Silly though it is, completions and wait_queue_heads use foo_ONSTACK
(COMPLETION_INITIALIZER_ONSTACK, DECLARE_COMPLETION_ONSTACK,
__WAIT_QUEUE_HEAD_INIT_ONSTACK and DECLARE_WAIT_QUEUE_HEAD_ONSTACK) so I
guess workqueues should do the same thing.
s/INIT_WORK_ON_STACK/INIT_WORK_ONSTACK/
s/INIT_DELAYED_WORK_ON_STACK/INIT_DELAYED_WORK_ONSTACK/
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
After all that's what they are intended for.
Signed-off-by: Jan Beulich <jbeulich@novell.com>
Cc: Miklos Szeredi <miklos@szeredi.hu>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: "Rafael J. Wysocki" <rjw@sisk.pl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Ensure kmap_atomic() usage is strictly nested
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Reviewed-by: Rik van Riel <riel@redhat.com>
Acked-by: Chris Metcalf <cmetcalf@tilera.com>
Cc: David Howells <dhowells@redhat.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Russell King <rmk@arm.linux.org.uk>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: David Miller <davem@davemloft.net>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
It's pointless to kill a task if another thread sharing its mm cannot be
killed to allow future memory freeing. A subsequent patch will prevent
kills in such cases, but first it's necessary to have a way to flag a task
that shares memory with an OOM_DISABLE task that doesn't incur an
additional tasklist scan, which would make select_bad_process() an O(n^2)
function.
This patch adds an atomic counter to struct mm_struct that follows how
many threads attached to it have an oom_score_adj of OOM_SCORE_ADJ_MIN.
They cannot be killed by the kernel, so their memory cannot be freed in
oom conditions.
This only requires task_lock() on the task that we're operating on, it
does not require mm->mmap_sem since task_lock() pins the mm and the
operation is atomic.
[rientjes@google.com: changelog and sys_unshare() code]
[rientjes@google.com: protect oom_disable_count with task_lock in fork]
[rientjes@google.com: use old_mm for oom_disable_count in exec]
Signed-off-by: Ying Han <yinghan@google.com>
Signed-off-by: David Rientjes <rientjes@google.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Rik van Riel <riel@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Allocate space from the top of a region first, then work downward,
if an architecture desires this.
When we allocate space from a resource, we look for gaps between children
of the resource. Previously, we always looked at gaps from the bottom up.
For example, given this:
[mem 0xbff00000-0xf7ffffff] PCI Bus 0000:00
[mem 0xbff00000-0xbfffffff] gap -- available
[mem 0xc0000000-0xdfffffff] PCI Bus 0000:02
[mem 0xe0000000-0xf7ffffff] gap -- available
we attempted to allocate from the [mem 0xbff00000-0xbfffffff] gap first,
then the [mem 0xe0000000-0xf7ffffff] gap.
With this patch an architecture can choose to allocate from the top gap
[mem 0xe0000000-0xf7ffffff] first.
We can't do this across the board because iomem_resource.end is initialized
to 0xffffffff_ffffffff on 64-bit architectures, and most machines can't
address the entire 64-bit physical address space. Therefore, we only
allocate top-down if the arch requests it by clearing
"resource_alloc_from_bottom".
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
If tmp.start is near ~0, ALIGN(tmp.start) may overflow, which would
make us think there's more available space than there really is. We
would likely return something that conflicts with a previous resource,
which would cause a failure when allocate_resource() requests the newly-
allocated region.
Reference: https://bugzilla.redhat.com/show_bug.cgi?id=646027
Reported-by: Fabrice Bellet <fabrice@bellet.info>
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
The alignment callback returns a proposed location, which may have been
adjusted to avoid ISA aliases or for other architecture-specific reasons.
We already had a check ("tmp.start < tmp.end") to make sure the callback
doesn't return an area that extends past the available area. This patch
reworks the check to make sure it doesn't return an area that extends
either below or above the available area.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
This factors out the min/max clipping to simplify find_resource().
No functional change.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
This removes a test from find_resource(), which is getting cluttered.
No functional change.
Signed-off-by: Bjorn Helgaas <bjorn.helgaas@hp.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
The nr_dentry stat is a globally touched cacheline and atomic operation
twice over the lifetime of a dentry. It is used for the benfit of userspace
only. Turn it into a per-cpu counter and always decrement it in d_free instead
of doing various batching operations to reduce lock hold times in the callers.
Based on an earlier patch from Nick Piggin <npiggin@suse.de>.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Instead of always assigning an increasing inode number in new_inode
move the call to assign it into those callers that actually need it.
For now callers that need it is estimated conservatively, that is
the call is added to all filesystems that do not assign an i_ino
by themselves. For a few more filesystems we can avoid assigning
any inode number given that they aren't user visible, and for others
it could be done lazily when an inode number is actually needed,
but that's left for later patches.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
The number of inodes allocated does not need to be tied to the
addition or removal of an inode to/from a list. If we are not tied
to a list lock, we could update the counters when inodes are
initialised or destroyed, but to do that we need to convert the
counters to be per-cpu (i.e. independent of a lock). This means that
we have the freedom to change the list/locking implementation
without needing to care about the counters.
Based on a patch originally from Eric Dumazet.
[AV: cleaned up a bit, fixed build breakage on weird configs
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Andrew,
Could you please review this patch, you probably are the right guy to
take it, because it crosses fs and net trees.
Note : /proc/sys/fs/file-nr is a read-only file, so this patch doesnt
depend on previous patch (sysctl: fix min/max handling in
__do_proc_doulongvec_minmax())
Thanks !
[PATCH V4] fs: allow for more than 2^31 files
Robin Holt tried to boot a 16TB system and found af_unix was overflowing
a 32bit value :
<quote>
We were seeing a failure which prevented boot. The kernel was incapable
of creating either a named pipe or unix domain socket. This comes down
to a common kernel function called unix_create1() which does:
atomic_inc(&unix_nr_socks);
if (atomic_read(&unix_nr_socks) > 2 * get_max_files())
goto out;
The function get_max_files() is a simple return of files_stat.max_files.
files_stat.max_files is a signed integer and is computed in
fs/file_table.c's files_init().
n = (mempages * (PAGE_SIZE / 1024)) / 10;
files_stat.max_files = n;
In our case, mempages (total_ram_pages) is approx 3,758,096,384
(0xe0000000). That leaves max_files at approximately 1,503,238,553.
This causes 2 * get_max_files() to integer overflow.
</quote>
Fix is to let /proc/sys/fs/file-nr & /proc/sys/fs/file-max use long
integers, and change af_unix to use an atomic_long_t instead of
atomic_t.
get_max_files() is changed to return an unsigned long.
get_nr_files() is changed to return a long.
unix_nr_socks is changed from atomic_t to atomic_long_t, while not
strictly needed to address Robin problem.
