = alpha_pmu_add,
+       .del            = alpha_pmu_del,
+       .start          = alpha_pmu_start,
+       .stop           = alpha_pmu_stop,
+       .read           = alpha_pmu_read,
+};
+
 
 /*
  * Main entry point - don't know when this is called but it
@@ -766,7 +809,7 @@ static void alpha_perf_event_irq_handler(unsigned long la_ptr,
        wrperfmon(PERFMON_CMD_DISABLE, cpuc->idx_mask);
 
        /* la_ptr is the counter that overflowed. */
-       if (unlikely(la_ptr >= perf_max_events)) {
+       if (unlikely(la_ptr >= alpha_pmu->num_pmcs)) {
                /* This should never occur! */
                irq_err_count++;
                pr_warning("PMI: silly index %ld\n", la_ptr);
@@ -807,7 +850,7 @@ static void alpha_perf_event_irq_handler(unsigned long la_ptr,
                        /* Interrupts coming too quickly; "throttle" the
                         * counter, i.e., disable it for a little while.
                         */
-                       cpuc->idx_mask &= ~(1UL<<idx);
+                       alpha_pmu_stop(event, 0);
                }
        }
        wrperfmon(PERFMON_CMD_ENABLE, cpuc->idx_mask);
@@ -837,6 +880,7 @@ void __init init_hw_perf_events(void)
 
        /* And set up PMU specification */
        alpha_pmu = &ev67_pmu;
-       perf_max_events = alpha_pmu->num_pmcs;
+
+       perf_pmu_register(&pmu);
 }
 
index 842dba308eab3065510857e50312496c674c1097..3ec35066f1dc51c09367c32bea106453a109953d 100644 (file)
@@ -356,7 +356,7 @@ dump_elf_thread(elf_greg_t *dest, struct pt_regs *pt, struct thread_info *ti)
        dest[27] = pt->r27;
        dest[28] = pt->r28;
        dest[29] = pt->gp;
-       dest[30] = rdusp();
+       dest[30] = ti == current_thread_info() ? rdusp() : ti->pcb.usp;
        dest[31] = pt->pc;
 
        /* Once upon a time this was the PS value.  Which is stupid
index 0f6b51ae865aa87d70c3095cee6cec7057e0834a..6f7feb5db27193f33e24e962e4d1be257d8464ef 100644 (file)
@@ -41,46 +41,20 @@ static void do_signal(struct pt_regs *, struct switch_stack *,
 /*
  * The OSF/1 sigprocmask calling sequence is different from the
  * C sigprocmask() sequence..
- *
- * how:
- * 1 - SIG_BLOCK
- * 2 - SIG_UNBLOCK
- * 3 - SIG_SETMASK
- *
- * We change the range to -1 .. 1 in order to let gcc easily
- * use the conditional move instructions.
- *
- * Note that we don't need to acquire the kernel lock for SMP
- * operation, as all of this is local to this thread.
  */
-SYSCALL_DEFINE3(osf_sigprocmask, int, how, unsigned long, newmask,
-               struct pt_regs *, regs)
+SYSCALL_DEFINE2(osf_sigprocmask, int, how, unsigned long, newmask)
 {
-       unsigned long oldmask = -EINVAL;
-
-       if ((unsigned long)how-1 <= 2) {
-               long sign = how-2;              /* -1 .. 1 */
-               unsigned long block, unblock;
-
-               newmask &= _BLOCKABLE;
-               spin_lock_irq(&current->sighand->siglock);
-               oldmask = current->blocked.sig[0];
-
-               unblock = oldmask & ~newmask;
-               block = oldmask | newmask;
-               if (!sign)
-                       block = unblock;
-               if (sign <= 0)
-                       newmask = block;
-               if (_NSIG_WORDS > 1 && sign > 0)
-                       sigemptyset(&current->blocked);
-               current->blocked.sig[0] = newmask;
-               recalc_sigpending();
-               spin_unlock_irq(&current->sighand->siglock);
-
-               regs->r0 = 0;           /* special no error return */
+       sigset_t oldmask;
+       sigset_t mask;
+       unsigned long res;
+
+       siginitset(&mask, newmask & _BLOCKABLE);
+       res = sigprocmask(how, &mask, &oldmask);
+       if (!res) {
+               force_successful_syscall_return();
+               res = oldmask.sig[0];
        }
-       return oldmask;
+       return res;
 }
 
 SYSCALL_DEFINE3(osf_sigaction, int, sig,
@@ -94,9 +68,9 @@ SYSCALL_DEFINE3(osf_sigaction, int, sig,
                old_sigset_t mask;
                if (!access_ok(VERIFY_READ, act, sizeof(*act)) ||
                    __get_user(new_ka.sa.sa_handler, &act->sa_handler) ||
-                   __get_user(new_ka.sa.sa_flags, &act->sa_flags))
+                   __get_user(new_ka.sa.sa_flags, &act->sa_flags) ||
+                   __get_user(mask, &act->sa_mask))
                        return -EFAULT;
-               __get_user(mask, &act->sa_mask);
                siginitset(&new_ka.sa.sa_mask, mask);
                new_ka.ka_restorer = NULL;
        }
@@ -106,9 +80,9 @@ SYSCALL_DEFINE3(osf_sigaction, int, sig,
        if (!ret && oact) {
                if (!access_ok(VERIFY_WRITE, oact, sizeof(*oact)) ||
                    __put_user(old_ka.sa.sa_handler, &oact->sa_handler) ||
-                   __put_user(old_ka.sa.sa_flags, &oact->sa_flags))
+                   __put_user(old_ka.sa.sa_flags, &oact->sa_flags) ||
+                   __put_user(old_ka.sa.sa_mask.sig[0], &oact->sa_mask))
                        return -EFAULT;
-               __put_user(old_ka.sa.sa_mask.sig[0], &oact->sa_mask);
        }
 
