Skip to contentSkip to conversation

posix timers fixlet

PATCH Superseded [email protected] 3 participants 17 messages May 3, 2013 – May 26, 2013

Subsystems (3)

v4 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

v4 cover posix timers fixlet

From: KOSAKI Motohiro <[email protected]>

Glibc's posix timer testcase found a lot of bugs in posix timer code. This series, hopefully, fixes all of them. All patches are independent each other logically.

Changes from v3
- task_sched_runtime() micro optimization add to care tsk->on_cpu. suggested Paul Turner.
- fixed several typo in changelogs.

KOSAKI Motohiro (7):
posix-cpu-timers: don't account cpu timer after stopped thread
runtime accounting
posix-cpu-timers: fix acounting delta_exec twice
posix-cpu-timers: fix wrong timer initialization
posix-cpu-timers: timer functions should use timer time instead of
clock time
posix-cpu-timers: check_thread_timers() uses task_sched_runtime()
sched: task_sched_runtime introduce micro optimization
posix-cpu-timers: cleanup cpu_{clock,timer}_sample{,_group}

 fs/binfmt_elf.c             |    2 +-
 fs/binfmt_elf_fdpic.c       |    2 +-
 include/linux/kernel_stat.h |    5 --
 include/linux/sched.h       |    5 +-
 kernel/posix-cpu-timers.c   |  143 ++++++++++++++++++++++++-------------------
 kernel/sched/core.c         |   34 ++++++-----
 kernel/sched/cputime.c      |   19 +++++-
 kernel/sched/stats.h        |    7 ++
 8 files changed, 125 insertions(+), 92 deletions(-)
#1 lore raw

PATCH v4 1/7 posix-cpu-timers: don't account cpu timer after stopped thread runtime accounting

From: KOSAKI Motohiro <[email protected]>

When tsk->signal->cputimer->running is 1, signal->cputimer and tsk->sum_sched_runtime increase at the same pace because update_curr() increases both accounting.

However, there is one exception. When thread exiting, __exit_signal() turns over task's sum_shced_runtime to sig->sum_sched_runtime, but it doesn't stop signal->cputimer accounting.

This inconsistency makes POSIX timer wake up too early. This patch fixes it.

---
 kernel/sched/stats.h |    7 +++++++
 1 files changed, 7 insertions(+), 0 deletions(-)
kernel/sched/stats.h +7 -0
index 2ef90a5..5a0cfc4 100644
@@ -225,6 +225,13 @@ static inline void account_group_exec_runtime(struct task_struct *tsk,
 	if (!cputimer->running)
 		return;
 
+	/*
+	 * After turning over se.sum_exec_runtime to sig->sum_sched_runtime
+	 * in __exit_signal(), we must not account exec_runtime for consistency.
+	 */
+	if (unlikely(!tsk->sighand))
+		return;
+
 	raw_spin_lock(&cputimer->lock);
 	cputimer->cputime.sum_exec_runtime += ns;
 	raw_spin_unlock(&cputimer->lock);
Signature · 1 line
1.7.1
#2 lore raw

Re: [PATCH 1/7] posix-cpu-timers: don't account cpu timer after stopped thread runtime accounting

Quoted 27 lines · On Fri, May 03, 2013 at 12:47:42AM -0400, [email protected] wrote:

From: KOSAKI Motohiro <[email protected]>

When tsk->signal->cputimer->running is 1, signal->cputimer and tsk->sum_sched_runtime increase at the same pace because update_curr() increases both accounting.

However, there is one exception. When thread exiting, __exit_signal() turns over task's sum_shced_runtime to sig->sum_sched_runtime, but it doesn't stop signal->cputimer accounting.

This inconsistency makes POSIX timer wake up too early. This patch fixes it.

---
 kernel/sched/stats.h |    7 +++++++
 1 files changed, 7 insertions(+), 0 deletions(-)
kernel/sched/stats.h +0 -0
index 2ef90a5..5a0cfc4 100644

kernel/sched/stats.h static inline void account_group_exec_runtime(struct task_struct *tsk, · line 233

Quoted 9 lines
kernel/sched/stats.h +6 -0
@@ -225,6 +225,13 @@ static inline void account_group_exec_runtime(struct task_struct *tsk,
 	if (!cputimer->running)
 		return;
 
+	/*
+	 * After turning over se.sum_exec_runtime to sig->sum_sched_runtime
+	 * in __exit_signal(), we must not account exec_runtime for consistency.
+	 */
+	if (unlikely(!tsk->sighand))
+		return;

Ok, if we want the clock and timer to be consistent, do we also want the same check in account_group_user_time() and account_group_system_time()? The task can still account a tick after autoreaping itself between release_task() and the final schedule().

