Re: clock drift in set_task_cpu()

From: Ingo Molnar
Date: Mon Sep 06 2010 - 02:53:12 EST



* Jack Daniel <wanders.thirst@xxxxxxxxx> wrote:

> Hi Ingo,
>
> [...]
>
> Any chance that you will be pulling in this fix ?

It's already upstream, as of v2.6.36-rc3. See the commit notification
email below from Aug 19.

Thanks,

Ingo

----- Forwarded message from tip-bot for Peter Zijlstra <peterz@xxxxxxxxxxxxx> -----

Date: Fri, 20 Aug 2010 14:16:14 GMT
From: tip-bot for Peter Zijlstra <peterz@xxxxxxxxxxxxx>
To: linux-tip-commits@xxxxxxxxxxxxxxx
Cc: linux-kernel@xxxxxxxxxxxxxxx, hpa@xxxxxxxxx, mingo@xxxxxxxxxx,
a.p.zijlstra@xxxxxxxxx, peterz@xxxxxxxxxxxxx, pjohn@xxxxxxxxxx,
wanders.thirst@xxxxxxxxx, tglx@xxxxxxxxxxxxx, mingo@xxxxxxx
Subject: [tip:sched/urgent] sched: Fix rq->clock synchronization when migrating tasks

Commit-ID: 861d034ee814917a83bd5de4b26e3b8336ddeeb8
Gitweb: http://git.kernel.org/tip/861d034ee814917a83bd5de4b26e3b8336ddeeb8
Author: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
AuthorDate: Thu, 19 Aug 2010 13:31:43 +0200
Committer: Ingo Molnar <mingo@xxxxxxx>
CommitDate: Fri, 20 Aug 2010 14:59:01 +0200

sched: Fix rq->clock synchronization when migrating tasks

sched_fork() -- we do task placement in ->task_fork_fair() ensure we
update_rq_clock() so we work with current time. We leave the vruntime
in relative state, so the time delay until wake_up_new_task() doesn't
matter.

wake_up_new_task() -- Since task_fork_fair() left p->vruntime in
relative state we can safely migrate, the activate_task() on the
remote rq will call update_rq_clock() and causes the clock to be
synced (enough).

Tested-by: Jack Daniel <wanders.thirst@xxxxxxxxx>
Tested-by: Philby John <pjohn@xxxxxxxxxx>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@xxxxxxxxx>
LKML-Reference: <1281002322.1923.1708.camel@laptop>
Signed-off-by: Ingo Molnar <mingo@xxxxxxx>
---
kernel/sched_fair.c | 2 ++
1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c
index 806d1b2..ab661eb 100644
--- a/kernel/sched_fair.c
+++ b/kernel/sched_fair.c
@@ -3752,6 +3752,8 @@ static void task_fork_fair(struct task_struct *p)

raw_spin_lock_irqsave(&rq->lock, flags);

+ update_rq_clock(rq);
+
if (unlikely(task_cpu(p) != this_cpu))
__set_task_cpu(p, this_cpu);


----- End forwarded message -----
--
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/