8 years agotty: fix endless work loop when the buffer fills up
Linus Torvalds [Mon, 4 Apr 2011 21:26:54 +0000]
tty: fix endless work loop when the buffer fills up

Commit f23eb2b2b285 ('tty: stop using "delayed_work" in the tty layer')
ended up causing hung machines on UP with no preemption, because the
work routine to flip the buffer data to the ldisc would endlessly re-arm
itself if the destination buffer had filled up.

With the delayed work, that only caused a timer-driving polling of the
tty state every timer tick, but without the delay we just ended up with
basically a busy loop instead.

Stop the insane polling, and instead make the code that opens up the
receive room re-schedule the buffer flip work.  That's what we should
have been doing anyway.

This same "poll for tty room" issue is almost certainly also the cause
of excessive kworker activity when idle reported by Dave Jones, who also
reported "flush_to_ldisc executing 2500 times a second" back in Nov 2010:

  http://lkml.org/lkml/2010/11/30/592

which is that silly flushing done every timer tick.  Wasting both power
and CPU for no good reason.

Reported-and-tested-by: Alexander Beregalov <a.beregalov@gmail.com>
Reported-and-tested-by: Sitsofe Wheeler <sitsofe@yahoo.com>
Cc: Greg KH <gregkh@suse.de>
Cc: Alan Cox <alan@lxorguk.ukuu.org.uk>
Cc: Dave Jones <davej@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agoMerge branch 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Mon, 4 Apr 2011 15:37:45 +0000]
Merge branch 'x86-fixes-for-linus' of git://git./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:
  x86, UV: Fix kdump reboot
  x86, amd-nb: Rename CPU PCI id define for F4
  sound: Add delay.h to sound/soc/codecs/sn95031.c
  x86, mtrr, pat: Fix one cpu getting out of sync during resume
  x86, microcode: Unregister syscore_ops after microcode unloaded
  x86: Stop including <linux/delay.h> in two asm header files

8 years agoMerge branch 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 4 Apr 2011 15:36:58 +0000]
Merge branch 'sched-fixes-for-linus' of git://git./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:
  sched: Fix rebalance interval calculation
  sched, doc: Beef up load balancing description
  sched: Leave sched_setscheduler() earlier if possible, do not disturb SCHED_FIFO tasks

8 years agoMerge branch 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 4 Apr 2011 15:36:40 +0000]
Merge branch 'perf-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  perf: Fix task_struct reference leak
  perf: Fix task context scheduling
  perf: mmap 512 kiB by default
  perf: Rebase max unprivileged mlock threshold on top of page size
  perf tools: Fix NO_NEWT=1 python build error
  perf symbols: Properly align symbol_conf.priv_size
  perf tools: Emit clearer message for sys_perf_event_open ENOENT return
  perf tools: Fixup exit path when not able to open events
  perf symbols: Fix vsyscall symbol lookup
  oprofile, x86: Allow setting EDGE/INV/CMASK for counter events

8 years agoMerge branch 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Mon, 4 Apr 2011 15:36:15 +0000]
Merge branch 'core-fixes-for-linus' of git://git./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:
  rcu: create new rcu_access_index() and use in mce
  WARN_ON_SMP(): Add comment to explain ({0;})

8 years agousb/serial: fix function args warnings, dropping *filp
Randy Dunlap [Sun, 3 Apr 2011 18:42:00 +0000]
usb/serial: fix function args warnings, dropping *filp

Fix build warnings caused by removal of *filp arg in struct
usb_serial_driver.

These changes were missed somehow in commits 00a0d0d65b61 ("tty: remove
filp from the USB tty ioctls") and 60b33c133ca0b ("tiocmget: kill off
the passing of the struct file")

  drivers/usb/serial/mct_u232.c:159: warning: initialization from incompatible pointer type
  drivers/usb/serial/opticon.c:627: warning: initialization from incompatible pointer type

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Alan Cox <alan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agoMerge branch 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab...
Linus Torvalds [Mon, 4 Apr 2011 15:33:53 +0000]
Merge branch 'v4l_for_linus' of git://git./linux/kernel/git/mchehab/linux-2.6

* 'v4l_for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-2.6:
  [media] radio: wl128x: Update registration process with ST
  [media] staging: altera-jtag needs delay.h

8 years agoMerge branch 'unicore32' of git://git.kernel.org/pub/scm/linux/kernel/git/epip/linux...
Linus Torvalds [Mon, 4 Apr 2011 15:33:21 +0000]
Merge branch 'unicore32' of git://git./linux/kernel/git/epip/linux-2.6-unicore32

* 'unicore32' of git://git.kernel.org/pub/scm/linux/kernel/git/epip/linux-2.6-unicore32:
  unicore32 framebuffer fix: get videomemory by __get_free_pages() and make it floatable
  unicore32 core architecture: remove duplicated #include
  unicore32 rtc driver fix: cleanup irq_set_freq and irq_set_state
  unicore32 fix: remove arch-specific futex support
  unicore32 ldscript fix: add cacheline parameter to PERCPU() macro

8 years agoMerge branch 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groec...
Linus Torvalds [Mon, 4 Apr 2011 15:32:56 +0000]
Merge branch 'hwmon-for-linus' of git://git./linux/kernel/git/groeck/staging

* 'hwmon-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/groeck/staging:
  hwmon: twl4030-madc-hwmon: add "platform:" prefix for platform modalias

8 years agontp: fix non privileged system time shifting
Richard Cochran [Mon, 4 Apr 2011 15:31:23 +0000]
ntp: fix non privileged system time shifting

The ADJ_SETOFFSET bit added in commit 094aa188 ("ntp: Add ADJ_SETOFFSET
mode bit") also introduced a way for any user to change the system time.
Sneaky or buggy calls to adjtimex() could set

    ADJ_OFFSET_SS_READ | ADJ_SETOFFSET

which would result in a successful call to timekeeping_inject_offset().
This patch fixes the issue by adding the capability check.

Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agortc-mrst: Fix section types
Major Lee [Mon, 4 Apr 2011 10:32:12 +0000]
rtc-mrst: Fix section types

Fix the following section mismatch warning.

  WARNING: drivers/rtc/built-in.o(.data+0xa0): Section mismatch in reference from the variable vrtc_mrst_platform_driver to the function .init.text:vrtc_mrst_platform_probe()
  The variable vrtc_mrst_platform_driver references the function __init vrtc_mrst_platform_probe()
  If the reference is valid then annotate the variable with __init* or __refdata (see linux/init.h) or name the variable:
    *_template, *_timer, *_sht, *_ops, *_probe, *_probe_one, *_console,

Signed-off-by: Major Lee <major_lee@wistron.com>
Signed-off-by: Alan Cox <alan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

8 years agounicore32 framebuffer fix: get videomemory by __get_free_pages() and make it floatable
Guan Xuetao [Fri, 1 Apr 2011 08:38:59 +0000]
unicore32 framebuffer fix: get videomemory by __get_free_pages() and make it floatable

1. get videomemory by __get_free_pages() in fb-puv3.c
2. remove resource reservation for old fixed UNIGFX_MMAP & UVC_MMAP space
3. remove unused macros: PKUNTIY_UNIGFX_MMAP_BASE, PKUNITY_UNIGFX_MMAP_SIZE,
PKUNITY_UVC_MMAP_BASE, PKUNITY_UVC_MMAP_SIZE and KUSER_UNIGFX_BASE
4. remove unused header linux/vmalloc.h in fb-puv3.h

Signed-off-by: Guan Xuetao <gxt@mprc.pku.edu.cn>
Acked-by: Arnd Bergmann <arnd@arndb.de>

8 years agounicore32 core architecture: remove duplicated #include
Huang Weiyi [Sun, 20 Mar 2011 08:56:55 +0000]
unicore32 core architecture: remove duplicated #include

Remove duplicated #include('s) in
  arch/unicore32/kernel/traps.c

Signed-off-by: Huang Weiyi <weiyi.huang@gmail.com>
Acked-by: Guan Xuetao <gxt@mprc.pku.edu.cn>

8 years agounicore32 rtc driver fix: cleanup irq_set_freq and irq_set_state
Guan Xuetao [Tue, 29 Mar 2011 12:38:51 +0000]
unicore32 rtc driver fix: cleanup irq_set_freq and irq_set_state

Signed-off-by: Guan Xuetao <gxt@mprc.pku.edu.cn>

8 years agounicore32 fix: remove arch-specific futex support
Guan Xuetao [Tue, 29 Mar 2011 14:17:42 +0000]
unicore32 fix: remove arch-specific futex support

The futex functions in unicore32 are not used and verified,
so just replaced by asm-generic version.

Signed-off-by: Guan Xuetao <gxt@mprc.pku.edu.cn>
Acked-by: Arnd Bergmann <arnd@arndb.de>

8 years agounicore32 ldscript fix: add cacheline parameter to PERCPU() macro
Guan Xuetao [Tue, 29 Mar 2011 13:30:04 +0000]
unicore32 ldscript fix: add cacheline parameter to PERCPU() macro

Also, adjust cacheline parameter of RW_DATA_SECTION and EXCEPTION_TABLE

Signed-off-by: Guan Xuetao <gxt@mprc.pku.edu.cn>
Acked-by: Arnd Bergmann <arnd@arndb.de>

8 years agoMerge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6
Linus Torvalds [Fri, 1 Apr 2011 20:29:22 +0000]
Merge branch 'release' of git://git./linux/kernel/git/aegl/linux-2.6

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
  [IA64] fix build warning: argument 1 of 'irqd_irq_disabled'

8 years agoMerge branch 'rcu/urgent' of git://git.kernel.org/pub/scm/linux/kernel/git/paulmck...
Ingo Molnar [Fri, 1 Apr 2011 19:14:04 +0000]
Merge branch 'rcu/urgent' of git://git./linux/kernel/git/paulmck/linux-2.6-rcu into core/urgent

8 years ago[IA64] fix build warning: argument 1 of 'irqd_irq_disabled'
Tony Luck [Fri, 1 Apr 2011 17:56:03 +0000]
[IA64] fix build warning: argument 1 of 'irqd_irq_disabled'

In commit f5e5bf088bd3d30990efb7429aaf9f1e5134ffd6
  ia64: Use irqd_irq_disabled() instead of desc->status access

Thomas forgot to convert from irq to *irq_data.

Add a call to irq_get_irq_data() to fix that.

Signed-off-by: Tony Luck <tony.luck@intel.com>

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
Linus Torvalds [Fri, 1 Apr 2011 16:14:52 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/tiwai/sound-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6:
  ALSA: pcm: fix infinite loop in snd_pcm_update_hw_ptr0()
  ALSA: HDA: Add dock mic quirk for Lenovo Thinkpad X220
  ALSA: ens1371: fix Creative Ectiva support
  ALSA: firewire-speakers: fix hang when unplugging a running device
  ASoC: Fix CODEC device name for Corgi
  ALSA: hda - Fix pin-config of Gigabyte mobo
  ASoC: imx: fix burstsize for DMA
  ASoC: imx: set watermarks for mx2-dma
  ASoC: twl6040: Return -ENOMEM if create_singlethread_workqueue fails
  ASoC: tlv320dac33: Restore L/R DAC power control register
  ASoC: Explicitly say registerless widgets have no register
  ASoC: tlv320dac33: Fix inconsistent spinlock usage

8 years agoMerge branch 'fix/asoc' into for-linus
Takashi Iwai [Fri, 1 Apr 2011 16:02:23 +0000]
Merge branch 'fix/asoc' into for-linus

8 years agoALSA: pcm: fix infinite loop in snd_pcm_update_hw_ptr0()
Kelly Anderson [Fri, 1 Apr 2011 09:58:25 +0000]
ALSA: pcm: fix infinite loop in snd_pcm_update_hw_ptr0()

When period interrupts are disabled, snd_pcm_update_hw_ptr0() compares
the current time against the time estimated for the current hardware
pointer to detect xruns.  The somewhat fuzzy threshold in the while loop
makes it possible that hdelta becomes negative; the comparison being
done with unsigned types then makes the loop go through the entire 263
negative range, and, depending on the value, never reach an unsigned
value that is small enough to stop the loop.  Doing this with interrupts
disabled results in the machine locking up.

To prevent this, ensure that the loop condition uses signed types for
both operands so that the comparison is correctly done.

Many thanks to Kelly Anderson for debugging this.

Reported-by: Nix <nix@esperi.org.uk>
Reported-by: "Christopher K." <c.krooss@googlemail.com>
Reported-and-tested-by: Kelly Anderson <kelly@silka.with-linux.com>
Signed-off-by: Kelly Anderson <kelly@silka.with-linux.com>
[cl: remove unneeded casts; use a temp variable]
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Cc: 2.6.38 <stable@kernel.org>

Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Linus Torvalds [Fri, 1 Apr 2011 15:57:02 +0000]
Merge branch 'merge' of git://git./linux/kernel/git/benh/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
  kdump: Allow shrinking of kdump region to be overridden
  powerpc/pmac/smp: Remove no-longer needed preempt workaround
  powerpc/smp: Increase vdso_data->processorCount, not just decrease it
  powerpc/smp: Create idle threads on demand and properly reset them
  powerpc/smp: Don't expose per-cpu "cpu_state" array
  powerpc/pmac/smp: Fix CPU hotplug crashes on some machines
  powerpc/smp: Add a smp_ops->bringup_up() done callback
  powerpc/pmac: Rename cpu_state in therm_pm72 to avoid collision
  powerpc/pmac/smp: Properly NAP offlined CPU on G5
  powerpc/pmac/smp: Remove HMT changes for PowerMac offline code
  powerpc/pmac/smp: Consolidate 32-bit and 64-bit PowerMac cpu_die in one file
  powerpc/pmac/smp: Fixup smp_core99_cpu_disable() and use it on 64-bit
  powerpc/pmac/smp: Rename fixup_irqs() to migrate_irqs() and use it on ppc32
  powerpc/pmac/smp: Fix 32-bit PowerMac cpu_die
  powerpc/smp: Remove unused smp_ops->cpu_enable()
  powerpc/smp: Remove unused generic_cpu_enable()
  powerpc/smp: Fix generic_mach_cpu_die()
  powerpc/smp: soft-replugged CPUs must go back to start_secondary
  powerpc: Make decrementer interrupt robust against offlined CPUs

8 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc-2.6
Linus Torvalds [Fri, 1 Apr 2011 15:54:14 +0000]
Merge git://git./linux/kernel/git/davem/sparc-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc-2.6:
  sparc32: Pass task_struct to schedule_tail() in ret_from_fork
  apbuart: Depend upon sparc.
  sparc64: Fix section mis-match errors.
  sparc32,leon: Fixed APBUART frequency detection
  sparc32, leon: APBUART driver must use archdata to get IRQ number
  sparc: Hook up syncfs system call.

