pandora-kernel.git
14 years agoMerge branch 'x86/apic' into auto-latest
Ingo Molnar [Mon, 23 May 2011 19:48:41 +0000 (21:48 +0200)]
Merge branch 'x86/apic' into auto-latest

14 years agox86, apic: Include module.h header in apic_flat_64.c
Randy Dunlap [Mon, 23 May 2011 17:43:00 +0000 (10:43 -0700)]
x86, apic: Include module.h header in apic_flat_64.c

apic_flat_64.c needs to include module.h because it uses
EXPORT_SYMBOL_GPL().

This fixes these warnings on some !SMP randconfigs:

  arch/x86/kernel/apic/apic_flat_64.c:31: warning: data definition has no type or storage class
  arch/x86/kernel/apic/apic_flat_64.c:31: warning: type defaults to 'int' in declaration of 'EXPORT_SYMBOL_GPL'
  arch/x86/kernel/apic/apic_flat_64.c:31: warning: parameter names (without types) in function declaration

Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Link: http://lkml.kernel.org/r/20110523104300.dd532a99.randy.dunlap@oracle.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoMerge branch 'perf/urgent' into auto-latest
Ingo Molnar [Mon, 23 May 2011 14:09:54 +0000 (16:09 +0200)]
Merge branch 'perf/urgent' into auto-latest

14 years agoMerge branch 'x86/urgent' into auto-latest
Ingo Molnar [Mon, 23 May 2011 14:09:41 +0000 (16:09 +0200)]
Merge branch 'x86/urgent' into auto-latest

14 years agoMerge branch 'x86/apic' into auto-latest
Ingo Molnar [Mon, 23 May 2011 14:09:39 +0000 (16:09 +0200)]
Merge branch 'x86/apic' into auto-latest

14 years agoMerge branch 'timers/core' into auto-latest
Ingo Molnar [Mon, 23 May 2011 14:09:37 +0000 (16:09 +0200)]
Merge branch 'timers/core' into auto-latest

14 years agoMerge branch 'sched/core' into auto-latest
Ingo Molnar [Mon, 23 May 2011 14:09:35 +0000 (16:09 +0200)]
Merge branch 'sched/core' into auto-latest

14 years agohrtimers: Reorder clock bases
Thomas Gleixner [Fri, 20 May 2011 21:14:04 +0000 (23:14 +0200)]
hrtimers: Reorder clock bases

The ordering of the clock bases is historical due to the
CLOCK_REALTIME and CLOCK_MONOTONIC constants. Now the hrtimer bases
have their own enumeration due to the gap between CLOCK_MONOTONIC and
CLOCK_BOOTTIME. So we can be more clever as most timers end up on the
CLOCK_MONOTONIC base due to the virtue of POSIX declaring that
relative CLOCK_REALTIME timers are not affected by time changes. In
desktop environments this is slowly changing as applications switch to
absolute timers, but I've observed empty CLOCK_REALTIME bases often
enough. There is no performance penalty or overhead when
CLOCK_REALTIME timers are active, but in case they are not we don't
skip over a full cache line.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Peter Zijlstra <peterz@infradead.org>
14 years agohrtimers: Avoid touching inactive timer bases
Thomas Gleixner [Fri, 20 May 2011 11:05:15 +0000 (13:05 +0200)]
hrtimers: Avoid touching inactive timer bases

Instead of iterating over all possible timer bases avoid it by marking
the active bases in the cpu base.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Peter Zijlstra <peterz@infradead.org>
14 years agohrtimers: Make struct hrtimer_cpu_base layout less stupid
Thomas Gleixner [Fri, 20 May 2011 11:02:58 +0000 (13:02 +0200)]
hrtimers: Make struct hrtimer_cpu_base layout less stupid

In the HIGHRES=y case we access the members at the end of struct
hrtimer_cpu_base first and then the one at the beginning. Move the
hrtimer data to front, so we have linear progressing access.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Peter Zijlstra <peterz@infradead.org>
14 years agotimerfd: Manage cancelable timers in timerfd
Thomas Gleixner [Fri, 20 May 2011 14:18:50 +0000 (16:18 +0200)]
timerfd: Manage cancelable timers in timerfd

Peter is concerned about the extra scan of CLOCK_REALTIME_COS in the
timer interrupt. Yes, I did not think about it, because the solution
was so elegant. I didn't like the extra list in timerfd when it was
proposed some time ago, but with a rcu based list the list walk it's
less horrible than the original global lock, which was held over the
list iteration.

Requested-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Peter Zijlstra <peterz@infradead.org>
14 years agoperf tools: Fix sample size bit operations
Frederic Weisbecker [Mon, 23 May 2011 11:06:27 +0000 (13:06 +0200)]
perf tools: Fix sample size bit operations

What we want is to count the number of bits in the mask,
not some other random operation written in the middle
of the night.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1306148788-6179-2-git-send-email-fweisbec@gmail.com
[ Fixed perf_event__names[] alignment which was nearby and hurting my eyes ... ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoperf tools: Fix ommitted mmap data update on remap
Frederic Weisbecker [Mon, 23 May 2011 11:06:28 +0000 (13:06 +0200)]
perf tools: Fix ommitted mmap data update on remap

Commit eac9eacee16 "perf tools: Check we are able to read the event
size on mmap" brought a check to ensure we can read the size of the
event before dereferencing it, and do a remap otherwise to move the
buffer forward.

However that remap was ommitting all the necessary work to
update the new page offset, head, and to unmap previous pages,
etc...

To fix this, gather all the code that fetches the event in a
seperate helper which does all the necessary checks about the
header/event size and tells us anytime a remap is needed.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/1306148788-6179-3-git-send-email-fweisbec@gmail.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agowatchdog: Change the default timeout and configure nmi watchdog period based on watch...
Mandeep Singh Baines [Mon, 23 May 2011 05:10:23 +0000 (22:10 -0700)]
watchdog: Change the default timeout and configure nmi watchdog period based on watchdog_thresh

Before the conversion of the NMI watchdog to perf event, the
watchdog timeout was 5 seconds. Now it is 60 seconds. For my
particular application, netbooks, 5 seconds was a better
timeout. With a short timeout, we catch faults earlier and are
able to send back a panic. With a 60 second timeout, the user is
unlikely to wait and will instead hit the power button, causing
us to lose the panic info.

This change configures the NMI period to watchdog_thresh and
sets the softlockup_thresh to watchdog_thresh * 2. In addition,
watchdog_thresh was reduced to 10 seconds as suggested by Ingo
Molnar.

Signed-off-by: Mandeep Singh Baines <msb@chromium.org>
Cc: Marcin Slusarz <marcin.slusarz@gmail.com>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Link: http://lkml.kernel.org/r/1306127423-3347-4-git-send-email-msb@chromium.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
LKML-Reference: <20110517071642.GF22305@elte.hu>

14 years agowatchdog: Disable watchdog when thresh is zero
Mandeep Singh Baines [Mon, 23 May 2011 05:10:22 +0000 (22:10 -0700)]
watchdog: Disable watchdog when thresh is zero

This restores the previous behavior of softlock_thresh.

Currently, setting watchdog_thresh to zero causes the watchdog
kthreads to consume a lot of CPU.

In addition, the logic of proc_dowatchdog_thresh and
proc_dowatchdog_enabled has been factored into proc_dowatchdog.

Signed-off-by: Mandeep Singh Baines <msb@chromium.org>
Cc: Marcin Slusarz <marcin.slusarz@gmail.com>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Link: http://lkml.kernel.org/r/1306127423-3347-3-git-send-email-msb@chromium.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
LKML-Reference: <20110517071018.GE22305@elte.hu>

14 years agowatchdog: Only disable/enable watchdog if neccessary
Mandeep Singh Baines [Mon, 23 May 2011 05:10:21 +0000 (22:10 -0700)]
watchdog: Only disable/enable watchdog if neccessary

Don't take any action on an unsuccessful write to /proc.

Signed-off-by: Mandeep Singh Baines <msb@chromium.org>
Cc: Marcin Slusarz <marcin.slusarz@gmail.com>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Link: http://lkml.kernel.org/r/1306127423-3347-2-git-send-email-msb@chromium.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agowatchdog: Fix rounding bug in get_sample_period()
Mandeep Singh Baines [Mon, 23 May 2011 05:10:20 +0000 (22:10 -0700)]
watchdog: Fix rounding bug in get_sample_period()

In get_sample_period(), softlockup_thresh is integer divided by
5 before the multiplication by NSEC_PER_SEC. This results in
softlockup_thresh being rounded down to the nearest integer
multiple of 5.

For example, a softlockup_thresh of 4 rounds down to 0.

