On Fri, Nov 13, 2026 at 2:10 AM Andrey Borodin <[email protected]> wrote:

> Hi all,
>
> On 13 Nov 2026, Michael Paquier wrote:
> > it's just a workaround due to the fact that wait
> > events are not handled in a nested way.
>
> I support moving towards nested waits. Keeping the previous event in
> the caller's frame is the approach I had in mind. Starting with the
> logging sites seems a useful step.
>
> The patch addresses the case that concerns me: logs on a slow boot disk
> not showed up in stats.
>
> For follow-up, I'd look at WaitEventSetWait() and FileReadV(). Both
> still clear an outer event when their wait ends.
>
> Could we reuse the debug nesting check to collect inner/outer call
> sites over a check-world run and identify most useful sites that can
> be converted to nested?
>
+1

  I like getting the waits a bit more accurate.  It might take a few passes.

Reply via email to