Re: [BUG] kernel freezes with latest tree

From: Eric Dumazet
Date: Tue Jan 10 2012 - 03:16:17 EST

Le mardi 10 janvier 2012 Ã 06:03 +0100, Eric Dumazet a Ãcrit :
> Le mardi 10 janvier 2012 Ã 05:57 +0100, Eric Dumazet a Ãcrit :
> > Hi Linus
> >
> > I got some freezes on two different machines, using latest kernel.
> >
> > while :; do hackbench 10 thread 4000; done
> >
> > Not sure I'll have time today to find the problem.
> >
> > It might be related to "perf top" also being run at least once.
> >
>
> Hmm, I can trigger the bug without ever using "perf".
>
>

OK I managed to bisect it, but I have to run now.

$ git bisect log
git bisect start
# bad: [31ae26ae7ba5ff6209b9ec5d1e1ba9442d6b87e6] Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6
git bisect bad 31ae26ae7ba5ff6209b9ec5d1e1ba9442d6b87e6
# good: [805a6af8dba5dfdd35ec35dc52ec0122400b2610] Linux 3.2
git bisect good 805a6af8dba5dfdd35ec35dc52ec0122400b2610
# bad: [e4e88f31bcb5f05f24b9ae518d4ecb44e1a7774d] Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc
git bisect bad e4e88f31bcb5f05f24b9ae518d4ecb44e1a7774d
# good: [ab9c17a009ee8eb8c667f22dc0be0709effceab9] mlx4_core: Modify driver initialization flow to accommodate SRIOV for Ethernet
git bisect good ab9c17a009ee8eb8c667f22dc0be0709effceab9
# good: [117ff42fd43e92d24c6aa6f3e4f0f1e1edada140] Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net
git bisect good 117ff42fd43e92d24c6aa6f3e4f0f1e1edada140
# bad: [67b0243131150391125d8d0beb5359d7aec78b55] Merge branch 'x86-apic-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
git bisect bad 67b0243131150391125d8d0beb5359d7aec78b55
# good: [423d091dfe58d3109d84c408810a7cfa82f6f184] Merge branch 'core-rcu-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
git bisect good 423d091dfe58d3109d84c408810a7cfa82f6f184
# good: [f1ac18af219835fd5b8e19c14d2dd75c55f78737] perf: Add support for PERF_HW_COUNT_REF_CPU_CYCLES
git bisect good f1ac18af219835fd5b8e19c14d2dd75c55f78737
# good: [1ac9bc6943edf7d181b4b1cc734981350d4f6bae] sched/tracing: Add a new tracepoint for sleeptime
git bisect good 1ac9bc6943edf7d181b4b1cc734981350d4f6bae
# bad: [0db49b72bce26341274b74fd968501489a361ae3] Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
git bisect bad 0db49b72bce26341274b74fd968501489a361ae3
# good: [1b5495043d5bc058def21f9b66fd8feaa794eb44] perf tools: Moving code in some files
git bisect good 1b5495043d5bc058def21f9b66fd8feaa794eb44
# good: [29c9862f1b818bf4caa4c48a30dbe5f25c84ee08] perf session: Remove impossible condition check
git bisect good 29c9862f1b818bf4caa4c48a30dbe5f25c84ee08
# good: [466e2876bcb9ddc9b92502c46689679bee7d72a0] perf script: Kill script_spec__delete
git bisect good 466e2876bcb9ddc9b92502c46689679bee7d72a0
# good: [35b740e4662ef386f0c60e1b60aaf5b44db9914c] Merge branch 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
git bisect good 35b740e4662ef386f0c60e1b60aaf5b44db9914c


Could it be a merge error ?

commit 0db49b72bce26341274b74fd968501489a361ae3
Merge: 35b740e 1ac9bc6
Author: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Date: Fri Jan 6 08:33:28 2012 -0800

Merge branch 'sched-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip

* '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

diff --cc arch/ia64/include/asm/cputime.h
index 5a274af,461e52f..3deac95
--- a/arch/ia64/include/asm/cputime.h
+++ b/arch/ia64/include/asm/cputime.h
@@@ -58,9 -46,10 +46,12 @@@ typedef u64 __nocast cputime64_t
/*
* Convert cputime <-> microseconds
*/
- #define cputime_to_usecs(__ct) ((__ct) / NSEC_PER_USEC)
- #define usecs_to_cputime(__usecs) ((__usecs) * NSEC_PER_USEC)
- #define usecs_to_cputime64(__usecs) usecs_to_cputime(__usecs)
+ #define cputime_to_usecs(__ct) \
+ ((__force u64)(__ct) / NSEC_PER_USEC)
+ #define usecs_to_cputime(__usecs) \
+ (__force cputime_t)((__usecs) * NSEC_PER_USEC)
++#define usecs_to_cputime64(__usecs) \
++ (__force cputime64_t)((__usecs) * NSEC_PER_USEC)

/*
* Convert cputime <-> seconds
diff --cc arch/powerpc/include/asm/cputime.h
index 98b7c4b,e94935c..6ec1c38
--- a/arch/powerpc/include/asm/cputime.h
+++ b/arch/powerpc/include/asm/cputime.h
@@@ -147,11 -131,9 +131,11 @@@ static inline cputime_t usecs_to_cputim
}
if (sec)
ct += (cputime_t) sec * tb_ticks_per_sec;
- return ct;
+ return (__force cputime_t) ct;
}

