Re: linux-next: Tree for Jun 28 [ BISECTED: rsyslog/imklog: High CPU usage ]

[Sedat Dilek <sedat.dilek@gmail.com>]

On Sun, Jun 30, 2013 at 8:19 PM, Oleg Nesterov <oleg@redhat.com> wrote:
> Andrew, please drop these
>
>         wait-introduce-wait_event_commonwq-condition-state-timeout.patch
>         wait-introduce-prepare_to_wait_event.patch
>
> patches again. I'll send v3 although it really looks like I should
> never try to touch wait.h.
>
> On 06/29, Sedat Dilek wrote:
>>
>> As this all did not show me what caused the problem I started a
>> git-bisect session.
>>
>> This revealed the following culprit commit:
>>
>>      commit bb1f30cb7d3ba21098f0ee7e0382160ba2599a43
>>      "wait: introduce wait_event_common(wq, condition, state, timeout)"
>
> Heh. First of all, I am really sorry.
>
> Not only "__wait_no_timeout(tout) ?:" was wrong, I didn't bother
> to recheck this logic even after I got the "warning: the omitted
> middle operand in ?:" reports.
>
> Sedat, thanks you very much! Any chance you can try the patch below?
>

[ CC linux-mm ML ]

Hi Oleg,

did you made a cleaned-up version?
AFAICS v3, I read that on linux-mm ML, sorry if I ask here in this thread.

Regards,
- Sedat -

> Oleg.
>
>
> --- a/include/linux/wait.h
> +++ b/include/linux/wait.h
> @@ -191,12 +191,8 @@ wait_queue_head_t *bit_waitqueue(void *, int);
>         for (;;) {                                                      \
>                 __ret = prepare_to_wait_event(&wq, &__wait, state);     \
>                 if (condition) {                                        \
> -                       __ret = __wait_no_timeout(tout);                \
> -                       if (!__ret) {                                   \
> -                               __ret = __tout;                         \
> -                               if (!__ret)                             \
> -                                       __ret = 1;                      \
> -                       }                                               \
> +                       __ret = __wait_no_timeout(tout) ? 0 :           \
> +                                                       (__tout ?: 1);  \
>                         break;                                          \
>                 }                                                       \
>                                                                         \
> @@ -217,16 +213,10 @@ wait_queue_head_t *bit_waitqueue(void *, int);
>  #define wait_event_common(wq, condition, state, tout)                  \
>  ({                                                                     \
>         long __ret;                                                     \
> -       if (condition) {                                                \
> -               __ret = __wait_no_timeout(tout);                        \
> -               if (!__ret) {                                           \
> -                       __ret = tout;                                   \
> -                       if (!__ret)                                     \
> -                               __ret = 1;                              \
> -               }                                                       \
> -       } else {                                                        \
> +       if (condition)                                                  \
> +               __ret = __wait_no_timeout(tout) ? 0 : ((tout) ?: 1);    \
> +       else                                                            \
>                 __ret = __wait_event_common(wq, condition, state, tout);\
> -       }                                                               \
>         __ret;                                                          \
>  })
>
>

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>