Re: [PATCH] Revert "random: Mix cputime from each thread that exitsto the pool"

[Russ Dill]

On Mon, Mar 4, 2013 at 9:05 AM, Theodore Ts'o <tytso@xxxxxxx> wrote:
> Russ, how about this patch instead?  Can you verify whether or not
> this fixes your issue?  If so, I'll push it to Linus.
>
>                                     - Ted

Yup, fixes the circular locking dependency.

Tested-by: Russ.Dill@xxxxxxxxx

> From 24c598b5abd466f2bab954d0cc8c225e0d3a2a74 Mon Sep 17 00:00:00 2001
> From: Theodore Ts'o <tytso@xxxxxxx>
> Date: Mon, 4 Mar 2013 11:59:12 -0500
> Subject: [PATCH] random: fix locking dependency with the tasklist_lock
>
> Commit 6133705494bb introduced a circular lock dependency because
> posix_cpu_timers_exit() is called by release_task(), which is holding
> a writer lock on tasklist_lock, and this can cause a deadlock since
> kill_fasync() gets called with nonblocking_pool.lock taken.
>
> There's no reason why kill_fasync() needs to be taken while the random
> pool is locked, so move it out to fix this locking dependency.
>
> Signed-off-by: "Theodore Ts'o" <tytso@xxxxxxx>
> Reported-by: Russ Dill <Russ.Dill@xxxxxxxxx>
> Cc: stable@xxxxxxxxxx
> ---
>  drivers/char/random.c | 12 ++++++++----
>  1 file changed, 8 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/char/random.c b/drivers/char/random.c
> index 85e81ec..57d4b15 100644
> --- a/drivers/char/random.c
> +++ b/drivers/char/random.c
> @@ -852,6 +852,7 @@ static size_t account(struct entropy_store *r, size_t nbytes, int min,
>                       int reserved)
>  {
>         unsigned long flags;
> +       int wakeup_write = 0;
>
>         /* Hold lock while accounting */
>         spin_lock_irqsave(&r->lock, flags);
> @@ -873,10 +874,8 @@ static size_t account(struct entropy_store *r, size_t nbytes, int min,
>                 else
>                         r->entropy_count = reserved;
>
> -               if (r->entropy_count < random_write_wakeup_thresh) {
> -                       wake_up_interruptible(&random_write_wait);
> -                       kill_fasync(&fasync, SIGIO, POLL_OUT);
> -               }
> +               if (r->entropy_count < random_write_wakeup_thresh)
> +                       wakeup_write = 1;
>         }
>
>         DEBUG_ENT("debiting %zu entropy credits from %s%s\n",
> @@ -884,6 +883,11 @@ static size_t account(struct entropy_store *r, size_t nbytes, int min,
>
>         spin_unlock_irqrestore(&r->lock, flags);
>
> +       if (wakeup_write) {
> +               wake_up_interruptible(&random_write_wait);
> +               kill_fasync(&fasync, SIGIO, POLL_OUT);
> +       }
> +
>         return nbytes;
>  }
>
> --
> 1.7.12.rc0.22.gcdd159b
>
--
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/