The current work request posting code writes the LSO segment before
writing any data segments.  This leaves a window where the LSO segment
overwrites the stamping in one cacheline that the HCA prefetches
before the rest of the cacheline is filled with the correct data
segments.  When the HCA processes this work request, a local
protection error may result.

Fix this by saving the LSO header size field off and writing it only
after all data segments are written.  This fix is a cleaned-up version
of a patch from Jack Morgenstein <jackm@dev.mellanox.co.il>.

This fixes <https://bugs.openfabrics.org/show_bug.cgi?id=1383>.
Reported-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Fix

    drivers/net/mlx4/profile.c: In function `mlx4_make_profile':
    drivers/net/mlx4/profile.c:110: warning: comparison of distinct pointer types lacks a cast

This happened because num_possible_cpus() was secretly changed by
commit ae7a47e7 ("cpumask: make cpumask.h eat its own dogfood.") from
returning "int" to (now) returning "unsigned int".  I think that was a
good change, so we should just swallow the fallout.

Cc: "David S. Miller" <davem@davemloft.net>
Cc: Roland Dreier <rolandd@cisco.com>
Cc: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Cc: Jack Morgenstein <jackm@dev.mellanox.co.il>
Cc: Vladimir Sokolovsky <vlad@mellanox.co.il>
Cc: Michael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

Merge branch 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  Revert "i386: add TRACE_IRQS_OFF for the nmi"

Merge branch 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  smp_call_function_single(): be slightly less stupid, fix #2
  lockdep, mm: fix might_fault() annotation

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband:
  IB/iser: Add dependency on INFINIBAND_ADDR_TRANS
  IPoIB: Do not join broadcast group if interface is brought down
  RDMA/nes: Fix for NIPQUAD removal
  IPoIB: Fix loss of connectivity after bonding failover on both sides
  IB/mlx4: Don't register IB device for adapters with no IB ports
  mlx4_core: Fix warning from min()
  IB/ehca: spin_lock_irqsave() takes an unsigned long

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  pata_it821x: Update RDC UDMA handling
  ata: fix wrong WARN_ON_ONCE

With some broken BIOSs when VT-d is enabled, the data structures are
filled incorrectly. This can cause a NULL pointer dereference in very
early boot.
Signed-off-by: Dirk Hohndel <hohndel@linux.intel.com>
Acked-by: Yu Zhao <yu.zhao@intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

The UDMA affliction is apparently specific to revision 0x11. Keeps us in sync
with drivers/ide current.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

This patch fixes a wrong WARN_ON that was triggered by 32bit PIO support:
WARNING: at drivers/ata/libata-sff.c:1017 ata_sff_hsm_move+0x45e/0x750()

__atapi_pio_bytes simply doesnt know enough to decide if there is a bug.
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Jeff Garzik <jgarzik@redhat.com>

Fix ib_iser build to depend on INFINIBAND_ADDR_TRANS; if INET=y but
IPV6=n, then the RDMA CM is not built but INFINIBAND_ISER can be
enabled, leading to:

    ERROR: "rdma_destroy_id" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
    ERROR: "rdma_connect" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
    ERROR: "rdma_destroy_qp" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
    ERROR: "rdma_create_id" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
    ERROR: "rdma_create_qp" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
    ERROR: "rdma_resolve_route" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
    ERROR: "rdma_disconnect" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
    ERROR: "rdma_resolve_addr" [drivers/infiniband/ulp/iser/ib_iser.ko] undefined!
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>

Because the ipoib_workqueue is not flushed when ipoib interface is
brought down, ipoib_mcast_join() may trigger a join to the broadcast
group after priv->broadcast was set to NULL (during cleanup).  This
will cause the system to be a member of the broadcast group when
interface is down.  As a side effect, this breaks the optimization of
setting the Q_key only when joining the broadcast group.
Signed-off-by: Yossi Etigin <yosefe@voltaire.com>
Signed-off-by: Roland Dreier <rolandd@cisco.com>

At 37000 feet somewhere near Greenland I woke up from a half-sleep with the
realisation that __lowest_in_progress() is buggy. After landing I checked
and there were indeed 2 problems with it; this patch fixes both:
* The order of the list checks was wrong
* The locking was not correct.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

There has been some light flamewar on lkml about decoding oopses
in modules (as part of the crashdump flamewar).

Now this isn't rocket science, just the markup_oops.pl script
cheaped out and didn't handle modules. But really; a flamewar
all about that?? What happened to C++ in the kernel or reading
files from inside the kernel?

