The function is_exported() with its helper function lookup_symbol() are used to
verify if a provided symbol is effectively exported by the kernel or by the
modules. Now that both have their symbols sorted we can replace a linear search
with a binary search which provide a considerably speed-up.
This work was supported by a hardware donation from the CE Linux Forum.
Signed-off-by: Alessio Igor Bogani <abogani@kernel.org>
Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Takes advantage of the order and locates symbols using binary search.
This work was supported by a hardware donation from the CE Linux Forum.
Signed-off-by: Alessio Igor Bogani <abogani@kernel.org>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Tested-by: Dirk Behme <dirk.behme@googlemail.com>
Instead of having a callback function for each symbol in the kernel,
have a callback for each array of symbols.
This eases the logic when we move to sorted symbols and binary search.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Alessio Igor Bogani <abogani@kernel.org>
Split the unprotect function into a function per section to make
the code more readable and add the missing static declaration.
Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
While debugging I stumbled over two problems in the code that protects module
pages.
First issue is that disabling the protection before freeing init or unload of
a module is not symmetric with the enablement. For instance, if pages are set
to RO the page range from module_core to module_core + core_ro_size is
protected. If a module is unloaded the page range from module_core to
module_core + core_size is set back to RW.
So pages that were not set to RO are also changed to RW.
This is not critical but IMHO it should be symmetric.
Second issue is that while set_memory_rw & set_memory_ro are used for
RO/RW changes only set_memory_nx is involved for NX/X. One would await that
the inverse function is called when the NX protection should be removed,
which is not the case here, unless I'm missing something.
Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Reset mod->init_ro_size to zero after the init part of a module is unloaded.
Otherwise we need to check if module->init is NULL in the unprotect functions
in the next patch.
Signed-off-by: Jan Glauber <jang@linux.vnet.ibm.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Fix function prototype to be ANSI-C compliant, consistent with other
function prototypes, addressing a sparse warning.
Signed-off-by: Daniel J Blueman <daniel.blueman@gmail.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
On m68k natural alignment is 2-byte boundary but we are trying to
align structures in __modver section on sizeof(void *) boundary.
This causes trouble when we try to access elements in this section
in array-like fashion when create "version" attributes for built-in
modules.
Moreover, as DaveM said, we can't reliably put structures into
independent objects, put them into a special section, and then expect
array access over them (via the section boundaries) after linking the
objects together to just "work" due to variable alignment choices in
different situations. The only solution that seems to work reliably
is to make an array of plain pointers to the objects in question and
put those pointers in the special section.
Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Dmitry Torokhov <dtor@vmware.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
The first cpu which switches from periodic to oneshot mode switches
also the broadcast device into oneshot mode. The broadcast device
serves as a backup for per cpu timers which stop in deeper
C-states. To avoid starvation of the cpus which might be in idle and
depend on broadcast mode it marks the other cpus as broadcast active
and sets the brodcast expiry value of those cpus to the next tick.
The oneshot mode broadcast bit for the other cpus is sticky and gets
only cleared when those cpus exit idle. If a cpu was not idle while
the bit got set in consequence the bit prevents that the broadcast
device is armed on behalf of that cpu when it enters idle for the
first time after it switched to oneshot mode.
In most cases that goes unnoticed as one of the other cpus has usually
a timer pending which keeps the broadcast device armed with a short
timeout. Now if the only cpu which has a short timer active has the
bit set then the broadcast device will not be armed on behalf of that
cpu and will fire way after the expected timer expiry. In the case of
Christians bug report it took ~145 seconds which is about half of the
wrap around time of HPET (the limit for that device) due to the fact
that all other cpus had no timers armed which expired before the 145
seconds timeframe.
The solution is simply to clear the broadcast active bit
unconditionally when a cpu switches to oneshot mode after the first
cpu switched the broadcast device over. It's not idle at that point
otherwise it would not be executing that code.
[ I fundamentally hate that broadcast crap. Why the heck thought some
folks that when going into deep idle it's a brilliant concept to
switch off the last device which brings the cpu back from that
state? ]
Thanks to Christian for providing all the valuable debug information!
Reported-and-tested-by: Christian Hoffmann <email@christianhoffmann.info>
Cc: John Stultz <johnstul@us.ibm.com>
Link: http://lkml.kernel.org/r/%3Calpine.LFD.2.02.1105161105170.3078%40ionos%3E
Cc: stable@kernel.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
If !CONFIG_USERNS, have current_user_ns() defined to (&init_user_ns).
Get rid of _current_user_ns. This requires nsown_capable() to be
defined in capability.c rather than as static inline in capability.h,
so do that.
