[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] [RFC PATCH 00/11] aio: Introduce handler type to fix ne
From: |
Fam Zheng |
Subject: |
Re: [Qemu-devel] [RFC PATCH 00/11] aio: Introduce handler type to fix nested aio_poll for dataplane |
Date: |
Mon, 27 Jul 2015 14:55:45 +0800 |
User-agent: |
Mutt/1.5.23 (2014-03-12) |
On Fri, 07/24 09:35, Paolo Bonzini wrote:
> > That way, aio_context_acquire can be dropped by:
> >
> > /* @pause_owner_thread: a callback which will be called when _main
> > thread_
> > * wants exclusively operate on the AioContext, for example with a
> > nested
> > * aio_poll().
> > * @resume_owner_thread: a callback which will be called when _main
> > thread_
> > * has done the exclusive operation.
> > */
> > AioContext *aio_context_new(AioContextPauseResumeFn *pause_owner_thread,
> > AioContextPauseResumeFn
> > *resume_owner_thread,
> > void *opaque,
> > Error **errp):
> >
> > /* Try to pause the owning thread */
> > void aio_context_pause(AioContext *ctx, Error **errp);
> >
> > /* Try to resume the paused owning thread, cannot fail */
> > void aio_context_resume(AioContext *ctx);
> >
> > Then, in iothread.c:
> >
> > iothread->ctx = aio_context_new(iothread_pause, iothread_resume,
> > iothread, &local_err);
> >
> > Where iothread_pause can block the thread with a QemuCond.
>
> Condition variables don't mix well with recursive mutexes... Once we
> have bottom halves outside the AioContext lock, however, we could use a
> separate lock for this condition variable. That would work.
Yes, I thought so.
>
> I like it, but I still ask myself... what's the difference between
> aio_context_pause/resume and aio_disable/enable_clients? :) There is
> still state, just in the iothread rather than in the AioContext.
I don't know, maybe this will make aio_context_acquire no longer necessary so
we get virtio-scsi dataplane fixed?
Fam
- [Qemu-devel] [RFC PATCH 00/11] aio: Introduce handler type to fix nested aio_poll for dataplane, Fam Zheng, 2015/07/23
- [Qemu-devel] [RFC PATCH 02/11] aio: Save type to AioHandler, Fam Zheng, 2015/07/23
- [Qemu-devel] [RFC PATCH 01/11] aio: Introduce "type" in aio_set_fd_handler and aio_set_event_notifier, Fam Zheng, 2015/07/23
- [Qemu-devel] [RFC PATCH 04/11] aio-win32: Implement aio_poll_clients, Fam Zheng, 2015/07/23
- [Qemu-devel] [RFC PATCH 08/11] dataplane: Mark host notifiers' client type as "dataplane", Fam Zheng, 2015/07/23
- [Qemu-devel] [RFC PATCH 11/11] block: Only poll block layer fds in bdrv_aio_poll, Fam Zheng, 2015/07/23
- Re: [Qemu-devel] [RFC PATCH 00/11] aio: Introduce handler type to fix nested aio_poll for dataplane, Paolo Bonzini, 2015/07/23
[Qemu-devel] [RFC PATCH 06/11] nbd: Mark fd handlers client type as "nbd server", Fam Zheng, 2015/07/23
[Qemu-devel] [RFC PATCH 10/11] block: Replace nested aio_poll with bdrv_aio_poll, Fam Zheng, 2015/07/23
[Qemu-devel] [RFC PATCH 09/11] block: Introduce bdrv_aio_poll, Fam Zheng, 2015/07/23
[Qemu-devel] [RFC PATCH 05/11] block: Mark fd handlers as "protocol", Fam Zheng, 2015/07/23
[Qemu-devel] [RFC PATCH 07/11] aio: Mark ctx->notifier's client type as "context", Fam Zheng, 2015/07/23
[Qemu-devel] [RFC PATCH 03/11] aio-posix: Introduce aio_poll_clients, Fam Zheng, 2015/07/23