This patch adds module support to markup_oops.pl; it's not the
most pretty perl but it works for my testcases...
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Add the Multi-Tech cellular modem firmware to the TI USB serial driver.
This firmware was extracted from:

ftp://ftp.multitech.com/wireless/wireless_linux.zip

Firmware licence: "all firmware components are redistributable in binary
	form" per support@multitech.com
	Copyright (C) 2005 Multi-Tech Systems, Inc.
Signed-off-by: Chris Adams <cmadams@hiwaay.net>
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Add Multi-Tech cellular modem support to the ti_usb_3410_5052 driver.
Signed-off-by: Chris Adams <cmadams@hiwaay.net>
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

The TI USB serial driver supports specifying alternate vendor and
product IDs (since the chips can and are used in devices under other
vendor/product IDs).  However, the alternate IDs were not loaded in the
combined product table.  This patch also adds support for loading
alternate firmware for alternate vendor/product IDs.
Signed-off-by: Chris Adams <cmadams@hiwaay.net>
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Julia Lawell found a case where a NULL check was misplaced in the
usb-serial code. However as the object in question cannot be NULL the
check can simply be removed.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Julia Lawall found an un-needed check in the neo driver. Her patch moves
the check to cover the code dereferencing it, however it cannot be NULL
anyway so remove the NULL check instead.
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

The pty changes and updates for window sizing forgot to correct the
kerneldoc
Signed-off-by: Alan Cox <alan@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

Merge branch 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  kernel/sched.c: add missing forward declaration for 'double_rq_lock'
  sched: partly revert "sched debug: remove NULL checking in print_cfs_rt_rq()"
  cpumask: fix CONFIG_NUMA=y sched.c

Merge branch 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  sparc64: Fix cpumask related build failure
  smp_call_function_single(): be slightly less stupid, fix
  smp_call_function_single(): be slightly less stupid
  rcu: fix bug in rcutorture system-shutdown code

* 'for-next' of git://git.o-hand.com/linux-mfd:
  mfd: Fix twl4030-core build
  mfd: Ensure sm501 GPIO pin mode is GPIO when configured
  mfd: dm355 evm MMC/SD card detection
  regulator: PCF50633 pmic driver
  input: PCF50633 input driver
  power_supply: PCF50633 battery charger driver
  rtc: PCF50633 rtc driver
  mfd: PCF50633 gpio support
  mfd: PCF50633 adc driver
  mfd: PCF50633 core driver

* 'upstream' of git://ftp.linux-mips.org/pub/scm/upstream-linus: (37 commits)
  MIPS: Only write c0_framemask on CPUs which have this register.
  MIPS: Alchemy: new userspace suspend interface for development boards.
  MIPS: Alchemy: dbdma suspend/resume support.
  MIPS: Alchemy: Fix up PM code on Au1550/Au1200
  MIPS: Alchemy: move calc_clock function.
  MIPS: Alchemy: RTC counter clocksource / clockevent support.
  MIPS: make cp0 counter clocksource/event usable as fallback.
  MIPS: Alchemy: remove cpu_table.
  MIPS: Alchemy: remove get/set_au1x00_lcd_clock().
  MIPS: Print irq handler description
  MIPS: Alchemy: pb1200: update CPLD cascade irq handler.
  MIPS: Alchemy: update core interrupt code.
  MIPS: Alchemy: move commandline mangling out of common code
  MIPS: Alchemy: devboards: consolidate files
  MIPS: Alchemy: Move development board code to common subdirectory
  MIPS: Add Cavium OCTEON to arch/mips/Kconfig
  MIPS: Add defconfig for Cavium OCTEON.
  MIPS: Adjust the dma-common.c platform hooks.
  MIPS: Add Cavium OCTEON slot into proper tlb category.
  MIPS:  Compute branch returns for Cavium OCTEON specific branch instructions.
  ...

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6: (37 commits)
  ucc_geth: use correct UCCE macros
  net_dma: acquire/release dma channels on ifup/ifdown
  cxgb3: Keep LRO off if disabled when interface is down
  sfc: SFT9001: Fix condition for LNPGA power-off
  dccp ccid-3: Fix RFC reference
  smsc911x: register irq with device name, not driver name
  smsc911x: fix smsc911x_reg_read compiler warning
  forcedeth: napi schedule lock fix
  net: fix section mismatch warnings in dccp/ccids/lib/tfrc.c
  forcedeth: remove mgmt unit for mcp79 chipset
  qlge: Remove dynamic alloc of rx ring control blocks.
  qlge: Fix schedule while atomic issue.
  qlge: Remove support for device ID 8000.
  qlge: Get rid of split addresses in hardware control blocks.
  qlge: Get rid of volatile usage for shadow register.
  forcedeth: version bump and copyright
  forcedeth: xmit lock fix
  netdev: missing validate_address hooks
  netdev: add missing set_mac_address hook
  netdev: gianfar: add MII ioctl handler
  ...

