Re: [RFC PATCH 5/5] md/raid1: introduce rectify action to repair badblocks

[Zheng Qixing]

Hi,

> > @@ -700,21 +738,34 @@ static enum sync_action __get_recovery_sync_action(struct mddev *mddev)
> >     static enum sync_action get_recovery_sync_action(struct mddev *mddev)
> >   {
> > +    if (test_bit(MD_RECOVERY_BADBLOCKS_RECTIFY, &mddev->recovery))
> > +        return ACTION_RECTIFY;
> >       return __get_recovery_sync_action(mddev);
> >   }
> >     static void init_recovery_position(struct mddev *mddev)
> >   {
> >       mddev->resync_min = 0;
> > +    mddev->rectify_min = 0;
> > +}
>
> As mentioned in patch 1, can we directly reuse resync_min?


To keep rectify's progress separate from check and repair, it's better to use
a dedicated variable to record it.

I'll update the other suggestions in v2.


Thanks,

Qixing