Git commit 09a1d34f85 "nohz: Make idle/iowait counter update
conditional" introduced a bug in regard to cpu hotplug. The effect is
that the number of idle ticks in the cpu summary line in /proc/stat is
still counting ticks for offline cpus.
Reproduction is easy, just start a workload that keeps all cpus busy,
switch off one or more cpus and then watch the idle field in top.
On a dual-core with one cpu 100% busy and one offline cpu you will get
something like this:
%Cpu(s): 48.7 us, 1.3 sy, 0.0 ni, 50.0 id, 0.0 wa, 0.0 hi, 0.0 si,
%0.0 st
The problem is that an offline cpu still has ts->idle_active == 1.
To fix this we should make sure that the cpu is online when calling
get_cpu_idle_time_us and get_cpu_iowait_time_us.
[Srivatsa: Rebased to current mainline]
Reported-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Reviewed-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Signed-off-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/20121010061820.8999.57245.stgit@srivatsabhat.in.ibm.com
Cc: deepthi@linux.vnet.ibm.com
Cc: stable@vger.kernel.org
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Git commit a25cac5198 "proc: Consider NO_HZ when printing idle and
iowait times" changes the code for /proc/stat to use get_cpu_idle_time_us
and get_cpu_iowait_time_us if the system is running with nohz enabled.
For architectures which define arch_idle_time (currently s390 only)
this is a change for the worse. The result of arch_idle_time is supposed
to be the exact sleep time of the target cpu and should be used instead
of the value kept by the scheduler.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Reviewed-by: Michal Hocko <mhocko@suse.cz>
Reviewed-by: Srivatsa S. Bhat <srivatsa.bhat@linux.vnet.ibm.com>
Link: http://lkml.kernel.org/r/20120330122308.18720283@de.ibm.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
== stat_check.py
num = 0
with open("/proc/stat") as f:
while num < 1000 :
data = f.read()
f.seek(0, 0)
num = num + 1
==
perf shows
20.39% stat_check.py [kernel.kallsyms] [k] format_decode
13.41% stat_check.py [kernel.kallsyms] [k] number
12.61% stat_check.py [kernel.kallsyms] [k] vsnprintf
10.85% stat_check.py [kernel.kallsyms] [k] memcpy
4.85% stat_check.py [kernel.kallsyms] [k] radix_tree_lookup
4.43% stat_check.py [kernel.kallsyms] [k] seq_printf
This patch removes most of calls to vsnprintf() by adding num_to_str()
and seq_print_decimal_ull(), which prints decimal numbers without rich
functions provided by printf().
On my 8cpu box.
== Before patch ==
[root@bluextal test]# time ./stat_check.py
real 0m0.150s
user 0m0.026s
sys 0m0.121s
== After patch ==
[root@bluextal test]# time ./stat_check.py
real 0m0.055s
user 0m0.022s
sys 0m0.030s
[akpm@linux-foundation.org: remove incorrect comment, use less statck in num_to_str(), move comment from .h to .c, simplify seq_put_decimal_ull()]
[andrea@betterlinux.com: avoid breaking the ABI in /proc/stat]
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Signed-off-by: Andrea Righi <andrea@betterlinux.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Glauber Costa <glommer@parallels.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Paul Turner <pjt@google.com>
Cc: Russell King <rmk@arm.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
On a typical 16 cpus machine, "cat /proc/stat" gives more than 4096 bytes,
and is slow :
# strace -T -o /tmp/STRACE cat /proc/stat | wc -c
5826
# grep "cpu " /tmp/STRACE
read(0, "cpu 1949310 19 2144714 12117253"..., 32768) = 5826 <0.001504>
Thats partly because show_stat() must be called twice since initial
buffer size is too small (4096 bytes for less than 32 possible cpus)
Fix this by :
1) Taking into account nr_irqs in the initial buffer sizing.
2) Using ksize() to allow better filling of initial buffer.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Glauber Costa <glommer@parallels.com>
Cc: Russell King - ARM Linux <linux@arm.linux.org.uk>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Paul Turner <pjt@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit 3292beb340 ("sched/accounting: Change cpustat fields to an array")
deleted the code which provides us with the sum of all interrupts in the
system, causing vmstat to report zero interrupts occuring in the system.
Fix this by restoring the code.
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Tested-by: Russell King <rmk+kernel@arm.linux.org.uk> # [on ARM]
Tested-by: Tony Luck <tony.luck@intel.com>
Tested-by: Steven Rostedt <rostedt@goodmis.org>
Cc: Glauber Costa <glommer@parallels.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Paul Tuner <pjt@google.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
* 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (40 commits)
sched/tracing: Add a new tracepoint for sleeptime
sched: Disable scheduler warnings during oopses
sched: Fix cgroup movement of waking process
sched: Fix cgroup movement of newly created process
sched: Fix cgroup movement of forking process
sched: Remove cfs bandwidth period check in tg_set_cfs_period()
sched: Fix load-balance lock-breaking
sched: Replace all_pinned with a generic flags field
sched: Only queue remote wakeups when crossing cache boundaries
sched: Add missing rcu_dereference() around ->real_parent usage
[S390] fix cputime overflow in uptime_proc_show
[S390] cputime: add sparse checking and cleanup
sched: Mark parent and real_parent as __rcu
sched, nohz: Fix missing RCU read lock
sched, nohz: Set the NOHZ_BALANCE_KICK flag for idle load balancer
sched, nohz: Fix the idle cpu check in nohz_idle_balance
sched: Use jump_labels for sched_feat
sched/accounting: Fix parameter passing in task_group_account_field
sched/accounting: Fix user/system tick double accounting
sched/accounting: Re-use scheduler statistics for the root cgroup
...
Fix up conflicts in
- arch/ia64/include/asm/cputime.h, include/asm-generic/cputime.h
usecs_to_cputime64() vs the sparse cleanups
- kernel/sched/fair.c, kernel/time/tick-sched.c
scheduler changes in multiple branches
Commit 2a95ea6c0d ("procfs: do not overflow get_{idle,iowait}_time
for nohz") did not take into account that one some architectures jiffies
and cputime use different units.
This causes get_idle_time() to return numbers in the wrong units, making
the idle time fields in /proc/stat wrong.
Instead of converting the usec value returned by
get_cpu_{idle,iowait}_time_us to units of jiffies, use the new function
usecs_to_cputime64 to convert it to the correct unit of cputime64_t.
Signed-off-by: Andreas Schwab <schwab@linux-m68k.org>
Acked-by: Michal Hocko <mhocko@suse.cz>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: "Artem S. Tashkinov" <t.artem@mailcity.com>
Cc: Dave Jones <davej@redhat.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "Luck, Tony" <tony.luck@intel.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Since commit a25cac5198 ("proc: Consider NO_HZ when printing idle and
iowait times") we are reporting idle/io_wait time also while a CPU is
tickless. We rely on get_{idle,iowait}_time functions to retrieve
proper data.
These functions, however, use usecs_to_cputime to translate micro
seconds time to cputime64_t. This is just an alias to usecs_to_jiffies
which reduces the data type from u64 to unsigned int and also checks
whether the given parameter overflows jiffies_to_usecs(MAX_JIFFY_OFFSET)
and returns MAX_JIFFY_OFFSET in that case.
When we overflow depends on CONFIG_HZ but especially for CONFIG_HZ_300
it is quite low (1431649781) so we are getting MAX_JIFFY_OFFSET for
>3000s! until we overflow unsigned int. Just for reference
CONFIG_HZ_100 has an overflow window around 20s, CONFIG_HZ_250 ~8s and
CONFIG_HZ_1000 ~2s.
