Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1199043 > unrolled thread

Re: [lkp] [net] 1fbe4b46cac: WARNING: CPU: 0 PID: 2245 at kernel/sched/core.c:7376 __might_sleep+0x8b/0xa8()

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [lkp] [net] 1fbe4b46cac: WARNING: CPU: 0 PID: 2245 at  kernel/sched/core.c:7376 __might_sleep+0x8b/0xa8() Oleg Nesterov <oleg@redhat.com> - 2015-08-03 19:00 +0200
    Re: [lkp] [net] 1fbe4b46cac: WARNING: CPU: 0 PID: 2245 at  kernel/sched/core.c:7376 __might_sleep+0x8b/0xa8() Oleg Nesterov <oleg@redhat.com> - 2015-08-04 18:40 +0200

#1199043 — Re: [lkp] [net] 1fbe4b46cac: WARNING: CPU: 0 PID: 2245 at kernel/sched/core.c:7376 __might_sleep+0x8b/0xa8()

On 07/30, Oleg Nesterov wrote:
>
> Thanks!
>
> I am travelling, I'll send the trivial fix on Monday.
>
> Just in case, this is not a bug, we only need to shut up the warning and we
> even have the special macro for that. I forgot its name, and I don't have the
> kernel sources right now.

Now that I can actually see the code, I think that we should simply remove
__set_current_state(TASK_RUNNING) and set_current_state(TASK_INTERRUPTIBLE).
But let me re-check this once again tomorrow, I simply can't understand why
pktgen_thread_worker() does this.

Unless I am totally confused they could be removed right after ef87979c273a2
"pktgen: better scheduler friendliness" which in particular changed the main
loop to use wait_event(). Then later baac167b "pktgen: avoid expensive
set_current_state() call in loop" changed the 1st __set_current_state() to
set RUNNING, and moved the 2nd set_current_state(TASK_INTERRUPTIBLE) outside
the main loop for absolutely no reason.

Oleg.

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

[toc] | [next] | [standalone]

#1200145

On 08/03, Oleg Nesterov wrote:
>
> Now that I can actually see the code, I think that we should simply remove
> __set_current_state(TASK_RUNNING) and set_current_state(TASK_INTERRUPTIBLE).
> But let me re-check this once again tomorrow, I simply can't understand why
> pktgen_thread_worker() does this.
>
> Unless I am totally confused they could be removed right after ef87979c273a2
> "pktgen: better scheduler friendliness" which in particular changed the main
> loop to use wait_event(). Then later baac167b "pktgen: avoid expensive
> set_current_state() call in loop" changed the 1st __set_current_state() to
> set RUNNING, and moved the 2nd set_current_state(TASK_INTERRUPTIBLE) outside
> the main loop for absolutely no reason.

Yes, I am sending the patch. Both set_current_state()'s look obviously unneeded,
and afaics we could hit the same warning even without 1fbe4b46caca. At least
pktgen_rem_thread()->remove_proc_entry() can sleep.

Oleg.

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

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web