Before patch (on a 64bit kernel) :
# echo 2147483648 >/proc/sys/fs/file-max
# cat /proc/sys/fs/file-max
-18446744071562067968
After patch:
# echo 2147483648 >/proc/sys/fs/file-max
# cat /proc/sys/fs/file-max
2147483648
# cat /proc/sys/fs/file-nr
704 0 2147483648
Reported-by: Robin Holt <holt@sgi.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: David Miller <davem@davemloft.net>
Reviewed-by: Robin Holt <holt@sgi.com>
Tested-by: Robin Holt <holt@sgi.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
In the MN10300 arch, we occasionally see an assertion being tripped in
alloc_cwqs() at the following line:
/* just in case, make sure it's actually aligned */
---> BUG_ON(!IS_ALIGNED(wq->cpu_wq.v, align));
return wq->cpu_wq.v ? 0 : -ENOMEM;
The values are:
wa->cpu_wq.v => 0x902776e0
align => 0x100
and align is calculated by the following:
const size_t align = max_t(size_t, 1 << WORK_STRUCT_FLAG_BITS,
__alignof__(unsigned long long));
This is because the pointer in question (wq->cpu_wq.v) loses some of its
lower bits to control flags, and so the object it points to must be
sufficiently aligned to avoid the need to use those bits for pointing to
things.
Currently, 4 control bits and 4 colour bits are used in normal
circumstances, plus a debugging bit if debugging is set. This requires
the cpu_workqueue_struct struct to be at least 256 bytes aligned (or 512
bytes aligned with debugging).
PERCPU() alignment on MN13000, however, is only 32 bytes as set in
vmlinux.lds.S. So we set this to PAGE_SIZE (4096) to match most other
arches and stick a comment in alloc_cwqs() for anyone else who triggers
the assertion.
Reported-by: Akira Takeuchi <takeuchi.akr@jp.panasonic.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Mark Salter <msalter@redhat.com>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Remove text_mutex locking in optimize_all_kprobes, because
this function doesn't modify text. It simply queues probes on
optimization list for kprobe_optimizer worker thread.
Signed-off-by: Masami Hiramatsu <mhiramat@redhat.com>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Anil S Keshavamurthy <anil.s.keshavamurthy@intel.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Jason Baron <jbaron@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20101025131801.19160.70939.stgit@ltc236.sdl.hitachi.co.jp>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Remove some sched_info_queue(), sched_info_dequeue() code comment.
We no longer belongs to the era of O(1) and we don't use active or expired
array anymore.
Signed-off-by: Rakib Mullick <rakib.mullick@gmail.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
LKML-Reference: <AANLkTi=REu0WzOp5N=nVT1=ZJ=ZA+MZFV+4CHSJ3Q-Yv@mail.gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* 'softirq-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
softirqs: Make wakeup_softirqd static
* 'x86-debug-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86, asm: Restore parentheses around one pushl_cfi argument
x86, asm: Fix ancient-GAS workaround
x86, asm: Fix CFI macro invocations to deal with shortcomings in gas
* 'x86-numa-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86, numa: Assign CPUs to nodes in round-robin manner on fake NUMA
* 'x86-quirks-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86: HPET force enable for CX700 / VIA Epia LT
* 'x86-setup-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86, setup: Use string copy operation to optimze copy in kernel compression
* 'x86-uv-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86, UV: Use allocated buffer in tlb_uv.c:tunables_read()
* 'x86-vm86-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86, vm86: Fix preemption bug for int1 debug and int3 breakpoint handlers.
* 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/linux-2.6-kgdb:
kdb,debug_core: adjust master cpu switch logic against new debug_core locking
debug_core: refactor locking for master/slave cpus
x86,kgdb: remove unnecessary call to kgdb_correct_hw_break()
debug_core: disable hw_breakpoints on all cores in kgdb_cpu_enter()
kdb,kgdb: fix sparse fixups
kdb: Fix oops in kdb_unregister
kdb,ftdump: Remove reference to internal kdb include
kdb: Allow kernel loadable modules to add kdb shell functions
debug_core: stop rcu warnings on kernel resume
debug_core: move all watch dog syncs to a single function
x86,kgdb: fix debugger hw breakpoint test regression in 2.6.35
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tj/wq:
workqueue: remove in_workqueue_context()
workqueue: Clarify that schedule_on_each_cpu is synchronous
memory_hotplug: drop spurious calls to flush_scheduled_work()
shpchp: update workqueue usage
pciehp: update workqueue usage
isdn/eicon: don't call flush_scheduled_work() from diva_os_remove_soft_isr()
workqueue: add and use WQ_MEM_RECLAIM flag
workqueue: fix HIGHPRI handling in keep_working()
workqueue: add queue_work and activate_work trace points
workqueue: prepare for more tracepoints
workqueue: implement flush[_delayed]_work_sync()
workqueue: factor out start_flush_work()
workqueue: cleanup flush/cancel functions
workqueue: implement alloc_ordered_workqueue()
Fix up trivial conflict in fs/gfs2/main.c as per Tejun
The kdb shell needs to enforce switching back to the original CPU that
took the exception before restoring normal kernel execution. Resuming
from a different CPU than what took the original exception will cause
problems with spin locks that are freed from the a different processor
than had taken the lock.
The special logic in dbg_cpu_switch() can go away entirely with
because the state of what cpus want to be masters or slaves will
remain unchanged between entry and exit of the debug_core exception
context.
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
For quite some time there have been problems with memory barriers and
various races with NMI on multi processor systems using the kernel
debugger. The algorithm for entering the kernel debug core and
resuming kernel execution was racy and had several known edge case
problems with attempting to debug something on a heavily loaded system
using breakpoints that are hit repeatedly and quickly.
The prior "locking" design entry worked as follows:
* The atomic counter kgdb_active was used with atomic exchange in
order to elect a master cpu out of all the cpus that may have
taken a debug exception.
* The master cpu increments all elements of passive_cpu_wait[].
* The master cpu issues the round up cpus message.
* Each "slave cpu" that enters the debug core increments its own
element in cpu_in_kgdb[].
* Each "slave cpu" spins on passive_cpu_wait[] until it becomes 0.
* The master cpu debugs the system.
The new scheme removes the two arrays of atomic counters and replaces
them with 2 single counters. One counter is used to count the number
of cpus waiting to become a master cpu (because one or more hit an
exception). The second counter is use to indicate how many cpus have
entered as slave cpus.
The new entry logic works as follows:
* One or more cpus enters via kgdb_handle_exception() and increments
the masters_in_kgdb. Each cpu attempts to get the spin lock called
dbg_master_lock.
* The master cpu sets kgdb_active to the current cpu.