This results in a bug when people saw [h]top going mad reporting 100%
CPU usage even though there was basically no CPU load. The reason was
simply that /proc/stat stopped reporting idle/io_wait changes (and
reported MAX_JIFFY_OFFSET) and so the only change happening was for user
system time.
Let's use nsecs_to_jiffies64 instead which doesn't reduce the precision
to 32b type and it is much more appropriate for cumulative time values
(unlike usecs_to_jiffies which intended for timeout calculations).
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Tested-by: Artem S. Tashkinov <t.artem@mailcity.com>
Cc: Dave Jones <davej@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch changes fields in cpustat from a structure, to an
u64 array. Math gets easier, and the code is more flexible.
Signed-off-by: Glauber Costa <glommer@parallels.com>
Reviewed-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Paul Tuner <pjt@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1322498719-2255-2-git-send-email-glommer@parallels.com
Signed-off-by: Ingo Molnar <mingo@elte.hu>
show_stat handler of the /proc/stat file relies on kstat_cpu(cpu)
statistics when priting information about idle and iowait times.
This is OK if we are not using tickless kernel (CONFIG_NO_HZ) because
counters are updated periodically.
With NO_HZ things got more tricky because we are not doing idle/iowait
accounting while we are tickless so the value might get outdated.
Users of /proc/stat will notice that by unchanged idle/iowait values
which is then interpreted as 0% idle/iowait time. From the user space
POV this is an unexpected behavior and a change of the interface.
Let's fix this by using get_cpu_{idle,iowait}_time_us which accounts the
total idle/iowait time since boot and it doesn't rely on sampling or any
other periodic activity. Fall back to the previous behavior if NO_HZ is
disabled or not configured.
Signed-off-by: Michal Hocko <mhocko@suse.cz>
Cc: Dave Jones <davej@redhat.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Link: http://lkml.kernel.org/r/39181366adac1b39cb6aa3cd53ff0f7c78d32676.1314172057.git.mhocko@suse.cz
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
There is a macro for the max size kmalloc can allocate, so use it instead
of a hardcoded number.
Signed-off-by: Yuanhan Liu <yuanhan.liu@linux.intel.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
For string without format specifiers, use seq_puts().
For seq_printf("\n"), use seq_putc('\n').
text data bss dec hex filename
61866 488 112 62466 f402 fs/proc/proc.o
61729 488 112 62329 f379 fs/proc/proc.o
----------------------------------------------------
-139
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
In /proc/stat, the number of per-IRQ event is shown by making a sum each
irq's events on all cpus. But we can make use of kstat_irqs().
kstat_irqs() do the same calculation, If !CONFIG_GENERIC_HARDIRQ,
it's not a big cost. (Both of the number of cpus and irqs are small.)
If a system is very big and CONFIG_GENERIC_HARDIRQ, it does
for_each_irq()
for_each_cpu()
- look up a radix tree
- read desc->irq_stat[cpu]
This seems not efficient. This patch adds kstat_irqs() for
CONFIG_GENRIC_HARDIRQ and change the calculation as
for_each_irq()
look up radix tree
for_each_cpu()
- read desc->irq_stat[cpu]
This reduces cost.
A test on (4096cpusp, 256 nodes, 4592 irqs) host (by Jack Steiner)
%time cat /proc/stat > /dev/null
Before Patch: 2.459 sec
After Patch : .561 sec
[akpm@linux-foundation.org: unexport kstat_irqs, coding-style tweaks]
[akpm@linux-foundation.org: fix unused variable 'per_irq_sum']
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Tested-by: Jack Steiner <steiner@sgi.com>
Acked-by: Jack Steiner <steiner@sgi.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
/proc/stat shows the total number of all interrupts to each cpu. But when
the number of IRQs are very large, it take very long time and 'cat
/proc/stat' takes more than 10 secs. This is because sum of all irq
events are counted when /proc/stat is read. This patch adds "sum of all
irq" counter percpu and reduce read costs.
The cost of reading /proc/stat is important because it's used by major
applications as 'top', 'ps', 'w', etc....