Signed-off-by: Mandeep Singh Baines <msb@chromium.org>
Cc: Marcin Slusarz <marcin.slusarz@gmail.com>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Link: http://lkml.kernel.org/r/1306127423-3347-1-git-send-email-msb@chromium.org
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, apic: Make apic drivers static
Suresh Siddha [Sat, 21 May 2011 00:51:20 +0000 (17:51 -0700)]
x86, apic: Make apic drivers static

Apic probe now looks at the apic drivers listed in the
.apicdrivers section. Remove apic_probe[] and make each apic
driver static.

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Tested-by: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: steiner@sgi.com
Cc: gorcunov@openvz.org
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110521005526.341718626@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, apic: Clean up bigsmp apic selection code
Suresh Siddha [Sat, 21 May 2011 00:51:19 +0000 (17:51 -0700)]
x86, apic: Clean up bigsmp apic selection code

Make generic_bigsmp_probe() return struct apic *. This will
avoid exporting apic_bigsmp, which will be consistent with
others.

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Tested-by: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: steiner@sgi.com
Cc: gorcunov@openvz.org
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110521005526.252703851@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, apic: Use .apicdrivers section for the apic drivers list
Suresh Siddha [Sat, 21 May 2011 00:51:18 +0000 (17:51 -0700)]
x86, apic: Use .apicdrivers section for the apic drivers list

This will eliminate the need for apic_probe[], as the probing
now will happen based on the apic drivers order in the
.apcidrivers section.

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Tested-by: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: steiner@sgi.com
Cc: gorcunov@openvz.org
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110521005526.164277071@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, apic: Introduce .apicdrivers section to find the list of apic drivers
Suresh Siddha [Sat, 21 May 2011 00:51:17 +0000 (17:51 -0700)]
x86, apic: Introduce .apicdrivers section to find the list of apic drivers

This will pave the way for each apic driver to be self-contained
and eliminate the need for apic_probe[].

Order in which apic drivers are listed in the .apicdrivers
section is important, as this determines the apic probe order.
And this is enforced by the ordering of apic driver files in the
Makefile and the macros apic_driver()/apic_drivers().

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Tested-by: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: steiner@sgi.com
Cc: gorcunov@openvz.org
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110521005526.068775085@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoMerge branch 'perf/core' of git://git.kernel.org/pub/scm/linux/kernel/git/frederic...
Ingo Molnar [Sun, 22 May 2011 08:07:37 +0000 (10:07 +0200)]
Merge branch 'perf/core' of git://git./linux/kernel/git/frederic/random-tracing into perf/urgent

Conflicts:
tools/perf/builtin-top.c

Semantic conflict:
util/include/linux/list.h        # fix prefetch.h removal fallout

Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoperf tools: Propagate event parse error handling
Frederic Weisbecker [Sun, 22 May 2011 00:17:22 +0000 (02:17 +0200)]
perf tools: Propagate event parse error handling

Better handle event parsing error by propagating the details
in upper layers or by dumping some failure message. So that
the user knows he has some crazy events in the batch.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
14 years agoperf tools: Robustify dynamic sample content fetch
Frederic Weisbecker [Sat, 21 May 2011 18:08:15 +0000 (20:08 +0200)]
perf tools: Robustify dynamic sample content fetch

Ensure the size of the dynamic fields such as callchains
or raw events don't overlap the whole event boundaries.

This prevents from dereferencing junk if the given size of
the callchain goes too eager.

Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Reported-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
14 years agoperf tools: Pre-check sample size before parsing
Frederic Weisbecker [Sat, 21 May 2011 17:33:04 +0000 (19:33 +0200)]
perf tools: Pre-check sample size before parsing

Check that the total size of the sample fields having a fixed
size do not exceed the one of the whole event. This robustifies
the sample parsing.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
14 years agoperf tools: Move evlist sample helpers to evlist area
Frederic Weisbecker [Sat, 21 May 2011 15:49:00 +0000 (17:49 +0200)]
perf tools: Move evlist sample helpers to evlist area

These APIs should belong to evlist.c as they may not be
exclusively tied to the headers.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com
14 years agoperf tools: Remove junk code in mmap size handling
Frederic Weisbecker [Sat, 21 May 2011 15:07:24 +0000 (17:07 +0200)]
perf tools: Remove junk code in mmap size handling

size is overriden later and used only then. Those
lines are only junk, probably a leftover.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
14 years agoperf tools: Check we are able to read the event size on mmap
Frederic Weisbecker [Sat, 21 May 2011 15:07:24 +0000 (17:07 +0200)]
perf tools: Check we are able to read the event size on mmap

Check we have enough mmaped space to read the current event
size from its headers, otherwise we may dereference some
hell there.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
14 years agoMerge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394...
Linus Torvalds [Sat, 21 May 2011 19:25:07 +0000 (12:25 -0700)]
Merge branch 'for-linus' of git://git./linux/kernel/git/ieee1394/linux1394-2.6

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6:
  firewire: sbp2: parallelize login, reconnect, logout
  firewire: sbp2: octlet AT payloads can be stack-allocated
  firewire: sbp2: omit Scsi_Host lock from queuecommand
  firewire: core: use non-reentrant workqueue with rescuer
  firewire: optimize iso queueing by setting wake only after the last packet
  firewire: octlet AT payloads can be stack-allocated
  firewire: ohci: optimize find_branch_descriptor()
  firewire: ohci: avoid separate DMA mapping for small AT payloads
  firewire: ohci: do not start DMA contexts before link is enabled

14 years agox86: Eliminate various 'set but not used' warnings
Gustavo F. Padovan [Fri, 20 May 2011 19:13:00 +0000 (16:13 -0300)]
x86: Eliminate various 'set but not used' warnings

Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
Cc: Joerg Roedel <joerg.roedel@amd.com> (supporter:AMD IOMMU (AMD-VI))
Cc: iommu@lists.linux-foundation.org (open list:AMD IOMMU (AMD-VI))
Link: http://lkml.kernel.org/r/1305918786-7239-3-git-send-email-padovan@profusion.mobi
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, SMEP: Fix section mismatch warnings
Fenghua Yu [Fri, 20 May 2011 22:33:17 +0000 (15:33 -0700)]
x86, SMEP: Fix section mismatch warnings

Fix these kernel compilation warnings:

 WARNING: arch/x86/built-in.o(.cpuinit.text+0x1e07): Section mismatch ...
 WARNING: arch/x86/built-in.o(.cpuinit.text+0x1b10): Section mismatch ...

introduced by:

  de5397ad5b9a: x86, cpu: Enable/disable Supervisor Mode Execution Protection

Change disable_smep from __initdata to __cpuinitdata.
Change setup_smep() from __init to __cpuinit.

Reported-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
Cc: Asit K Mallick <asit.k.mallick@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/1305930797-11409-1-git-send-email-fenghua.yu@intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
Linus Torvalds [Sat, 21 May 2011 07:13:03 +0000 (00:13 -0700)]
Merge git://git./linux/kernel/git/davem/net-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6:
  sctp: Fix build failure.
  garp: use kfree_rcu()
  ipv6: copy prefsrc setting when copying route entry

14 years agoVFS: move BUG_ON test for symlink nd->depth after current->link_count test
Erez Zadok [Sat, 21 May 2011 05:19:59 +0000 (01:19 -0400)]
VFS: move BUG_ON test for symlink nd->depth after current->link_count test

This solves a serious VFS-level bug in nested_symlink (which was
rewritten from do_follow_link), and follows the order of depth tests
that existed before.

The bug triggers a BUG_ON in fs/namei.c:1381, when running racer with
symlink and rename ops.

Signed-off-by: Erez Zadok <ezk@cs.sunysb.edu>
Acked-by: Miklos Szeredi <mszeredi@suse.cz>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agosctp: Fix build failure.
David S. Miller [Sat, 21 May 2011 06:10:23 +0000 (02:10 -0400)]
sctp: Fix build failure.

Commit c182f90bc1f22ce5039b8722e45621d5f96862c2 ("SCTP: fix race
between sctp_bind_addr_free() and sctp_bind_addr_conflict()") and
commit 1231f0baa547a541a7481119323b7f964dda4788 ("net,rcu: convert
call_rcu(sctp_local_addr_free) to kfree_rcu()"), happening in
different trees, introduced a build failure.

Simply make the SCTP race fix use kfree_rcu() too.

Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agogarp: use kfree_rcu()
Eric Dumazet [Fri, 20 May 2011 18:31:30 +0000 (18:31 +0000)]
garp: use kfree_rcu()

Use kfree_rcu() instead of call_rcu(), remove garp_cleanup_module()

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoipv6: copy prefsrc setting when copying route entry
Florian Westphal [Fri, 20 May 2011 11:27:24 +0000 (11:27 +0000)]
ipv6: copy prefsrc setting when copying route entry

commit c3968a857a6b6c3d2ef4ead35776b055fb664d74
('ipv6: RTA_PREFSRC support for ipv6 route source address selection')
added support for ipv6 prefsrc as an alternative to ipv6 addrlabels,
but it did not work because the prefsrc entry was not copied.

