11:03 <dj-death> anybody knows the magic to run piglit replay

11:03 <dj-death> ?

11:06 <dj-death> traces-db$ ../piglit/piglit run -l verbose --timeout 300 -j9 replay /tmp/results

11:06 <dj-death> [0/0]

11:06 <dj-death> User error: No tests were run, not writing a result file

12:11 <dj-death> made my way through hardcoded paths in piglit

12:11 <dj-death> now still have to get the piglit summary to work

15:05 <pac85> zmike: I've got a present for you, in the form of a commit hash and a trace. So yesterday I ran Serious Sam 2 on zink compiled from main and it stuttered any time an enemy showed up on screen. bisecting points to 52f27cda05b9ff2eb93cbaf1cfae15198ed5a3d6 which is yours. The trace doesn't reproduce the issue so I don't know if it is worth sending it to you.

15:08 <zmike> 🤔

15:09 <pac85> I hoped you would have some idea of what could be going on, I'm looking at the MSAA thing for now so I can't work on it

15:12 <zmike> uhhhhh

15:12 <zmike> pac85: is this on ANV?

15:12 <pac85> yes

15:12 <pac85> no sorry

15:12 <pac85> I read radv

15:12 <pac85> it was on radv

15:12 <zmike> uhhhh

15:12 <zmike> so it doesn't happen in the trace but it does happen in the main game?

15:12 <pac85> yes

15:13 <pac85> also it seems to happen when particles show up? I'm not completely sure, maybe xfb could be involved?

15:13 <zmike> can you check whether it's hitting the threadsafe path?

15:13 <zmike> seems unlikely

15:13 <zmike> but maybe

15:33 <pac85> You mean that lock wrapped conditiuoned by `if (usage & PIPE_MAP_THREAD_SAFE) ` right? nope that's not being hit

15:34 <zmike> hmm