* The master cpu takes the spinlock dbg_slave_lock.
* The master cpu asks to round up all the other cpus.
* Each slave cpu that is not already in kgdb_handle_exception()
will enter and increment slaves_in_kgdb. Each slave will now spin
try_locking on dbg_slave_lock.
* The master cpu waits for the sum of masters_in_kgdb and slaves_in_kgdb
to be equal to the sum of the online cpus.
* The master cpu debugs the system.
In the new design the kgdb_active can only be changed while holding
dbg_master_lock. Stress testing has not turned up any further
entry/exit races that existed in the prior locking design. The prior
locking design suffered from atomic variables not being truly atomic
(in the capacity as used by kgdb) along with memory barrier races.
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Acked-by: Dongdong Deng <dongdong.deng@windriver.com>
The slave cpus do not have the hw breakpoints disabled upon entry to
the debug_core and as a result could cause unrecoverable recursive
faults on badly placed breakpoints, or get out of sync with the arch
specific hw breakpoint operations.
This patch addresses the problem by invoking kgdb_disable_hw_debug()
earlier in kgdb_enter_cpu for each cpu that enters the debug core.
The hw breakpoint dis/enable flow should be:
master_debug_cpu slave_debug_cpu
\ /
kgdb_cpu_enter
|
kgdb_disable_hw_debug --> uninstall pre-enabled hw_breakpoint
|
do add/rm dis/enable operates to hw_breakpoints on master_debug_cpu..
|
correct_hw_break --> correct/install the enabled hw_breakpoint
|
leave_kgdb
Signed-off-by: Dongdong Deng <dongdong.deng@windriver.com>
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Fix the following sparse warnings:
kdb_main.c:328:5: warning: symbol 'kdbgetu64arg' was not declared. Should it be static?
kgdboc.c:246:12: warning: symbol 'kgdboc_early_init' was not declared. Should it be static?
kgdb.c:652:26: warning: incorrect type in argument 1 (different address spaces)
kgdb.c:652:26: expected void const *ptr
kgdb.c:652:26: got struct perf_event *[noderef] <asn:3>*pev
The one in kgdb.c required the (void * __force) because of the return
code from register_wide_hw_breakpoint looking like:
return (void __percpu __force *)ERR_PTR(err);
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Nothing should try to use kdb_commands directly as sometimes it is
null. Instead, use the for_each_kdbcmd() iterator.
This particular problem dates back to the initial kdb merge (2.6.35),
but at that point nothing was dynamically unregistering commands from
the kdb shell.
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Now that include/linux/kdb.h properly exports all the functions
required to dynamically add a kdb shell command, the reference to the
private kdb header can be removed.
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
In order to allow kernel modules to dynamically add a command to the
kdb shell the kdb_register, kdb_register_repeat, kdb_unregister, and
kdb_printf need to be exported as GPL symbols.
Any kernel module that adds a dynamic kdb shell function should only
need to include linux/kdb.h.
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
When returning from the kernel debugger reset the rcu jiffies_stall
value to prevent the rcu stall detector from sending NMI events which
invoke a stack dump for each cpu in the system.
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
Move the various clock and watch dog syncs to a single function in
advance of adding another sync for the rcu stall detector.
Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
* 'llseek' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/bkl:
vfs: make no_llseek the default
vfs: don't use BKL in default_llseek
llseek: automatically add .llseek fop
libfs: use generic_file_llseek for simple_attr
mac80211: disallow seeks in minstrel debug code
lirc: make chardev nonseekable
viotape: use noop_llseek
raw: use explicit llseek file operations
ibmasmfs: use generic_file_llseek
spufs: use llseek in all file operations
arm/omap: use generic_file_llseek in iommu_debug
lkdtm: use generic_file_llseek in debugfs
net/wireless: use generic_file_llseek in debugfs
drm: use noop_llseek
* 'vfs' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/bkl: (30 commits)
BKL: remove BKL from freevxfs
BKL: remove BKL from qnx4
autofs4: Only declare function when CONFIG_COMPAT is defined
autofs: Only declare function when CONFIG_COMPAT is defined
ncpfs: Lock socket in ncpfs while setting its callbacks
fs/locks.c: prepare for BKL removal
BKL: Remove BKL from ncpfs
BKL: Remove BKL from OCFS2
BKL: Remove BKL from squashfs
BKL: Remove BKL from jffs2
BKL: Remove BKL from ecryptfs
BKL: Remove BKL from afs
BKL: Remove BKL from USB gadgetfs
BKL: Remove BKL from autofs4
BKL: Remove BKL from isofs
BKL: Remove BKL from fat
BKL: Remove BKL from ext2 filesystem
BKL: Remove BKL from do_new_mount()
BKL: Remove BKL from cgroup
BKL: Remove BKL from NTFS
...
* 'config' of git://git.kernel.org/pub/scm/linux/kernel/git/arnd/bkl:
BKL: introduce CONFIG_BKL.
dabusb: remove the BKL
sunrpc: remove the big kernel lock
init/main.c: remove BKL notations
blktrace: remove the big kernel lock
rtmutex-tester: make it build without BKL
dvb-core: kill the big kernel lock
dvb/bt8xx: kill the big kernel lock
tlclk: remove big kernel lock
fix rawctl compat ioctls breakage on amd64 and itanic
uml: kill big kernel lock
parisc: remove big kernel lock
cris: autoconvert trivial BKL users
alpha: kill big kernel lock
isapnp: BKL removal
s390/block: kill the big kernel lock
hpet: kill BKL, add compat_ioctl
This new version (see commit 8e5fc1a) is much simpler and ensures that
in case of error in group_sched_in() during event_sched_in(), the
events up to the failed event go through regular event_sched_out().
But the failed event and the remaining events in the group have their
timings adjusted as if they had also gone through event_sched_in() and
event_sched_out(). This ensures timing uniformity across all events in
a group. This also takes care of the tstamp_stopped problem in case
the group could never be scheduled. The tstamp_stopped is updated as
if the event had actually run.
With this patch, the following now reports correct time_enabled,
in case the NMI watchdog is active:
$ task -e unhalted_core_cycles,instructions_retired,baclears,baclears
noploop 1
noploop for 1 seconds
0 unhalted_core_cycles (100.00% scaling, ena=997,552,872, run=0)
0 instructions_retired (100.00% scaling, ena=997,552,872, run=0)
0 baclears (100.00% scaling, ena=997,552,872, run=0)
0 baclears (100.00% scaling, ena=997,552,872, run=0)
And the older test case also works:
$ task -einstructions_retired,baclears,baclears -e
unhalted_core_cycles,baclears,baclears sleep 5
1680885 instructions_retired (69.39% scaling, ena=950756, run=291006)
10735 baclears (69.39% scaling, ena=950756, run=291006)
10735 baclears (69.39% scaling, ena=950756, run=291006)
0 unhalted_core_cycles (100.00% scaling, ena=817932, run=0)
0 baclears (100.00% scaling, ena=817932, run=0)
0 baclears (100.00% scaling, ena=817932, run=0)
Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <4cbeeebc.8ee7d80a.5a28.0d5f@mx.google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch reverts commit 8e5fc1a (perf_events: Fix transaction
recovery in group_sched_in()) because it had one flaw in case the
group could never be scheduled. It would cause time_enabled to get
negative.
Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <4cbeeeb7.0aefd80a.6e40.0e2f@mx.google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Dima noticed that we fail to correct the ->vruntime of sleeping tasks
when we move them between cgroups.
Reported-by: Dima Zavin <dima@android.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Tested-by: Mike Galbraith <efault@gmx.de>
LKML-Reference: <1287150604.29097.1513.camel@twins>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc: (71 commits)
powerpc/44x: Update ppc44x_defconfig
powerpc/watchdog: Make default timeout for Book-E watchdog a Kconfig option
fsl_rio: Add comments for sRIO registers.
powerpc/fsl-booke: Add e55xx (64-bit) smp defconfig
powerpc/fsl-booke: Add p5020 DS board support
powerpc/fsl-booke64: Use TLB CAMs to cover linear mapping on FSL 64-bit chips
powerpc/fsl-booke: Add support for FSL Arch v1.0 MMU in setup_page_sizes
powerpc/fsl-booke: Add support for FSL 64-bit e5500 core
powerpc/85xx: add cache-sram support
powerpc/85xx: add ngPIXIS FPGA device tree node to the P1022DS board
powerpc: Fix compile error with paca code on ppc64e
powerpc/fsl-booke: Add p3041 DS board support
oprofile/fsl emb: Don't set MSR[PMM] until after clearing the interrupt.
powerpc/fsl-booke: Add PCI device ids for P2040/P3041/P5010/P5020 QoirQ chips
powerpc/mpc8xxx_gpio: Add support for 'qoriq-gpio' controllers
powerpc/fsl_booke: Add support to boot from core other than 0
powerpc/p1022: Add probing for individual DMA channels
powerpc/fsl_soc: Search all global-utilities nodes for rstccr
powerpc: Fix invalid page flags in create TLB CAM path for PTE_64BIT
powerpc/mpc83xx: Support for MPC8308 P1M board
...
Fix up conflict with the generic irq_work changes in arch/powerpc/kernel/time.c
* 'core-memblock-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (74 commits)
x86-64: Only set max_pfn_mapped to 512 MiB if we enter via head_64.S
xen: Cope with unmapped pages when initializing kernel pagetable
memblock, bootmem: Round pfn properly for memory and reserved regions
memblock: Annotate memblock functions with __init_memblock
memblock: Allow memblock_init to be called early
memblock/arm: Fix memblock_region_is_memory() typo
x86, memblock: Remove __memblock_x86_find_in_range_size()
memblock: Fix wraparound in find_region()
x86-32, memblock: Make add_highpages honor early reserved ranges
x86, memblock: Fix crashkernel allocation
arm, memblock: Fix the sparsemem build
memblock: Fix section mismatch warnings
powerpc, memblock: Fix memblock API change fallout
memblock, microblaze: Fix memblock API change fallout
x86: Remove old bootmem code
x86, memblock: Use memblock_memory_size()/memblock_free_memory_size() to get correct dma_reserve
x86: Remove not used early_res code
x86, memblock: Replace e820_/_early string with memblock_
x86: Use memblock to replace early_res
x86, memblock: Use memblock_debug to control debug message print out
...
Fix up trivial conflicts in arch/x86/kernel/setup.c and kernel/Makefile
* 'devel' of master.kernel.org:/home/rmk/linux-2.6-arm: (278 commits)
arm: remove machine_desc.io_pg_offst and .phys_io
arm: use addruart macro to establish debug mappings
arm: return both physical and virtual addresses from addruart
arm/debug: consolidate addruart macros for CONFIG_DEBUG_ICEDCC
ARM: make struct machine_desc definition coherent with its comment
eukrea_mbimxsd-baseboard: Pass the correct GPIO to gpio_free
cpuimx27: fix compile when ULPI is selected
mach-pcm037_eet: fix compile errors
Fixing ethernet driver compilation error for i.MX31 ADS board
cpuimx51: update board support
mx5: add cpuimx51sd module and its baseboard
iomux-mx51: fix GPIO_1_xx 's IOMUX configuration
imx-esdhc: update devices registration
mx51: add resources for SD/MMC on i.MX51
iomux-mx51: fix SD1 and SD2's iomux configuration
clock-mx51: rename CLOCK1 to CLOCK_CCGR for better readability
clock-mx51: factorize clk_set_parent and clk_get_rate
eukrea_mbimxsd: add support for DVI displays
cpuimx25 & cpuimx35: fix OTG port registration in host mode
i.MX31 and i.MX35 : fix errate TLSbo65953 and ENGcm09472
...
* 'irq-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (96 commits)
apic, x86: Use BIOS settings for IBS and MCE threshold interrupt LVT offsets
apic, x86: Check if EILVT APIC registers are available (AMD only)
x86: ioapic: Call free_irte only if interrupt remapping enabled
arm: Use ARCH_IRQ_INIT_FLAGS
genirq, ARM: Fix boot on ARM platforms
genirq: Fix CONFIG_GENIRQ_NO_DEPRECATED=y build
x86: Switch sparse_irq allocations to GFP_KERNEL
genirq: Switch sparse_irq allocator to GFP_KERNEL
genirq: Make sparse_lock a mutex
x86: lguest: Use new irq allocator
genirq: Remove the now unused sparse irq leftovers
genirq: Sanitize dynamic irq handling
genirq: Remove arch_init_chip_data()
x86: xen: Sanitise sparse_irq handling
x86: Use sane enumeration
x86: uv: Clean up the direct access to irq_desc
x86: Make io_apic.c local functions static
genirq: Remove irq_2_iommu
x86: Speed up the irq_remapped check in hot pathes
intr_remap: Simplify the code further
...
Fix up trivial conflicts in arch/x86/Kconfig
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (29 commits)
sched: Export account_system_vtime()
sched: Call tick_check_idle before __irq_enter
sched: Remove irq time from available CPU power
sched: Do not account irq time to current task
x86: Add IRQ_TIME_ACCOUNTING
sched: Add IRQ_TIME_ACCOUNTING, finer accounting of irq time
sched: Add a PF flag for ksoftirqd identification
sched: Consolidate account_system_vtime extern declaration
sched: Fix softirq time accounting
sched: Drop group_capacity to 1 only if local group has extra capacity
sched: Force balancing on newidle balance if local group has capacity
sched: Set group_imb only a task can be pulled from the busiest cpu
sched: Do not consider SCHED_IDLE tasks to be cache hot
sched: Drop all load weight manipulation for RT tasks
sched: Create special class for stop/migrate work
sched: Unindent labels
sched: Comment updates: fix default latency and granularity numbers
tracing/sched: Add sched_pi_setprio tracepoint
sched: Give CPU bound RT tasks preference
sched: Try not to migrate higher priority RT tasks
...
