Re: [PATCH 7.1.y, 6.18.y, 6.12.y] futex: Prevent robust futex exit race some more

On Fri, Aug 14 2026 at 22:33, Sasha Levin wrote:

>> commit 6d4514ca9cdf61fec4ec634cf50386f6f7e69748 upstream.
>>
>> A robust futex unlock stores 0 over the whole futex value - wiping
>> FUTEX_WAITERS - and wakes a single waiter. That wakeup is a one-shot
>> notification: the protocol relies on its recipient to either acquire the
>> futex (and eventually unlock while aware of the remaining contention) or
>> re-arm FUTEX_WAITERS before sleeping again.
>
> Queued for all the active stable trees (7.1 down to 5.10), thanks. The older
> ones needed futex_wake()'s signature adapted; the logic is unchanged.

That's why I sent two batches. One for 7.1 - 6.12 and one for 6.6 and
older :)