qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH v2 05/17] aio: stop using .io_flush()


From: Stefan Hajnoczi
Subject: Re: [Qemu-devel] [PATCH v2 05/17] aio: stop using .io_flush()
Date: Wed, 17 Apr 2013 16:53:41 +0200
User-agent: Mutt/1.5.21 (2010-09-15)

On Wed, Apr 17, 2013 at 03:20:02PM +0200, Paolo Bonzini wrote:
> Il 17/04/2013 15:07, Stefan Hajnoczi ha scritto:
> > 
> > Due to this change we need to update tests/test-aio.c which asserts
> > aio_poll() return values.  Note that QEMU doesn't actually rely on these
> > return values so only tests/test-aio.c cares.
> > 
> > Note that ctx->notifier, the EventNotifier fd used for aio_notify(), is
> > now handled as a special case.  This is a little ugly but maintains
> > aio_poll() semantics, i.e. aio_notify() does not count as 'progress' and
> > aio_poll() avoids blocking when the user has not set any fd handlers yet.
> > 
> > Patches after this remove .io_flush() handler code until we can finally
> > drop the io_flush arguments to aio_set_fd_handler() and friends.
> 
> Dropping aio_poll()'s return value is ok (as a follow up).  The test
> generally test for progress anyway with assertions on the results of the
> notify callbacks.

I thought about dropping aio_poll()'s return value but decided to leave
return values tested.

Although only test-aio.c cares about the return values, it's actually
handy.  It means we'll notice any changes to aio_poll() semantics.

At least it was useful to me while writing these patches - it forced me
to think carefully about how aio_poll() behaved previously.

Stefan



reply via email to

[Prev in Thread] Current Thread [Next in Thread]