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

Groups > linux.kernel > #1683199 > unrolled thread

blk-mq timeout question

Back to article view | Back to linux.kernel

Contents

  blk-mq timeout question Stefan Haberland <sth@linux.vnet.ibm.com> - 2017-07-07 15:50 +0200
    Re: blk-mq timeout question Ming Lei <tom.leiming@gmail.com> - 2017-07-10 04:30 +0200

#1683199 — blk-mq timeout question

While changing the DASD device driver to use the blk-mq interfaces I 
found the following unexpected behavior:

In case of a timeout our complete callback is never called. Here is the 
sequence of events as I understood:

- timeout occurs
- blk_mq_check_expired() calls and checks blk_mark_rq_complete()
- our .timeout callback is called which returns BLK_EH_NOT_HANDLED and 
schedules delayed work
- our worker calls blk_mq_complete_request()
- this also checks blk_mark_rq_complete() and therefore never calls our 
complete callback

Question:
Should blk_clear_rq_complete() also be called for the BLK_EH_NOT_HANDLED 
case?

Regards,
Stefan

[toc] | [next] | [standalone]

#1683905

On Fri, Jul 7, 2017 at 9:45 PM, Stefan Haberland <sth@linux.vnet.ibm.com> wrote:
> While changing the DASD device driver to use the blk-mq interfaces I found
> the following unexpected behavior:
>
> In case of a timeout our complete callback is never called. Here is the
> sequence of events as I understood:
>
> - timeout occurs
> - blk_mq_check_expired() calls and checks blk_mark_rq_complete()
> - our .timeout callback is called which returns BLK_EH_NOT_HANDLED and
> schedules delayed work
> - our worker calls blk_mq_complete_request()
> - this also checks blk_mark_rq_complete() and therefore never calls our
> complete callback
>
> Question:
> Should blk_clear_rq_complete() also be called for the BLK_EH_NOT_HANDLED
> case?

From comment in blk_rq_timed_out():

        case BLK_EH_NOT_HANDLED:
                /*
                 * LLD handles this for now but in the future
                 * we can send a request msg to abort the command
                 * and we can move more of the generic scsi eh code to
                 * the blk layer.

Looks you can/should handle the case inside DASD, and not do that in blk layer.


-- 
Ming Lei

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web