A test on a mechin (4096cpu, 256 nodes, 4592 irqs) shows
%time cat /proc/stat > /dev/null
Before Patch: 12.627 sec
After Patch: 2.459 sec
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Tested-by: Jack Steiner <steiner@sgi.com>
Acked-by: Jack Steiner <steiner@sgi.com>
Cc: Yinghai Lu <yinghai@kernel.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
CPU time of a guest is always accounted in 'user' time
without concern for the nice value of its counterpart
process although the guest is scheduled under the nice
value.
This patch fixes the defect and accounts cpu time of
a niced guest in 'nice' time as same as a niced process.
And also the patch adds 'guest_nice' to cpuacct. The
value provides niced guest cpu time which is like 'nice'
to 'user'.
The original discussions can be found here:
http://www.mail-archive.com/kvm@vger.kernel.org/msg23982.htmlhttp://www.mail-archive.com/kvm@vger.kernel.org/msg23860.html
Signed-off-by: Ryota Ozaki <ozaki.ryota@gmail.com>
Acked-by: Avi Kivity <avi@redhat.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <1256314810-7897-1-git-send-email-ozaki.ryota@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Export statistics for softirq in /proc/softirqs and /proc/stat.
1. /proc/softirqs
Implement /proc/softirqs which shows the number of softirq
for each CPU like /proc/interrupts.
2. /proc/stat
Add the "softirq" line to /proc/stat.
This line shows the number of softirq for all cpu.
The first column is the total of all softirqs and
each subsequent column is the total for particular softirq.
[kosaki.motohiro@jp.fujitsu.com: remove redundant for_each_possible_cpu() loop]
Signed-off-by: Keika Kobayashi <kobayashi.kk@ncos.nec.co.jp>
Reviewed-by: Hiroshi Shimamoto <h-shimamoto@ct.jp.nec.com>
Cc: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Eric Dumazet <dada1@cosmosbay.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The cpu idle field in the output of /proc/stat is too small for cpus
that have been idle for more than a tick. Add the architecture hook
arch_idle_time that allows to add the not accounted idle time of a
sleeping cpu without waking the cpu.
The s390 implementation of arch_idle_time uses the already existing
s390_idle_data per_cpu variable to find the sleep time of a neighboring
idle cpu.
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Impact: restructure code to fix compiler warning
commit 240d367b4e moved desc usage point
into #ifdef CONFIG_SPARSE_IRQ.
Eliminate the desc variable, otherwise following warning happens:
fs/proc/stat.c: In function 'show_stat':
fs/proc/stat.c:31: warning: unused variable 'desc'
[ akpm: cleaned up the patch to remove #ifdef ]
Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: build fix on Alpha
-tip testing found this build failure on the Alpha defconfig:
/home/mingo/tip/fs/proc/stat.c: In function 'show_stat':
/home/mingo/tip/fs/proc/stat.c:48: error: implicit declaration of function 'for_each_irq_desc'
/home/mingo/tip/fs/proc/stat.c:48: error: expected ';' before '{' token
can not use irq_desc() in stat.c on older architectures.
Signed-off-by: Yinghai Lu <yinghai@kernel.orgg>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: new feature
Problem on distro kernels: irq_desc[NR_IRQS] takes megabytes of RAM with
NR_CPUS set to large values. The goal is to be able to scale up to much
larger NR_IRQS value without impacting the (important) common case.
To solve this, we generalize irq_desc[NR_IRQS] to an (optional) array of
irq_desc pointers.
When CONFIG_SPARSE_IRQ=y is used, we use kzalloc_node to get irq_desc,
this also makes the IRQ descriptors NUMA-local (to the site that calls
request_irq()).
This gets rid of the irq_cfg[] static array on x86 as well: irq_cfg now
uses desc->chip_data for x86 to store irq_cfg.
Signed-off-by: Yinghai Lu <yinghai@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Michal Hocko