From: Peter Zijlstra <a.p.zijlstra@chello.nl>
To: Ari Savolainen <ari.m.savolainen@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>,
linux-kernel@vger.kernel.org, Dave Jones <davej@redhat.com>
Subject: Re: sig->cputimer.lock/rq->lock lockdep report
Date: Wed, 12 Oct 2011 22:50:51 +0200 [thread overview]
Message-ID: <1318452651.14236.9.camel@twins> (raw)
In-Reply-To: <CAEbykaX=HWMJmTB0VCFkfW2v2G9=FcnFZBFwrYk8bFTT_FrD-Q@mail.gmail.com>
On Wed, 2011-10-12 at 23:13 +0300, Ari Savolainen wrote:
> I've got the same problem that Dave reported earlier. I bisected it to
> commit d670ec13178d "posix-cpu-timers: Cure SMP wobbles".
https://lkml.org/lkml/2011/10/7/282
Does the below work for you? (still not proud of it)
---
include/linux/sched.h | 3 +++
kernel/posix-cpu-timers.c | 6 +++++-
kernel/sched_stats.h | 4 ++--
3 files changed, 10 insertions(+), 3 deletions(-)
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 41d0237..ad9eafc 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -503,6 +503,7 @@ struct task_cputime {
* @running: non-zero when there are timers running and
* @cputime receives updates.
* @lock: lock for fields in this struct.
+ * @runtime_lock: lock for cputime.sum_exec_runtime
*
* This structure contains the version of task_cputime, above, that is
* used for thread group CPU timer calculations.
@@ -511,6 +512,7 @@ struct thread_group_cputimer {
struct task_cputime cputime;
int running;
spinlock_t lock;
+ spinlock_t runtime_lock;
};
#include <linux/rwsem.h>
@@ -2566,6 +2568,7 @@ void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times);
static inline void thread_group_cputime_init(struct signal_struct *sig)
{
spin_lock_init(&sig->cputimer.lock);
+ spin_lock_init(&sig->cputimer.runtime_lock);
}
/*
diff --git a/kernel/posix-cpu-timers.c b/kernel/posix-cpu-timers.c
index c8008dd..fa189a6 100644
--- a/kernel/posix-cpu-timers.c
+++ b/kernel/posix-cpu-timers.c
@@ -284,9 +284,13 @@ void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times)
* it.
*/
thread_group_cputime(tsk, &sum);
+ spin_lock(&cputimer->runtime_lock);
update_gt_cputime(&cputimer->cputime, &sum);
- }
+ } else
+ spin_lock(&cputimer->runtime_lock);
+
*times = cputimer->cputime;
+ spin_unlock(&cputimer->runtime_lock);
spin_unlock_irqrestore(&cputimer->lock, flags);
}
diff --git a/kernel/sched_stats.h b/kernel/sched_stats.h
index 331e01b..a7e2c1a 100644
--- a/kernel/sched_stats.h
+++ b/kernel/sched_stats.h
@@ -330,7 +330,7 @@ static inline void account_group_exec_runtime(struct task_struct *tsk,
if (!cputimer->running)
return;
- spin_lock(&cputimer->lock);
+ spin_lock(&cputimer->runtime_lock);
cputimer->cputime.sum_exec_runtime += ns;
- spin_unlock(&cputimer->lock);
+ spin_unlock(&cputimer->runtime_lock);
}
next prev parent reply other threads:[~2011-10-12 20:51 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-10-12 20:13 sig->cputimer.lock/rq->lock lockdep report Ari Savolainen
2011-10-12 20:50 ` Peter Zijlstra [this message]
2011-10-13 14:44 ` Ari Savolainen
2011-10-14 19:01 ` Ari Savolainen
-- strict thread matches above, loose matches on Subject: below --
2011-10-06 15:55 Dave Jones
2011-10-20 15:47 ` Paul Bolle
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1318452651.14236.9.camel@twins \
--to=a.p.zijlstra@chello.nl \
--cc=ari.m.savolainen@gmail.com \
--cc=davej@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox