Re: BUG: scheduling while atomic - Linux 2.6.20-rc2-ga3d89517

["Fabio Comolli" <fabio.comolli@gmail.com>]

OK, I applied your patch to yesterday's Linus' GIT. I will run it for
some days and will let you know if the problem represents. Please note
that it happened only twice and I don't have any clue on how to
reproduce it.

I added Pavel and Rafael to CC-list because for the first time in at
least six months my laptop failed to resume after suspend-to-disk
(userland tools) with this kernel. Guys, do you think that this
failure could be related to this BUG?

Best regards and Happy Holidays,
Fabio




On 12/24/06, OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> wrote:
> "Fabio Comolli" <fabio.comolli@gmail.com> writes:
>
> > Just found this in syslog. It was during normal activity, about 6
> > minutes after resume-from-ram. I never saw this before.
>
> It seems someone missed to check PREEMPT_ACTIVE in __resched_legal().
> Could you please test the following patch?
> --
> OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
>
>
>
> Signed-off-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
> ---
>
>  kernel/sched.c |    5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
>
> diff -puN kernel/sched.c~__resched_legal kernel/sched.c
> --- linux-2.6/kernel/sched.c~__resched_legal    2006-12-24 22:40:19.000000000 +0900
> +++ linux-2.6-hirofumi/kernel/sched.c   2006-12-24 23:54:01.000000000 +0900
> @@ -4619,10 +4619,11 @@ asmlinkage long sys_sched_yield(void)
>
>  static inline int __resched_legal(int expected_preempt_count)
>  {
> -#ifdef CONFIG_PREEMPT
> +#ifndef CONFIG_PREEMPT
> +       expected_preempt_count = 0;
> +#endif
>         if (unlikely(preempt_count() != expected_preempt_count))
>                 return 0;
> -#endif
>         if (unlikely(system_state != SYSTEM_RUNNING))
>                 return 0;
>         return 1;
> _
>