8 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Fri, 1 Apr 2011 15:53:50 +0000]
Merge git://git./linux/kernel/git/davem/net-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6:
  appletalk: Fix OOPS in atalk_release().
  mlx4: Fixing bad size of event queue buffer
  mlx4: Fixing use after free
  bonding:typo in comment
  sctp: Pass __GFP_NOWARN to hash table allocation attempts.
  connector: convert to synchronous netlink message processing
  fib: add rtnl locking in ip_fib_net_exit
  atm/solos-pci: Don't flap VCs when carrier state changes
  atm/solos-pci: Don't include frame pseudo-header on transmit hex-dump
  atm/solos-pci: Use VPI.VCI notation uniformly.
  Atheros, atl2: Fix mem leaks in error paths of atl2_set_eeprom
  netdev: fix mtu check when TSO is enabled
  net/usb: Ethernet quirks for the LG-VL600 4G modem
  phylib: phy_attach_direct: phy_init_hw can fail, add cleanup
  bridge: mcast snooping, fix length check of snooped MLDv1/2
  via-ircc: Pass PCI device pointer to dma_{alloc, free}_coherent()
  via-ircc: Use pci_{get, set}_drvdata() instead of static pointer variable
  net: gre: provide multicast mappings for ipv4 and ipv6
  bridge: Fix compilation warning in function br_stp_recalculate_bridge_id()
  net: Fix warnings caused by MAX_SKB_FRAGS change.

8 years agorcu: create new rcu_access_index() and use in mce
Paul E. McKenney [Fri, 1 Apr 2011 14:15:14 +0000]
rcu: create new rcu_access_index() and use in mce

The MCE subsystem needs to sample an RCU-protected index outside of
any protection for that index.  If this was a pointer, we would use
rcu_access_pointer(), but there is no corresponding rcu_access_index().
This commit therefore creates an rcu_access_index() and applies it
to MCE.

Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Tested-by: Zdenek Kabelac <zkabelac@redhat.com>

8 years agokdump: Allow shrinking of kdump region to be overridden
Anton Blanchard [Wed, 25 Aug 2010 00:22:58 +0000]
kdump: Allow shrinking of kdump region to be overridden

On ppc64 the crashkernel region almost always overlaps an area of firmware.
This works fine except when using the sysfs interface to reduce the kdump
region. If we free the firmware area we are guaranteed to crash.

Rename free_reserved_phys_range to crash_free_reserved_phys_range and make
it a weak function so we can override it.

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/pmac/smp: Remove no-longer needed preempt workaround
Benjamin Herrenschmidt [Tue, 8 Mar 2011 03:57:26 +0000]
powerpc/pmac/smp: Remove no-longer needed preempt workaround

The generic code properly re-initializes the preempt count in the
idle thread now

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/smp: Increase vdso_data->processorCount, not just decrease it
Benjamin Herrenschmidt [Tue, 8 Mar 2011 03:49:33 +0000]
powerpc/smp: Increase vdso_data->processorCount, not just decrease it

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/smp: Create idle threads on demand and properly reset them
Benjamin Herrenschmidt [Tue, 8 Mar 2011 03:40:04 +0000]
powerpc/smp: Create idle threads on demand and properly reset them

Instead of creating idle threads at boot for all possible CPUs, we
create them on demand, like x86 or ARM, and we properly call init_idle
to re-initialize an idle thread when a CPU was unplugged and is now
re-plugged.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/smp: Don't expose per-cpu "cpu_state" array
Benjamin Herrenschmidt [Thu, 31 Mar 2011 22:23:37 +0000]
powerpc/smp: Don't expose per-cpu "cpu_state" array

Instead, keep it static, expose an accessor and use that from
the PowerMac code. Avoids easy namespace collisions and will
make it easier to consolidate with other implementations.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/pmac/smp: Fix CPU hotplug crashes on some machines
Benjamin Herrenschmidt [Tue, 8 Mar 2011 02:54:50 +0000]
powerpc/pmac/smp: Fix CPU hotplug crashes on some machines

On some machines that use i2c to synchronize the timebases (such
as PowerMac7,2/7,3 G5 machines), hotplug CPU would crash when
putting back a new CPU online due to the underlying i2c bus being
closed.

This uses the newly added bringup_done() callback to move the close
along with other housekeeping calls, and adds a CPU notifier to
re-open the i2c bus around subsequent hotplug operations

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/smp: Add a smp_ops->bringup_up() done callback
Benjamin Herrenschmidt [Tue, 8 Mar 2011 02:50:37 +0000]
powerpc/smp: Add a smp_ops->bringup_up() done callback

This allows us to stop abusing smp_ops->setup_cpu() for cleanup
tasks that have to take place after the initial boot time CPU
bringup.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/pmac: Rename cpu_state in therm_pm72 to avoid collision
Benjamin Herrenschmidt [Mon, 7 Mar 2011 06:02:49 +0000]
powerpc/pmac: Rename cpu_state in therm_pm72 to avoid collision

This collides with the cpu_state in our SMP code, use processor_state
instead.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/pmac/smp: Properly NAP offlined CPU on G5
Benjamin Herrenschmidt [Mon, 21 Feb 2011 05:49:58 +0000]
powerpc/pmac/smp: Properly NAP offlined CPU on G5

The current code soft-disables, and then goes to NAP mode which
turns interrupts on. That means that if an interrupt occurs, we
will hit the masked interrupt code path which isn't what we want,
as it will return with EE off, which will either get us out of
NAP mode, or fail to enter it (according to spec).

Instead, let's just rely on the fact that it is safe to take
decrementer interrupts on an offline CPU and leave interrupts
enabled. We can also get rid of the special case in asm for
power4_cpu_offline_powersave() and just use power4_idle().

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/pmac/smp: Remove HMT changes for PowerMac offline code
Benjamin Herrenschmidt [Fri, 11 Feb 2011 03:55:42 +0000]
powerpc/pmac/smp: Remove HMT changes for PowerMac offline code

Those instructions do nothing on non-threaded processors such
as 970's used on those machines.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/pmac/smp: Consolidate 32-bit and 64-bit PowerMac cpu_die in one file
Benjamin Herrenschmidt [Fri, 11 Feb 2011 03:03:20 +0000]
powerpc/pmac/smp: Consolidate 32-bit and 64-bit PowerMac cpu_die in one file

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/pmac/smp: Fixup smp_core99_cpu_disable() and use it on 64-bit
Benjamin Herrenschmidt [Mon, 21 Feb 2011 05:31:49 +0000]
powerpc/pmac/smp: Fixup smp_core99_cpu_disable() and use it on 64-bit

Use the generic code, just add the MPIC priority setting,

I don't see any use in mucking around with the decrementer,
as 32-bit will have EE off all along, and 64-bit will be able
to deal with it.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/pmac/smp: Rename fixup_irqs() to migrate_irqs() and use it on ppc32
Benjamin Herrenschmidt [Fri, 11 Feb 2011 02:05:17 +0000]
powerpc/pmac/smp: Rename fixup_irqs() to migrate_irqs() and use it on ppc32

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/pmac/smp: Fix 32-bit PowerMac cpu_die
Benjamin Herrenschmidt [Fri, 11 Feb 2011 03:09:32 +0000]
powerpc/pmac/smp: Fix 32-bit PowerMac cpu_die

Use generic cpu_state, call idle_task_exit() properly, and
remove smp_core99_cpu_die() which isn't useful, the generic
function does the job just fine.

8 years agopowerpc/smp: Remove unused smp_ops->cpu_enable()
Benjamin Herrenschmidt [Fri, 11 Feb 2011 01:49:01 +0000]
powerpc/smp: Remove unused smp_ops->cpu_enable()

Remove the last remnants of cpu_enable(), everybody uses the normal
__cpu_up() path now

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/smp: Remove unused generic_cpu_enable()
Benjamin Herrenschmidt [Fri, 11 Feb 2011 01:46:41 +0000]
powerpc/smp: Remove unused generic_cpu_enable()

Nobody uses it, besides we should always use the normal __cpu_up
path anyways

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/smp: Fix generic_mach_cpu_die()
Benjamin Herrenschmidt [Thu, 10 Feb 2011 07:46:50 +0000]
powerpc/smp: Fix generic_mach_cpu_die()

This is used by some "soft" hotplug implementations. I needs to
call idle_task_exit() when the CPU is going away, and we remove
the now no-longer needed set_cpu_online() and local_irq_enable()
which are handled by the return to start_secondary

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc/smp: soft-replugged CPUs must go back to start_secondary
Benjamin Herrenschmidt [Thu, 10 Feb 2011 07:45:24 +0000]
powerpc/smp: soft-replugged CPUs must go back to start_secondary

Various thing are torn down when a CPU is hot-unplugged. That CPU
is expected to go back to start_secondary when re-plugged to re
initialize everything, such as clock sources, maps, ...

