On Sat, 21 Apr 2012 09:12:17 -0700
Ben Widawsky <ben at bwidawsk.net> wrote:
> On Sat, 21 Apr 2012 10:41:55 +0100
> Chris Wilson <chris at chris-wilson.co.uk> wrote:
>
> > On Fri, 20 Apr 2012 18:23:31 -0700, Ben Widawsky <ben at bwidawsk.net>
> > wrote:
> > > + ret = i915_seqno_wait_timed(ring, seqno, true, &timeout);
> > > + if (ret == -ERESTARTSYS)
> > > + ret = -EINTR;
> > Don't convert it here, pass ERESTARTSYS to the system call handler
> > which decides how to handle it.
> >
> > > + else if (ret == -ETIME) {
> > > + ret = -EBUSY;
> >
> > Why the semantic change? ETIME for timer timed out still seems
> > appropriate.
>
> I must be missing something. Can you point to me where the system call
> handler converts these? The only reason for the change was to prevent
> passing the internal return types to user space.
>
> >
> > I think this whole interface is a stop-gap solution for pollable sync
> > objects without a clear use case. Do we have a spec for a feature
> > wishing to build upon this interface?
>
> Good point. This is primarily for glClientWaitSync. I will update the
> commit message with this information.
>
> > -Chris
> >
NVM, got them both. Thanks.