Cc: Daniel Walter <sahne@0x90.at>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Linus Torvalds [Sat, 21 May 2011 00:24:14 +0000 (17:24 -0700)]
Merge git://git./linux/kernel/git/herbert/crypto-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6: (45 commits)
  crypto: caam - add support for sha512 variants of existing AEAD algorithms
  crypto: caam - remove unused authkeylen from caam_ctx
  crypto: caam - fix decryption shared vs. non-shared key setting
  crypto: caam - platform_bus_type migration
  crypto: aesni-intel - fix aesni build on i386
  crypto: aesni-intel - Merge with fpu.ko
  crypto: mv_cesa - make count_sgs() null-pointer proof
  crypto: mv_cesa - copy remaining bytes to SRAM only when needed
  crypto: mv_cesa - move digest state initialisation to a better place
  crypto: mv_cesa - fill inner/outer IV fields only in HMAC case
  crypto: mv_cesa - refactor copy_src_to_buf()
  crypto: mv_cesa - no need to save digest state after the last chunk
  crypto: mv_cesa - print a warning when registration of AES algos fail
  crypto: mv_cesa - drop this call to mv_hash_final from mv_hash_finup
  crypto: mv_cesa - the descriptor pointer register needs to be set just once
  crypto: mv_cesa - use ablkcipher_request_cast instead of the manual container_of
  crypto: caam - fix printk recursion for long error texts
  crypto: caam - remove unused keylen from session context
  hwrng: amd - enable AMD hw rnd driver for Maple PPC boards
  hwrng: amd - manage resource allocation
  ...

14 years agoFix for buffer overflow in ldm_frag_add not sufficient
Timo Warns [Thu, 19 May 2011 07:24:17 +0000 (09:24 +0200)]
Fix for buffer overflow in ldm_frag_add not sufficient

As Ben Hutchings discovered [1], the patch for CVE-2011-1017 (buffer
overflow in ldm_frag_add) is not sufficient.  The original patch in
commit c340b1d64000 ("fs/partitions/ldm.c: fix oops caused by corrupted
partition table") does not consider that, for subsequent fragments,
previously allocated memory is used.

[1] http://lkml.org/lkml/2011/5/6/407

Reported-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Timo Warns <warns@pre-sense.de>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agotmpfs: fix highmem swapoff crash regression
Hugh Dickins [Fri, 20 May 2011 22:47:33 +0000 (15:47 -0700)]
tmpfs: fix highmem swapoff crash regression

Commit 778dd893ae78 ("tmpfs: fix race between umount and swapoff")
forgot the new rules for strict atomic kmap nesting, causing

  WARNING: at arch/x86/mm/highmem_32.c:81

from __kunmap_atomic(), then

  BUG: unable to handle kernel paging request at fffb9000

from shmem_swp_set() when shmem_unuse_inode() is handling swapoff with
highmem in use.  My disgrace again.

See
  https://bugzilla.kernel.org/show_bug.cgi?id=35352

Reported-by: Witold Baryluk <baryluk@smp.if.uj.edu.pl>
Signed-off-by: Hugh Dickins <hughd@google.com>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agopowerpc: Fix 32-bit SMP build
Josh Boyer [Fri, 20 May 2011 20:22:25 +0000 (16:22 -0400)]
powerpc: Fix 32-bit SMP build

Commit 69e3cea8d5fd526 ("powerpc/smp: Make start_secondary_resume
available to all CPU variants") introduced start_secondary_resume to
misc_32.S, however it uses a 64-bit instruction which is not valid on
32-bit platforms.  Use 'stw' instead.

Reported-by: Richard Cochran <richardcochran@gmail.com>
Tested-by: Richard Cochran <richardcochran@gmail.com>
Signed-off-by: Josh Boyer <jwboyer@linux.vnet.ibm.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoMerge branch 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzi...
Linus Torvalds [Fri, 20 May 2011 21:31:27 +0000 (14:31 -0700)]
Merge branch 'upstream-linus' of git://git./linux/kernel/git/jgarzik/libata-dev

* 'upstream-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev:
  libata: Power off empty ports
  libata-pmp: add support for Thermaltake BlackX Duet esata drive dock
  ATA: Don't powerdown Compaq Triflex IDE device on suspend
  libata: Use Maximum Write Same Length to report discard size limit
  drivers/ata/acard-ahci.c: fix enum warning
  pata_at91: SMC settings calculation bugfixes, support for t6z and IORDY
  libata-sff: prevent irq descriptions for dummy ports
  pata_cm64x: fix boot crash on parisc

14 years agoipc: Add missing sys_ni entries for ipc/compat.c functions
Kevin Cernekee [Tue, 17 May 2011 17:39:58 +0000 (10:39 -0700)]
ipc: Add missing sys_ni entries for ipc/compat.c functions

When building with:

  CONFIG_64BIT=y
  CONFIG_MIPS32_COMPAT=y
  CONFIG_COMPAT=y
  CONFIG_MIPS32_O32=y
  CONFIG_MIPS32_N32=y
  CONFIG_SYSVIPC is not set
  (and implicitly: CONFIG_SYSVIPC_COMPAT is not set)

the final link fails with unresolved symbols for:

  compat_sys_semctl, compat_sys_msgsnd, compat_sys_msgrcv,
  compat_sys_shmctl, compat_sys_msgctl, compat_sys_semtimedop

The fix is to add cond_syscall declarations for all syscalls in
ipc/compat.c

Signed-off-by: Kevin Cernekee <cernekee@gmail.com>
Acked-by: Ralf Baechle <ralf@linux-mips.org>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6
Linus Torvalds [Fri, 20 May 2011 20:43:21 +0000 (13:43 -0700)]
Merge git://git./linux/kernel/git/davem/net-next-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6: (1446 commits)
  macvlan: fix panic if lowerdev in a bond
  tg3: Add braces around 5906 workaround.
  tg3: Fix NETIF_F_LOOPBACK error
  macvlan: remove one synchronize_rcu() call
  networking: NET_CLS_ROUTE4 depends on INET
  irda: Fix error propagation in ircomm_lmp_connect_response()
  irda: Kill set but unused variable 'bytes' in irlan_check_command_param()
  irda: Kill set but unused variable 'clen' in ircomm_connect_indication()
  rxrpc: Fix set but unused variable 'usage' in rxrpc_get_transport()
  be2net: Kill set but unused variable 'req' in lancer_fw_download()
  irda: Kill set but unused vars 'saddr' and 'daddr' in irlan_provider_connect_indication()
  atl1c: atl1c_resume() is only used when CONFIG_PM_SLEEP is defined.
  rxrpc: Fix set but unused variable 'usage' in rxrpc_get_peer().
  rxrpc: Kill set but unused variable 'local' in rxrpc_UDP_error_handler()
  rxrpc: Kill set but unused variable 'sp' in rxrpc_process_connection()
  rxrpc: Kill set but unused variable 'sp' in rxrpc_rotate_tx_window()
  pkt_sched: Kill set but unused variable 'protocol' in tc_classify()
  isdn: capi: Use pr_debug() instead of ifdefs.
  tg3: Update version to 3.119
  tg3: Apply rx_discards fix to 5719/5720
  ...

Fix up trivial conflicts in arch/x86/Kconfig and net/mac80211/agg-tx.c
as per Davem.

14 years agoMerge branch 'release' of git://git.kernel.org/pub/scm/linux/kernel/git/aegl/linux-2.6
Linus Torvalds [Fri, 20 May 2011 20:39:00 +0000 (13:39 -0700)]
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] define "_sdata" symbol
  pstore: Fix Kconfig dependencies for apei->pstore
  pstore: fix potential logic issue in pstore read interface
  pstore: fix pstore filesystem mount/remount issue
  pstore: fix one type of return value in pstore
  [IA64] fix build warning in arch/ia64/oprofile/backtrace.c

14 years agoMerge branch 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 20 May 2011 20:38:28 +0000 (13:38 -0700)]
Merge branch 'timers-fixes-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  MAINTAINERS: Add drivers/clocksource to TIMEKEEPING
  clockevents/source: Use u64 to make 32bit happy

14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6
Linus Torvalds [Fri, 20 May 2011 20:37:49 +0000 (13:37 -0700)]
Merge git://git./linux/kernel/git/sfrench/cifs-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/sfrench/cifs-2.6: (32 commits)
  [CIFS] Fix to problem with getattr caused by invalidate simplification patch
  [CIFS] Remove sparse warning
  [CIFS] Update cifs to version 1.72
  cifs: Change key name to cifs.idmap, misc. clean-up
  cifs: Unconditionally copy mount options to superblock info
  cifs: Use kstrndup for cifs_sb->mountdata
  cifs: Simplify handling of submount options in cifs_mount.
  cifs: cifs_parse_mount_options: do not tokenize mount options in-place
  cifs: Add support for mounting Windows 2008 DFS shares
  cifs: Extract DFS referral expansion logic to separate function
  cifs: turn BCC into a static inlined function
  cifs: keep BCC in little-endian format
  cifs: fix some unused variable warnings in id_rb_search
  CIFS: Simplify invalidate part (try #5)
  CIFS: directio read/write cleanups
  consistently use smb_buf_length as be32 for cifs (try 3)
  cifs: Invoke id mapping functions (try #17 repost)
  cifs: Add idmap key and related data structures and functions (try #17 repost)
  CIFS: Add launder_page operation (try #3)
  Introduce smb2 mounts as vers=2
  ...

