lkml.org 
[lkml]   [2019]   [Apr]   [4]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
Date
SubjectRe: [PATCH 1/1] blk-mq: do not reset plug->rq_count before the list is sorted
On Thu, Apr 4, 2019 at 10:54 AM Dongli Zhang <dongli.zhang@oracle.com> wrote:
>
> We would never be able to sort the list if we first reset plug->rq_count
> which is used in conditional check later.
>
> Fixes: ce5b009cff19 ("block: improve logic around when to sort a plug list")
> Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com>
> ---
> block/blk-mq.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index 3ff3d7b..c162718 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -1711,11 +1711,12 @@ void blk_mq_flush_plug_list(struct blk_plug *plug, bool from_schedule)
> unsigned int depth;
>
> list_splice_init(&plug->mq_list, &list);
> - plug->rq_count = 0;
>
> if (plug->rq_count > 2 && plug->multiple_queues)
> list_sort(NULL, &list, plug_rq_cmp);
>
> + plug->rq_count = 0;
> +
> this_q = NULL;
> this_hctx = NULL;
> this_ctx = NULL;
> --
> 2.7.4
>

Reviewed-by: Ming Lei <ming.lei@redhat.com>

Thanks,
Ming Lei

\
 
 \ /
  Last update: 2019-04-04 08:55    [W:0.035 / U:0.068 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site