[dm-devel] [PATCH 8/9] dm: Fix two race conditions related to stopping and starting queues

[Bart Van Assche]

On 08/31/16 20:14, Mike Snitzer wrote:
> On Wed, Aug 31 2016 at  6:18pm -0400,
> Bart Van Assche <bart.vanassche at sandisk.com> wrote:
>
>> Ensure that all ongoing dm_mq_queue_rq() and dm_mq_requeue_request()
>> calls have stopped before setting the "queue stopped" flag. This
>> allows to remove the "queue stopped" test from dm_mq_queue_rq() and
>> dm_mq_requeue_request(). Use BLK_MQ_S_STOPPED instead of
>> QUEUE_FLAG_STOPPED.
>
> At first glance, at a minimum this patch needs a better header.  It
> seems you're doing 2 things:
>
> 1) using blk_mq_{freeze,unfreeze}_queue() actually makes dm_stop_queue()
>    work for blk-mq?  Whereby fixing blk-mq race(s)?
>
> 2) switching away from QUEUE_FLAG_STOPPED to BLK_MQ_S_STOPPED (via
>    blk_mq_queue_stopped)
>    - not clear to me that dm-mq's use of QUEUE_FLAG_STOPPED wasn't fine;
>      NVMe also uses it for blk-mq

Hello Mike,

Adding the blk_mq_{freeze,unfreeze}_queue() calls is indeed what fixes 
the race conditions related to stopping dm queues and what makes 
dm_stop_queue() work.

If other blk-mq users and developers agree that QUEUE_FLAG_STOPPED 
should be set for stopped blk-mq queues then I think the code to set 
that flag should be moved into the blk-mq core. However, setting that 
flag for blk-mq queues seems redundant to me. Hence my proposal to 
introduce blk_mq_queue_stopped() instead.

Bart.