Quoted 7 lines
kernel/sched/stats.h +1 -0
+
 	raw_spin_lock(&cputimer->lock);
 	cputimer->cputime.sum_exec_runtime += ns;
 	raw_spin_unlock(&cputimer->lock);
Signature · 1 line
1.7.1
#3 lore raw

Re: [PATCH 1/7] posix-cpu-timers: don't account cpu timer after stopped thread runtime accounting

Quoted 10 lines

kernel/sched/stats.h static inline void account_group_exec_runtime(struct task_struct *tsk, · line 233

Quoted 6 lines
kernel/sched/stats.h +6 -0
+     /*
+      * After turning over se.sum_exec_runtime to sig->sum_sched_runtime
+      * in __exit_signal(), we must not account exec_runtime for consistency.
+      */
+     if (unlikely(!tsk->sighand))
+             return;

Ok, if we want the clock and timer to be consistent, do we also want the same check in account_group_user_time() and account_group_system_time()? The task can still account a tick after autoreaping itself between release_task() and the final schedule().

You are right.

That said, current the man pages don't describe this linux specific extensions. So, nobody
(glibc, ltp, and me) tested them. Please give me a couple of days. I'll test and fix this features
too.

timer_create(2): http://man7.org/linux/man-pages/man2/timer_create.2.html

#4 lore raw

Re: [PATCH 1/7] posix-cpu-timers: don't account cpu timer after stopped thread runtime accounting

2013/5/7 KOSAKI Motohiro <[email protected]>:

Quoted 13 lines
Quoted 10 lines

kernel/sched/stats.h static inline void account_group_exec_runtime(struct task_struct *tsk, · line 233

Quoted 6 lines
kernel/sched/stats.h +6 -0
+     /*
+      * After turning over se.sum_exec_runtime to sig->sum_sched_runtime
+      * in __exit_signal(), we must not account exec_runtime for consistency.
+      */
+     if (unlikely(!tsk->sighand))
+             return;

Ok, if we want the clock and timer to be consistent, do we also want the same check in account_group_user_time() and account_group_system_time()? The task can still account a tick after autoreaping itself between release_task() and the final schedule().

You are right.

Quoted 7 lines

That said, current the man pages don't describe this linux specific extensions. So, nobody
(glibc, ltp, and me) tested them. Please give me a couple of days. I'll test and fix this features
too.

timer_create(2): http://man7.org/linux/man-pages/man2/timer_create.2.html

Ah, indeed timer_create() seem to only create CPUCLOCK_SCHED timers. So that issue with timer_gettime becoming asynchonous with clock_gettime can't happen with PROF and VIRT clocks

I see itimers can use those clocks. But there don't seem to be a similar issue with
getitimer/setitimer as they don't have matching clock reads.

Thanks.

#5 lore raw

Re: [PATCH 1/7] posix-cpu-timers: don't account cpu timer after stopped thread runtime accounting

Quoted 22 lines · Frederic Weisbecker

2013/5/7 KOSAKI Motohiro <[email protected]>:

Quoted 13 lines
Quoted 10 lines

kernel/sched/stats.h static inline void account_group_exec_runtime(struct task_struct *tsk, · line 233

Quoted 6 lines
kernel/sched/stats.h +6 -0
+     /*
+      * After turning over se.sum_exec_runtime to sig->sum_sched_runtime
+      * in __exit_signal(), we must not account exec_runtime for consistency.
+      */
+     if (unlikely(!tsk->sighand))
+             return;

Ok, if we want the clock and timer to be consistent, do we also want the same check in account_group_user_time() and account_group_system_time()? The task can still account a tick after autoreaping itself between release_task() and the final schedule().

You are right.

Quoted 7 lines

That said, current the man pages don't describe this linux specific extensions. So, nobody
(glibc, ltp, and me) tested them. Please give me a couple of days. I'll test and fix this features
too.

timer_create(2): http://man7.org/linux/man-pages/man2/timer_create.2.html

Quoted 7 lines

Ah, indeed timer_create() seem to only create CPUCLOCK_SCHED timers. So that issue with timer_gettime becoming asynchonous with clock_gettime can't happen with PROF and VIRT clocks

I see itimers can use those clocks. But there don't seem to be a similar issue with
getitimer/setitimer as they don't have matching clock reads.

OK. I've found PROF and VIRT clock of posix timer have a bug and I could narrow down and fixed it. Please see my next iteration.

Thanks.

#6 lore raw

Re: [PATCH 1/7] posix-cpu-timers: don't account cpu timer after stopped thread runtime accounting

In reply to Frederic Weisbecker
Quoted 31 lines · Frederic Weisbecker
Quoted 30 lines · On Fri, May 03, 2013 at 12:47:42AM -0400, [email protected] wrote:

From: KOSAKI Motohiro <[email protected]>

When tsk->signal->cputimer->running is 1, signal->cputimer and tsk->sum_sched_runtime increase at the same pace because update_curr() increases both accounting.

However, there is one exception. When thread exiting, __exit_signal() turns over task's sum_shced_runtime to sig->sum_sched_runtime, but it doesn't stop signal->cputimer accounting.

This inconsistency makes POSIX timer wake up too early. This patch fixes it.

