Re: [PATCH] task_work: add a scheduling point in task_work_run()

[Mimi Zohar <zohar@linux.vnet.ibm.com>]

On Tue, 2012-08-21 at 15:05 +0200, Eric Dumazet wrote:
> From: Eric Dumazet <edumazet@google.com>
> 
> It seems commit 4a9d4b02 (switch fput to task_work_add) reintroduced
> the problem addressed in commit 944be0b2 (close_files(): add scheduling
> point)
> 
> If a server process with a lot of files (say 2 million tcp sockets)
> is killed, we can spend a lot of time in task_work_run() and trigger
> a soft lockup.
> 
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> ---
>  kernel/task_work.c |    1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/kernel/task_work.c b/kernel/task_work.c
> index 91d4e17..d320d44 100644
> --- a/kernel/task_work.c
> +++ b/kernel/task_work.c
> @@ -75,6 +75,7 @@ void task_work_run(void)
>  			p = q->next;
>  			q->func(q);
>  			q = p;
> +			cond_resched();
>  		}
>  	}
>  }

We're here, because fput() called schedule_work() to delay the last
fput().  The execution needs to take place before the syscall returns to
userspace.  Need to read __schedule()...  Do you know if cond_resched()
can guarantee that it will be executed before the return to userspace? 

thanks,

Mimi