This is the mail archive of the
mailing list for the GCC project.
Re: Re: [GSoC'19, libgomp work-stealing] Task parallelism runtime
- From: 김규래<msca8h at naver dot com>
- To: Jakub Jelinek<jakub at redhat dot com>
- Cc: gcc general<gcc at gcc dot gnu dot org>
- Date: Mon, 05 Aug 2019 20:01:12 +0900
- Subject: Re: Re: [GSoC'19, libgomp work-stealing] Task parallelism runtime
- References: <firstname.lastname@example.org> <email@example.com> <20190713062848.GE2125@tucnak> <firstname.lastname@example.org> <20190722185413.GA13123@laptop.zalov.cz> <email@example.com> <firstname.lastname@example.org> <20190805103216.GQ2726@tucnak>
> I thought we don't want to go lock-free, the queue operations > aren't easily
> implementable lock-free, but instead with a lock for each of > the queues,
By lock-free I meant to use locks only for the queues,
But my terminology was indeed confusing sorry about that.
> mean we don't in some cases wake anybody, so there will be > threads idling
> instead of doing useful work, but at least one thread
> probably should handle
> it later.
I was personally worried about this case
Since this could result in huge inefficiencies, but maybe it'll be fine.
I'll first try to implement it.