---
 kernel/sched/stats.h |    7 +++++++
 1 files changed, 7 insertions(+), 0 deletions(-)
kernel/sched/stats.h +0 -0
index 2ef90a5..5a0cfc4 100644
@@ -225,6 +225,13 @@ static inline void account_group_exec_runtime(struct task_struct *tsk,
      if (!cputimer->running)
              return;
 
Quoted 11 lines

kernel/sched/stats.h static inline void account_group_exec_runtime(struct task_struct *tsk, · line 233

Quoted 6 lines
kernel/sched/stats.h +6 -0
+     /*
+      * After turning over se.sum_exec_runtime to sig->sum_sched_runtime
+      * in __exit_signal(), we must not account exec_runtime for consistency.
+      */
+     if (unlikely(!tsk->sighand))
+             return;

Ok, if we want the clock and timer to be consistent, do we also want the same check in account_group_user_time() and account_group_system_time()? The task can still account a tick after autoreaping itself between release_task() and the final schedule().

If I can reply, I believe that Kosaki refactoring is superior to my initial proposal because:

1. Condition will only be evaluated when cputimer is on (and not slow down scheduler when it is off)
2. It handles all the cases when this function is called. There are few other places outside update_curr() that this function is called.

#7 lore raw

Re: [PATCH 1/7] posix-cpu-timers: don't account cpu timer after stopped thread runtime accounting

Quoted 32 lines · Olivier Langlois
Quoted 31 lines · Frederic Weisbecker
Quoted 30 lines · On Fri, May 03, 2013 at 12:47:42AM -0400, [email protected] wrote:

From: KOSAKI Motohiro <[email protected]>

When tsk->signal->cputimer->running is 1, signal->cputimer and tsk->sum_sched_runtime increase at the same pace because update_curr() increases both accounting.

However, there is one exception. When thread exiting, __exit_signal() turns over task's sum_shced_runtime to sig->sum_sched_runtime, but it doesn't stop signal->cputimer accounting.

This inconsistency makes POSIX timer wake up too early. This patch fixes it.

---
 kernel/sched/stats.h |    7 +++++++
 1 files changed, 7 insertions(+), 0 deletions(-)
kernel/sched/stats.h +0 -0
index 2ef90a5..5a0cfc4 100644
@@ -225,6 +225,13 @@ static inline void account_group_exec_runtime(struct task_struct *tsk,
      if (!cputimer->running)
              return;
 
Quoted 4 lines
Quoted 4 lines

kernel/sched/stats.h static inline void account_group_exec_runtime(struct task_struct *tsk, · line 231

Quoted 4 lines
kernel/sched/stats.h +4 -0
+     /*
+      * After turning over se.sum_exec_runtime to sig->sum_sched_runtime
+      * in __exit_signal(), we must not account exec_runtime for consistency.
+      */

Please just precise the nature of that consistency: the fact we want CLOCK_PROCESS_CPUTIME_ID clock and timer to be consistent.

Quoted 7 lines
Quoted 7 lines

kernel/sched/stats.h static inline void account_group_exec_runtime(struct task_struct *tsk, · line 233

Quoted 2 lines
kernel/sched/stats.h +2 -0
+     if (unlikely(!tsk->sighand))
+             return;

Ok, if we want the clock and timer to be consistent, do we also want the same check in account_group_user_time() and account_group_system_time()? The task can still account a tick after autoreaping itself between release_task() and the final schedule().

Quoted 7 lines

If I can reply, I believe that Kosaki refactoring is superior to my initial proposal because:

1. Condition will only be evaluated when cputimer is on (and not slow down scheduler when it is off)
2. It handles all the cases when this function is called. There are few other places outside update_curr() that this function is called.

Agreed. Thanks!

#8 lore raw

Re: [PATCH 1/7] posix-cpu-timers: don't account cpu timer after stopped thread runtime accounting

Quoted 33 lines · Frederic Weisbecker
Quoted 32 lines · Olivier Langlois
Quoted 31 lines · Frederic Weisbecker
Quoted 30 lines · On Fri, May 03, 2013 at 12:47:42AM -0400, [email protected] wrote:

From: KOSAKI Motohiro <[email protected]>

When tsk->signal->cputimer->running is 1, signal->cputimer and tsk->sum_sched_runtime increase at the same pace because update_curr() increases both accounting.

However, there is one exception. When thread exiting, __exit_signal() turns over task's sum_shced_runtime to sig->sum_sched_runtime, but it doesn't stop signal->cputimer accounting.

This inconsistency makes POSIX timer wake up too early. This patch fixes it.

---
 kernel/sched/stats.h |    7 +++++++
 1 files changed, 7 insertions(+), 0 deletions(-)
kernel/sched/stats.h +0 -0
index 2ef90a5..5a0cfc4 100644
@@ -225,6 +225,13 @@ static inline void account_group_exec_runtime(struct task_struct *tsk,
      if (!cputimer->running)
              return;
 
Quoted 7 lines
Quoted 4 lines
Quoted 4 lines

kernel/sched/stats.h static inline void account_group_exec_runtime(struct task_struct *tsk, · line 231

Quoted 4 lines
kernel/sched/stats.h +4 -0
+     /*
+      * After turning over se.sum_exec_runtime to sig->sum_sched_runtime
+      * in __exit_signal(), we must not account exec_runtime for consistency.
+      */

Please just precise the nature of that consistency: the fact we want CLOCK_PROCESS_CPUTIME_ID clock and timer to be consistent.

typo? This patch fixes an inconsistency between a thread cputime and a process cputime.