Request_key needs init_user_ns defined at current_user_ns if
!CONFIG_USERNS, so forward-declare that in cred.h if !CONFIG_USERNS
at current_user_ns() define.
Compile-tested with and without CONFIG_USERNS.
Signed-off-by: Serge E. Hallyn <serge.hallyn@canonical.com>
[ This makes a huge performance difference for acl_permission_check(),
up to 30%. And that is one of the hottest kernel functions for loads
that are pathname-lookup heavy. ]
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The SNAPSHOT_S2RAM ioctl used for implementing the feature allowing
one to suspend to RAM after creating a hibernation image is currently
broken, because it doesn't clear the "ready" flag in the struct
snapshot_data object handled by it. As a result, the
SNAPSHOT_UNFREEZE doesn't work correctly after SNAPSHOT_S2RAM has
returned and the user space hibernate task cannot thaw the other
processes as appropriate. Make SNAPSHOT_S2RAM clear data->ready
to fix this problem.
Tested-by: Alexandre Felipe Muller de Souza <alexandrefm@mandriva.com.br>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Cc: stable@kernel.org
If the process using the hibernate user space interface closes
/dev/snapshot after creating a hibernation image without thawing
tasks, snapshot_release() should call pm_restore_gfp_mask() to
restore the GFP mask used before the creation of the image. Make
that happen.
Tested-by: Alexandre Felipe Muller de Souza <alexandrefm@mandriva.com.br>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Cc: stable@kernel.org
A warning is printed by pm_restrict_gfp_mask() while the
SNAPSHOT_S2RAM ioctl is being executed after creating a hibernation
image, because pm_restrict_gfp_mask() has been called once already
before the image creation and suspend_devices_and_enter() calls it
once again. This happens after commit 452aa6999e
(mm/pm: force GFP_NOIO during suspend/hibernation and resume).
To avoid this issue, move pm_restrict_gfp_mask() and
pm_restore_gfp_mask() from suspend_devices_and_enter() to its caller
in kernel/power/suspend.c.
Reported-by: Alexandre Felipe Muller de Souza <alexandrefm@mandriva.com.br>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Cc: stable@kernel.org
This partially reverts commit e6e1e25935.
That commit changed the structure layout of the trace structure, which
in turn broke PowerTOP (1.9x generation) quite badly.
I appreciate not wanting to expose the variable in question, and
PowerTOP was not using it, so I've replaced the variable with just a
padding field - that way if in the future a new field is needed it can
just use this padding field.
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Christian Hoffmann reported that the command line clocksource override
with acpi_pm timer fails:
Kernel command line: <SNIP> clocksource=acpi_pm
hpet clockevent registered
Switching to clocksource hpet
Override clocksource acpi_pm is not HRT compatible.
Cannot switch while in HRT/NOHZ mode.
The watchdog code is what enables CLOCK_SOURCE_VALID_FOR_HRES, but we
actually end up selecting the clocksource before we enqueue it into
the watchdog list, so that's why we see the warning and fail to switch
to acpi_pm timer as requested. That's particularly bad when we want to
debug timekeeping related problems in early boot.
Put the selection call last.
Reported-by: Christian Hoffmann <email@christianhoffmann.info>
Signed-off-by: John Stultz <johnstul@us.ibm.com>
Cc: stable@kernel.org # 32...
Link: http://lkml.kernel.org/r/%3C1304558210.2943.24.camel%40work-vm%3E
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
commit ab7798ffcf ("genirq: Expand generic
show_interrupts()") added the Kconfig option GENERIC_IRQ_SHOW_LEVEL to
accomodate PowerPC, but this doesn't actually enable the functionality due
to a typo in the #ifdef check.
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: Linux/PPC Development <linuxppc-dev@lists.ozlabs.org>
Link: http://lkml.kernel.org/r/%3Calpine.DEB.2.00.1104302251370.19068%40ayla.of.borg%3E
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
* 'fixes-2.6.39' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq:
workqueue: fix deadlock in worker_maybe_bind_and_lock()
workqueue: Document debugging tricks
Fix up trivial spelling conflict in kernel/workqueue.c
* 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
perf, x86, nmi: Move LVT un-masking into irq handlers
perf events, x86: Work around the Nehalem AAJ80 erratum
perf, x86: Fix BTS condition
ftrace: Build without frame pointers on Microblaze
If a rescuer and stop_machine() bringing down a CPU race with each
other, they may deadlock on non-preemptive kernel. The CPU won't
accept a new task, so the rescuer can't migrate to the target CPU,
while stop_machine() can't proceed because the rescuer is holding one
of the CPU retrying migration. GCWQ_DISASSOCIATED is never cleared
and worker_maybe_bind_and_lock() retries indefinitely.