+#define usecs_to_cputime64(us) usecs_to_cputime(us)
+
/*
* Convert cputime <-> seconds
*/
diff --cc arch/s390/include/asm/cputime.h
index b9acaaa,0887a04..c23c390
--- a/arch/s390/include/asm/cputime.h
+++ b/arch/s390/include/asm/cputime.h
@@@ -75,20 -62,16 +62,18 @@@ static inline cputime64_t jiffies64_to_
/*
* Convert cputime to microseconds and back.
*/
- static inline unsigned int
- cputime_to_usecs(const cputime_t cputime)
+ static inline unsigned int cputime_to_usecs(const cputime_t cputime)
{
- return cputime_div(cputime, 4096);
+ return (__force unsigned long long) cputime >> 12;
}

- static inline cputime_t
- usecs_to_cputime(const unsigned int m)
+ static inline cputime_t usecs_to_cputime(const unsigned int m)
{
- return (cputime_t) m * 4096;
+ return (__force cputime_t)(m * 4096ULL);
}

+#define usecs_to_cputime64(m) usecs_to_cputime(m)
+
/*
* Convert cputime to milliseconds and back.
*/
diff --cc fs/proc/stat.c
index 0855e6f,2527a68..d76ca6a
--- a/fs/proc/stat.c
+++ b/fs/proc/stat.c
@@@ -29,10 -28,10 +28,10 @@@ static u64 get_idle_time(int cpu

if (idle_time == -1ULL) {
/* !NO_HZ so we can rely on cpustat.idle */
- idle = kstat_cpu(cpu).cpustat.idle;
- idle = cputime64_add(idle, arch_idle_time(cpu));
+ idle = kcpustat_cpu(cpu).cpustat[CPUTIME_IDLE];
+ idle += arch_idle_time(cpu);
} else
- idle = nsecs_to_jiffies64(1000 * idle_time);
+ idle = usecs_to_cputime64(idle_time);

return idle;
}
@@@ -44,9 -42,9 +42,9 @@@ static u64 get_iowait_time(int cpu

if (iowait_time == -1ULL)
/* !NO_HZ so we can rely on cpustat.iowait */
- iowait = kstat_cpu(cpu).cpustat.iowait;
+ iowait = kcpustat_cpu(cpu).cpustat[CPUTIME_IOWAIT];
else
- iowait = nsecs_to_jiffies64(1000 * iowait_time);
+ iowait = usecs_to_cputime64(iowait_time);

return iowait;
}
diff --cc include/asm-generic/cputime.h
index 12a1764,77202e2..9a62937
--- a/include/asm-generic/cputime.h
+++ b/include/asm-generic/cputime.h
@@@ -38,9 -23,10 +23,12 @@@ typedef u64 __nocast cputime64_t
/*
* Convert cputime to microseconds and back.
*/
- #define cputime_to_usecs(__ct) jiffies_to_usecs(__ct)
- #define usecs_to_cputime(__msecs) usecs_to_jiffies(__msecs)
- #define usecs_to_cputime64(__msecs) nsecs_to_jiffies64((__msecs) * 1000)
+ #define cputime_to_usecs(__ct) \
- jiffies_to_usecs(cputime_to_jiffies(__ct));
-#define usecs_to_cputime(__msecs) \
- jiffies_to_cputime(usecs_to_jiffies(__msecs));
++ jiffies_to_usecs(cputime_to_jiffies(__ct))
++#define usecs_to_cputime(__usec) \
++ jiffies_to_cputime(usecs_to_jiffies(__usec))
++#define usecs_to_cputime64(__usec) \
++ jiffies64_to_cputime64(nsecs_to_jiffies64((__usec) * 1000))

/*
* Convert cputime to seconds and back.
diff --cc kernel/time/tick-sched.c
index 0ec8b83,31cc061..7656642
--- a/kernel/time/tick-sched.c
+++ b/kernel/time/tick-sched.c
@@@ -446,56 -481,8 +446,64 @@@ out
ts->next_jiffies = next_jiffies;
ts->last_jiffies = last_jiffies;
ts->sleep_length = ktime_sub(dev->next_event, now);
-end:
- local_irq_restore(flags);
+}
+
+/**
+ * tick_nohz_idle_enter - stop the idle tick from the idle task
+ *
+ * When the next event is more than a tick into the future, stop the idle tick
+ * Called when we start the idle loop.
+ *
+ * The arch is responsible of calling:
+ *
+ * - rcu_idle_enter() after its last use of RCU before the CPU is put
+ * to sleep.
+ * - rcu_idle_exit() before the first use of RCU after the CPU is woken up.
+ */
+void tick_nohz_idle_enter(void)
+{
+ struct tick_sched *ts;
+
+ WARN_ON_ONCE(irqs_disabled());
+
++ /*
++ * Update the idle state in the scheduler domain hierarchy
++ * when tick_nohz_stop_sched_tick() is called from the idle loop.
++ * State will be updated to busy during the first busy tick after
++ * exiting idle.
++ */
++ set_cpu_sd_state_idle();
++
+ local_irq_disable();
+
+ ts = &__get_cpu_var(tick_cpu_sched);
+ /*
+ * set ts->inidle unconditionally. even if the system did not
+ * switch to nohz mode the cpu frequency governers rely on the
+ * update of the idle time accounting in tick_nohz_start_idle().
+ */
+ ts->inidle = 1;
+ tick_nohz_stop_sched_tick(ts);
+
+ local_irq_enable();
+}
+
+/**
+ * tick_nohz_irq_exit - update next tick event from interrupt exit
+ *
+ * When an interrupt fires while we are idle and it doesn't cause
+ * a reschedule, it may still add, modify or delete a timer, enqueue
+ * an RCU callback, etc...
+ * So we need to re-calculate and reprogram the next tick event.
+ */
+void tick_nohz_irq_exit(void)
+{
+ struct tick_sched *ts = &__get_cpu_var(tick_cpu_sched);
+
+ if (!ts->inidle)
+ return;
+
+ tick_nohz_stop_sched_tick(ts);
}

/**


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/