How is this?

/*
* After turning over se.sum_exec_runtime to sig->sum_sched_runtime
* in __exit_signal(), a per-thread cputime of the thread will be lost. We
* must not account exec_runtime here too because we need to keep
* consistent cputime between per-thread and per-process. Otherwise,
* the inconsistency is observable when single thread program run. */

#9 lore raw

PATCH v4 2/7 posix-cpu-timers: fix acounting delta_exec twice

In reply to [email protected]

From: KOSAKI Motohiro <[email protected]>

Currently glibc rt/tst-cpuclock2 test(*) sporadically fails because scheduler delta can be accounted twice from thread_group_cputimer() and account_group_exec_runtime().

Finally, clock_nanosleep() wakes up before an argument. This is posix violation. This issue was introduced by commit d670ec1317 (posix-cpu-timers:
Cure SMP wobbles).

(*) http://sourceware.org/git/?p=glibc.git;a=blob;f=rt/tst-cpuclock2.c;h=6752721717f959e89c0d692b3f1ee082d507eec2;hb=HEAD

Cc Olivier Langlois <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Frederic Weisbecker <[email protected]>
Cc Ingo Molnar <[email protected]>
---
 fs/binfmt_elf.c           |    2 +-
 fs/binfmt_elf_fdpic.c     |    2 +-
 include/linux/sched.h     |    4 ++--
 kernel/posix-cpu-timers.c |   15 ++++++++++-----
 kernel/sched/core.c       |    6 ++++--
 kernel/sched/cputime.c    |    8 ++++----
 6 files changed, 22 insertions(+), 15 deletions(-)
fs/binfmt_elf.c +1 -1
index 86af964..fea51e7 100644
@@ -1322,7 +1322,7 @@ static void fill_prstatus(struct elf_prstatus *prstatus,
 		 * This is the record for the group leader.  It shows the
 		 * group-wide total, not its individual thread total.
 		 */
-		thread_group_cputime(p, &cputime);
+		thread_group_cputime(p, true, &cputime);
 		cputime_to_timeval(cputime.utime, &prstatus->pr_utime);
 		cputime_to_timeval(cputime.stime, &prstatus->pr_stime);
 	} else {
fs/binfmt_elf_fdpic.c +1 -1
index 9c13e02..ab5b508 100644
@@ -1371,7 +1371,7 @@ static void fill_prstatus(struct elf_prstatus *prstatus,
 		 * This is the record for the group leader.  It shows the
 		 * group-wide total, not its individual thread total.
 		 */
-		thread_group_cputime(p, &cputime);
+		thread_group_cputime(p, true, &cputime);
 		cputime_to_timeval(cputime.utime, &prstatus->pr_utime);
 		cputime_to_timeval(cputime.stime, &prstatus->pr_stime);
 	} else {
include/linux/sched.h +2 -2
index e692a02..7863d4b 100644
@@ -2002,7 +2002,7 @@ static inline void disable_sched_clock_irqtime(void) {}
 #endif
 
 extern unsigned long long
-task_sched_runtime(struct task_struct *task);
+task_sched_runtime(struct task_struct *task, bool add_delta);
 
 /* sched_exec is called by processes performing an exec */
 #ifdef CONFIG_SMP
@@ -2625,7 +2625,7 @@ static inline int spin_needbreak(spinlock_t *lock)
 /*
  * Thread group CPU time accounting.
  */
-void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times);
+void thread_group_cputime(struct task_struct *tsk, bool add_delta, struct task_cputime *times);
 void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times);
 
 static inline void thread_group_cputime_init(struct signal_struct *sig)
kernel/posix-cpu-timers.c +10 -5
index 8fd709c..e56be4c 100644
@@ -220,7 +220,7 @@ static int cpu_clock_sample(const clockid_t which_clock, struct task_struct *p,
 		cpu->cpu = virt_ticks(p);
 		break;
 	case CPUCLOCK_SCHED:
-		cpu->sched = task_sched_runtime(p);
+		cpu->sched = task_sched_runtime(p, true);
 		break;
 	}
 	return 0;
@@ -250,8 +250,13 @@ void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times)
 		 * values through the TIMER_ABSTIME flag, therefore we have
 		 * to synchronize the timer to the clock every time we start
 		 * it.
+		 *
+		 * Do not add the current delta, because
+		 * account_group_exec_runtime() will also this delta and we
+		 * wouldn't want to double account time and get ahead of
+		 * ourselves.
 		 */
-		thread_group_cputime(tsk, &sum);
+		thread_group_cputime(tsk, false, &sum);
 		raw_spin_lock_irqsave(&cputimer->lock, flags);
 		cputimer->running = 1;
 		update_gt_cputime(&cputimer->cputime, &sum);
@@ -275,15 +280,15 @@ static int cpu_clock_sample_group(const clockid_t which_clock,
 	default:
 		return -EINVAL;
 	case CPUCLOCK_PROF:
-		thread_group_cputime(p, &cputime);
+		thread_group_cputime(p, true, &cputime);
 		cpu->cpu = cputime.utime + cputime.stime;
 		break;
 	case CPUCLOCK_VIRT:
-		thread_group_cputime(p, &cputime);
+		thread_group_cputime(p, true, &cputime);
 		cpu->cpu = cputime.utime;
 		break;
 	case CPUCLOCK_SCHED:
-		thread_group_cputime(p, &cputime);
+		thread_group_cputime(p, true, &cputime);
 		cpu->sched = cputime.sum_exec_runtime;
 		break;
 	}
kernel/sched/core.c +4 -2
index 67d0465..ad3339f 100644
@@ -2664,14 +2664,16 @@ unsigned long long task_delta_exec(struct task_struct *p)
  * In case the task is currently running, return the runtime plus current's
  * pending runtime that have not been accounted yet.
  */
-unsigned long long task_sched_runtime(struct task_struct *p)
+unsigned long long task_sched_runtime(struct task_struct *p, bool add_delta)
 {
 	unsigned long flags;
 	struct rq *rq;
 	u64 ns = 0;
 