* 'core-locking-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
lockdep: Check the depth of subclass
lockdep: Add improved subclass caching
affs: Use sema_init instead of init_MUTEX
hfs: Convert tree_lock to mutex
arm: Bcmring: semaphore cleanup
printk: Make console_sem a semaphore not a pseudo mutex
drivers/macintosh/adb: Do not claim that the semaphore is a mutex
parport: Semaphore cleanup
irda: Semaphore cleanup
net: Wan/cosa.c: Convert "mutex" to semaphore
net: Ppp_async: semaphore cleanup
hamradio: Mkiss: semaphore cleanup
hamradio: 6pack: semaphore cleanup
net: 3c527: semaphore cleanup
input: Serio/hp_sdc: semaphore cleanup
input: Serio/hil_mlc: semaphore cleanup
input: Misc/hp_sdc_rtc: semaphore cleanup
lockup_detector: Make callback function static
lockup detector: Fix grammar by adding a missing "to" in the comments
lockdep: Remove __debug_show_held_locks
With the addition of trace_softirq_raise() the softirq tracepoint got
even more convoluted. Why the tracepoints take two pointers to assign
an integer is beyond my comprehension.
But adding an extra case which treats the first pointer as an unsigned
long when the second pointer is NULL including the back and forth
type casting is just horrible.
Convert the softirq tracepoints to take a single unsigned int argument
for the softirq vector number and fix the call sites.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
LKML-Reference: <alpine.LFD.2.00.1010191428560.6815@localhost6.localdomain6>
Acked-by: Peter Zijlstra <peterz@infradead.org>
Acked-by: mathieu.desnoyers@efficios.com
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
The tracing per_cpu buffers were limited to 999 CPUs for a mear
savings in stack space of a char array. Up the array to 30 characters
which is more than enough to hold a 64 bit number.
Reported-by: Robin Holt <holt@sgi.com>
Suggested-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
All security modules shouldn't change sched_param parameter of
security_task_setscheduler(). This is not only meaningless, but also
make a harmful result if caller pass a static variable.
This patch remove policy and sched_param parameter from
security_task_setscheduler() becuase none of security module is
using it.
Cc: James Morris <jmorris@namei.org>
Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: James Morris <jmorris@namei.org>
With the binding of time extends to events we no longer need to use
the macro RB_TIMESTAMPS_PER_PAGE. Remove it.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
There's a condition to check if we should add a time extend or
not in the fast path. But this condition is racey (in the sense
that we can add a unnecessary time extend, but nothing that
can break anything). We later check if the time or event time
delta should be zero or have real data in it (not racey), making
this first check redundant.
This check may help save space once in a while, but really is
not worth the hassle to try to save some space that happens at
most 134 ms at a time.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
When the time between two timestamps is greater than
2^27 nanosecs (~134 ms) a time extend event is added that extends
the time difference to 59 bits (~18 years). This is due to
events only having a 27 bit field to store time.
Currently this time extend is a separate event. We add it just before
the event data that is being written to the buffer. But before
the event data is committed, the event data can also be discarded (as
with the case of filters). But because the time extend has already been
committed, it will stay in the buffer.
If lots of events are being filtered and no event is being
written, then every 134ms a time extend can be added to the buffer
without any data attached. To keep from filling the entire buffer
with time extends, a time extend will never be the first event
in a page because the page timestamp can be used. Time extends can
only fill the rest of a page with some data at the beginning.
This patch binds the time extend with the data. The difference here
is that the time extend is not committed before the data is added.
Instead, when a time extend is needed, the space reserved on
the ring buffer is the time extend + the data event size. The
time extend is added to the first part of the reserved block and
the data is added to the second. The time extend event is passed
back to the reserver, but since the reserver also uses a function
to find the data portion of the reserved block, no changes to the
ring buffer interface need to be made.
When a commit is discarded, we now remove both the time extend and
the event. With this approach no more than one time extend can
be in the buffer in a row. Data must always follow a time extend.
Thanks to Mathieu Desnoyers for suggesting this idea.
Suggested-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The delta between events is passed to the timestamp code by reference
and the timestamp code will reset the value. But it can be reset
from the caller. No need to pass it in by reference.
By changing the call to pass by value, lets gcc optimize the code
a bit more where it can store the delta in a register and not
worry about updating the reference.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The original code for the ring buffer had locations that modified
the timestamp and that change was used by the callers. Now,
the timestamp is not reused by the callers and there is no reason
to pass it by reference.
By changing the call to pass by value, lets gcc optimize the code
a bit more where it can store the timestamp in a register and not
worry about updating the reference.
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Gcc inlines the slow path of the ring buffer write which can
hurt performance. This patch simply forces the slow path function
rb_move_tail() to always be a function.
The ring_buffer_benchmark module with reader_disabled=1 shows that
this patch changes the time to record an event from 135 ns to
132 ns. (3 ns or 2.22% improvement)
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
The function start_func_tracer() was incorrectly added in the
#ifdef CONFIG_FUNCTION_TRACER condition, but is still used even
when function tracing is not enabled.
The calls to register_ftrace_function() and register_ftrace_graph()
become nops (and their arguments are even ignored), thus there is
no reason to hide start_func_tracer() when function tracing is
not enabled.
Reported-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
futex_wait() is leaking key references due to futex_wait_setup()
acquiring an additional reference via the queue_lock() routine. The
nested key ref-counting has been masking bugs and complicating code
analysis. queue_lock() is only called with a previously ref-counted
key, so remove the additional ref-counting from the queue_(un)lock()
functions.
Also futex_wait_requeue_pi() drops one key reference too many in
unqueue_me_pi(). Remove the key reference handling from
unqueue_me_pi(). This was paired with a queue_lock() in
futex_lock_pi(), so the count remains unchanged.
Document remaining nested key ref-counting sites.
Signed-off-by: Darren Hart <dvhart@linux.intel.com>
Reported-and-tested-by: Matthieu Fertré<matthieu.fertre@kerlabs.com>
Reported-by: Louis Rilling<louis.rilling@kerlabs.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: John Kacur <jkacur@redhat.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
LKML-Reference: <4CBB17A8.70401@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: stable@kernel.org
According to Jens, this code does not need the BKL at all,
it is sufficiently serialized by bd_mutex.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Jens Axboe <jaxboe@fusionio.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
The big kernel lock is going away, so make sure
that if it is disabled by Kconfig, we do not
try to validate it, which would result in
compile errors.
