Re: [PATCH v2 5/5] locking/ww_mutex/test: Time an ABBA with both owners running
From: Mathieu Desnoyers
Date: Wed Nov 18 2026 - 15:36:35 EST
Hi Sebastian,
> The timeout that was introduced in commit cfa92b6d52071
> ("locking/ww_mutex/test: Make sure we bail out instead of livelock")
> triggers. That is not related to your patch.
Thanks for tracking it down.
> I don't call the shots here, I just explained why I don't see this
> needed. We might want to cover this code path if the paths divert in
> future. I don't know.
I would keep it. stress_inorder_work shows the bug only through its
timeout, and its margin depends on the machine. This case times the
ABBA itself: 950ms without the fix, microseconds with it. It would also
catch a change in the ww_mutex path that the futex cases cannot see.
The next version keeps it, with your Tested-by.
Thanks,
Mathieu
--
Mathieu Desnoyers
EfficiOS Inc.
https://www.efficios.com