Re: [PATCH] mm/memory-failure: fix hardware poison check in unpoison_memory()

[Sidhartha Kumar]

On 7/17/23 12:00 PM, Andrew Morton wrote:
> On Mon, 17 Jul 2023 11:18:12 -0700 Sidhartha Kumar <sidhartha.kumar@xxxxxxxxxx> wrote:
>
> > It was pointed out[1] that using folio_test_hwpoison() is wrong
> > as we need to check the indiviual page that has poison.
> > folio_test_hwpoison() only checks the head page so go back to using
> > PageHWPoison().
>
> Please describe the user-visible effects of the bug, especially
> when proposing a -stable backport.

User-visible effects include existing hwpoison-inject tests possibly 
failing as unpoisoning a single subpage could lead to unpoisoning an 
entire folio. Memory unpoisoning could also not work as expected as the 
function will break early due to only checking the head page and not the 
actually poisoned subpage.