<html><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8" /></head><body style='font-size: 10pt; font-family: Verdana,Geneva,sans-serif'> We know that it loads the radeonsi driver, but.. it doesn't use the driver to render anything by itself, as far as I can understand it... Starting dolphin (or konsole) makes the screen a mess, but I don't know if libGL is to blame for it yet... but what do I know... I got a trace of some sorts from running apitrace, and I've attached it to this posting, but replaying it does not trigger the corruption, and the trace does not contain any frames either. It seems dolphin just loads the radeonsi driver, but what it does with it that triggers the corruption beats me. I have not had time to look into it more today, but I'll try to trace it later like Nicolai suggested on tomorrow (or monday). - Mads On 2016-06-18 21:58, StDenis, Tom wrote: <blockquote type="cite" style="padding: 0 0.4em; border-left: #1010ff 2px solid; margin: 0">  <div id="divtagdefaultwrapper" style="font-size: 12pt; color: #000000; background-color: #ffffff; font-family: Calibri,Arial,Helvetica,sans-serif;"> Hey all, Wouldn't "any" apitrace from that version of Dolphin work? Can you get it at all to work with your version of mesa? Once you capture a valid trace once you can replay it in the circumstances you're seeing the VM fault in right? tom <div style="color: #000000;"> <div><hr style="display: inline-block; width: 98%;" /> <div id="x_divRplyFwdMsg" dir="ltr">From: amd-gfx <amd-gfx-bounces@lists.freedesktop.org> on behalf of Mads <mads@ab3.no> Sent: Saturday, June 18, 2016 09:46 To: Nicolai Hähnle Cc: amd-gfx@lists.freedesktop.org Subject: Re: [amd-gfx] AMD Carrizo - GPU fault detected: 146 0x0842b714 <div> </div> </div> </div> <div class="PlainText">On 2016-06-18 14:30, Nicolai Hähnle wrote: > Okay, so since dolphin uses OpenGL for rendering as well, the problem > now is to figure out whether the VM fault comes from dolphin or from > the compositor. > > There are two approaches. The first one is to just try your luck and > capture an apitrace of dolphin, and then see whether playing that > apitrace back also produces VM faults. If it does, great - upload the > apitrace somewhere, and we can hopefully get it fixed. apitrace was genious, didn't know about that one! But unfortunately there's no frames rendered by dolphin, so apitrace replay does not trigger the bug... So I'll try the second option you mentioned. I'll post again when I find something. Thank you very much for your helpful input! - Mads _______________________________________________ amd-gfx mailing list amd-gfx@lists.freedesktop.org <a id="LPlnk243126" href="https://lists.freedesktop.org/mailman/listinfo/amd-gfx">https://lists.freedesktop.org/mailman/listinfo/amd-gfx</a></div> </div> </div> </blockquote> </body></html>