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Commit a25909a4 (lockdep: Add an in_workqueue_context() lockdep-based
test function) added in_workqueue_context() but there hasn't been any
in-kernel user and the lockdep annotation in workqueue is scheduled to
change. Remove the unused function.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
The documentation for schedule_on_each_cpu() states that it calls a
function on each online CPU from keventd. This can easily be
interpreted as an asyncronous call because the description does not
mention that flush_work is called. Clarify that it is synchronous.
tj: rephrased a bit
Signed-off-by: Mel Gorman <mel@csn.ul.ie>
Reviewed-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
When CPU is idle and on first interrupt, irq_enter calls tick_check_idle()
to notify interruption from idle. But, there is a problem if this call
is done after __irq_enter, as all routines in __irq_enter may find
stale time due to yet to be done tick_check_idle.
Specifically, trace calls in __irq_enter when they use global clock and also
account_system_vtime change in this patch as it wants to use sched_clock_cpu()
to do proper irq timing.
But, tick_check_idle was moved after __irq_enter intentionally to
prevent problem of unneeded ksoftirqd wakeups by the commit ee5f80a:
irq: call __irq_enter() before calling the tick_idle_check
Impact: avoid spurious ksoftirqd wakeups
Moving tick_check_idle() before __irq_enter and wrapping it with
local_bh_enable/disable would solve both the problems.
Fixed-by: Yong Zhang <yong.zhang0@gmail.com>
Signed-off-by: Venkatesh Pallipadi <venki@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1286237003-12406-9-git-send-email-venki@google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The idea was suggested by Peter Zijlstra here:
http://marc.info/?l=linux-kernel&m=127476934517534&w=2
irq time is technically not available to the tasks running on the CPU.
This patch removes irq time from CPU power piggybacking on
sched_rt_avg_update().
Tested this by keeping CPU X busy with a network intensive task having 75%
oa a single CPU irq processing (hard+soft) on a 4-way system. And start seven
cycle soakers on the system. Without this change, there will be two tasks on
each CPU. With this change, there is a single task on irq busy CPU X and
remaining 7 tasks are spread around among other 3 CPUs.
Signed-off-by: Venkatesh Pallipadi <venki@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1286237003-12406-8-git-send-email-venki@google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Scheduler accounts both softirq and interrupt processing times to the
currently running task. This means, if the interrupt processing was
for some other task in the system, then the current task ends up being
penalized as it gets shorter runtime than otherwise.
Change sched task accounting to acoount only actual task time from
currently running task. Now update_curr(), modifies the delta_exec to
depend on rq->clock_task.
Note that this change only handles CONFIG_IRQ_TIME_ACCOUNTING case. We can
extend this to CONFIG_VIRT_CPU_ACCOUNTING with minimal effort. But, thats
for later.
This change will impact scheduling behavior in interrupt heavy conditions.
Tested on a 4-way system with eth0 handled by CPU 2 and a network heavy
task (nc) running on CPU 3 (and no RSS/RFS). With that I have CPU 2
spending 75%+ of its time in irq processing. CPU 3 spending around 35%
time running nc task.
Now, if I run another CPU intensive task on CPU 2, without this change
/proc/<pid>/schedstat shows 100% of time accounted to this task. With this
change, it rightly shows less than 25% accounted to this task as remaining
time is actually spent on irq processing.
Signed-off-by: Venkatesh Pallipadi <venki@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1286237003-12406-7-git-send-email-venki@google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
s390/powerpc/ia64 have support for CONFIG_VIRT_CPU_ACCOUNTING which does
the fine granularity accounting of user, system, hardirq, softirq times.
Adding that option on archs like x86 will be challenging however, given the
state of TSC reliability on various platforms and also the overhead it will
add in syscall entry exit.
Instead, add a lighter variant that only does finer accounting of
hardirq and softirq times, providing precise irq times (instead of timer tick
based samples). This accounting is added with a new config option
CONFIG_IRQ_TIME_ACCOUNTING so that there won't be any overhead for users not
interested in paying the perf penalty.
This accounting is based on sched_clock, with the code being generic.
So, other archs may find it useful as well.
This patch just adds the core logic and does not enable this logic yet.
Signed-off-by: Venkatesh Pallipadi <venki@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1286237003-12406-5-git-send-email-venki@google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
To account softirq time cleanly in scheduler, we need to identify whether
softirq is invoked in ksoftirqd context or softirq at hardirq tail context.
Add PF_KSOFTIRQD for that purpose.
As all PF flag bits are currently taken, create space by moving one of the
infrequently used bits (PF_THREAD_BOUND) down in task_struct to be along
with some other state fields.
Signed-off-by: Venkatesh Pallipadi <venki@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1286237003-12406-4-git-send-email-venki@google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Peter Zijlstra found a bug in the way softirq time is accounted in
VIRT_CPU_ACCOUNTING on this thread:
http://lkml.indiana.edu/hypermail//linux/kernel/1009.2/01366.html
The problem is, softirq processing uses local_bh_disable internally. There
is no way, later in the flow, to differentiate between whether softirq is
being processed or is it just that bh has been disabled. So, a hardirq when bh
is disabled results in time being wrongly accounted as softirq.
Looking at the code a bit more, the problem exists in !VIRT_CPU_ACCOUNTING
as well. As account_system_time() in normal tick based accouting also uses
softirq_count, which will be set even when not in softirq with bh disabled.
Peter also suggested solution of using 2*SOFTIRQ_OFFSET as irq count
for local_bh_{disable,enable} and using just SOFTIRQ_OFFSET while softirq
processing. The patch below does that and adds API in_serving_softirq() which
returns whether we are currently processing softirq or not.
Also changes one of the usages of softirq_count in net/sched/cls_cgroup.c
to in_serving_softirq.
Looks like many usages of in_softirq really want in_serving_softirq. Those
changes can be made individually on a case by case basis.
Signed-off-by: Venkatesh Pallipadi <venki@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1286237003-12406-2-git-send-email-venki@google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
When SD_PREFER_SIBLING is set on a sched domain, drop group_capacity to 1
only if the local group has extra capacity. The extra check prevents the case
where you always pull from the heaviest group when it is already under-utilized
(possible with a large weight task outweighs the tasks on the system).
For example, consider a 16-cpu quad-core quad-socket machine with MC and NUMA
scheduling domains. Let's say we spawn 15 nice0 tasks and one nice-15 task,
and each task is running on one core. In this case, we observe the following
events when balancing at the NUMA domain:
- find_busiest_group() will always pick the sched group containing the niced
task to be the busiest group.
- find_busiest_queue() will then always pick one of the cpus running the
nice0 task (never picks the cpu with the nice -15 task since
weighted_cpuload > imbalance).
- The load balancer fails to migrate the task since it is the running task
and increments sd->nr_balance_failed.
- It repeats the above steps a few more times until sd->nr_balance_failed > 5,
at which point it kicks off the active load balancer, wakes up the migration
thread and kicks the nice 0 task off the cpu.
The load balancer doesn't stop until we kick out all nice 0 tasks from
the sched group, leaving you with 3 idle cpus and one cpu running the
nice -15 task.
When balancing at the NUMA domain, we drop sgs.group_capacity to 1 if the child
domain (in this case MC) has SD_PREFER_SIBLING set. Subsequent load checks are
not relevant because the niced task has a very large weight.
In this patch, we add an extra condition to the "if(prefer_sibling)" check in
update_sd_lb_stats(). We drop the capacity of a group only if the local group
has extra capacity, ie. nr_running < group_capacity. This patch preserves the
original intent of the prefer_siblings check (to spread tasks across the system
in low utilization scenarios) and fixes the case above.
It helps in the following ways:
- In low utilization cases (where nr_tasks << nr_cpus), we still drop
group_capacity down to 1 if we prefer siblings.
- On very busy systems (where nr_tasks >> nr_cpus), sgs.nr_running will most
likely be > sgs.group_capacity.
- When balancing large weight tasks, if the local group does not have extra
capacity, we do not pick the group with the niced task as the busiest group.
This prevents failed balances, active migration and the under-utilization
described above.
Signed-off-by: Nikhil Rao <ncrao@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1287173550-30365-5-git-send-email-ncrao@google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch forces a load balance on a newly idle cpu when the local group has
extra capacity and the busiest group does not have any. It improves system
utilization when balancing tasks with a large weight differential.
Under certain situations, such as a niced down task (i.e. nice = -15) in the
presence of nr_cpus NICE0 tasks, the niced task lands on a sched group and
kicks away other tasks because of its large weight. This leads to sub-optimal
utilization of the machine. Even though the sched group has capacity, it does
not pull tasks because sds.this_load >> sds.max_load, and f_b_g() returns NULL.
With this patch, if the local group has extra capacity, we shortcut the checks
in f_b_g() and try to pull a task over. A sched group has extra capacity if the
group capacity is greater than the number of running tasks in that group.
Thanks to Mike Galbraith for discussions leading to this patch and for the
insight to reuse SD_NEWIDLE_BALANCE.
Signed-off-by: Nikhil Rao <ncrao@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1287173550-30365-4-git-send-email-ncrao@google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
When cycling through sched groups to determine the busiest group, set
group_imb only if the busiest cpu has more than 1 runnable task. This patch
fixes the case where two cpus in a group have one runnable task each, but there
is a large weight differential between these two tasks. The load balancer is
unable to migrate any task from this group, and hence do not consider this
group to be imbalanced.
Signed-off-by: Nikhil Rao <ncrao@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1286996978-7007-3-git-send-email-ncrao@google.com>
[ small code readability edits ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
This patch adds a check in task_hot to return if the task has SCHED_IDLE
policy. SCHED_IDLE tasks have very low weight, and when run with regular
workloads, are typically scheduled many milliseconds apart. There is no
need to consider these tasks hot for load balancing.
Signed-off-by: Nikhil Rao <ncrao@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1287173550-30365-2-git-send-email-ncrao@google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Trades a call + conditional + ret for an unconditional jmp.
Acked-by: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20101014203625.501657727@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Now that there's still only a few users around, rename things to make
them more consistent.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20101014203625.448565169@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
hw_breakpoint creation needs to account stuff per-task to ensure there
is always sufficient hardware resources to back these things due to
ptrace.
With the perf per pmu context changes the event initialization no
longer has access to the event context, for the simple reason that we
need to first find the pmu (result of initialization) before we can
find the context.
This makes hw_breakpoints unhappy, because it can no longer do per
task accounting, cure this by frobbing a task pointer in the event::hw
bits for now...
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <20101014203625.391543667@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
So that we can pass the task pointer to the event allocation, so that
we can use task associated data during event initialization.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20101014203625.340789919@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Currently it looks like find_lively_task_by_vpid() takes a task ref
and relies on find_get_context() to drop it.
The problem is that perf_event_create_kernel_counter() shouldn't be
dropping task refs.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Frederic Weisbecker <fweisbec@gmail.com>
Acked-by: Matt Helsley <matthltc@us.ibm.com>
LKML-Reference: <20101014203625.278436085@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Matt found we trigger the WARN_ON_ONCE() in perf_group_attach() when we take
the move_group path in perf_event_open().
Since we cannot de-construct the group (we rely on it to move the events), we
have to simply ignore the double attach. The group state is context invariant
and doesn't need changing.
Reported-by: Matt Fleming <matt@console-pimps.org>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1287135757.29097.1368.camel@twins>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Provide a mechanism that allows running code in IRQ context. It is
most useful for NMI code that needs to interact with the rest of the
system -- like wakeup a task to drain buffers.
Perf currently has such a mechanism, so extract that and provide it as
a generic feature, independent of perf so that others may also
benefit.
The IRQ context callback is generated through self-IPIs where
possible, or on architectures like powerpc the decrementer (the
built-in timer facility) is set to generate an interrupt immediately.
Architectures that don't have anything like this get to do with a
callback from the timer tick. These architectures can call
irq_work_run() at the tail of any IRQ handlers that might enqueue such
work (like the perf IRQ handler) to avoid undue latencies in
processing the work.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Kyle McMartin <kyle@mcmartin.ca>
Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
[ various fixes ]
Signed-off-by: Huang Ying <ying.huang@intel.com>
LKML-Reference: <1287036094.7768.291.camel@yhuang-dev>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
The group_sched_in() function uses a transactional approach to schedule
a group of events. In a group, either all events can be scheduled or
none are. To schedule each event in, the function calls event_sched_in().
In case of error, event_sched_out() is called on each event in the group.
The problem is that event_sched_out() does not completely cancel the
effects of event_sched_in(). Furthermore event_sched_out() changes the
state of the event as if it had run which is not true is this particular
case.
Those inconsistencies impact time tracking fields and may lead to events
in a group not all reporting the same time_enabled and time_running values.
This is demonstrated with the example below:
$ task -eunhalted_core_cycles,baclears,baclears -e unhalted_core_cycles,baclears,baclears sleep 5
1946101 unhalted_core_cycles (32.85% scaling, ena=829181, run=556827)
11423 baclears (32.85% scaling, ena=829181, run=556827)
7671 baclears (0.00% scaling, ena=556827, run=556827)
2250443 unhalted_core_cycles (57.83% scaling, ena=962822, run=405995)
11705 baclears (57.83% scaling, ena=962822, run=405995)
11705 baclears (57.83% scaling, ena=962822, run=405995)
Notice that in the first group, the last baclears event does not
report the same timings as its siblings.
