On Mon, Sep 16, 2013 at 9:22 AM, Juan Casse <jcasse@xxxxxxxxxxxx> wrote:
> Problem:
> When specifying iodepth > 1 while doing random async io
> without a random map (and using the default random generator), the
> main while loop in do_io() will issue extra io in the amount of io
> allowed to accumulate in the queue. For example, a job with size=24k
> and bs=8k will perform 4 io instead of the expected 3.
>
> Reason:
> The above behavior occurs because the while loop in do_io()
> will continue until the amount of "completed io" >= size, but io
> in the queue are not accounted for because they have not been
> completed by io_u_queued_complete().
>
> Exceptions:
> The above behavior does not manifest itself when:
> using random_generator=lfsr bacause lfsr knows when to stop
> generating, and get_io_u() inside the while loop returns NULL,
> breaking out of the loop.
> using random map because random map ensures that all blocks
> are hit, and after they are all hit, no more are generated.
>
> Proposed Solution:
> Stop while loop based on bytes issued instead of bytes completed.
(before Jen's complains about it. :)
Signed-off-by: Juan Casse <jcasse@xxxxxxxxxxxx> (he's standing next to me :)
Reviewed-by: Grant Grundler <grundler@xxxxxxxxxxxx>
Good Catch! :)
thanks,
grant
> ---
> backend.c | 7 +++++++
> 1 file changed, 7 insertions(+)
>
> diff --git a/backend.c b/backend.c
> index b9c1c12..180a487 100644
> --- a/backend.c
> +++ b/backend.c
> @@ -642,6 +642,7 @@ static uint64_t do_io(struct thread_data *td)
> uint64_t bytes_done[DDIR_RWDIR_CNT] = { 0, 0, 0 };
> unsigned int i;
> int ret = 0;
> + uint64_t bytes_issued = 0;
>
> if (in_ramp_time(td))
> td_set_runstate(td, TD_RAMP);
> @@ -675,6 +676,9 @@ static uint64_t do_io(struct thread_data *td)
> if (flow_threshold_exceeded(td))
> continue;
>
> + if (bytes_issued >= (uint64_t) td->o.size)
> + break;
> +
> io_u = get_io_u(td);
> if (!io_u)
> break;
> @@ -708,6 +712,7 @@ static uint64_t do_io(struct thread_data *td)
> int bytes = io_u->xfer_buflen - io_u->resid;
> struct fio_file *f = io_u->file;
>
> + bytes_issued += bytes;
> /*
> * zero read, fail
> */
> @@ -738,6 +743,7 @@ sync_done:
> ret = io_u_sync_complete(td, io_u, bytes_done);
> if (ret < 0)
> break;
> + bytes_issued += io_u->xfer_buflen;
> }
> break;
> case FIO_Q_QUEUED:
> @@ -748,6 +754,7 @@ sync_done:
> */
> if (td->io_ops->commit == NULL)
> io_u_queued(td, io_u);
> + bytes_issued += io_u->xfer_buflen;
> break;
> case FIO_Q_BUSY:
> requeue_io_u(td, &io_u);
> --
> 1.7.12.4
>
--
To unsubscribe from this list: send the line "unsubscribe fio" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