This problem can be reproduced semi reliably while the system is
entering suspend.
http://thread.gmane.org/gmane.linux.kernel/1122051
A lot of kudos to Thilo-Alexander for reporting this tricky issue and
painstaking testing.
stable: This affects all kernels with cmwq, so all kernels since and
including v2.6.36 need this fix.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Thilo-Alexander Ginkel <thilo@ginkel.com>
Tested-by: Thilo-Alexander Ginkel <thilo@ginkel.com>
Cc: stable@kernel.org
Sedat and Bruno reported RCU stalls which turned out to be caused by
the following;
sched_init() calls init_rt_bandwidth() which calls hrtimer_init()
_BEFORE_ hrtimers_init() is called. While not entirely correct this
worked because hrtimer_init() only accessed statically initialized
data (hrtimer_bases.clock_base[CLOCK_MONOTONIC])
Commit e06383db9 (hrtimers: extend hrtimer base code to handle more
then 2 clockids) added an indirection to the hrtimer_bases.clock_base
lookup to avoid gap handling in the hot path. The table which is used
for the translataion from CLOCK_ID to HRTIMER_BASE index is
initialized at runtime in hrtimers_init(). So the early call of the
scheduler code translates CLOCK_MONOTONIC to HRTIMER_BASE_REALTIME.
Thus the rt_bandwith timer ends up on CLOCK_REALTIME. If the timer is
armed and the wall clock time is set (e.g. ntpdate in the early boot
process - which also gives the problem deterministic behaviour
i.e. magic recovery after N hours), then the timer ends up with an
expiry time far into the future. That breaks the RT throttler
mechanism as rt runtime is accumulated and never cleared, so the rt
throttler detects a false cpu hog condition and blocks all RT tasks
until the timer finally expires. That in turn stalls the RCU thread of
TINYRCU which leads to an huge amount of RCU callbacks piling up.
Make the translation table statically initialized, so we are back to
the status of <= 2.6.39.
Reported-and-tested-by: Sedat Dilek <sedat.dilek@gmail.com>
Reported-by: Bruno Prémont <bonbons@linux-vserver.org>
Cc: John stultz <johnstul@us.ibm.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/%3Calpine.LFD.2.02.1104282353140.3005%40ionos%3E
Reviewed-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
In corner cases where softlockup watchdog is not setup successfully, the
relevant nmi perf event for hardlockup watchdog could be disabled, then
the status of the underlying hardware remains unchanged.
Also, if the kthread doesn't start then the hrtimer won't run and the
hardlockup detector will falsely fire.
Signed-off-by: Hillf Danton <dhillf@gmail.com>
Signed-off-by: Don Zickus <dzickus@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When a task is traced and is in a stopped state, the tracer
may execute a ptrace request to examine the tracee state and
get its task struct. Right after, the tracee can be killed
and thus its breakpoints released.
This can happen concurrently when the tracer is in the middle
of reading or modifying these breakpoints, leading to dereferencing
a freed pointer.
Hence, to prepare the fix, create a generic breakpoint reference
holding API. When a reference on the breakpoints of a task is
held, the breakpoints won't be released until the last reference
is dropped. After that, no more ptrace request on the task's
breakpoints can be serviced for the tracer.
Reported-by: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Prasad <prasad@linux.vnet.ibm.com>
Cc: Paul Mundt <lethal@linux-sh.org>
Cc: v2.6.33.. <stable@kernel.org>
Link: http://lkml.kernel.org/r/1302284067-7860-2-git-send-email-fweisbec@gmail.com
Device suspend/resume infrastructure is used not only by the suspend
and hibernate code in kernel/power, but also by APM, Xen and the
kexec jump feature. However, commit 40dc166cb5
(PM / Core: Introduce struct syscore_ops for core subsystems PM)
failed to add syscore_suspend() and syscore_resume() calls to that
code, which generally leads to breakage when the features in question
are used.
To fix this problem, add the missing syscore_suspend() and
syscore_resume() calls to arch/x86/kernel/apm_32.c, kernel/kexec.c
and drivers/xen/manage.c.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
* 'timer-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
RTC: rtc-omap: Fix a leak of the IRQ during init failure
posix clocks: Replace mutex with reader/writer semaphore
If syscore_suspend() fails in suspend_enter(), create_image() or
resume_target_kernel(), it is necessary to call sysdev_resume(),
because sysdev_suspend() has been called already and succeeded
and we are going to abort the transition.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Acked-by: Greg Kroah-Hartman <gregkh@suse.de>
next_pidmap() just quietly accepted whatever 'last' pid that was passed
in, which is not all that safe when one of the users is /proc.