This issue comes from the fact that tstamp_stopped is updated
by event_sched_out() as if the event had actually run.
To solve the issue, we must ensure that, in case of error, there is
no change in the event state whatsoever. That means timings must
remain as they were when entering group_sched_in().
To do this we defer updating tstamp_running until we know the
transaction succeeded. Therefore, we have split event_sched_in()
in two parts separating the update to tstamp_running.
Similarly, in case of error, we do not want to update tstamp_stopped.
Therefore, we have split event_sched_out() in two parts separating
the update to tstamp_stopped.
With this patch, we now get the following output:
$ task -eunhalted_core_cycles,baclears,baclears -e unhalted_core_cycles,baclears,baclears sleep 5
2492050 unhalted_core_cycles (71.75% scaling, ena=1093330, run=308841)
11243 baclears (71.75% scaling, ena=1093330, run=308841)
11243 baclears (71.75% scaling, ena=1093330, run=308841)
1852746 unhalted_core_cycles (0.00% scaling, ena=784489, run=784489)
9253 baclears (0.00% scaling, ena=784489, run=784489)
9253 baclears (0.00% scaling, ena=784489, run=784489)
Note that the uneven timing between groups is a side effect of
the process spending most of its time sleeping, i.e., not enough
event rotations (but that's a separate issue).
Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <4cb86b4c.41e9d80a.44e9.3e19@mx.google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
You can only call update_context_time() when the context
is active, i.e., the thread it is attached to is still running.
However, perf_event_read() can be called even when the context
is inactive, e.g., user read() the counters. The call to
update_context_time() must be conditioned on the status of
the context, otherwise, bogus time_enabled, time_running may
be returned. Here is an example on AMD64. The task program
is an example from libpfm4. The -p prints deltas every 1s.
$ task -p -e cpu_clk_unhalted sleep 5
2,266,610 cpu_clk_unhalted (0.00% scaling, ena=2,158,982, run=2,158,982)
0 cpu_clk_unhalted (0.00% scaling, ena=2,158,982, run=2,158,982)
0 cpu_clk_unhalted (0.00% scaling, ena=2,158,982, run=2,158,982)
0 cpu_clk_unhalted (0.00% scaling, ena=2,158,982, run=2,158,982)
0 cpu_clk_unhalted (0.00% scaling, ena=2,158,982, run=2,158,982)
5,242,358,071 cpu_clk_unhalted (99.95% scaling, ena=5,000,359,984, run=2,319,270)
Whereas if you don't read deltas, e.g., no call to perf_event_read() until
the process terminates:
$ task -e cpu_clk_unhalted sleep 5
2,497,783 cpu_clk_unhalted (0.00% scaling, ena=2,376,899, run=2,376,899)
Notice that time_enable, time_running are bogus in the first example
causing bogus scaling.
This patch fixes the problem, by conditionally calling update_context_time()
in perf_event_read().
Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: stable@kernel.org
LKML-Reference: <4cb856dc.51edd80a.5ae0.38fb@mx.google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Current look_up_lock_class() doesn't check the parameter "subclass".
This rarely rises problems because the main caller of this function,
register_lock_class(), checks it.
But register_lock_class() is not the only function which calls
look_up_lock_class(). lock_set_class() and its callees also call it.
And lock_set_class() doesn't check this parameter.
This will rise problems when the the value of subclass is larger than
MAX_LOCKDEP_SUBCLASSES. Because the address (used as the key of class)
caliculated with too large subclass has a probability to point
another key in different lock_class_key.
Of course this problem depends on the memory layout and
occurs with really low probability.
Signed-off-by: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
Cc: Dmitry Torokhov <dtor@mail.ru>
Cc: Vojtech Pavlik <vojtech@ucw.cz>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1286958626-986-1-git-send-email-mitake@dcl.info.waseda.ac.jp>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Current lockdep_map only caches one class with subclass == 0,
and looks up hash table of classes when subclass != 0.
It seems that this has no problem because the case of
subclass != 0 is rare. But locks of struct rq are
acquired with subclass == 1 when task migration is executed.
Task migration is high frequent event, so I modified lockdep
to cache subclasses.
I measured the score of perf bench sched messaging.
This patch has slightly but certain (order of milli seconds
or 10 milli seconds) effect when lots of tasks are running.
I'll show the result in the tail of this description.
NR_LOCKDEP_CACHING_CLASSES specifies how many classes can be
cached in the instances of lockdep_map.
I discussed with Peter Zijlstra in LinuxCon Japan about
this approach and he taught me that caching every subclasses(8)
is cleary waste of memory. So number of cached classes
should be configurable.
=== Score comparison of benchmarks ===
# "min" means best score, and "max" means worst score
for i in `seq 1 10`; do ./perf bench -f simple sched messaging; done
before: min: 0.565000, max: 0.583000, avg: 0.572500
after: min: 0.559000, max: 0.568000, avg: 0.563300
# with more processes
for i in `seq 1 10`; do ./perf bench -f simple sched messaging -g 40; done
before: min: 2.274000, max: 2.298000, avg: 2.286300
after: min: 2.242000, max: 2.270000, avg: 2.259700
Signed-off-by: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1286269311-28336-2-git-send-email-mitake@dcl.info.waseda.ac.jp>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Load weights are for the CFS, they do not belong in the RT task. This makes all
RT scheduling classes leave the CFS weights alone.
This fixes a real bug as well: I noticed the following phonomena: a process
elevated to SCHED_RR forks with SCHED_RESET_ON_FORK set, and the child is
indeed SCHED_OTHER, and the niceval is indeed reset to 0. However the weight
inserted by set_load_weight() remains at 0, giving the task insignificat
priority.
With this fix, the weight is reset to what the task had before being elevated
to SCHED_RR/SCHED_FIFO.
Cc: Lennart Poettering <lennart@poettering.net>
Cc: stable@kernel.org
Signed-off-by: Linus Walleij <linus.walleij@stericsson.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1286807811-10568-1-git-send-email-linus.walleij@stericsson.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
In order to separate the stop/migrate work thread from the SCHED_FIFO
implementation, create a special class for it that is of higher priority than
SCHED_FIFO itself.
This currently solves a problem where cpu-hotplug consumes so much cpu-time
that the SCHED_FIFO class gets throttled, but has the bandwidth replenishment
timer pending on the now dead cpu.
It is also required for when we add the planned deadline scheduling class above
SCHED_FIFO, as the stop/migrate thread still needs to transcent those tasks.
Tested-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1285165776.2275.1022.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Labels should be on column 0.
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
