I noticed a regression with the game 0ad: it crashes with swrastg and it stucks at game start with r300g. Bisecting, with both r300g and swrastg: # good: [815e79e72c1f4aa849c0ee6103621685b678bc9d] draw: re-order optimization passes depending on LLVM version, 32/64-bit git bisect good 815e79e72c1f4aa849c0ee6103621685b678bc9d # bad: [7b7845a076c933e096ac511b4184141ba194449a] glx: Move WaitGL, WaitX, UseXFont to context vtable functions git bisect bad 7b7845a076c933e096ac511b4184141ba194449a With ab434f6b7641a64d30725a9ac24929240362d466 it works with swrastg but it stucks with r300g. Trying to bisect in between I get other crashes that are apparently fixed later.
The game works fine here with r300g.
Note: with current mesa git the problem appears only with r300g, not with swrastg. I have a RV530. After starting the game (7732 release, with default config, i.e. no ~/.config/0ad/config/local.cfg file) I get: TIMER| InitVfs: 53.8333 ms TIMER| InitScripting: 1.18771 ms TIMER| CONFIG_Init: 4.14799 ms and the game blocks.
(In reply to comment #2) > Note: with current mesa git the problem appears only with r300g, not with > swrastg. I have a RV530. After starting the game (7732 release, with default > config, i.e. no ~/.config/0ad/config/local.cfg file) I get: > > TIMER| InitVfs: 53.8333 ms > TIMER| InitScripting: 1.18771 ms > TIMER| CONFIG_Init: 4.14799 ms > > and the game blocks. I don't have an r300 here to test with, could you run it under gdb and Ctrl-C it when it hangs and the type 'bt' to get a backtrace? Also, did you do 'make clean' first? Mesa has trouble building cleanly when jumping between revisions.
Make sure you have both the driver and libGL*.so from git properly installed, and please let me know if you still have problems even with latest libGL*.so.
This was fixed with 1f1928db001527c3dcf1d78d6a5d2ef8f519327b .
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.