03:31orbea: soreau: this might be related https://gitlab.freedesktop.org/mesa/mesa/-/issues/11760
03:31orbea: proton-GE should have the fix in vkd3d-proton already, also maybe RADV_DEBUG=nohiz might fix it, will try both
03:33soreau: nice find
03:33soreau: I can't ever get the search results to find anything useful on gitlab.. it's almost better to google the crawlers :P
03:46HdkR: Note, that SGDT log is non-fatal. That's just anti-tamper/anti-cheat doing things allowed on Windows. Linux emulates the access
03:49HdkR: (I think it is just ensuring the GDT stays inside of kernel memory mappings or something, I didn't look too hard at it last time I saw it)
06:00orbea: HdkR: thanks for explaining that, proton-GE seems to have fixed it, or at least made crashes much harder to trigger
06:06HdkR: Well Proton-GE wouldn't do anything about that SGDT log at least. Since that's purely in game code and can't be patched out
06:06orbea: i meant it fixed the crashes
06:06orbea: the umip dmesg output is still there
06:06HdkR: cool
18:58orbea: seems proton-ge doesn't entirely fix FFXVI, it did for my first play session after it compiled shaders internally, but when I started it again today *crash*, guess will try RADV_DEBUG=nohiz now
20:46orbea: nohiz seems to have worked, but I'm not sure if that caused a performance impact or its just the game itself doing that
23:01kode54: orbea: is this new to the full version of the game?
23:01kode54: I had no problems with the demo
23:10orbea: kode54: i never tried the demo so im not sure
23:11orbea: but there were similar issues reported for the demo
23:11orbea: https://gitlab.freedesktop.org/mesa/mesa/-/issues/11760