14 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 20 May 2011 20:37:22 +0000 (13:37 -0700)]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  extable, core_kernel_data(): Make sure all archs define _sdata
  core_kernel_data(): Fix architectures that do not define _sdata

14 years agoMerge branch 'ptrace' of git://git.kernel.org/pub/scm/linux/kernel/git/oleg/misc
Linus Torvalds [Fri, 20 May 2011 20:33:21 +0000 (13:33 -0700)]
Merge branch 'ptrace' of git://git./linux/kernel/git/oleg/misc

* 'ptrace' of git://git.kernel.org/pub/scm/linux/kernel/git/oleg/misc: (41 commits)
  signal: trivial, fix the "timespec declared inside parameter list" warning
  job control: reorganize wait_task_stopped()
  ptrace: fix signal->wait_chldexit usage in task_clear_group_stop_trapping()
  signal: sys_sigprocmask() needs retarget_shared_pending()
  signal: cleanup sys_sigprocmask()
  signal: rename signandsets() to sigandnsets()
  signal: do_sigtimedwait() needs retarget_shared_pending()
  signal: introduce do_sigtimedwait() to factor out compat/native code
  signal: sys_rt_sigtimedwait: simplify the timeout logic
  signal: cleanup sys_rt_sigprocmask()
  x86: signal: sys_rt_sigreturn() should use set_current_blocked()
  x86: signal: handle_signal() should use set_current_blocked()
  signal: sigprocmask() should do retarget_shared_pending()
  signal: sigprocmask: narrow the scope of ->siglock
  signal: retarget_shared_pending: optimize while_each_thread() loop
  signal: retarget_shared_pending: consider shared/unblocked signals only
  signal: introduce retarget_shared_pending()
  ptrace: ptrace_check_attach() should not do s/STOPPED/TRACED/
  signal: Turn SIGNAL_STOP_DEQUEUED into GROUP_STOP_DEQUEUED
  signal: do_signal_stop: Remove the unneeded task_clear_group_stop_pending()
  ...

14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6
Linus Torvalds [Fri, 20 May 2011 20:29:52 +0000 (13:29 -0700)]
Merge git://git./linux/kernel/git/jejb/scsi-misc-2.6

* git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi-misc-2.6: (110 commits)
  [SCSI] qla2xxx: Refactor call to qla2xxx_read_sfp for thermal temperature.
  [SCSI] qla2xxx: Unify the read/write sfp mailbox command routines.
  [SCSI] qla2xxx: Clear complete initialization control block.
  [SCSI] qla2xxx: Allow an override of the registered maximum LUN.
  [SCSI] qla2xxx: Add host number in reset and quiescent message logs.
  [SCSI] qla2xxx: Correctly read sfp single byte mailbox register.
  [SCSI] qla2xxx: Add qla82xx_rom_unlock() function.
  [SCSI] qla2xxx: Log if qla82xx firmware fails to load from flash.
  [SCSI] qla2xxx: Use passed in host to initialize local scsi_qla_host in queuecommand function
  [SCSI] qla2xxx: Correct buffer start in edc sysfs debug print.
  [SCSI] qla2xxx: Update firmware version after flash update for ISP82xx.
  [SCSI] qla2xxx: Fix hang during driver unload when vport is active.
  [SCSI] qla2xxx: Properly set the dsd_list_len for dsd_chaining in cmd type 6.
  [SCSI] qla2xxx: Fix virtual port failing to login after chip reset.
  [SCSI] qla2xxx: Fix vport delete hang when logins are outstanding.
  [SCSI] hpsa: Change memset using sizeof(ptr) to sizeof(*ptr)
  [SCSI] ipr: Rate limit DMA mapping errors
  [SCSI] hpsa: add P2000 to list of shared SAS devices
  [SCSI] hpsa: do not attempt PCI power management reset method if we know it won't work.
  [SCSI] hpsa: remove superfluous sleeps around reset code
  ...

14 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-2.6-nmw
Linus Torvalds [Fri, 20 May 2011 20:28:45 +0000 (13:28 -0700)]
Merge git://git./linux/kernel/git/steve/gfs2-2.6-nmw

* git://git.kernel.org/pub/scm/linux/kernel/git/steve/gfs2-2.6-nmw: (32 commits)
  GFS2: Move all locking inside the inode creation function
  GFS2: Clean up symlink creation
  GFS2: Clean up mkdir
  GFS2: Use UUID field in generic superblock
  GFS2: Rename ops_inode.c to inode.c
  GFS2: Inode.c is empty now, remove it
  GFS2: Move final part of inode.c into super.c
  GFS2: Move most of the remaining inode.c into ops_inode.c
  GFS2: Move gfs2_refresh_inode() and friends into glops.c
  GFS2: Remove gfs2_dinode_print() function
  GFS2: When adding a new dir entry, inc link count if it is a subdir
  GFS2: Make gfs2_dir_del update link count when required
  GFS2: Don't use gfs2_change_nlink in link syscall
  GFS2: Don't use a try lock when promoting to a higher mode
  GFS2: Double check link count under glock
  GFS2: Improve bug trap code in ->releasepage()
  GFS2: Fix ail list traversal
  GFS2: make sure fallocate bytes is a multiple of blksize
  GFS2: Add an AIL writeback tracepoint
  GFS2: Make writeback more responsive to system conditions
  ...

14 years agoMerge branch 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
Linus Torvalds [Fri, 20 May 2011 20:28:01 +0000 (13:28 -0700)]
Merge branch 'merge' of git://git./linux/kernel/git/benh/powerpc

* 'merge' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc: (152 commits)
  powerpc: Fix hard CPU IDs detection
  powerpc/pmac: Update via-pmu to new syscore_ops
  powerpc/kvm: Fix the build for 32-bit Book 3S (classic) processors
  powerpc/kvm: Fix kvmppc_core_pending_dec
  powerpc: Remove last piece of GEMINI
  powerpc: Fix for Pegasos keyboard and mouse
  powerpc: Make early memory scan more resilient to out of order nodes
  powerpc/pseries/iommu: Cleanup ddw naming
  powerpc/pseries/iommu: Find windows after kexec during boot
  powerpc/pseries/iommu: Remove ddw property when destroying window
  powerpc/pseries/iommu: Add additional checks when changing iommu mask
  powerpc/pseries/iommu: Use correct return type in dupe_ddw_if_already_created
  powerpc: Remove unused/obsolete CONFIG_XICS
  misc: Add CARMA DATA-FPGA Programmer support
  misc: Add CARMA DATA-FPGA Access Driver
  powerpc: Make IRQ_NOREQUEST last to clear, first to set
  powerpc: Integrated Flash controller device tree bindings
  powerpc/85xx: Create dts of each core in CAMP mode for P1020RDB
  powerpc/85xx: Fix PCIe IDSEL for Px020RDB
  powerpc/85xx: P2020 DTS: re-organize dts files
  ...

14 years agospinlock_up.h: include asm/processor.h in for cpu_relax
Stephen Rothwell [Fri, 20 May 2011 05:48:17 +0000 (15:48 +1000)]
spinlock_up.h: include asm/processor.h in for cpu_relax

Commit e66eed651fd1 ("list: remove prefetching from regular list
iterators") removed the include of prefetch.h from list.h and this was a
path to including asm/processor.h.  We need to include it excplicitly
now.

Fixes this build error on sparc32 (at least):

  In file included from include/linux/seqlock.h:29,
                   from include/linux/time.h:8,
                   from include/linux/timex.h:56,
                   from include/linux/sched.h:57,
                   from arch/sparc/kernel/asm-offsets.c:13:
  include/linux/spinlock.h: In function 'spin_unlock_wait':
  include/linux/spinlock.h:360: error: implicit declaration of function 'cpu_relax

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agosanitize <linux/prefetch.h> usage
Linus Torvalds [Fri, 20 May 2011 19:50:29 +0000 (12:50 -0700)]
sanitize <linux/prefetch.h> usage

Commit e66eed651fd1 ("list: remove prefetching from regular list
iterators") removed the include of prefetch.h from list.h, which
uncovered several cases that had apparently relied on that rather
obscure header file dependency.