Some implementations just return from cpu_die() callback
in the idle loop when the CPU is "re-plugged". This is not enough.

We fix it using a little asm trampoline which resets the stack
and calls back into start_secondary as if we were all fresh from
boot. The trampoline already existed on ppc64, but we add it for
ppc32

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agopowerpc: Make decrementer interrupt robust against offlined CPUs
Benjamin Herrenschmidt [Tue, 29 Mar 2011 03:51:10 +0000]
powerpc: Make decrementer interrupt robust against offlined CPUs

With some implementations, it is possible that a timer interrupt
occurs every few seconds on an offline CPU. In this case, just
re-arm the decrementer and return immediately

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

8 years agoappletalk: Fix OOPS in atalk_release().
David S. Miller [Fri, 1 Apr 2011 01:59:10 +0000]
appletalk: Fix OOPS in atalk_release().

Commit 60d9f461a20ba59219fdcdc30cbf8e3a4ad3f625 ("appletalk: remove
the BKL") added a dereference of "sk" before checking for NULL in
atalk_release().

Guard the code block completely, rather than partially, with the
NULL check.

Reported-by: Dave Jones <davej@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years ago[media] radio: wl128x: Update registration process with ST
Manjunatha Halli [Wed, 23 Mar 2011 10:44:30 +0000]
[media] radio: wl128x: Update registration process with ST

As underlying ST driver registration API's have changed with
latest 2.6.38-rc8 kernel this patch will update the FM driver
accordingly.

Signed-off-by: Manjunatha Halli <manjunatha_halli@ti.com>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

8 years ago[media] staging: altera-jtag needs delay.h
Randy Dunlap [Mon, 28 Mar 2011 15:23:05 +0000]
[media] staging: altera-jtag needs delay.h

altera-jtag.c needs to include <linux/delay.h> to fix a build error:

drivers/staging/altera-stapl/altera-jtag.c:398: error: implicit declaration of function 'udelay'

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Igor M. Liplianin <liplianin@netup.ru>
Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

8 years agoALSA: HDA: Add dock mic quirk for Lenovo Thinkpad X220
David Henningsson [Thu, 31 Mar 2011 07:36:19 +0000]
ALSA: HDA: Add dock mic quirk for Lenovo Thinkpad X220

This quirk is needed for the docking station mic of
Lenovo Thinkpad X220 to function correctly.

BugLink: http://bugs.launchpad.net/bugs/746259
Cc: stable@kernel.org
Tested-by: James Ferguson <james.ferguson@canonical.com>
Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agox86, UV: Fix kdump reboot
Cliff Wickman [Thu, 31 Mar 2011 14:32:02 +0000]
x86, UV: Fix kdump reboot

After a crash dump on an SGI Altix UV system the crash kernel
fails to cause a reboot.  EFI mode is disabled in the kdump
kernel, so only the reboot_type of BOOT_ACPI works.

Signed-off-by: Cliff Wickman <cpw@sgi.com>
Cc: rja@sgi.com
LKML-Reference: <E1Q5Iuo-00013b-UK@eag09.americas.sgi.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agoperf: Fix task_struct reference leak
Peter Zijlstra [Mon, 28 Mar 2011 11:13:56 +0000]
perf: Fix task_struct reference leak

sys_perf_event_open() had an imbalance in the number of task refs it
took causing memory leakage

Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: stable@kernel.org # .37+
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agoperf: Fix task context scheduling
Peter Zijlstra [Thu, 31 Mar 2011 08:29:26 +0000]
perf: Fix task context scheduling

Jiri reported:

 |
 | - once an event is created by sys_perf_event_open, task context
 |   is created and it stays even if the event is closed, until the
 |   task is finished ... thats what I see in code and I assume it's
 |   correct
 |
 | - when the task opens event, perf_sched_events jump label is
 |   incremented and following callbacks are started from scheduler
 |
 |         __perf_event_task_sched_in
 |         __perf_event_task_sched_out
 |
 |   These callback *in/out set/unset cpuctx->task_ctx value to the
 |   task context.
 |
 | - close is called on event on CPU 0:
 |         - the task is scheduled on CPU 0
 |         - __perf_event_task_sched_in is called
 |         - cpuctx->task_ctx is set
 |         - perf_sched_events jump label is decremented and == 0
 |         - __perf_event_task_sched_out is not called
 |         - cpuctx->task_ctx on CPU 0 stays set
 |
 | - exit is called on CPU 1:
 |         - the task is scheduled on CPU 1
 |         - perf_event_exit_task is called
 |         - task_ctx_sched_out unsets cpuctx->task_ctx on CPU 1
 |         - put_ctx destroys the context
 |
 | - another call of perf_rotate_context on CPU 0 will use invalid
 |   task_ctx pointer, and eventualy panic.
 |

Cure this the simplest possibly way by partially reverting the
jump_label optimization for the sched_out case.

Reported-and-tested-by: Jiri Olsa <jolsa@redhat.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: <stable@kernel.org> # .37+
LKML-Reference: <1301520405.4859.213.camel@twins>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agoperf: mmap 512 kiB by default
Frederic Weisbecker [Thu, 31 Mar 2011 01:35:24 +0000]
perf: mmap 512 kiB by default

The default setting of perf record is to mmap 128 pages if the user
did not override with -m.

However the page size may vary accross different architecture
settings, giving different default size between each.

Moreover the kernel side still has a default max number of mlocked
pages of 512 kiB + 1 page for unprivileged users. 128 + 1 pages
with page size > 4096 overlaps this threshold.

Thus, better adapt to this limitation and set the default number of
pages to fit those 512 kiB + 1 page.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Stephane Eranian <eranian@google.com>
LKML-Reference: <1301535324-9735-1-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agoperf: Rebase max unprivileged mlock threshold on top of page size
Frederic Weisbecker [Thu, 31 Mar 2011 01:33:29 +0000]
perf: Rebase max unprivileged mlock threshold on top of page size

Ensure we allow 512 kiB + 1 page for user control without
assuming a 4096 bytes page size.

Reported-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: <stable@kernel.org>
LKML-Reference: <1301535209-9679-1-git-send-email-fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agosched: Fix rebalance interval calculation
Sisir Koppaka [Sat, 26 Mar 2011 12:52:55 +0000]
sched: Fix rebalance interval calculation

The interval for checking scheduling domains if they are due to be
balanced currently depends on boot state NR_CPUS, which may not
accurately reflect the number of online CPUs at the time of check.

Thus replace NR_CPUS with num_online_cpus().

 (ed: Should only affect those who set NR_CPUS really high, such as 4096
      or so :-)

Signed-off-by: Sisir Koppaka <sisir.koppaka@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <AANLkTikqHWid2Q93F5U5Qw5snJH8C5PXoa7J6=6hYO94@mail.gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agosched, doc: Beef up load balancing description
Borislav Petkov [Sun, 27 Mar 2011 15:57:13 +0000]
sched, doc: Beef up load balancing description

Correct all function names pertaining to load balancing and explain
shortly how load balancing is performed.