 	rq = task_rq_lock(p, &flags);
-	ns = p->se.sum_exec_runtime + do_task_delta_exec(p, rq);
+	ns = p->se.sum_exec_runtime;
+	if (add_delta)
+		ns += do_task_delta_exec(p, rq);
 	task_rq_unlock(rq, p, &flags);
 
 	return ns;
kernel/sched/cputime.c +4 -4
index e93cca9..69d3f6c 100644
@@ -293,7 +293,7 @@ static __always_inline bool steal_account_process_tick(void)
  * Accumulate raw cputime values of dead tasks (sig->[us]time) and live
  * tasks (sum on group iteration) belonging to @tsk's group.
  */
-void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times)
+void thread_group_cputime(struct task_struct *tsk, bool add_delta, struct task_cputime *times)
 {
 	struct signal_struct *sig = tsk->signal;
 	cputime_t utime, stime;
@@ -313,7 +313,7 @@ void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times)
 		task_cputime(t, &utime, &stime);
 		times->utime += utime;
 		times->stime += stime;
-		times->sum_exec_runtime += task_sched_runtime(t);
+		times->sum_exec_runtime += task_sched_runtime(t, add_delta);
 	} while_each_thread(tsk, t);
 out:
 	rcu_read_unlock();
@@ -459,7 +459,7 @@ void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime
 {
 	struct task_cputime cputime;
 
-	thread_group_cputime(p, &cputime);
+	thread_group_cputime(p, true, &cputime);
 
 	*ut = cputime.utime;
 	*st = cputime.stime;
@@ -594,7 +594,7 @@ void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime
 {
 	struct task_cputime cputime;
 
-	thread_group_cputime(p, &cputime);
+	thread_group_cputime(p, true, &cputime);
 	cputime_adjust(&cputime, &p->signal->prev_cputime, ut, st);
 }
 #endif /* !CONFIG_VIRT_CPU_ACCOUNTING */
Signature · 1 line
1.7.1
#10 lore raw

Re: [PATCH 2/7] posix-cpu-timers: fix acounting delta_exec twice

Quoted 18 lines · On Fri, May 03, 2013 at 12:47:43AM -0400, [email protected] wrote:

From: KOSAKI Motohiro <[email protected]>

Currently glibc rt/tst-cpuclock2 test(*) sporadically fails because scheduler delta can be accounted twice from thread_group_cputimer() and account_group_exec_runtime().

Finally, clock_nanosleep() wakes up before an argument. This is posix violation. This issue was introduced by commit d670ec1317 (posix-cpu-timers:
Cure SMP wobbles).

(*) http://sourceware.org/git/?p=glibc.git;a=blob;f=rt/tst-cpuclock2.c;h=6752721717f959e89c0d692b3f1ee082d507eec2;hb=HEAD

Cc Olivier Langlois <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Frederic Weisbecker <[email protected]>
Cc Ingo Molnar <[email protected]>

[...]

Quoted 16 lines
kernel/posix-cpu-timers.c +1 -1
index 8fd709c..e56be4c 100644
@@ -220,7 +220,7 @@ static int cpu_clock_sample(const clockid_t which_clock, struct task_struct *p,
 		cpu->cpu = virt_ticks(p);
 		break;
 	case CPUCLOCK_SCHED:
-		cpu->sched = task_sched_runtime(p);
+		cpu->sched = task_sched_runtime(p, true);
 		break;
 	}
 	return 0;
@@ -250,8 +250,13 @@ void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times)
 		 * values through the TIMER_ABSTIME flag, therefore we have
 		 * to synchronize the timer to the clock every time we start

kernel/posix-cpu-timers.c void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times) · line 260

Quoted 10 lines
kernel/posix-cpu-timers.c +6 -1
 		 * it.
+		 *
+		 * Do not add the current delta, because
+		 * account_group_exec_runtime() will also this delta and we
+		 * wouldn't want to double account time and get ahead of
+		 * ourselves.
 		 */
-		thread_group_cputime(tsk, &sum);
+		thread_group_cputime(tsk, false, &sum);
 		raw_spin_lock_irqsave(&cputimer->lock, flags);

I wonder if we should move thread_group_cputime() inside this lock. Otherwise we can miss some updates in-between.

