On Wed, 2006-05-31 at 16:40 +0200, Roland Scheidegger wrote: > Michel Dänzer wrote: > > Indeed. The X server's GetTimeInMillis() might be more convenient > > than gettimeofday() (and would also work with an elfloader X server > > ;). > Ah yes. That was just a quick hack :-). The question is, what should the > timeout value be?
Not sure, the time the current code took on the machine it was written on? ;) Seriously though, I think one second should be plenty. > >> but I'm not sure there's a better reg to monitor activity - ring > >> ptr might be another possibility, but I think this can point at the > >> same location for a long time too (for instance when drawing using > >> the idx buf command). > > > > One of the performance counters might be better. > Which one did you have in mind? One of the MC counters, e.g. But these may be different since R300 as well. > Don't they need to be configured first? No idea. > Sounds like a good idea though, or maybe that ring ptr would do too, I > think chances aren't that great that it is really stuck at the same > place for a long time. You're probably right. Maybe check CP_CSQ_STAT in addition, in order to also cover indirect buffers, but that may be tricky due to the wraparounds. -- Earthling Michel Dänzer | http://tungstengraphics.com Libre software enthusiast | Debian, X and DRI developer -- _______________________________________________ Dri-devel mailing list Dri-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel