Niu Danny wrote, on 20 May 2024:
> 
> I see the bug's still not closed. Would standard maintainers disclose 
> if there's still discussion over this issue?

By coincidence we reached that bug in today's teleconference, and
have processed it as withdrawn based on your comments in the bug.

> Another issue I thought of, but refrained from bringing onto the issue 
> tracker,
> is that in older (not necessarily Unix/POSIX) systems, when there's no
> condition variable, threads would wait for predicates to become true by
> yielding, or sleeping if expiration is needed. Should yield/sleep synchronize
> memory? What's the group's opinion on this?

My personal take is that I only see yield as affecting scheduling.
I.e. it's purely a way for the thread to ask the scheduler to stop
running it earlier than it otherwise would.  So I would not expect
its effect on memory to be any different than if the scheduler
happened to stop running the thread at that point in the code.

-- 
Geoff Clare <[email protected]>
The Open Group, Apex Plaza, Forbury Road, Reading, RG1 1AX, England

Reply via email to