So this fixes things up a bit, using

   grep -L linux/prefetch.h $(git grep -l '[^a-z_]prefetchw*(' -- '*.[ch]')
   grep -L 'prefetchw*(' $(git grep -l 'linux/prefetch.h' -- '*.[ch]')

to guide us in finding files that either need <linux/prefetch.h>
inclusion, or have it despite not needing it.

There are more of them around (mostly network drivers), but this gets
many core ones.

Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agomacvlan: fix panic if lowerdev in a bond
Eric Dumazet [Fri, 20 May 2011 18:59:23 +0000 (14:59 -0400)]
macvlan: fix panic if lowerdev in a bond

commit a35e2c1b6d905 (macvlan: use rx_handler_data pointer to store
macvlan_port pointer V2) added a bug in macvlan_port_create()

Steps to reproduce the bug:

# ifenslave bond0 eth0 eth1

# ip link add link eth0 up name eth0#1 type macvlan
->error EBUSY

# ip link add link eth0 up name eth0#1 type macvlan
->panic

Fix: Dont set IFF_MACVLAN_PORT in error case.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoMerge branch 'timers/urgent' into timers/core
Thomas Gleixner [Fri, 20 May 2011 18:06:24 +0000 (20:06 +0200)]
Merge branch 'timers/urgent' into timers/core

Reason: Get upstream fixes and kfree_rcu which is necessary for a
follow up patch.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
14 years ago[IA64] define "_sdata" symbol
Tony Luck [Fri, 20 May 2011 17:38:53 +0000 (10:38 -0700)]
[IA64] define "_sdata" symbol

core_kernel_data() wants to know if an address looks like kernel
data. IA64 has had _edata forever, but never needed _sdata until
now.

Signed-off-by: Tony Luck <tony.luck@intel.com>
14 years agoPull misc-2.6.40 into release branch
Tony Luck [Fri, 20 May 2011 17:35:03 +0000 (10:35 -0700)]
Pull misc-2.6.40 into release branch

14 years agoPull pstore into release branch
Tony Luck [Fri, 20 May 2011 17:34:50 +0000 (10:34 -0700)]
Pull pstore into release branch

14 years agopstore: Fix Kconfig dependencies for apei->pstore
Luck, Tony [Wed, 13 Apr 2011 17:48:12 +0000 (10:48 -0700)]
pstore: Fix Kconfig dependencies for apei->pstore

Geert Uytterhoeven ran a dependency checker which kicked out this warning:

+ warning: (ACPI_APEI) selects PSTORE which has unmet direct dependencies (MISC_FILESYSTEMS):  => N/A

Randy confirmed that the fix was to "select MISC_FILESYSTEMS" too.

Tested-by: Randy Dunlap <randy.dunlap@oracle.com>
Signed-off-by: Tony Luck <tony.luck@intel.com>
14 years ago[CIFS] Fix to problem with getattr caused by invalidate simplification patch
Steve French [Fri, 20 May 2011 17:00:01 +0000 (17:00 +0000)]
[CIFS] Fix to problem with getattr caused by invalidate simplification patch

Fix to earlier "Simplify invalidate part (try #6)" patch
That patch caused problems with connectathon test 5.

Reviewed-by: Jeff Layton <jlayton@samba.org>
Signed-off-by: Pavel Shilovsky <piastry@etersoft.ru>
Signed-off-by: Steve French <sfrench@us.ibm.com>
14 years agosched: Increase SCHED_LOAD_SCALE resolution
Nikhil Rao [Wed, 18 May 2011 21:37:48 +0000 (14:37 -0700)]
sched: Increase SCHED_LOAD_SCALE resolution

Introduce SCHED_LOAD_RESOLUTION, which scales is added to
SCHED_LOAD_SHIFT and increases the resolution of
SCHED_LOAD_SCALE. This patch sets the value of
SCHED_LOAD_RESOLUTION to 10, scaling up the weights for all
sched entities by a factor of 1024. With this extra resolution,
we can handle deeper cgroup hiearchies and the scheduler can do
better shares distribution and load load balancing on larger
systems (especially for low weight task groups).

This does not change the existing user interface, the scaled
weights are only used internally. We do not modify
prio_to_weight values or inverses, but use the original weights
when calculating the inverse which is used to scale execution
time delta in calc_delta_mine(). This ensures we do not lose
accuracy when accounting time to the sched entities. Thanks to
Nikunj Dadhania for fixing an bug in c_d_m() that broken fairness.

Below is some analysis of the performance costs/improvements of
this patch.

1. Micro-arch performance costs:

Experiment was to run Ingo's pipe_test_100k 200 times with the
task pinned to one cpu. I measured instruction, cycles and
stalled-cycles for the runs. See:

   http://thread.gmane.org/gmane.linux.kernel/1129232/focus=1129389

for more info.

-tip (baseline):

 Performance counter stats for '/root/load-scale/pipe-test-100k' (200 runs):

       964,991,769 instructions             #    0.82  insns per cycle
                                            #    0.33  stalled cycles per insn
                                            #    ( +-  0.05% )
     1,171,186,635 cycles                   #    0.000 GHz                      ( +-  0.08% )
       306,373,664 stalled-cycles-backend   #   26.16% backend  cycles idle     ( +-  0.28% )
       314,933,621 stalled-cycles-frontend  #   26.89% frontend cycles idle     ( +-  0.34% )

        1.122405684  seconds time elapsed  ( +-  0.05% )

-tip+patches:

 Performance counter stats for './load-scale/pipe-test-100k' (200 runs):

       963,624,821 instructions             #    0.82  insns per cycle
                                            #    0.33  stalled cycles per insn
                                            #    ( +-  0.04% )
     1,175,215,649 cycles                   #    0.000 GHz                      ( +-  0.08% )
       315,321,126 stalled-cycles-backend   #   26.83% backend  cycles idle     ( +-  0.28% )
       316,835,873 stalled-cycles-frontend  #   26.96% frontend cycles idle     ( +-  0.29% )

        1.122238659  seconds time elapsed  ( +-  0.06% )

With this patch, instructions decrease by ~0.10% and cycles
increase by 0.27%. This doesn't look statistically significant.
The number of stalled cycles in the backend increased from
26.16% to 26.83%. This can be attributed to the shifts we do in
c_d_m() and other places. The fraction of stalled cycles in the
frontend remains about the same, at 26.96% compared to 26.89% in -tip.

2. Balancing low-weight task groups

Test setup: run 50 tasks with random sleep/busy times (biased
around 100ms) in a low weight container (with cpu.shares = 2).
Measure %idle as reported by mpstat over a 10s window.

-tip (baseline):

06:47:48 PM  CPU    %usr   %nice    %sys %iowait    %irq   %soft  %steal  %guest   %idle    intr/s
06:47:49 PM  all   94.32    0.00    0.06    0.00    0.00    0.00    0.00    0.00    5.62  15888.00
06:47:50 PM  all   94.57    0.00    0.62    0.00    0.00    0.00    0.00    0.00    4.81  16180.00
06:47:51 PM  all   94.69    0.00    0.06    0.00    0.00    0.00    0.00    0.00    5.25  15966.00
06:47:52 PM  all   95.81    0.00    0.00    0.00    0.00    0.00    0.00    0.00    4.19  16053.00
06:47:53 PM  all   94.88    0.06    0.00    0.00    0.00    0.00    0.00    0.00    5.06  15984.00
06:47:54 PM  all   93.31    0.00    0.00    0.00    0.00    0.00    0.00    0.00    6.69  15806.00
06:47:55 PM  all   94.19    0.00    0.06    0.00    0.00    0.00    0.00    0.00    5.75  15896.00
06:47:56 PM  all   92.87    0.00    0.00    0.00    0.00    0.00    0.00    0.00    7.13  15716.00
06:47:57 PM  all   94.88    0.00    0.00    0.00    0.00    0.00    0.00    0.00    5.12  15982.00
06:47:58 PM  all   95.44    0.00    0.00    0.00    0.00    0.00    0.00    0.00    4.56  16075.00
Average:     all   94.49    0.01    0.08    0.00    0.00    0.00    0.00    0.00    5.42  15954.60

-tip+patches:

06:47:03 PM  CPU    %usr   %nice    %sys %iowait    %irq   %soft  %steal  %guest   %idle    intr/s
06:47:04 PM  all  100.00    0.00    0.00    0.00    0.00    0.00    0.00    0.00    0.00  16630.00
06:47:05 PM  all   99.69    0.00    0.00    0.00    0.00    0.00    0.00    0.00    0.31  16580.20
06:47:06 PM  all   99.69    0.00    0.06    0.00    0.00    0.00    0.00    0.00    0.25  16596.00
06:47:07 PM  all   99.20    0.00    0.74    0.00    0.00    0.06    0.00    0.00    0.00  17838.61
06:47:08 PM  all  100.00    0.00    0.00    0.00    0.00    0.00    0.00    0.00    0.00  16540.00
06:47:09 PM  all  100.00    0.00    0.00    0.00    0.00    0.00    0.00    0.00    0.00  16575.00
06:47:10 PM  all  100.00    0.00    0.00    0.00    0.00    0.00    0.00    0.00    0.00  16614.00
06:47:11 PM  all   99.94    0.00    0.00    0.00    0.00    0.00    0.00    0.00    0.06  16588.00
06:47:12 PM  all   99.94    0.00    0.06    0.00    0.00    0.00    0.00    0.00    0.00  16593.00
06:47:13 PM  all   99.94    0.00    0.06    0.00    0.00    0.00    0.00    0.00    0.00  16551.00
Average:     all   99.84    0.00    0.09    0.00    0.00    0.01    0.00    0.00    0.06  16711.58

We see an improvement in idle% on the system (drops from 5.42% on -tip to 0.06%
with the patches).