Signed-off-by: Borislav Petkov <bp@alien8.de>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1301241433-3790-1-git-send-email-bp@alien8.de>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agosched: Leave sched_setscheduler() earlier if possible, do not disturb SCHED_FIFO...
Dario Faggioli [Thu, 24 Mar 2011 13:00:18 +0000]
sched: Leave sched_setscheduler() earlier if possible, do not disturb SCHED_FIFO tasks

sched_setscheduler() (in sched.c) is called in order of changing the
scheduling policy and/or the real-time priority of a task. Thus,
if we find out that neither of those are actually being modified, it
is possible to return earlier and save the overhead of a full
deactivate+activate cycle of the task in question.

Beside that, if we have more than one SCHED_FIFO task with the same
priority on the same rq (which means they share the same priority queue)
having one of them changing its position in the priority queue because of
a sched_setscheduler (as it happens by means of the deactivate+activate)
that does not actually change the priority violates POSIX which states,
for SCHED_FIFO:

  "If a thread whose policy or priority has been modified by
   pthread_setschedprio() is a running thread or is runnable, the effect on
   its position in the thread list depends on the direction of the
   modification, as follows: a. <...> b. If the priority is unchanged, the
   thread does not change position in the thread list. c. <...>"

     http://pubs.opengroup.org/onlinepubs/009695399/functions/xsh_chap02_08.html

 (ed: And the POSIX specification here does, briefly and somewhat unexpectedly,
      match what common sense tells us as well. )

Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1300971618.3960.82.camel@Palantir>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agomlx4: Fixing bad size of event queue buffer
Yevgeny Petrilin [Wed, 30 Mar 2011 23:30:17 +0000]
mlx4: Fixing bad size of event queue buffer

We should reduce the number of reserved completion queues from the total
number of entries. Since the queue size is power of two, not reducing the
reserved entries, caused a double queue size, which may lead to allocation
failures in some cases.

Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agomlx4: Fixing use after free
Yevgeny Petrilin [Wed, 30 Mar 2011 23:28:52 +0000]
mlx4: Fixing use after free

In case of allocation failure, tried to use the promiscuous QP
entry that was previously freed.
Now freeing this entry only in case we will not put it back to the list
of promiscuous entries.

Reported-by: Dan Carpenter <error27@gmail.com>
Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobonding:typo in comment
Peter Pan(潘卫平) [Wed, 30 Mar 2011 20:46:19 +0000]
bonding:typo in comment

use accumulates instead of acumulates.

Signed-off-by: Pan Weiping <panweiping3@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agosparc32: Pass task_struct to schedule_tail() in ret_from_fork
Tkhai Kirill [Thu, 31 Mar 2011 07:52:38 +0000]
sparc32: Pass task_struct to schedule_tail() in ret_from_fork

We have to pass task_struct of previous process to function
schedule_tail(). Currently in ret_from_fork previous thread_info
is passed:

switch_to: mov %g6, %g3 /* previous thread_info in g6 */

ret_from_fork: call    schedule_tail
                mov    %g3, %o0 /* previous thread_info is passed */

void schedule_tail(struct task_struct *prev);

Signed-off-by: Tkhai Kirill <tkhai@yandex.ru>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agox86, amd-nb: Rename CPU PCI id define for F4
Borislav Petkov [Wed, 30 Mar 2011 18:34:47 +0000]
x86, amd-nb: Rename CPU PCI id define for F4

With increasing number of PCI function ids, add the PCI function
id in the define name instead of its symbolic name in the BKDG
for more clarity. This renames function 4 define.

Signed-off-by: Borislav Petkov <borislav.petkov@amd.com>
Cc: Jesse Barnes <jbarnes@virtuousgeek.org>
LKML-Reference: <20110330183447.GA3668@aftab>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agosound: Add delay.h to sound/soc/codecs/sn95031.c
Ingo Molnar [Thu, 13 Nov 2008 10:51:57 +0000]
sound: Add delay.h to sound/soc/codecs/sn95031.c

This is further fallout from delay.h removal from asm/apic.h and asm/dma.h:

  ca444564a947: x86: Stop including <linux/delay.h> in two asm header files

Which caused this build failure:

  sound/soc/codecs/sn95031.c: In function ‘sn95031_get_mic_bias’:
  sound/soc/codecs/sn95031.c:153:2: error: implicit declaration of function ‘msleep’ [-Werror=implicit-function-declaration]

Cc: Jean Delvare <khali@linux-fr.org>
Cc: James E.J. Bottomley <James.Bottomley@suse.de>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
LKML-Reference: <20110325152014.297890ec@endymion.delvare>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agoapbuart: Depend upon sparc.
David S. Miller [Thu, 31 Mar 2011 04:11:35 +0000]
apbuart: Depend upon sparc.

It absolutely needs to be able to get at pdev_archdata members
which are sparc specific.

Signed-off-by: David S. Miller <davem@davemloft.net>
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>

8 years agosctp: Pass __GFP_NOWARN to hash table allocation attempts.
David S. Miller [Thu, 31 Mar 2011 00:51:36 +0000]
sctp: Pass __GFP_NOWARN to hash table allocation attempts.

Like DCCP and other similar pieces of code, there are mechanisms
here to try allocating smaller hash tables if the allocation
fails.  So pass in __GFP_NOWARN like the others do instead of
emitting a scary message.

Reported-by: Dave Jones <davej@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agosparc64: Fix section mis-match errors.
David S. Miller [Thu, 31 Mar 2011 00:37:56 +0000]
sparc64: Fix section mis-match errors.

Fix all of the problems spotted by CONFIG_DEBUG_SECTION_MISMATCH under
arch/sparc during a 64-bit defconfig build.

They fall into two categorites:

1) of_device_id is marked as __initdata, and we can never do this
   since these objects sit in the device core data structures way
   past boot.  So even if a driver will never be reloaded, we have
   to keep the device ID table around.

   Mark such cases const instead.

2) The bootmem alloc/free handling code in mdesc.c was not fully
   marked __init as it should be, thus generating a reference
   to free_bootmem_late() (which is __init) from non-__init code.

Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoconnector: convert to synchronous netlink message processing
Patrick McHardy [Mon, 28 Mar 2011 08:39:36 +0000]
connector: convert to synchronous netlink message processing

Commits 01a16b21 (netlink: kill eff_cap from struct netlink_skb_parms)
and c53fa1ed (netlink: kill loginuid/sessionid/sid members from struct
netlink_skb_parms) removed some members from struct netlink_skb_parms
that depend on the current context, all netlink users are now required
to do synchronous message processing.

connector however queues received messages and processes them in a work
queue, which is not valid anymore. This patch converts connector to do
synchronous message processing by invoking the registered callback handler
directly from the netlink receive function.

In order to avoid invoking the callback with connector locks held, a
reference count is added to struct cn_callback_entry, the reference
is taken when finding a matching callback entry on the device's queue_list
and released after the callback handler has been invoked.

Signed-off-by: Patrick McHardy <kaber@trash.net>
Acked-by: Evgeniy Polyakov <zbr@ioremap.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agofib: add rtnl locking in ip_fib_net_exit
Eric Dumazet [Wed, 30 Mar 2011 23:57:46 +0000]
fib: add rtnl locking in ip_fib_net_exit

Daniel J Blueman reported a lockdep splat in trie_firstleaf(), caused by
RTNL being not locked before a call to fib_table_flush()

Reported-by: Daniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoatm/solos-pci: Don't flap VCs when carrier state changes
Philip A. Prindeville [Wed, 30 Mar 2011 13:17:04 +0000]
atm/solos-pci: Don't flap VCs when carrier state changes

Don't flap VCs when carrier state changes; higher-level protocols
can detect loss of connectivity and act accordingly. This is more
consistent with how other network interfaces work.

We no longer use release_vccs() so we can delete it.

release_vccs() was duplicated from net/atm/common.c; make the
corresponding function exported, since other code duplicates it
and could leverage it if it were public.

