The patch titled
tick: fix update_ts_time_stat idle accounting
has been added to the -mm tree. Its filename is
tick-fix-update_ts_time_stat-idle-accounting.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
See http://userweb.kernel.org/~akpm/stuff/added-to-mm.txt to find
out what to do about this
The current -mm tree may be found at http://userweb.kernel.org/~akpm/mmotm/
------------------------------------------------------
Subject: tick: fix update_ts_time_stat idle accounting
From: Michal Hocko <mhocko@xxxxxxx>
update_ts_time_stat currently updates idle time even if we are in iowait
loop at the moment. The only real users of the idle counter (via
get_cpu_idle_time_us) are CPU governors and they expect to get cumulative
time for both idle and iowait times. The value (idle_sleeptime) is also
printed to userspace by print_cpu but it prints both idle and iowait times
so the idle part is misleading.
Let's clean this up and fix update_ts_time_stat to account both counters
properly and update consumers of idle to consider iowait time as well. If
we do this we might use get_cpu_{idle,iowait}_time_us from other contexts
as well and we will get expected values.
Signed-off-by: Michal Hocko <mhocko@xxxxxxx>
Cc: Dave Jones <davej@xxxxxxxxxx>
Cc: Arnd Bergmann <arnd@xxxxxxxx>
Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx>
Cc: Alexey Dobriyan <adobriyan@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
drivers/cpufreq/cpufreq_conservative.c | 4 +++-
drivers/cpufreq/cpufreq_ondemand.c | 4 +++-
kernel/time/tick-sched.c | 8 ++++----
3 files changed, 10 insertions(+), 6 deletions(-)
diff -puN drivers/cpufreq/cpufreq_conservative.c~tick-fix-update_ts_time_stat-idle-accounting drivers/cpufreq/cpufreq_conservative.c
--- a/drivers/cpufreq/cpufreq_conservative.c~tick-fix-update_ts_time_stat-idle-accounting
+++ a/drivers/cpufreq/cpufreq_conservative.c
@@ -120,10 +120,12 @@ static inline cputime64_t get_cpu_idle_t
static inline cputime64_t get_cpu_idle_time(unsigned int cpu, cputime64_t *wall)
{
- u64 idle_time = get_cpu_idle_time_us(cpu, wall);
+ u64 idle_time = get_cpu_idle_time_us(cpu, NULL);
if (idle_time == -1ULL)
return get_cpu_idle_time_jiffy(cpu, wall);
+ else
+ idle_time += get_cpu_iowait_time_us(cpu, wall);
return idle_time;
}
diff -puN drivers/cpufreq/cpufreq_ondemand.c~tick-fix-update_ts_time_stat-idle-accounting drivers/cpufreq/cpufreq_ondemand.c
--- a/drivers/cpufreq/cpufreq_ondemand.c~tick-fix-update_ts_time_stat-idle-accounting
+++ a/drivers/cpufreq/cpufreq_ondemand.c
@@ -144,10 +144,12 @@ static inline cputime64_t get_cpu_idle_t
static inline cputime64_t get_cpu_idle_time(unsigned int cpu, cputime64_t *wall)
{
- u64 idle_time = get_cpu_idle_time_us(cpu, wall);
+ u64 idle_time = get_cpu_idle_time_us(cpu, NULL);
if (idle_time == -1ULL)
return get_cpu_idle_time_jiffy(cpu, wall);
+ else
+ idle_time += get_cpu_iowait_time_us(cpu, wall);
return idle_time;
}
diff -puN kernel/time/tick-sched.c~tick-fix-update_ts_time_stat-idle-accounting kernel/time/tick-sched.c
--- a/kernel/time/tick-sched.c~tick-fix-update_ts_time_stat-idle-accounting
+++ a/kernel/time/tick-sched.c
@@ -158,9 +158,10 @@ update_ts_time_stats(int cpu, struct tic
if (ts->idle_active) {
delta = ktime_sub(now, ts->idle_entrytime);
- ts->idle_sleeptime = ktime_add(ts->idle_sleeptime, delta);
if (nr_iowait_cpu(cpu) > 0)
ts->iowait_sleeptime = ktime_add(ts->iowait_sleeptime, delta);
+ else
+ ts->idle_sleeptime = ktime_add(ts->idle_sleeptime, delta);
ts->idle_entrytime = now;
}
@@ -199,8 +200,7 @@ static ktime_t tick_nohz_start_idle(int
* @last_update_time: variable to store update time in
*
* Return the cummulative idle time (since boot) for a given
- * CPU, in microseconds. The idle time returned includes
- * the iowait time (unlike what "top" and co report).
+ * CPU, in microseconds.
*
* This time is measured via accounting rather than sampling,
* and is as accurate as ktime_get() is.
@@ -220,7 +220,7 @@ u64 get_cpu_idle_time_us(int cpu, u64 *l
}
EXPORT_SYMBOL_GPL(get_cpu_idle_time_us);
-/*
+/**
* get_cpu_iowait_time_us - get the total iowait time of a cpu
* @cpu: CPU number to query
* @last_update_time: variable to store update time in
_
Patches currently in -mm which might be from mhocko@xxxxxxx are
memcg-pin-execution-to-current-cpu-while-draining-stock.patch
memcg-make-oom_lock-0-and-1-based-rather-than-coutner.patch
linux-next.patch
cputime-clean-up-cputime_to_usecs-and-usecs_to_cputime-macros.patch
tick-fix-update_ts_time_stat-idle-accounting.patch
nohz-do-not-update-idle-iowait-counters-from-get_cpu_idleiowait_time_us-if-not-asked.patch
proc-consider-no_hz-when-printing-idle-and-iowait-times.patch
mm-compaction-trivial-clean-up-in-acct_isolated.patch
mm-change-isolate-mode-from-define-to-bitwise-type.patch
mm-compaction-make-isolate_lru_page-filter-aware.patch
mm-zone_reclaim-make-isolate_lru_page-filter-aware.patch
mm-migration-clean-up-unmap_and_move.patch
mm-page-writebackc-make-determine_dirtyable_memory-static-again.patch
oom-avoid-killing-kthreads-if-they-assume-the-oom-killed-threads-mm.patch
mm-vmscan-fix-force-scanning-small-targets-without-swap.patch
mm-vmscan-drop-nr_force_scan-from-get_scan_count.patch
memcg-rename-mem-variable-to-memcg.patch
memcg-fix-oom-schedule_timeout.patch
memcg-do-not-expose-uninitialized-mem_cgroup_per_node-to-world.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
