Re: [PATCH v5 0/7] locking: contended_release tracepoint instrumentation

From: Dmitry Ilvokhin

Date: Fri May 01 2026 - 09:33:07 EST


I plan to rebase this series on top of Linus' tree to pick up Vineeth's
patch [1], and send an updated version next week.

In the meantime, I would appreciate feedback, especially on:

- tracepoint semantics across different lock types
- overhead concerns in hot paths (e.g. qspinlock)

As a follow-up, I am also working on an RFC to extend perf lock
contention to make use of the contended_release tracepoint, so feedback
in that context would also be helpful.

Feedback from locking and tracing maintainers would be particularly
appreciated before respinning.

[1]: https://lore.kernel.org/all/20260323160052.17528-1-vineeth@xxxxxxxxxxxxxxx/