Message ID | 20190222031413.20250-1-peterx@redhat.com |
---|---|
Headers | show |
Series | iothread: create gcontext unconditionally | expand |
On 22/02/19 04:14, Peter Xu wrote: > And if this patchset can survive... how about running gcontext > directly in iothread_run()? I believe there could be a bit more > things to clean but I'll see. Do you mean instead of aio_poll? The problem is that GMainContext is quite a bit slower than aio_poll. Frediano and I tried to bring some of the optimizations of aio_poll to GMainContext (https://github.com/GNOME/glib/commit/e91c11841808ccca408da96136f433a82b2e2145), but they broke webkit. :( Paolo
On Fri, Feb 22, 2019 at 10:28:57AM +0100, Paolo Bonzini wrote: > On 22/02/19 04:14, Peter Xu wrote: > > And if this patchset can survive... how about running gcontext > > directly in iothread_run()? I believe there could be a bit more > > things to clean but I'll see. > > Do you mean instead of aio_poll? Yes. > The problem is that GMainContext is > quite a bit slower than aio_poll. That's really what I wanted to know; so it's about performance. We should mention it somewhere in iothread_run. I can do that after I know how this series will go. > > Frediano and I tried to bring some of the optimizations of aio_poll to > GMainContext > (https://github.com/GNOME/glib/commit/e91c11841808ccca408da96136f433a82b2e2145), > but they broke webkit. :( What a pity!
On Fri, Feb 22, 2019 at 11:14:09AM +0800, Peter Xu wrote: > Comments welcomed. Thanks, > > Peter Xu (4): > iothread: replace init_done_cond with a semaphore > iothread: create the gcontext onconditionally > iothread: create main loop unconditionally > iothread: push gcontext earlier in the thread_fn What is the status of this series? Will you send another revision? Stefan
On Wed, Mar 06, 2019 at 10:19:59AM +0000, Stefan Hajnoczi wrote: > On Fri, Feb 22, 2019 at 11:14:09AM +0800, Peter Xu wrote: > > Comments welcomed. Thanks, > > > > Peter Xu (4): > > iothread: replace init_done_cond with a semaphore > > iothread: create the gcontext onconditionally > > iothread: create main loop unconditionally > > iothread: push gcontext earlier in the thread_fn > > What is the status of this series? Will you send another revision? I'll add some comment to patch 4 and repost, probably with another patch to document why we can't drop aio_poll in iothread_run. Regards,