On Tue, Aug 16, 2016 at 03:59:24PM +0100, Robert Bragg wrote: > On Mon, Aug 15, 2016 at 3:57 PM, Chris Wilson > Alternatively you could follow the standard pattern for read. Dare I ask > what is going to go into state that needs the obfuscation? > > I had dug around a bit when I was trying to decide how to handle the > corner cases here and found some precedent for prioritize reporting any > data copied over an error for a read(). Reporting completed bytes before the error is correct. I was referring to going between passing the return value as a mixture of state and return, when it just appears to be following the usual pattern of read(). i.e. I could find anything to support why your internal read callback required a different signature. It looks unusual, so I am expecting it to do unusual things. -Chris -- Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list Intel-gfx@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/intel-gfx