#11 lore raw

Re: [PATCH 2/7] posix-cpu-timers: fix acounting delta_exec twice

Quoted 8 lines
Quoted 8 lines
kernel/posix-cpu-timers.c +4 -0
@@ -250,8 +250,13 @@ void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times)
               * values through the TIMER_ABSTIME flag, therefore we have
               * to synchronize the timer to the clock every time we start
               * it.
+              *
+              * Do not add the current delta, because
+              * account_group_exec_runtime() will also this delta and we
+              * wouldn't want to double account time and get ahead of
Quoted 8 lines

kernel/posix-cpu-timers.c void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times) · line 260

Quoted 5 lines
kernel/posix-cpu-timers.c +2 -1
+              * ourselves.
               */
-             thread_group_cputime(tsk, &sum);
+             thread_group_cputime(tsk, false, &sum);
              raw_spin_lock_irqsave(&cputimer->lock, flags);

I wonder if we should move thread_group_cputime() inside this lock. Otherwise we can miss some updates in-between.

Hmm..

I don't agree with this. Right, we can miss some updates. But 1) cputimer->lock doesn't
prevent any update update_curr() only take rq_lock, and 2) POSIX timer and sleeping
semantics allow longer sleep than an argument. Then, the missing is safe, nobody
can observe which of the timer_setime() syscall and update_curr() happened earlier.

Ah, I'm now finding when update_gt_cputime() effectively work. It helps to avoid timer_settime() vs timer_settime() mess.

#12 lore raw

PATCH v4 3/7 posix-cpu-timers: fix wrong timer initialization

In reply to [email protected]

From: KOSAKI Motohiro <[email protected]>

Currently glibc's rt/tst-cputimer1 testcase sporadically fails because a timer created by timer_create() may fire earlier than specified.

posix_cpu_timer_set() uses "val" as current time for three purpose. 1) initialize sig->cputimer. 2) calculation "old" val. 3) calculations an expires.

(1) and (2) should only use committed time (i.e. without delta_exec) because run_posix_cpu_timers() don't care of delta_exec and we need consistency, but (3) need exact current time (aka cpu clock time) because an expires should be "now + timeout" by definition.

This patch distinguishes between two kinds of "now".

Cc Olivier Langlois <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Frederic Weisbecker <[email protected]>
Cc Ingo Molnar <[email protected]>
---
 include/linux/kernel_stat.h |    5 -----
 kernel/posix-cpu-timers.c   |   14 ++++++++++++--
 kernel/sched/core.c         |   13 -------------
 3 files changed, 12 insertions(+), 20 deletions(-)
include/linux/kernel_stat.h +0 -5
index ed5f6ed..f5d4fdf 100644
@@ -117,11 +117,6 @@ static inline unsigned int kstat_cpu_irqs_sum(unsigned int cpu)
 	return kstat_cpu(cpu).irqs_sum;
 }
 
-/*
- * Lock/unlock the current runqueue - to extract task statistics:
- */
-extern unsigned long long task_delta_exec(struct task_struct *);
-
 extern void account_user_time(struct task_struct *, cputime_t, cputime_t);
 extern void account_system_time(struct task_struct *, int, cputime_t, cputime_t);
 extern void account_steal_time(cputime_t);
kernel/posix-cpu-timers.c +12 -2
index e56be4c..98f354e 100644
@@ -635,7 +635,7 @@ static int cpu_timer_sample_group(const clockid_t which_clock,
 		cpu->cpu = cputime.utime;
 		break;
 	case CPUCLOCK_SCHED:
-		cpu->sched = cputime.sum_exec_runtime + task_delta_exec(p);
+		cpu->sched = cputime.sum_exec_runtime;
 		break;
 	}
 	return 0;
@@ -749,7 +749,17 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int flags,
 	}
 
 	if (new_expires.sched != 0 && !(flags & TIMER_ABSTIME)) {
-		cpu_time_add(timer->it_clock, &new_expires, val);
+		union cpu_time_count now;
+
+		/*
+		 * The expires is "now + timeout" by definition. So,
+		 * we need exact current time.
+		 */
+		if (CPUCLOCK_PERTHREAD(timer->it_clock))
+			now = val;
+		else
+			cpu_clock_sample_group(timer->it_clock, p, &now);
+		cpu_time_add(timer->it_clock, &new_expires, now);
 	}
 
 	/*
kernel/sched/core.c +0 -13
index ad3339f..b817e6d 100644
@@ -2646,19 +2646,6 @@ static u64 do_task_delta_exec(struct task_struct *p, struct rq *rq)
 	return ns;
 }
 
-unsigned long long task_delta_exec(struct task_struct *p)
-{
-	unsigned long flags;
-	struct rq *rq;
-	u64 ns = 0;
-
-	rq = task_rq_lock(p, &flags);
-	ns = do_task_delta_exec(p, rq);
-	task_rq_unlock(rq, p, &flags);
-
-	return ns;
-}
-
 /*
  * Return accounted runtime for the task.
  * In case the task is currently running, return the runtime plus current's
Signature · 1 line
1.7.1
#13 lore raw

PATCH v4 4/7 posix-cpu-timers: timer functions should use timer time instead of clock time

In reply to [email protected]

From: KOSAKI Motohiro <[email protected]>

For process timers, we use cpu_clock_sample_group() and cpu_timer_sample_group() correctly. However, for thread timers, we always use cpu_clock_sample(). This is wrong because a cpu_clock_sample() accounts uncommitted delta_exec too. And this is inconsistent against run_posix_cpu_tiemrs().

