Hi
> On 5. Nov 2024, at 02:20, Yu Kuai <yukuai1@xxxxxxxxxxxxxxx> wrote:
>
> Thanks for the test, and sorry that I just realized my patch doesn't fix
> the problem I noticed. :(
>
> Please test again with the following patch on the top. I'm confident
> that this is a real problem at least.
>
> Thanks for the testing!
> Kuai
>
> diff --git a/drivers/md/raid5.c b/drivers/md/raid5.c
> index 04f32173839a..7ec6a5d2d166 100644
> --- a/drivers/md/raid5.c
> +++ b/drivers/md/raid5.c
> @@ -4042,7 +4042,7 @@ static void handle_stripe_clean_event(struct r5conf *conf,
> test_bit(R5_SkipCopy, &dev->flags))) {
> /* We can return any write requests */
> struct bio *wbi, *wbi2;
> - bool written = false;
> + bool written;
>
> pr_debug("Return write for disc %d\n", i);
> if (test_and_clear_bit(R5_Discard, &dev->flags))
> @@ -4053,6 +4053,7 @@ static void handle_stripe_clean_event(struct r5conf *conf,
> do_endio = true;
>
> returnbi:
> + written = false;
> dev->page = dev->orig_page;
> wbi = dev->written;
> dev->written = NULL;
Heh - I stared at your first patch earlier and had somewhat of a hunch that the looping part would need a reset but I didn’t quite trust my intuition as I don’t know the code well enough … ;)
I’m booting with this additional change now and will report back again later.
Christian
--
Christian Theune · ct@xxxxxxxxxxxxxxx · +49 345 219401 0
Flying Circus Internet Operations GmbH · https://flyingcircus.io
Leipziger Str. 70/71 · 06108 Halle (Saale) · Deutschland
HR Stendal HRB 21169 · Geschäftsführer: Christian Theune, Christian Zagrodnick
