Re: [PATCH 02/10] Introduce mm_has_pending_aio() helper

[Nathan Lynch <ntl@pobox.com>]

On Tue, 2011-03-01 at 10:40 -0500, Jeff Moyer wrote:
> 
> ntl@pobox.com writes:
> 
> > From: Nathan Lynch <ntl@pobox.com>
> >
> > Support for AIO is on the to-do list, but until that is implemented,
> > checkpoint will have to fail if a mm_struct has outstanding AIO
> > contexts.  Add a mm_has_pending_aio() helper function for this
> > purpose.
> 
> Just because a process has an io context, doesn't mean that the process
> has active outstanding requests.  So, is this really what you wanted to
> test?

As a temporary measure, yeah.  We haven't settled on code to
record/restore the io context objects themselves, so we do want to bail
if we encounter any.  I realize now the name of the function doesn't
actually express this well.  Will try to come up with something better
for the next round.

Thanks!