On Fri, Nov 24, 2023 at 3:18 PM <junxiao.bi@xxxxxxxxxx> wrote:
>
> On 11/24/23 9:29 AM, Song Liu wrote:
>
> > On Wed, Nov 8, 2023 at 10:22 AM Junxiao Bi <junxiao.bi@xxxxxxxxxx> wrote:
> >> This reverts commit 5e2cf333b7bd5d3e62595a44d598a254c697cd74.
> >>
> >> That commit introduced the following race and can cause system hung.
> >>
> >> md_write_start: raid5d:
> >> // mddev->in_sync == 1
> >> set "MD_SB_CHANGE_PENDING"
> >> // running before md_write_start wakeup it
> >> waiting "MD_SB_CHANGE_PENDING" cleared
> >> >>>>>>>>> hung
> >> wakeup mddev->thread
> >> ...
> >> waiting "MD_SB_CHANGE_PENDING" cleared
> >> >>>> hung, raid5d should clear this flag
> >> but get hung by same flag.
> >>
> >> The issue reverted commit fixing is fixed by last patch in a new way.
> >>
> >> Fixes: 5e2cf333b7bd ("md/raid5: Wait for MD_SB_CHANGE_PENDING in raid5d")
> >> Signed-off-by: Junxiao Bi <junxiao.bi@xxxxxxxxxx>
> > The set looks good to me. Thanks!
> Thanks for the review.
> >
> > Quick question: from the earlier thread, the issue was observed in
> > production. Have you reproduced the issue and thus verified the fix
> > works as expected?
>
> I didn't try reproducing this since the system hung on the code where
> the bad commit added, after revert it, this issue will not reproduce any
> more.
Thanks for the information. I applied the set to md-next.
Song
