[PATCH] block: re-introduce blk_mq_complete_request_sync

Chao Leng lengchao at huawei.com
Tue Oct 13 21:25:43 EDT 2020



On 2020/10/14 6:31, Sagi Grimberg wrote:
> 
>>>>>>> -- 
>>>>>>> diff --git a/drivers/nvme/host/tcp.c b/drivers/nvme/host/tcp.c
>>>>>>> index 629b025685d1..46428ff0b0fc 100644
>>>>>>> --- a/drivers/nvme/host/tcp.c
>>>>>>> +++ b/drivers/nvme/host/tcp.c
>>>>>>> @@ -2175,7 +2175,7 @@ static void nvme_tcp_complete_timed_out(struct request *rq)
>>>>>>>          /* fence other contexts that may complete the command */
>>>>>>>          mutex_lock(&to_tcp_ctrl(ctrl)->teardown_lock);
>>>>>>>          nvme_tcp_stop_queue(ctrl, nvme_tcp_queue_id(req->queue));
>>>>>>> -       if (!blk_mq_request_completed(rq)) {
>>>>>>> +       if (blk_mq_request_started(rq) && !blk_mq_request_completed(rq)) {
>>>>>>>                  nvme_req(rq)->status = NVME_SC_HOST_ABORTED_CMD;
>>>>>>>                  blk_mq_complete_request_sync(rq);
>>>>>>>          }
>>> This may just reduce the probability. The concurrency of timeout and teardown will cause the same request
>>> be treated repeatly, this is not we expected.
>>
>> That is right, not like SCSI, NVME doesn't apply atomic request completion, so
>> request may be completed/freed from both timeout & nvme_cancel_request().
>>
>> .teardown_lock still may cover the race with Sagi's patch because teardown
>> actually cancels requests in sync style.
>>
>>> In the teardown process, after quiesced queues delete the timer and cancel the timeout work maybe a better option.
>>
>> Seems better solution, given it is aligned with NVME PCI's reset
>> handling. nvme_sync_queues() may be called in nvme_tcp_teardown_io_queues() to
>> avoid this race.
> 
> We can't call nvme_sync_queues, that flushes the timeout work that is
> serializing with the teardown, it will deadlock.
yes, the teardown_lock is not needed if we sync_queue. We is testing the patch with nvme over roce.
> .



More information about the Linux-nvme mailing list