This is not big matter because the following workaround code in posix_cpu_timer_get() hides the timer miscalculation issue. However, it makes rq lock held wrongly and it would be better to fix it.

  if (cpu_time_before(timer->it_clock, now, timer->it.cpu.expires)) {
    ...
  } else {
    /*
     * The timer should have expired already, but the firing
     * hasn't taken place yet.  Say it's just about to expire.
     */
     itp->it_value.tv_nsec = 1;
     itp->it_value.tv_sec = 0;
Cc Olivier Langlois <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Frederic Weisbecker <[email protected]>
Cc Ingo Molnar <[email protected]>
---
 kernel/posix-cpu-timers.c |   37 +++++++++++++++++++++++++++----------
 1 files changed, 27 insertions(+), 10 deletions(-)
kernel/posix-cpu-timers.c +27 -10
index 98f354e..c69b8d8 100644
@@ -204,11 +204,10 @@ posix_cpu_clock_set(const clockid_t which_clock, const struct timespec *tp)
 }
 
 
-/*
- * Sample a per-thread clock for the given task.
- */
-static int cpu_clock_sample(const clockid_t which_clock, struct task_struct *p,
-			    union cpu_time_count *cpu)
+static int do_cpu_clock_timer_sample(const clockid_t which_clock,
+				     struct task_struct *p,
+				     bool add_delta,
+				     union cpu_time_count *cpu)
 {
 	switch (CPUCLOCK_WHICH(which_clock)) {
 	default:
@@ -220,12 +219,30 @@ static int cpu_clock_sample(const clockid_t which_clock, struct task_struct *p,
 		cpu->cpu = virt_ticks(p);
 		break;
 	case CPUCLOCK_SCHED:
-		cpu->sched = task_sched_runtime(p, true);
+		cpu->sched = task_sched_runtime(p, add_delta);
 		break;
 	}
 	return 0;
 }
 
+/*
+ * Sample a per-thread clock for the given task.
+ */
+static int cpu_clock_sample(const clockid_t which_clock, struct task_struct *p,
+			    union cpu_time_count *cpu)
+{
+	return do_cpu_clock_timer_sample(which_clock, p, true, cpu);
+}
+
+/*
+ * Sample a per-thread timer clock for the given task.
+ */
+static int cpu_timer_sample(const clockid_t which_clock, struct task_struct *p,
+			    union cpu_time_count *cpu)
+{
+	return do_cpu_clock_timer_sample(which_clock, p, false, cpu);
+}
+
 static void update_gt_cputime(struct task_cputime *a, struct task_cputime *b)
 {
 	if (b->utime > a->utime)
@@ -700,7 +717,7 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int flags,
 	 * check if it's already passed.  In short, we need a sample.
 	 */
 	if (CPUCLOCK_PERTHREAD(timer->it_clock)) {
-		cpu_clock_sample(timer->it_clock, p, &val);
+		cpu_timer_sample(timer->it_clock, p, &val);
 	} else {
 		cpu_timer_sample_group(timer->it_clock, p, &val);
 	}
@@ -756,7 +773,7 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int flags,
 		 * we need exact current time.
 		 */
 		if (CPUCLOCK_PERTHREAD(timer->it_clock))
-			now = val;
+			cpu_clock_sample(timer->it_clock, p, &now);
 		else
 			cpu_clock_sample_group(timer->it_clock, p, &now);
 		cpu_time_add(timer->it_clock, &new_expires, now);
@@ -844,7 +861,7 @@ static void posix_cpu_timer_get(struct k_itimer *timer, struct itimerspec *itp)
 	 * Sample the clock to take the difference with the expiry time.
 	 */
 	if (CPUCLOCK_PERTHREAD(timer->it_clock)) {
-		cpu_clock_sample(timer->it_clock, p, &now);
+		cpu_timer_sample(timer->it_clock, p, &now);
 		clear_dead = p->exit_state;
 	} else {
 		read_lock(&tasklist_lock);
@@ -1168,7 +1185,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer)
 	 * Fetch the current sample and update the timer's expiry time.
 	 */
 	if (CPUCLOCK_PERTHREAD(timer->it_clock)) {
-		cpu_clock_sample(timer->it_clock, p, &now);
+		cpu_timer_sample(timer->it_clock, p, &now);
 		bump_cpu_timer(timer, now);
 		if (unlikely(p->exit_state)) {
 			clear_dead_task(timer, now);
Signature · 1 line
1.7.1
#14 lore raw

PATCH v4 5/7 posix-cpu-timers: check_thread_timers() uses task_sched_runtime()

In reply to [email protected]

From: KOSAKI Motohiro <[email protected]>

A type of tsk->se.sum_exec_runtime is u64. Thus, reading it is racy when running 32bit. We should use task_sched_runtime().