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc-2.6:
  sparc: Fix asm/signal.h for 32-bit.
  sparc: Eliminate PROMLIB_INTERNAL as it does nothing
  sparc: Kill exports of prom internal functions
  sparc64: move EXPORT_SYMBOL to the symbols definition
  sparc: move EXPORT_SYMBOL to the symbols definition
  sparc: Create a new file lib/ksyms.c and add export of all symbols defined in assembler in lib/ to this file.
  sparc: Most unaligned_64.c tweaks for branch tracer warnings.
  sparc: Fix sun4d_irq.c build.
  sparc: Update 32-bit defconfig.
  sparc64: fix warnings in psycho_common after ull conversion

* 'for_2.6.29' of git://git.kernel.org/pub/scm/linux/kernel/git/kkeil/ISDN-2.6:
  Fix small typo
  misdn: indentation and braces disagree - add braces
  misdn: one handmade ARRAY_SIZE converted
  drivers/isdn/hardware/mISDN: move a dereference below a NULL test
  indentation & braces disagree - add braces
  Make parameter debug writable
  BUGFIX: used NULL pointer at ioctl(sk,IMGETDEVINFO,&devinfo) when devinfo.id not registered

* git://git.kernel.org/pub/scm/linux/kernel/git/wim/linux-2.6-watchdog:
  [WATCHDOG] Pika Warp appliance watchdog timer
  [WATCHDOG] Enable watchdog timer on GE Fanuc's SBC610
  [WATCHDOG] Basic support for GE Fanuc's FPGA based watchdog timer
  [WATCHDOG] wm8350: Fix section annotations

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/geert/linux-m68k: (30 commits)
  m68k: Fix --build-id breakage for sun3
  m68k: Wire up sys_restart_syscall
  fbdev: Kill Atari vblank cursor blinking
  m68k: zorro - Use %pR to print resources
  m68k: dio - Kill resource_size_t format warnings
  m68k: dmasound - Kill warn_unused_result warnings
  m68k: zorro - Kill warn_unused_result warnings
  m68k: dio - Kill warn_unused_result warnings
  m68k: atafb - Kill warn_unused_result warnings
  m68k: amiserial - Kill warn_unused_result warnings
  m68k: ser_a2232 - Kill warn_unused_result warnings
  m68k: vme_scc - Kill warn_unused_result warnings
  m68k: sun3 core - Kill warn_unused_result warnings
  m68k: mvme147 core - Kill warn_unused_result warnings
  m68k: mac core - Kill warn_unused_result warnings
  m68k: hp300 core - Kill warn_unused_result warnings
  m68k: atari core - Kill warn_unused_result warnings
  m68k: apollo core - Kill warn_unused_result warnings
  m68k: amiga core - Kill warn_unused_result warnings
  m68k: Kill several external declarations in source files
  ...

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/teigland/dlm:
  dlm: change rsbtbl rwlock to spinlock
  dlm: fix seq_file usage in debugfs lock dump

Merge branch 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  Revert "cpumask: use work_on_cpu in acpi-cpufreq.c for drv_read and drv_write"
  x86: fix apic.c build error on latest git
  x86: fix mpparse.c build error on latest git
  x86: avoid theoretical vmalloc fault loop
  x86, mtrr: fix types used in userspace exported header

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  ALSA: hda - Use own workqueue
  ALSA: hda - add support for Intel DX58SO board
  ASoC: TWL4030: Module unloading fix
  ALSA: hda - create hda_codec.control_mutex for kcontrol->private_value
  ALSA: caiaq - Version 1.3.10
  ALSA: hda - Add quirk for Dell Inspiron Mini9
  ALSA: caiaq - Fix Oops with MIDI
  ASoC: TWL4030: Change the soc_value_enum back to soc_enum
  ASoC: Merge the soc_value_enum to soc_enum struct
  ALSA: hda - Add quirks for Acer Aspire 5930G and 6930G
  ALSA: hda - Add codec ID for MCP73 HDMI
  ALSA: hda - Fix typos for AD1882 codecs
  ALSA: hda - Add quirk for HP 2230s

The FPGA based watchdog timer used by the Pika Warp appliance.
Signed-off-by: Sean MacLennan <smaclennan@pikatech.com>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>

Support for the FPGA based watchdog timer on GE Fanuc's SBC610.

This patch enables one of the watchdog timers found on the SBC610. There are
two identical watchdog timers at different offsets in the above mentioned
boards, however the current driver is only capable of supporting one of them.

The watchdog timers are also capable of generating interrupts at a
user-configurable threshold, though support for this operation is currently
not supported by the driver.
Signed-off-by: Martyn Welch <martyn.welch@gefanuc.com>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>

GE Fanuc SBC610

Support for the FPGA based watchdog timer as found on GE Fanuc's SBC310,
SBC610 and PPC9A Single Board Computers.

This patch adds support for the watchdog timer found in one of the devices
FPGAs. There are two identical watchdog timers at different offsets in the
above mentioned boards, this driver is capable of supporting one of them.
The watchdog timers are also capable of generating interrupts at a
user-configurable threshold, though support for this operation is currently
not supported by the driver.
Signed-off-by: Martyn Welch <martyn.welch@gefanuc.com>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>

The probe and remove functions were incorrectly annotated, with the
misannotation of the remove function causing build failures when built
in.
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Wim Van Sebroeck <wim@iguana.be>

Counterpart of commit 08a3db94 ("m68k: Add
NOTES to init data so its discarded at boot") for sun3 build.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>

Make restart blocks working, required for proper syscall restarting.
Signed-off-by: Andreas Schwab <schwab@suse.de>
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>

Kill the last remaining vblank cursor blinking user
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>

Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>