On 7/12/23 3:25?AM, Peter Zijlstra wrote:
> On Tue, Jul 11, 2023 at 06:47:03PM -0600, Jens Axboe wrote:
>
>> Since we now provide a way to pass in a wake handler and data, ensure we
>> use __futex_queue() to avoid having futex_queue() overwrite our wait
>> data.
>
>> diff --git a/kernel/futex/waitwake.c b/kernel/futex/waitwake.c
>> index 3471af87cb7d..dfd02ca5ecfa 100644
>> --- a/kernel/futex/waitwake.c
>> +++ b/kernel/futex/waitwake.c
>> @@ -446,7 +446,8 @@ static int futex_wait_multiple_setup(struct futex_vector *vs, int count, int *wo
>> * next futex. Queue each futex at this moment so hb can
>> * be unlocked.
>> */
>> - futex_queue(q, hb);
>> + __futex_queue(q, hb);
>> + spin_unlock(&hb->lock);
>> continue;
>> }
>
> I'm not following; I even applied all your patches up to this point, but
> futex_queue() still reads:
>
> static inline void futex_queue(struct futex_q *q, struct futex_hash_bucket *hb)
> __releases(&hb->lock)
> {
> __futex_queue(q, hb);
> spin_unlock(&hb->lock);
> }
>
> How would it be different and overwrite anything ?!?
Good catch, this is a leftover from storing the task/wakeup data
separately. But I got rid of that, so it's stale comment at this point
and we can certainly use futex_queue() here again and drop this hunk.
Will make that edit.
--
Jens Axboe