Cc Olivier Langlois <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Frederic Weisbecker <[email protected]>
Cc Ingo Molnar <[email protected]>
---
 kernel/posix-cpu-timers.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)
kernel/posix-cpu-timers.c +2 -1
index c69b8d8..4043282 100644
@@ -958,7 +958,8 @@ static void check_thread_timers(struct task_struct *tsk,
 		struct cpu_timer_list *t = list_first_entry(timers,
 						      struct cpu_timer_list,
 						      entry);
-		if (!--maxfire || tsk->se.sum_exec_runtime < t->expires.sched) {
+		unsigned long long runtime = task_sched_runtime(tsk, false);
+		if (!--maxfire || runtime < t->expires.sched) {
 			tsk->cputime_expires.sched_exp = t->expires.sched;
 			break;
 		}
Signature · 1 line
1.7.1
#15 lore raw

PATCH v4 6/7 sched: task_sched_runtime introduce micro optimization

In reply to [email protected]

From: KOSAKI Motohiro <[email protected]>

rq lock in task_sched_runtime() is necessary for two reasons. 1) accessing se.sum_exec_runtime is not atomic on 32bit and 2) do_task_delta_exec() require it.

So, 64bit can avoid holding rq lock when add_delta is false and delta_exec is 0.

Cc Olivier Langlois <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Frederic Weisbecker <[email protected]>
Cc Ingo Molnar <[email protected]>
---
 kernel/sched/core.c |   15 +++++++++++++++
 1 files changed, 15 insertions(+), 0 deletions(-)
kernel/sched/core.c +15 -0
index b817e6d..75872c3 100644
@@ -2657,6 +2657,21 @@ unsigned long long task_sched_runtime(struct task_struct *p, bool add_delta)
 	struct rq *rq;
 	u64 ns = 0;
 
+#ifdef CONFIG_64BIT
+	/*
+	 * 64-bit doesn't need locks to atomically read a 64bit value. So we
+	 * have two optimization chances, 1) when caller doesn't need
+	 * delta_exec and 2) when the task's delta_exec is 0. The former is
+	 * obvious. The latter is complicated. reading ->on_cpu is racy, but
+	 * this is ok. If we race with it leaving cpu, we'll take a lock. So
+	 * we're correct. If we race with it entering cpu, unaccounted time
+	 * is 0. This is indistinguishable from the read occurring a few
+	 * cycles earlier.
+	 */
+	if (!add_delta || !p->on_cpu)
+		return p->se.sum_exec_runtime;
+#endif
+
 	rq = task_rq_lock(p, &flags);
 	ns = p->se.sum_exec_runtime;
 	if (add_delta)
Signature · 1 line
1.7.1
#16 lore raw

PATCH v4 7/7 posix-cpu-timers: cleanup cpu_{clock,timer}_sample{,_group}

In reply to [email protected]

From: KOSAKI Motohiro <[email protected]>

Now we have four similar timer related functions, cpu_clock_sample(), cpu_clock_sample_group(), cpu_timer_sample() and cpu_timer_sample_group().

For readability, make do_cpu_clock_timer_sample() and thread_cputime() helper functions and all *_sample functions use these.

Cc Olivier Langlois <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Frederic Weisbecker <[email protected]>
Cc Ingo Molnar <[email protected]>
---
 include/linux/sched.h     |    1 +
 kernel/posix-cpu-timers.c |  132 +++++++++++++++++++-------------------------
 kernel/sched/cputime.c    |   11 ++++
 3 files changed, 69 insertions(+), 75 deletions(-)
include/linux/sched.h +1 -0
index 7863d4b..73ac8fa 100644
@@ -2622,6 +2622,7 @@ static inline int spin_needbreak(spinlock_t *lock)
 #endif
 }
 
+void thread_cputime(struct task_struct *tsk, bool add_delta, struct task_cputime *times);
 /*
  * Thread group CPU time accounting.
  */
kernel/posix-cpu-timers.c +57 -75

170 lines not on this page · raw message

kernel/sched/cputime.c +11 -0
index 69d3f6c..29ef7d7 100644
@@ -289,6 +289,17 @@ static __always_inline bool steal_account_process_tick(void)
 	return false;
 }
 
+void thread_cputime(struct task_struct *tsk, bool add_delta, struct task_cputime *times)
+{
+	struct signal_struct *sig = tsk->signal;
+	cputime_t utime, stime;
+
+	task_cputime(tsk, &utime, &stime);
+	times->utime = utime;
+	times->stime = stime;
+	times->sum_exec_runtime = task_sched_runtime(tsk, add_delta);
+}
+
 /*
  * Accumulate raw cputime values of dead tasks (sig->[us]time) and live
  * tasks (sum on group iteration) belonging to @tsk's group.
Signature · 1 line
1.7.1