        return ret;
index ce594ef533cc5b3d410b3403a8ca0467e3ebe47f..a6a1de9db16fd6e57d96c4e0b1f76db70f27147f 100644 (file)
@@ -58,7 +58,7 @@ sys_call_table:
        .quad sys_open                          /* 45 */
        .quad alpha_ni_syscall
        .quad sys_getxgid
-       .quad osf_sigprocmask
+       .quad sys_osf_sigprocmask
        .quad alpha_ni_syscall
        .quad alpha_ni_syscall                  /* 50 */
        .quad sys_acct
index 396af1799ea44e3a80bd758226f31f0c3bced799..0f1d8493cfca9c498a5187b97ae7c9dff8195b71 100644 (file)
@@ -41,7 +41,7 @@
 #include <linux/init.h>
 #include <linux/bcd.h>
 #include <linux/profile.h>
-#include <linux/perf_event.h>
+#include <linux/irq_work.h>
 
 #include <asm/uaccess.h>
 #include <asm/io.h>
@@ -83,25 +83,25 @@ static struct {
 
 unsigned long est_cycle_freq;
 
-#ifdef CONFIG_PERF_EVENTS
+#ifdef CONFIG_IRQ_WORK
 
-DEFINE_PER_CPU(u8, perf_event_pending);
+DEFINE_PER_CPU(u8, irq_work_pending);
 
-#define set_perf_event_pending_flag()  __get_cpu_var(perf_event_pending) = 1
-#define test_perf_event_pending()      __get_cpu_var(perf_event_pending)
-#define clear_perf_event_pending()     __get_cpu_var(perf_event_pending) = 0
+#define set_irq_work_pending_flag()  __get_cpu_var(irq_work_pending) = 1
+#define test_irq_work_pending()      __get_cpu_var(irq_work_pending)
+#define clear_irq_work_pending()     __get_cpu_var(irq_work_pending) = 0
 
-void set_perf_event_pending(void)
+void set_irq_work_pending(void)
 {
-       set_perf_event_pending_flag();
+       set_irq_work_pending_flag();
 }
 
-#else  /* CONFIG_PERF_EVENTS */
+#else  /* CONFIG_IRQ_WORK */
 
-#define test_perf_event_pending()      0
-#define clear_perf_event_pending()
+#define test_irq_work_pending()      0
+#define clear_irq_work_pending()
 
-#endif /* CONFIG_PERF_EVENTS */
+#endif /* CONFIG_IRQ_WORK */
 
 
 static inline __u32 rpcc(void)
@@ -191,9 +191,9 @@ irqreturn_t timer_interrupt(int irq, void *dev)
 
        write_sequnlock(&xtime_lock);
 
-       if (test_perf_event_pending()) {
-               clear_perf_event_pending();
-               perf_event_do_pending();
+       if (test_irq_work_pending()) {
+               clear_irq_work_pending();
+               irq_work_run();
        }
 
 #ifndef CONFIG_SMP
index 553b7cf17bfb0bac057eaedf0402a0bca2aa47e8..9103904b3daba63e73815bb2980923bdc3cb045d 100644 (file)
@@ -23,6 +23,7 @@ config ARM
        select HAVE_KERNEL_GZIP
        select HAVE_KERNEL_LZO
        select HAVE_KERNEL_LZMA
+       select HAVE_IRQ_WORK
        select HAVE_PERF_EVENTS
        select PERF_USE_VMALLOC
        select HAVE_REGS_AND_STACK_ACCESS_API
@@ -271,7 +272,6 @@ config ARCH_AT91
        bool "Atmel AT91"
        select ARCH_REQUIRE_GPIOLIB
        select HAVE_CLK
-       select ARCH_USES_GETTIMEOFFSET
        help
          This enables support for systems based on the Atmel AT91RM9200,
          AT91SAM9 and AT91CAP9 processors.
@@ -1051,6 +1051,32 @@ config ARM_ERRATA_460075
          ACTLR register. Note that setting specific bits in the ACTLR register
          may not be available in non-secure mode.
 
+config ARM_ERRATA_742230
+       bool "ARM errata: DMB operation may be faulty"
+       depends on CPU_V7 && SMP
+       help
+         This option enables the workaround for the 742230 Cortex-A9
+         (r1p0..r2p2) erratum. Under rare circumstances, a DMB instruction
+         between two write operations may not ensure the correct visibility
+         ordering of the two writes. This workaround sets a specific bit in
+         the diagnostic register of the Cortex-A9 which causes the DMB
+         instruction to behave as a DSB, ensuring the correct behaviour of
+         the two writes.
+
+config ARM_ERRATA_742231
+       bool "ARM errata: Incorrect hazard handling in the SCU may lead to data corruption"
+       depends on CPU_V7 && SMP
+       help