Re: arm64: fp-stress: BUG: KFENCE: memory corruption in fpsimd_release_task

[Mark Brown]

On Tue, May 16, 2023 at 02:44:49PM +0100, Will Deacon wrote:

> Mark -- given that this is an SME allocation, please can you take a look?

I'm on holiday.

> I think the implication of the kfence report is that we're writing beyond
> the end of 'task->thread.sme_state' at some point and corrupting the
> redzone.

> There are two reports here, so hopefully it's not too hard to repro.

I think I *once* saw something that might be this but I've never
reproduced it, and I suspect that if this just suddenly came up with
LKFT in stable kernels when there's been no relevant changes AFAIR it's
not showing up terribly reliably there either.

Attachment: signature.asc
Description: PGP signature