Admittedly the proc code should do some sanity checking on the range
(and that will be the next commit), but that doesn't mean that the
helper functions should just do that pidmap pointer arithmetic without
checking the range of its arguments.
So clamp 'last' to PID_MAX_LIMIT. The fact that we then do "last+1"
doesn't really matter, the for-loop does check against the end of the
pidmap array properly (it's only the actual pointer arithmetic overflow
case we need to worry about, and going one bit beyond isn't going to
overflow).
[ Use PID_MAX_LIMIT rather than pid_max as per Eric Biederman ]
Reported-by: Tavis Ormandy <taviso@cmpxchg8b.com>
Analyzed-by: Robert Święcki <robert@swiecki.net>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
A dynamic posix clock is protected from asynchronous removal by a mutex.
However, using a mutex has the unwanted effect that a long running clock
operation in one process will unnecessarily block other processes.
For example, one process might call read() to get an external time stamp
coming in at one pulse per second. A second process calling clock_gettime
would have to wait for almost a whole second.
This patch fixes the issue by using a reader/writer semaphore instead of
a mutex.
Signed-off-by: Richard Cochran <richard.cochran@omicron.at>
Cc: John Stultz <john.stultz@linaro.org>
Link: http://lkml.kernel.org/r/%3C20110330132421.GA31771%40riccoc20.at.omicron.at%3E
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
block: make unplug timer trace event correspond to the schedule() unplug
block: let io_schedule() flush the plug inline
* 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
futex: Set FLAGS_HAS_TIMEOUT during futex_wait restart setup
* 'perf-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
perf_event: Fix cgrp event scheduling bug in perf_enable_on_exec()
perf: Fix a build error with some GCC versions
* 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
sched: Fix erroneous all_pinned logic
sched: Fix sched-domain avg_load calculation
* 'timer-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
RTC: rtc-mrst: follow on to the change of rtc_device_register()
RTC: add missing "return 0" in new alarm func for rtc-bfin.c
RTC: Fix s3c compile error due to missing s3c_rtc_setpie
RTC: Fix early irqs caused by calling rtc_set_alarm too early
* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86, amd: Disable GartTlbWlkErr when BIOS forgets it
x86, NUMA: Fix fakenuma boot failure
x86/mrst: Fix boot crash caused by incorrect pin to irq mapping
x86/ce4100: Add reg property to bridges
It's a pretty close match to what we had before - the timer triggering
would mean that nobody unplugged the plug in due time, in the new
scheme this matches very closely what the schedule() unplug now is.
It's essentially the difference between an explicit unplug (IO unplug)
or an implicit unplug (timer unplug, we scheduled with pending IO
queued).
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Linus correctly observes that the most important dispatch cases
are now done from kblockd, this isn't ideal for latency reasons.
The original reason for switching dispatches out-of-line was to
avoid too deep a stack, so by _only_ letting the "accidental"
flush directly in schedule() be guarded by offload to kblockd,
we should be able to get the best of both worlds.
So add a blk_schedule_flush_plug() that offloads to kblockd,
and only use that from the schedule() path.
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
* 'for-linus' of git://git.kernel.dk/linux-2.6-block:
block: only force kblockd unplugging from the schedule() path
block: cleanup the block plug helper functions
block, blk-sysfs: Use the variable directly instead of a function call
block: move queue run on unplug to kblockd
block: kill queue_sync_plugs()
block: readd plug trace event
block: add callback function for unplug notification
block: add comment on why we save and disable interrupts in flush_plug_list()
block: fixup block IO unplug trace call
block: remove block_unplug_timer() trace point
block: splice plug list to local context
The FLAGS_HAS_TIMEOUT flag was not getting set, causing the restart_block to
restart futex_wait() without a timeout after a signal.
Commit b41277dc7a in 2.6.38 introduced the regression by accidentally
removing the the FLAGS_HAS_TIMEOUT assignment from futex_wait() during the setup
of the restart block. Restore the originaly behavior.
Fixes: https://bugzilla.kernel.org/show_bug.cgi?id=32922
Reported-by: Tim Smith <tsmith201104@yahoo.com>
Reported-by: Torsten Hilbrich <torsten.hilbrich@secunet.com>
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: John Kacur <jkacur@redhat.com>
Cc: stable@kernel.org
Link: http://lkml.kernel.org/r/%3Cdaac0eb3af607f72b9a4d3126b2ba8fb5ed3b883.1302820917.git.dvhart%40linux.intel.com%3E
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
We really only want to unplug the pending IO when the process actually
goes to sleep. So move the test for flushing the plug up to the place
where we actually deactivate the task - where we have properly checked
for preemption and for the process really sleeping.