Signed-off-by: Philip A. Prindeville <philipp@redfish-solutions.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoatm/solos-pci: Don't include frame pseudo-header on transmit hex-dump
Philip A. Prindeville [Wed, 30 Mar 2011 12:59:26 +0000]
atm/solos-pci: Don't include frame pseudo-header on transmit hex-dump

Omit pkt_hdr preamble when dumping transmitted packet as hex-dump;
we can pull this up because the frame has already been sent, and
dumping it is the last thing we do with it before freeing it.

Also include the size, vpi, and vci in the debug as is done on
receive.

Use "port" consistently instead of "device" intermittently.

Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoatm/solos-pci: Use VPI.VCI notation uniformly.
Philip A. Prindeville [Wed, 30 Mar 2011 13:22:45 +0000]
atm/solos-pci: Use VPI.VCI notation uniformly.

Use VPI.VCI notation consistently throughout the module. This is the
one remaining place where the VCI is used before the VPI in any output.

Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoMerge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6
Linus Torvalds [Wed, 30 Mar 2011 19:22:15 +0000]
Merge branch 'release' of git://git./linux/kernel/git/aegl/linux-2.6

* 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6:
  [IA64] Fix build breakage - sn_force_interrupt_flag: undefined

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke...
Linus Torvalds [Wed, 30 Mar 2011 16:46:36 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/ryusuke/nilfs2

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ryusuke/nilfs2:
  nilfs2: fix whitespace coding style issues
  nilfs2: fix oops due to a bad aops initialization
  nilfs2: fix data loss in mmap page write for hole blocks

8 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client
Linus Torvalds [Wed, 30 Mar 2011 16:46:09 +0000]
Merge git://git./linux/kernel/git/sage/ceph-client

* git://git.kernel.org/pub/scm/linux/kernel/git/sage/ceph-client:
  libceph: Create a new key type "ceph".
  libceph: Get secret from the kernel keys api when mounting with key=NAME.
  ceph: Move secret key parsing earlier.
  libceph: fix null dereference when unregistering linger requests
  ceph: unlock on error in ceph_osdc_start_request()
  ceph: fix possible NULL pointer dereference
  ceph: flush msgr_wq during mds_client shutdown

8 years agoMerge branch 'irq-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Wed, 30 Mar 2011 16:35:52 +0000]
Merge branch 'irq-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'irq-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  avr32: Fix missing irq namespace conversion
  powerpc: qe_ic: Rename get_irq_desc_data and get_irq_desc_chip
  genirq: Remove the now obsolete config options and select statements
  arm: versatile : Fix typo introduced in irq namespace cleanup
  sound: Fixup the last user of the old irq functions
  genirq: Remove obsolete comment
  genirq: Remove now obsolete set_irq_wake()
  sh: Fix irq cleanup fallout
  x86: apb_timer: Fixup genirq fallout
  genirq: Fix misnamed label in handle_edge_eoi_irq

Fix up crazy conflict in arch/powerpc/include/asm/qe_ic.h:

 - commit eead4d5c63ff ("powerpc: qe_ic: Rename get_irq_desc_data and
   get_irq_desc_chip") made the helper functions use
   irq_desc_get_handler_data() instead of the legacy (and no longer
   existing) get_irq_desc_data.

 - commit d4db35e8dc96 ("powerpc/qe_ic: Fix another breakage from the
   irq_data conversion") used irq_desc_get_chip_data() instead.

According to Thomas, the former is the correct direct conversion, but it
does look like both should work (arch/powerpc/sysdev/qe_lib/qe_ic.c
seems to initialize both to the same thing), and the chip data in some
ways is the more logical.  Somebody should really decide on one of the
other.

This merge picks irq_desc_get_handler_data() as the straightforward pure
conversion to new names, as per Thomas.

8 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Linus Torvalds [Wed, 30 Mar 2011 14:53:17 +0000]
Merge branch 'merge' of git://git./linux/kernel/git/benh/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc:
  powerpc/qe_ic: Fix another breakage from the irq_data conversion
  powerpc/8xx: Fix another breakage from the irq_data conversion
  powerpc/cell: Use handle_edge_eoi_irq for real
  powerpc/pseries: Enable Chelsio network and iWARP drivers
  powerpc/mm: Move the STAB0 location to 0x8000 to make room in low memory
  powerpc: Fix accounting of softirq time when idle
  powerpc/pseries/smp: query-cpu-stopped-state support won't change
  powerpc/xics: Use hwirq for xics domain irq number
  powerpc/xics: Fix numberspace mismatch from irq_desc conversion
  powerpc: Wire up new syscalls
  powerpc/booke: Correct the SPRN_MAS5 definition.
  powerpc: ARCH_PFN_OFFSET should be unsigned long
  powerpc: Implement dma_mmap_coherent()
  powerpc/nvram: Don't overwrite oops/panic report on normal shutdown
  powerpc: Restore some misc devices to our configs

8 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/vapier...
Linus Torvalds [Wed, 30 Mar 2011 14:51:29 +0000]
Merge branch 'for-linus' of git://git./linux/kernel/git/vapier/blackfin

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/vapier/blackfin:
  Blackfin: bitops: fix include order after little endian inclusion
  Blackfin: defconfigs: update after misc devices defaulted to N
  Blackfin: use more standard pr_fmt in the module loader

8 years agoMerge branch 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm
Linus Torvalds [Wed, 30 Mar 2011 14:50:45 +0000]
Merge branch 'for-linus' of /home/rmk/linux-2.6-arm

* 'for-linus' of master.kernel.org:/home/rmk/linux-2.6-arm: (65 commits)
  ARM: 6826/1: Merge v6 and v7 DEBUG_LL DCC support
  ARM: 6838/1: etm: fix section mismatch warning
  ARM: 6837/1: remove unused pci_fixup_prpmc1100
  ARM: 6836/1: kprobes/fix emulation of LDR/STR instruction when Rn == PC
  Fix the broken build for Marvell Dove platform.
  ARM: 6835/1: perf: ensure overflows aren't missed due to IRQ latency
  ARM: 6834/1: perf: reset counters on all CPUs during initialisation
  ARM: 6833/1: perf: add required isbs() to ARMv7 backend
  ARM: 6825/1: kernel/sleep.S: fix Thumb2 compilation issues
  ARM: 6807/1: realview: Fix secondary GIC initialisation for EB with MPCore tile
  arm: mach-mx3: pcm043: add write-protect and card-detect for SD1
  eukrea_mbimxsd51: add SD Card detect
  eukrea_mbimxsd25-baseboard: add SD card detect
  mx3/eukrea_mbimxsd-baseboard: add SD card detect support
  mx3/eukrea_mbimxsd-baseboard: fix gpio request
  ARM: mxs/mx28evk: add mmc device
  ARM: mxs/mx23evk: add mmc device
  ARM: mxs: dynamically allocate mmc device
  ARM: mx51_efika: update platform data for new mfd changes
  mx2/iomux: Set direction for CSPI2 pins
  ...

8 years agoavr32: Fix missing irq namespace conversion
Thomas Gleixner [Wed, 30 Mar 2011 13:33:51 +0000]
avr32: Fix missing irq namespace conversion

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

8 years agopowerpc: qe_ic: Rename get_irq_desc_data and get_irq_desc_chip
Richard Cochran [Wed, 30 Mar 2011 13:25:04 +0000]
powerpc: qe_ic: Rename get_irq_desc_data and get_irq_desc_chip

These two functions disappeared in commit

    0c6f8a8b917ad361319c8ace3e9f28e69bfdb4c1
    "genirq: Remove compat code"

but they still exist in qe_ic.h.
This patch renames the function to their new names.

Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Cc: linuxppc-dev@lists.ozlabs.org
Cc: Lennert Buytenhek <buytenh@secretlab.ca>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
LKML-Reference: <20110330132504.GA31832@riccoc20.at.omicron.at>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

8 years agogenirq: Remove the now obsolete config options and select statements
Thomas Gleixner [Wed, 30 Mar 2011 12:13:23 +0000]
genirq: Remove the now obsolete config options and select statements

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

8 years agosparc32,leon: Fixed APBUART frequency detection
Daniel Hellstrom [Wed, 30 Mar 2011 01:12:41 +0000]
sparc32,leon: Fixed APBUART frequency detection

The UARTs may be located on different APB buses, thus have

different UART clock frequency. The system frequency is not
the same (but often) as the UART frequency, rather the APB bus
frequency that the APBUART is located at has the same
frequency, so this looks at the "freq" property instead.

Signed-off-by: Daniel Hellstrom <daniel@gaisler.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agosparc32, leon: APBUART driver must use archdata to get IRQ number
Daniel Hellstrom [Wed, 30 Mar 2011 01:12:40 +0000]
sparc32, leon: APBUART driver must use archdata to get IRQ number

See Commit id 1636f8ac2b08410df4766449f7c86b912443cd99 (sparc/of:
Move of_device fields into struct pdev_archdata), this patch
is similar to 19e4875fb21a69fbf620e84769a74d189c69c58d (of/sparc:
fix build regression from of_device changes)

Signed-off-by: Daniel Hellstrom <daniel@gaisler.com>
Acked-by: Sam Ravnborg <sam@ravnborg.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoarm: versatile : Fix typo introduced in irq namespace cleanup
Grant Likely [Wed, 30 Mar 2011 06:02:29 +0000]
arm: versatile : Fix typo introduced in irq namespace cleanup

Commit 6845664(arm: Cleanup the irq namespace) introduces a typo
causing a build failure for the versatile platform.

[ tglx: Sorry, my bad. Have no idea how I fatfingered that ]

Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: linux-arm-kernel@lists.infradead.org
LKML-Reference: <20110330060229.27397.7628.stgit@ponder>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

8 years agoAtheros, atl2: Fix mem leaks in error paths of atl2_set_eeprom
Jesper Juhl [Sun, 27 Mar 2011 09:16:12 +0000]
Atheros, atl2: Fix mem leaks in error paths of atl2_set_eeprom

We leak in some error paths of drivers/net/atlx/atl2.c:atl2_set_eeprom().
The memory allocated to 'eeprom_buff' is not freed when we return -EIO.
This patch fixes that up and also removes a pointless explicit cast.

Signed-off-by: Jesper Juhl <jj@chaosbits.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonetdev: fix mtu check when TSO is enabled
Daniel Lezcano [Wed, 30 Mar 2011 09:42:17 +0000]
netdev: fix mtu check when TSO is enabled

In case the device where is coming from the packet has TSO enabled,
we should not check the mtu size value as this one could be bigger
than the expected value.

This is the case for the macvlan driver when the lower device has
TSO enabled. The macvlan inherit this feature and forward the packets
without fragmenting them. Then the packets go through dev_forward_skb
and are dropped. This patch fix this by checking TSO is not enabled
when we want to check the mtu size.

Signed-off-by: Daniel Lezcano <daniel.lezcano@free.fr>
Acked-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonet/usb: Ethernet quirks for the LG-VL600 4G modem
Andrzej Zaborowski [Mon, 28 Mar 2011 12:56:33 +0000]
net/usb: Ethernet quirks for the LG-VL600 4G modem

This adds a driver for the CDC Ethernet part of this modem.  The
device's ID is blacklisted in cdc_ether.c and is white-listed in
this new driver because of the quirks needed to make it useful.
The modem's firmware exposes a CDC ACM port for modem control and a
CDC Ethernet port for network data.  The descriptors look fine but
both ports actually are some sort of multiplexers requiring non-
standard headers added/removed from every packet or they get
ignored.  All information is based on a usb traffic log from a
Windows machine.

On the Verizon 4G network I've seen speeds up to 1.1MB/s so far with
this driver, a speed-o-meter site reports 16.2Mbps/10.5Mbps.
Userspace scripts are required to talk to the CDC ACM port.

Signed-off-by: Andrzej Zaborowski <balrogg@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agophylib: phy_attach_direct: phy_init_hw can fail, add cleanup
Marc Kleine-Budde [Mon, 28 Mar 2011 14:54:08 +0000]
phylib: phy_attach_direct: phy_init_hw can fail, add cleanup

The function phy_attach_direct attaches the phy and calls phy_init_hw.
phy_init_hw can fail, but the phy is still marked as attached. Successive
calls to phy_attach_direct will fail because the phy is busy.

[    1.020000] eth0: Freescale FEC PHY driver [Generic PHY] (mii_bus:phy_addr=1:00, irq=-1)
[    1.030000] eth1: Freescale FEC PHY driver [Generic PHY] (mii_bus:phy_addr=1:01, irq=-1)
[    2.050000] Sending DHCP requests .
[    3.020000] PHY: 1:00 - Link is Up - 100/Full
[    5.110000] ..... timed out!
[   87.660000] IP-Config: Reopening network devices...
[   88.190000] FEC: MDIO read timeout
[   88.190000] eth0: could not attach to PHY
[   88.190000] IP-Config: Failed to open eth0
[   88.210000] FEC: MDIO read timeout
[   88.210000] eth1: could not attach to PHY
[   88.210000] IP-Config: Failed to open eth1
[   88.220000] IP-Config: No network devices available.
[   88.220000] Freeing init memory: 6968K

[...]

starting network interfaces...
ip: RTNETLINK answers: File exists
[   94.000000] net eth0: PHY already attached
[   94.010000] eth0: could not attach to PHY
ip: SIOCSIFFLAGS: Device or resource busy

This patch adds phy_detach to clean up if phy_init_hw fails.

Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agobridge: mcast snooping, fix length check of snooped MLDv1/2
Linus Lüssing [Sat, 26 Mar 2011 20:27:24 +0000]
bridge: mcast snooping, fix length check of snooped MLDv1/2

"len = ntohs(ip6h->payload_len)" does not include the length of the ipv6
header itself, which the rest of this function assumes, though.

This leads to a length check less restrictive as it should be in the
following line for one thing. For another, it very likely leads to an
integer underrun when substracting the offset and therefore to a very
high new value of 'len' due to its unsignedness. This will ultimately
lead to the pskb_trim_rcsum() practically never being called, even in
the cases where it should.

Signed-off-by: Linus Lüssing <linus.luessing@web.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonilfs2: fix whitespace coding style issues
Nicolas Kaiser [Sat, 19 Mar 2011 15:47:54 +0000]
nilfs2: fix whitespace coding style issues

Fixes whitespace coding style issues.

Signed-off-by: Nicolas Kaiser <nikai@nikai.net>
Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>

8 years agonilfs2: fix oops due to a bad aops initialization
Ryusuke Konishi [Wed, 30 Mar 2011 02:49:20 +0000]
nilfs2: fix oops due to a bad aops initialization

Nilfs in 2.6.39-rc1 hit the following oops:

 BUG: unable to handle kernel NULL pointer dereference at 0000000000000048
 IP: [<ffffffff810ac235>] try_to_release_page+0x2a/0x3d
 PGD 234cb6067 PUD 234c72067 PMD 0
 Oops: 0000 [#1] SMP
 <snip>
 Process truncate (pid: 10995, threadinfo ffff8802353c2000, task ffff880234cfa000)
 Stack:
  ffff8802333c77b8 ffffffff810b64b0 0000000000003802 ffffffffa0052cca
  0000000000000000 ffff8802353c3b58 0000000000000000 ffff8802353c3b58
  0000000000000001 0000000000000000 ffffea0007b92308 ffffea0007b92308
 Call Trace:
  [<ffffffff810b64b0>] ? invalidate_inode_pages2_range+0x15f/0x273
  [<ffffffffa0052cca>] ? nilfs_palloc_get_block+0x2d/0xaf [nilfs2]
  [<ffffffff810589e7>] ? bit_waitqueue+0x14/0xa1
  [<ffffffff81058ab1>] ? wake_up_bit+0x10/0x20
  [<ffffffffa00433fd>] ? nilfs_forget_buffer+0x66/0x7a [nilfs2]
  [<ffffffffa00467b8>] ? nilfs_btree_concat_left+0x5c/0x77 [nilfs2]
  [<ffffffffa00471fc>] ? nilfs_btree_delete+0x395/0x3cf [nilfs2]
  [<ffffffffa00449a3>] ? nilfs_bmap_do_delete+0x6e/0x79 [nilfs2]
  [<ffffffffa0045845>] ? nilfs_btree_last_key+0x14b/0x15e [nilfs2]
  [<ffffffffa00449dd>] ? nilfs_bmap_truncate+0x2f/0x83 [nilfs2]
  [<ffffffffa0044ab2>] ? nilfs_bmap_last_key+0x35/0x62 [nilfs2]
  [<ffffffffa003e99b>] ? nilfs_truncate_bmap+0x6b/0xc7 [nilfs2]
  [<ffffffffa003ee4a>] ? nilfs_truncate+0x79/0xe4 [nilfs2]
  [<ffffffff810b6c00>] ? vmtruncate+0x33/0x3b
  [<ffffffffa003e8f1>] ? nilfs_setattr+0x4d/0x8c [nilfs2]
  [<ffffffff81026106>] ? do_page_fault+0x31b/0x356
  [<ffffffff810f9d61>] ? notify_change+0x17d/0x262
  [<ffffffff810e5046>] ? do_truncate+0x65/0x80
  [<ffffffff810e52af>] ? sys_ftruncate+0xf1/0xf6
  [<ffffffff8132c012>] ? system_call_fastpath+0x16/0x1b
 Code: c3 48 83 ec 08 48 8b 17 48 8b 47 18 80 e2 01 75 04 0f 0b eb fe 48 8b 17 80 e6 20 74 05 31 c0 41 59 c3 48 85 c0 74 11 48 8b 40 58
  8b 40 48 48 85 c0 74 04 41 58 ff e0 59 e9 b1 b5 05 00 41 54
 RIP  [<ffffffff810ac235>] try_to_release_page+0x2a/0x3d
  RSP <ffff8802353c3b08>
 CR2: 0000000000000048

This oops was brought in by the change "block: remove per-queue
plugging" (commit: 7eaceaccab5f40bb).  It initializes mapping->a_ops
with a NULL pointer for some pages in nilfs (e.g. btree node pages),
but mm code doesn't NULL pointer checks against mapping->a_ops. (the
check is done for each callback function)

This corrects the aops initialization and fixes the oops.

Signed-off-by: Ryusuke Konishi <konishi.ryusuke@lab.ntt.co.jp>
Acked-by: Jens Axboe <jaxboe@fusionio.com>

8 years agovia-ircc: Pass PCI device pointer to dma_{alloc, free}_coherent()
Ben Hutchings [Mon, 28 Mar 2011 17:12:52 +0000]
via-ircc: Pass PCI device pointer to dma_{alloc, free}_coherent()

via-ircc has been passing a NULL pointer to DMA allocation functions,
which is completely invalid and results in a BUG on PowerPC.  Now
that we always have the device pointer available, pass it in.

Reference: http://bugs.debian.org/619450
Reported-by: Andrew Buckeridge <andrewb@bgc.com.au>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Tested-by: Andrew Buckeridge <andrewb@bgc.com.au> [against 2.6.32]
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agovia-ircc: Use pci_{get, set}_drvdata() instead of static pointer variable
Ben Hutchings [Mon, 28 Mar 2011 17:10:43 +0000]
via-ircc: Use pci_{get, set}_drvdata() instead of static pointer variable

via-ircc still maintains its own array of device pointers in Linux 2.4
style.  Worse, it always uses index 0, so it will crash if there are
multiple suitable devices in the system.

Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agonet: gre: provide multicast mappings for ipv4 and ipv6
Timo Teräs [Mon, 28 Mar 2011 22:40:53 +0000]
net: gre: provide multicast mappings for ipv4 and ipv6

My commit 6d55cb91a0020ac0 (gre: fix hard header destination
address checking) broke multicast.

The reason is that ip_gre used to get ipgre_header() calls with
zero destination if we have NOARP or multicast destination. Instead
the actual target was decided at ipgre_tunnel_xmit() time based on
per-protocol dissection.

Instead of allowing the "abuse" of ->header() calls with invalid
destination, this creates multicast mappings for ip_gre. This also
fixes "ip neigh show nud noarp" to display the proper multicast
mappings used by the gre device.

Reported-by: Doug Kehn <rdkehn@yahoo.com>
Signed-off-by: Timo Teräs <timo.teras@iki.fi>
Acked-by: Doug Kehn <rdkehn@yahoo.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoMerge commits 'ca6a42586fae' and 'c286c419c784' into perf/urgent
Ingo Molnar [Wed, 30 Mar 2011 07:08:20 +0000]
Merge commits 'ca6a42586fae' and 'c286c419c784' into perf/urgent

Pick up these two commits from Arnaldo's perf/core tree:

  ca6a42586fae: perf tools: Emit clearer message for sys_perf_event_open ENOENT return
  c286c419c784: perf tools: Fixup exit path when not able to open events

As they are really fixes we want to have sooner than laer.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agoMerge commit 'v2.6.39-rc1' into perf/urgent
Ingo Molnar [Wed, 30 Mar 2011 07:07:39 +0000]
Merge commit 'v2.6.39-rc1' into perf/urgent

Merge reason: use the post-merge-window tree.

Signed-off-by: Ingo Molnar <mingo@elte.hu>

8 years agoALSA: ens1371: fix Creative Ectiva support
Clemens Ladisch [Wed, 30 Mar 2011 06:24:25 +0000]
ALSA: ens1371: fix Creative Ectiva support

To make the EV1938 chip work, add a magic bit and an extra delay.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Tested-by: Tino Schmidt <mailtinoshomepage@gmx.net>
Cc: all 2.6.x <stable@kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agoALSA: firewire-speakers: fix hang when unplugging a running device
Clemens Ladisch [Wed, 30 Mar 2011 06:24:00 +0000]
ALSA: firewire-speakers: fix hang when unplugging a running device

When aborting a PCM stream, the xrun is signaled only if the stream is
running.  When disconnecting a PCM stream, calling snd_card_disconnect()
too early would change the stream into a non-running state and thus
prevent the xrun from being noticed by user space.

To prevent this, move the snd_card_disconnect() call after the xrun.

Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>

8 years agobridge: Fix compilation warning in function br_stp_recalculate_bridge_id()
Balaji G [Tue, 29 Mar 2011 06:20:04 +0000]
bridge: Fix compilation warning in function br_stp_recalculate_bridge_id()

net/bridge/br_stp_if.c: In function ‘br_stp_recalculate_bridge_id’:
net/bridge/br_stp_if.c:216:3: warning: ‘return’ with no value, in function returning non-void

Signed-off-by: G.Balaji <balajig81@gmail.com>
Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

8 years agoMerge branch 'for-2.6.39' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie...
Takashi Iwai [Wed, 30 Mar 2011 06:35:18 +0000]
Merge branch 'for-2.6.39' of git://git./linux/kernel/git/broonie/sound-2.6 into fix/asoc