We see an improvement in idle% on the system (drops from 5.42%
on -tip to 0.06% with the patches).

Signed-off-by: Nikhil Rao <ncrao@google.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Nikunj A. Dadhania <nikunj@linux.vnet.ibm.com>
Cc: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Cc: Stephan Barwolf <stephan.baerwolf@tu-ilmenau.de>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Link: http://lkml.kernel.org/r/1305754668-18792-1-git-send-email-ncrao@google.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agosched: Introduce SCHED_POWER_SCALE to scale cpu_power calculations
Nikhil Rao [Wed, 18 May 2011 17:09:39 +0000 (10:09 -0700)]
sched: Introduce SCHED_POWER_SCALE to scale cpu_power calculations

SCHED_LOAD_SCALE is used to increase nice resolution and to
scale cpu_power calculations in the scheduler. This patch
introduces SCHED_POWER_SCALE and converts all uses of
SCHED_LOAD_SCALE for scaling cpu_power to use SCHED_POWER_SCALE
instead.

This is a preparatory patch for increasing the resolution of
SCHED_LOAD_SCALE, and there is no need to increase resolution
for cpu_power calculations.

Signed-off-by: Nikhil Rao <ncrao@google.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Nikunj A. Dadhania <nikunj@linux.vnet.ibm.com>
Cc: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Cc: Stephan Barwolf <stephan.baerwolf@tu-ilmenau.de>
Cc: Mike Galbraith <efault@gmx.de>
Link: http://lkml.kernel.org/r/1305738580-9924-3-git-send-email-ncrao@google.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agosched: Cleanup set_load_weight()
Nikhil Rao [Wed, 18 May 2011 17:09:38 +0000 (10:09 -0700)]
sched: Cleanup set_load_weight()

Avoid using long repetitious names; make this simpler and nicer
to read. No functional change introduced in this patch.

Signed-off-by: Nikhil Rao <ncrao@google.com>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Cc: Nikunj A. Dadhania <nikunj@linux.vnet.ibm.com>
Cc: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Cc: Stephan Barwolf <stephan.baerwolf@tu-ilmenau.de>
Cc: Mike Galbraith <efault@gmx.de>
Link: http://lkml.kernel.org/r/1305738580-9924-2-git-send-email-ncrao@google.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, x2apic: Move the common bits to x2apic.h
Cyrill Gorcunov [Thu, 19 May 2011 23:45:50 +0000 (16:45 -0700)]
x86, x2apic: Move the common bits to x2apic.h

To eliminate code duplication.

Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: steiner@sgi.com
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110519234637.591426753@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, x2apic: Minimize IPI register writes using cluster groups
Cyrill Gorcunov [Thu, 19 May 2011 23:45:49 +0000 (16:45 -0700)]
x86, x2apic: Minimize IPI register writes using cluster groups

In the case of x2apic cluster mode we can group IPI register
writes based on the cluster group instead of individual per-cpu
destination messages.

This reduces the apic register writes and reduces the amount of
IPI messages (in the best case we can reduce it by a factor of
16).

With this change, the cost of flush_tlb_others(), with the flush
tlb IPI being sent from a cpu in the socket-1 to all the logical
cpus in socket-2 (on a Westmere-EX system that has 20 logical
cpus in a socket) is 3x times better now (compared to the former
'send one-by-one' algorithm).

Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: steiner@sgi.com
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110519234637.512271057@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, x2apic: Track the x2apic cluster sibling map
Cyrill Gorcunov [Thu, 19 May 2011 23:45:48 +0000 (16:45 -0700)]
x86, x2apic: Track the x2apic cluster sibling map

In the case of x2apic cluster mode, we can group IPI register
writes based on the cluster group instead of individual per-cpu
destination messages.

For this purpose, track the cpu's that belong to the same x2apic
cluster.

Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: steiner@sgi.com
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110519234637.421800999@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, x2apic: Remove duplicate code for IPI mask routines
Suresh Siddha [Thu, 19 May 2011 23:45:47 +0000 (16:45 -0700)]
x86, x2apic: Remove duplicate code for IPI mask routines

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Acked-by: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: steiner@sgi.com
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110519234637.337024125@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, apic: Use probe routines to simplify apic selection
Suresh Siddha [Thu, 19 May 2011 23:45:46 +0000 (16:45 -0700)]
x86, apic: Use probe routines to simplify apic selection

Use the unused probe routine in the apic driver to finalize the
apic model selection. This cleans up the
default_setup_apic_routing() and this probe routine in future
can also be used for doing any apic model specific
initialisation.

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Acked-by: Cyrill Gorcunov <gorcunov@openvz.org>
Cc: steiner@sgi.com
Cc: yinghai@kernel.org
Link: http://lkml.kernel.org/r/20110519234637.247458931@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, ioapic: Consolidate mp_ioapic_routing[] into 'struct ioapic'
Suresh Siddha [Wed, 18 May 2011 23:31:39 +0000 (16:31 -0700)]
x86, ioapic: Consolidate mp_ioapic_routing[] into 'struct ioapic'

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: daniel.blueman@gmail.com
Link: http://lkml.kernel.org/r/20110518233158.089978277@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, ioapic: Consolidate gsi routing info into 'struct ioapic'
Suresh Siddha [Wed, 18 May 2011 23:31:38 +0000 (16:31 -0700)]
x86, ioapic: Consolidate gsi routing info into 'struct ioapic'

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: daniel.blueman@gmail.com
Link: http://lkml.kernel.org/r/20110518233157.994002011@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, ioapic: Consolidate mp_ioapics[] into 'struct ioapic'
Suresh Siddha [Wed, 18 May 2011 23:31:37 +0000 (16:31 -0700)]
x86, ioapic: Consolidate mp_ioapics[] into 'struct ioapic'

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: daniel.blueman@gmail.com
Link: http://lkml.kernel.org/r/20110518233157.909013179@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, ioapic: Consolidate ioapic_saved_data[] into 'struct ioapic'
Suresh Siddha [Wed, 18 May 2011 23:31:36 +0000 (16:31 -0700)]
x86, ioapic: Consolidate ioapic_saved_data[] into 'struct ioapic'

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: daniel.blueman@gmail.com
Link: http://lkml.kernel.org/r/20110518233157.830697056@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, ioapic: Add struct ioapic
Suresh Siddha [Wed, 18 May 2011 23:31:35 +0000 (16:31 -0700)]
x86, ioapic: Add struct ioapic

Introduce struct ioapic with nr_registers field.

This will pave way for consolidating different MAX_IO_APICS
arrays into it.

Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: daniel.blueman@gmail.com
Link: http://lkml.kernel.org/r/20110518233157.744315519@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, ioapic: Remove duplicate code for saving/restoring RTEs
Suresh Siddha [Wed, 18 May 2011 23:31:34 +0000 (16:31 -0700)]
x86, ioapic: Remove duplicate code for saving/restoring RTEs

Code flow for enabling interrupt-remapping has its own routines
for saving and restoring io-apic RTE's. ioapic suspend/resume
code flow also has similar routines. Remove the duplicate code.

Tested-by: Daniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/20110518233157.673130611@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, ioapic: Use ioapic_saved_data while enabling intr-remapping
Suresh Siddha [Wed, 18 May 2011 23:31:33 +0000 (16:31 -0700)]
x86, ioapic: Use ioapic_saved_data while enabling intr-remapping

Code flow for enabling interrupt-remapping was
allocating/freeing buffers for saving/restoring io-apic RTE's.
ioapic suspend/resume code uses boot time allocated
ioapic_saved_data that is a perfect match for reuse here.

This will remove the unnecessary allocation/free of the
temporary buffers during suspend/resume of interrupt-remapping
enabled platforms aswell as paving the way for further code
consolidation.

