orbea

This issue is caused by a combination of `-O1` or higher and `-march=native`. It can also be reproduced with `-march=bdver2` and `-march=amdfam10`, but doesn't occur with `-march=athlon-xp` or `-march=i686`. I'll...

> Additionally when nir is enabled, wine itself will crash with this very unhelpful backtrace. This issue with nir is now fixed in mesa 19.2 and wine 4.15, but I...

Is this fixed with the mesa master now?

A nouveau developer was looking at this the other day as discussed in irc, but he also ran into crashes. Maybe it was the same issue?

The crashes when replaying the trace were also discussed in irc, I think diagnosing and potentially fixing the crashes when replaying the trace or starting the game would make it...

There are different `kill(1)` signals, presumably metalog could pay attention to how its being killed and give the correct exit code accordingly.

Please tag your commit message with `Upstream-PR:` and `Upstream-Commit:`, you can look at the commit history for other examples. Additionally its helpful to add URLs for the Gentoo issue and...

POSIX says: > For broader portability, the [sigsuspend()](https://pubs.opengroup.org/onlinepubs/9699919799.2018edition/functions/sigsuspend.html) function should be used instead of the obsolescent sigpause() function. > These functions may be removed in a future version. https://pubs.opengroup.org/onlinepubs/9699919799.2018edition/functions/sigpause.html

> Yeah, this has to do with a mismatch in the libc version in the snap vs the one used to build the core. Fwiw the buildbot runs this script...

I'd guess the less obvious issues are a consequence of the statically linked glibc.