Re: [PATCH v2] mm/migrate: fix softlockup in do_pages_move() with huge page arrays
From: David Hildenbrand (Arm)
Date: Sun Nov 15 2026 - 12:02:54 EST
On 11/14/26 21:33, Gregory Price (Meta) wrote:
> On 2026-11-14 17:51 +0000, Lorenzo Stoakes (ARM) wrote:
>> On Sun, Nov 08, 2026 at 02:28:31PM +0800, shao.mingyin@xxxxxxxxxx wrote:
>>> From: Shao Mingyin <shao.mingyin@xxxxxxxxxx>
>>>
>>> do_pages_stat() had the same problem; it was fixed by 8e934ab44389. In
>>> current mainline only alpha, hexagon and classic m68k can select
>>> PREEMPT_NONE and they cannot enable CONFIG_NUMA, so this mainly targets
>>> pre-v7.0 stable kernels.
>>>
>>> Fixes: a49bd4d71637 ("mm, numa: rework do_pages_move")
>>> Cc: stable@xxxxxxxxxxxxxxx
>>> Signed-off-by: Shao Mingyin <shao.mingyin@xxxxxxxxxx>
>>
>> NAK.
> ...
>> And absolutely not to this person/script doing backports, thanks.
>
> Have to agree here. This is just creating backport busywork.
>
> This might matter if we intended to do automated cross-arch testing
> on QEMU TCG on these platforms... on those LTS kernels... but also
> mehhhhhh
>
> Yeah we shouldn't encourage this kind of patch spew.
>
More than happy to drop this and wait for an actual user report -- and likely
find a different way to fix this stuff in stable without involving upstream that
doesn't need these changes.
--
Cheers,
David