Acked-by: Jens Axboe <jaxboe@fusionio.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
It was removed with the on-stack plugging, readd it and track the
depth of requests added when flushing the plug.
Signed-off-by: Jens Axboe <jaxboe@fusionio.com>
Make XEN_SAVE_RESTORE select HIBERNATE_CALLBACKS.
Remove XEN_SAVE_RESTORE dependency from PM_SLEEP.
Signed-off-by: Shriram Rajagopalan <rshriram@cs.ubc.ca>
Acked-by: Ian Campbell <ian.campbell@citrix.com>
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Xen save/restore is going to use hibernate device callbacks for
quiescing devices and putting them back to normal operations and it
would need to select CONFIG_HIBERNATION for this purpose. However,
that also would cause the hibernate interfaces for user space to be
enabled, which might confuse user space, because the Xen kernels
don't support hibernation. Moreover, it would be wasteful, as it
would make the Xen kernels include a substantial amount of code that
they would never use.
To address this issue introduce new power management Kconfig option
CONFIG_HIBERNATE_CALLBACKS, such that it will only select the code
that is necessary for the hibernate device callbacks to work and make
CONFIG_HIBERNATION select it. Then, Xen save/restore will be able to
select CONFIG_HIBERNATE_CALLBACKS without dragging the entire
hibernate code along with it.
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl>
Tested-by: Shriram Rajagopalan <rshriram@cs.ubc.ca>
The scheduler load balancer has specific code to deal with cases of
unbalanced system due to lots of unmovable tasks (for example because of
hard CPU affinity). In those situation, it excludes the busiest CPU that
has pinned tasks for load balance consideration such that it can perform
second 2nd load balance pass on the rest of the system.
This all works as designed if there is only one cgroup in the system.
However, when we have multiple cgroups, this logic has false positives and
triggers multiple load balance passes despite there are actually no pinned
tasks at all.
The reason it has false positives is that the all pinned logic is deep in
the lowest function of can_migrate_task() and is too low level:
load_balance_fair() iterates each task group and calls balance_tasks() to
migrate target load. Along the way, balance_tasks() will also set a
all_pinned variable. Given that task-groups are iterated, this all_pinned
variable is essentially the status of last group in the scanning process.
Task group can have number of reasons that no load being migrated, none
due to cpu affinity. However, this status bit is being propagated back up
to the higher level load_balance(), which incorrectly think that no tasks
were moved. It kick off the all pinned logic and start multiple passes
attempt to move load onto puller CPU.
To fix this, move the all_pinned aggregation up at the iterator level.
This ensures that the status is aggregated over all task-groups, not just
last one in the list.
Signed-off-by: Ken Chen <kenchen@google.com>
Cc: stable@kernel.org
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/BANLkTi=ernzNawaR5tJZEsV_QVnfxqXmsQ@mail.gmail.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
In function find_busiest_group(), the sched-domain avg_load isn't
calculated at all if there is a group imbalance within the domain. This
will cause erroneous imbalance calculation.
The reason is that calculate_imbalance() sees sds->avg_load = 0 and it
will dump entire sds->max_load into imbalance variable, which is used
later on to migrate entire load from busiest CPU to the puller CPU.
This has two really bad effect:
1. stampede of task migration, and they won't be able to break out
of the bad state because of positive feedback loop: large load
delta -> heavier load migration -> larger imbalance and the cycle
goes on.
2. severe imbalance in CPU queue depth. This causes really long
scheduling latency blip which affects badly on application that
has tight latency requirement.
The fix is to have kernel calculate domain avg_load in both cases. This
will ensure that imbalance calculation is always sensible and the target
is usually half way between busiest and puller CPU.
Signed-off-by: Ken Chen <kenchen@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: <stable@kernel.org>
Link: http://lkml.kernel.org/r/20110408002322.3A0D812217F@elm.corp.google.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
There is a bug in perf_event_enable_on_exec() when cgroup events are
active on a CPU: the cgroup events may be scheduled twice causing event
state corruptions which eventually may lead to kernel panics.
The reason is that the function needs to first schedule out the cgroup
events, just like for the per-thread events. The cgroup event are
scheduled back in automatically from the perf_event_context_sched_in()
function.
The patch also adds a WARN_ON_ONCE() is perf_cgroup_switch() to catch any
bogus state.
Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/20110406005454.GA1062@quad
Signed-off-by: Ingo Molnar <mingo@elte.hu>