Tested-by: Daniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/20110518233157.574469296@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, ioapic: Allocate ioapic_saved_data early
Suresh Siddha [Wed, 18 May 2011 23:31:32 +0000 (16:31 -0700)]
x86, ioapic: Allocate ioapic_saved_data early

This allows re-using this buffer for enabling
interrupt-remapping during boot and resume. And thus allow for
consolidating the code between ioapic suspend/resume and
interrupt-remapping.

Tested-by: Daniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Link: http://lkml.kernel.org/r/20110518233157.481404505@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, ioapic: Fix potential resume deadlock
Daniel J Blueman [Wed, 18 May 2011 23:31:31 +0000 (16:31 -0700)]
x86, ioapic: Fix potential resume deadlock

Fix a potential deadlock when resuming; here the calling
function has disabled interrupts, so we cannot sleep.

Change the memory allocation flag from GFP_KERNEL to GFP_ATOMIC.

TODO: We can do away with this memory allocation during resume
      by reusing the ioapic suspend/resume code that uses boot time
      allocated buffers, but we want to keep this -stable patch
      simple.

Signed-off-by: Daniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
Cc: <stable@kernel.org> # v2.6.38/39
Link: http://lkml.kernel.org/r/20110518233157.385970138@sbsiddha-MOBL3.sc.intel.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agox86, amd: Use _safe() msr access for GartTlbWlk disable code
Roedel, Joerg [Thu, 19 May 2011 09:13:39 +0000 (11:13 +0200)]
x86, amd: Use _safe() msr access for GartTlbWlk disable code

The workaround for Bugzilla:

https://bugzilla.kernel.org/show_bug.cgi?id=33012

introduced a read and a write to the MC4 mask msr.

Unfortunatly this MSR is not emulated by the KVM hypervisor
so that the kernel will get a #GP and crashes when applying
this workaround when running inside KVM.

This issue was reported as:

https://bugzilla.kernel.org/show_bug.cgi?id=35132

and is fixed with this patch. The change just let the kernel
ignore any #GP it gets while accessing this MSR by using the
_safe msr access methods.

Reported-by: Török Edwin <edwintorok@gmail.com>
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
Cc: Rafael J. Wysocki <rjw@sisk.pl>
Cc: Maciej Rutecki <maciej.rutecki@gmail.com>
Cc: Avi Kivity <avi@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: <stable@kernel.org> # .39.x
Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agoMAINTAINERS: Add drivers/clocksource to TIMEKEEPING
Thomas Gleixner [Fri, 20 May 2011 09:38:24 +0000 (11:38 +0200)]
MAINTAINERS: Add drivers/clocksource to TIMEKEEPING

It's not a random dump ground and we care about it.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Ralf Baechle <ralf@linux-mips.org>
14 years agoclockevents/source: Use u64 to make 32bit happy
Thomas Gleixner [Fri, 20 May 2011 08:50:52 +0000 (10:50 +0200)]
clockevents/source: Use u64 to make 32bit happy

unsigned long is not 64bit on 32bit machine.

Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
14 years agopowerpc: Fix hard CPU IDs detection
Benjamin Herrenschmidt [Fri, 20 May 2011 07:50:18 +0000 (17:50 +1000)]
powerpc: Fix hard CPU IDs detection

commit 9d07bc841c9779b4d7902e417f4e509996ce805d
"powerpc: Properly handshake CPUs going out of boot spin loop"

Would cause a miscalculation of the hard CPU ID. It removes breaking
out of the loop when finding a match with a processor, thus the "i"
used as an index in the intserv array is always incorrect

This broke interrupt on my PowerMac laptop.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
14 years agoextable, core_kernel_data(): Make sure all archs define _sdata
Steven Rostedt [Fri, 20 May 2011 01:34:58 +0000 (21:34 -0400)]
extable, core_kernel_data(): Make sure all archs define _sdata

A new utility function (core_kernel_data()) is used to determine if a
passed in address is part of core kernel data or not. It may or may not
return true for RO data, but this utility must work for RW data.

Thus both _sdata and _edata must be defined and continuous,
without .init sections that may later be freed and replaced by
volatile memory (memory that can be freed).

This utility function is used to determine if data is safe from
ever being freed. Thus it should return true for all RW global
data that is not in a module or has been allocated, or false
otherwise.

Also change core_kernel_data() back to the more precise _sdata condition
and document the function.

Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Acked-by: Ralf Baechle <ralf@linux-mips.org>
Acked-by: Hirokazu Takata <takata@linux-m32r.org>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Roman Zippel <zippel@linux-m68k.org>
Cc: linux-m68k@lists.linux-m68k.org
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: Helge Deller <deller@gmx.de>
Cc: JamesE.J.Bottomley <jejb@parisc-linux.org>
Link: http://lkml.kernel.org/r/1305855298.1465.19.camel@gandalf.stny.rr.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
----
 arch/alpha/kernel/vmlinux.lds.S   |    1 +
 arch/m32r/kernel/vmlinux.lds.S    |    1 +
 arch/m68k/kernel/vmlinux-std.lds  |    2 ++
 arch/m68k/kernel/vmlinux-sun3.lds |    1 +
 arch/mips/kernel/vmlinux.lds.S    |    1 +
 arch/parisc/kernel/vmlinux.lds.S  |    3 +++
 kernel/extable.c                  |   12 +++++++++++-
 7 files changed, 20 insertions(+), 1 deletion(-)

14 years agoMerge branch 'perf/core' into perf/urgent
Ingo Molnar [Fri, 20 May 2011 06:54:05 +0000 (08:54 +0200)]
Merge branch 'perf/core' into perf/urgent

Merge reason: One pending commit was left in perf/core after Linus merged
              perf/core - continue v2.6.40 work in the perf/urgent reason.

Signed-off-by: Ingo Molnar <mingo@elte.hu>
14 years agopowerpc/pmac: Update via-pmu to new syscore_ops
Benjamin Herrenschmidt [Fri, 20 May 2011 05:37:22 +0000 (15:37 +1000)]
powerpc/pmac: Update via-pmu to new syscore_ops

This was left as a sysdev, breaking the build

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
14 years agoMerge remote branch 'origin/master' into merge
Benjamin Herrenschmidt [Fri, 20 May 2011 05:36:52 +0000 (15:36 +1000)]
Merge remote branch 'origin/master' into merge

Manual merge of arch/powerpc/kernel/smp.c and add missing scheduler_ipi()
call to arch/powerpc/platforms/cell/interrupt.c

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
14 years agotg3: Add braces around 5906 workaround.
Matt Carlson [Thu, 19 May 2011 16:02:44 +0000 (16:02 +0000)]
tg3: Add braces around 5906 workaround.

Commit dabc5c670d3f86d15ee4f42ab38ec5bd2682487d, entitled
"tg3: Move TSO_CAPABLE assignment", moved some TSO flagging code around.
In the process it failed to add braces around an exceptional 5906
condition.  This patch fixes the problem.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agotg3: Fix NETIF_F_LOOPBACK error
Matt Carlson [Thu, 19 May 2011 16:02:43 +0000 (16:02 +0000)]
tg3: Fix NETIF_F_LOOPBACK error

Mahesh Bandewar noticed that the features cleanup in commit
0da0606f493c5cdab74bdcc96b12f4305ad94085, entitled
"tg3: Consolidate all netdev feature assignments", mistakenly sets
NETIF_F_LOOPBACK by default.  This patch corrects the error.

Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Signed-off-by: Mahesh Bandewar <maheshb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agomacvlan: remove one synchronize_rcu() call
Eric Dumazet [Thu, 19 May 2011 12:24:16 +0000 (12:24 +0000)]
macvlan: remove one synchronize_rcu() call

When one macvlan device is dismantled, we can avoid one
synchronize_rcu() call done after deletion from hash list, since caller
will perform a synchronize_net() call after its ndo_stop() call.

Add a new netdev->dismantle field to signal this dismantle intent.

Reduces RTNL hold time.

Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Patrick McHardy <kaber@trash.net>
CC: Ben Greear <greearb@candelatech.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
14 years agoselinux: avoid unnecessary avc cache stat hit count
Linus Torvalds [Fri, 20 May 2011 04:22:53 +0000 (21:22 -0700)]
selinux: avoid unnecessary avc cache stat hit count

There is no point in counting hits - we can calculate it from the number
of lookups and misses.

This makes the avc statistics a bit smaller, and makes the code
generation better too.

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoMerge remote branch 'kumar/next' into next
Benjamin Herrenschmidt [Fri, 20 May 2011 03:43:47 +0000 (13:43 +1000)]
Merge remote branch 'kumar/next' into next

14 years agopowerpc/kvm: Fix the build for 32-bit Book 3S (classic) processors
Paul Mackerras [Wed, 11 May 2011 00:39:50 +0000 (00:39 +0000)]
powerpc/kvm: Fix the build for 32-bit Book 3S (classic) processors

Commits a5d4f3ad3a ("powerpc: Base support for exceptions using
HSRR0/1") and 673b189a2e ("powerpc: Always use SPRN_SPRG_HSCRATCH0
when running in HV mode") cause compile and link errors for 32-bit
classic Book 3S processors when KVM is enabled.  This fixes these
errors.

Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
14 years agopowerpc/kvm: Fix kvmppc_core_pending_dec
Paul Mackerras [Wed, 11 May 2011 00:38:50 +0000 (00:38 +0000)]
powerpc/kvm: Fix kvmppc_core_pending_dec

The vcpu->arch.pending_exceptions field is a bitfield indexed by
interrupt priority number as returned by kvmppc_book3s_vec2irqprio.
However, kvmppc_core_pending_dec was using an interrupt vector shifted
by 7 as the bit index.  Fix it to use the irqprio value for the
decrementer interrupt instead.  This problem was found by code
inspection.

Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
14 years agoselinux: de-crapify avc cache stat code generation
Linus Torvalds [Fri, 20 May 2011 01:59:47 +0000 (18:59 -0700)]
selinux: de-crapify avc cache stat code generation

You can turn off the avc cache stats, but distributions seem to not do
that (perhaps because several performance tuning how-to's talk about the
avc cache statistics).

Which is sad, because the code it generates is truly horrendous, with
the statistics update being sandwitched between get_cpu/put_cpu which in
turn causes preemption disables etc.  We're talking ten+ instructions
just to increment a per-cpu variable in some pretty hot code.

Fix the craziness by just using 'this_cpu_inc()' instead.  Suddenly we
only need a single 'inc' instruction to increment the statistics.  This
is quite noticeable in the incredibly hot avc_has_perm_noaudit()
function (which triggers all the statistics by virtue of doing an
avc_lookup() call).

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoMerge branch 'driver-core-next' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 20 May 2011 01:24:11 +0000 (18:24 -0700)]
Merge branch 'driver-core-next' of git://git./linux/kernel/git/gregkh/driver-core-2.6

* 'driver-core-next' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/driver-core-2.6: (44 commits)
  debugfs: Silence DEBUG_STRICT_USER_COPY_CHECKS=y warning
  sysfs: remove "last sysfs file:" line from the oops messages
  drivers/base/memory.c: fix warning due to "memory hotplug: Speed up add/remove when blocks are larger than PAGES_PER_SECTION"
  memory hotplug: Speed up add/remove when blocks are larger than PAGES_PER_SECTION
  SYSFS: Fix erroneous comments for sysfs_update_group().
  driver core: remove the driver-model structures from the documentation
  driver core: Add the device driver-model structures to kerneldoc
  Translated Documentation/email-clients.txt
  RAW driver: Remove call to kobject_put().
  reboot: disable usermodehelper to prevent fs access
  efivars: prevent oops on unload when efi is not enabled
  Allow setting of number of raw devices as a module parameter
  Introduce CONFIG_GOOGLE_FIRMWARE
  driver: Google Memory Console
  driver: Google EFI SMI
  x86: Better comments for get_bios_ebda()
  x86: get_bios_ebda_length()
  misc: fix ti-st build issues
  params.c: Use new strtobool function to process boolean inputs
  debugfs: move to new strtobool
  ...

Fix up trivial conflicts in fs/debugfs/file.c due to the same patch
being applied twice, and an unrelated cleanup nearby.

14 years agosignal.h need a definition of struct task_struct
Stephen Rothwell [Fri, 20 May 2011 01:11:53 +0000 (11:11 +1000)]
signal.h need a definition of struct task_struct

This fixes these build errors on powerpc:

  In file included from arch/powerpc/mm/fault.c:18:
  include/linux/signal.h:239: error: 'struct task_struct' declared inside parameter list
  include/linux/signal.h:239: error: its scope is only this definition or declaration, which is probably not what you want
  include/linux/signal.h:240: error: 'struct task_struct' declared inside parameter list
  ..

Exposed by commit e66eed651fd1 ("list: remove prefetching from regular
list iterators"), which removed the include of <linux/prefetch.h> from
<linux/list.h>.

Without that, linux/signal.h no longer accidentally got the declaration
of 'struct task_struct'.

Fix by properly declaring the struct, rather than introducing any new
header file dependency.

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
14 years agoMerge branch 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 20 May 2011 01:14:34 +0000 (18:14 -0700)]
Merge branch 'core-rcu-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (78 commits)
  Revert "rcu: Decrease memory-barrier usage based on semi-formal proof"
  net,rcu: convert call_rcu(prl_entry_destroy_rcu) to kfree
  batman,rcu: convert call_rcu(softif_neigh_free_rcu) to kfree_rcu
  batman,rcu: convert call_rcu(neigh_node_free_rcu) to kfree()
  batman,rcu: convert call_rcu(gw_node_free_rcu) to kfree_rcu
  net,rcu: convert call_rcu(kfree_tid_tx) to kfree_rcu()
  net,rcu: convert call_rcu(xt_osf_finger_free_rcu) to kfree_rcu()
  net/mac80211,rcu: convert call_rcu(work_free_rcu) to kfree_rcu()
  net,rcu: convert call_rcu(wq_free_rcu) to kfree_rcu()
  net,rcu: convert call_rcu(phonet_device_rcu_free) to kfree_rcu()
  perf,rcu: convert call_rcu(swevent_hlist_release_rcu) to kfree_rcu()
  perf,rcu: convert call_rcu(free_ctx) to kfree_rcu()
  net,rcu: convert call_rcu(__nf_ct_ext_free_rcu) to kfree_rcu()
  net,rcu: convert call_rcu(net_generic_release) to kfree_rcu()
  net,rcu: convert call_rcu(netlbl_unlhsh_free_addr6) to kfree_rcu()
  net,rcu: convert call_rcu(netlbl_unlhsh_free_addr4) to kfree_rcu()
  security,rcu: convert call_rcu(sel_netif_free) to kfree_rcu()
  net,rcu: convert call_rcu(xps_dev_maps_release) to kfree_rcu()
  net,rcu: convert call_rcu(xps_map_release) to kfree_rcu()
  net,rcu: convert call_rcu(rps_map_release) to kfree_rcu()
  ...

14 years agoMerge branch 'x86-smep-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 20 May 2011 01:10:17 +0000 (18:10 -0700)]
Merge branch 'x86-smep-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'x86-smep-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86, cpu: Enable/disable Supervisor Mode Execution Protection
  x86, cpu: Add SMEP CPU feature in CR4
  x86, cpufeature: Add cpufeature flag for SMEP

14 years agoMerge branches 'x86-reboot-for-linus' and 'x86-setup-for-linus' of git://git.kernel...
Linus Torvalds [Fri, 20 May 2011 01:09:45 +0000 (18:09 -0700)]
Merge branches 'x86-reboot-for-linus' and 'x86-setup-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'x86-reboot-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86: Reorder reboot method preferences

* 'x86-setup-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86, setup: Fix EDD3.0 data verification.

14 years agoMerge branch 'x86-platform-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Fri, 20 May 2011 01:08:06 +0000 (18:08 -0700)]
Merge branch 'x86-platform-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'x86-platform-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  x86: Introduce pci_map_biosrom()
  x86, olpc: Use device tree for platform identification

14 years agoMerge branch 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git...
Linus Torvalds [Fri, 20 May 2011 01:07:31 +0000 (18:07 -0700)]
Merge branch 'x86-mm-for-linus' of git://git./linux/kernel/git/tip/linux-2.6-tip

* 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (50 commits)
  x86, mm: Allow ZONE_DMA to be configurable
  x86, NUMA: Trim numa meminfo with max_pfn in a separate loop
  x86, NUMA: Rename setup_node_bootmem() to setup_node_data()
  x86, NUMA: Enable emulation on 32bit too
  x86, NUMA: Enable CONFIG_AMD_NUMA on 32bit too
  x86, NUMA: Rename amdtopology_64.c to amdtopology.c
  x86, NUMA: Make numa_init_array() static
  x86, NUMA: Make 32bit use common NUMA init path
  x86, NUMA: Initialize and use remap allocator from setup_node_bootmem()
  x86-32, NUMA: Add @start and @end to init_alloc_remap()
  x86, NUMA: Remove long 64bit assumption from numa.c
  x86, NUMA: Enable build of generic NUMA init code on 32bit
  x86, NUMA: Move NUMA init logic from numa_64.c to numa.c
  x86-32, NUMA: Update numaq to use new NUMA init protocol
  x86-32, NUMA: Replace srat_32.c with srat.c
  x86-32, NUMA: implement temporary NUMA init shims
  x86, NUMA: Move numa_nodes_parsed to numa.[hc]
  x86-32, NUMA: Move get_memcfg_numa() into numa_32.c
  x86, NUMA: make srat.c 32bit safe
  x86, NUMA: rename srat_64.c to srat.c
  ...