Mesa 18.2-RC4 Released Due To Open Graphics Driver Bugs

[phoronix]

Phoronix: Mesa 18.2-RC4 Released Due To Open Graphics Driver Bugs

The fourth release candidate of Mesa 18.2 is out today rather than the final release due to open blocker bugs still persisting...

http://www.phoronix.com/scan.php?page=news_item&px=Mesa-18.2-RC4-Released

[aufkrawall]

When I read something about blocking Mesa regressions on Phoronix, it's in very most cases Intel. One more reason for them to ditch i965 in favor of Gallium?

[puleglot]

When I read something about blocking Mesa regressions on Phoronix, it's in very most cases Intel. One more reason for them to ditch i965 in favor of Gallium?

AMD/RADV devs probably have different release process and don't use a tracker bug. That doesn't mean they have less regressions.

[DanL]

AMD/RADV devs probably have different release process and don't use a tracker bug. That doesn't mean they have less regressions.

Mesa is mesa. If radeonsi or RADV have Piglit regressions, those should also be blocker bugs.

[smitty3268]

Mesa is mesa. If radeonsi or RADV have Piglit regressions, those should also be blocker bugs.

Intel has a very rigorous testing environment that sends large numbers of tests through large numbers of varying hardware. AMD doesn't have that, at least for radeonsi and I imagine there's even less done on the radv side since it's not even supported by AMD.

So it's not really surprising that most of these types of blocker bugs get reported by Intel, because they're the ones doing most of the testing.

[swizzler]

Intel has a very rigorous testing environment that sends large numbers of tests through large numbers of varying hardware. AMD doesn't have that, at least for radeonsi and I imagine there's even less done on the radv side since it's not even supported by AMD.

So it's not really surprising that most of these types of blocker bugs get reported by Intel, because they're the ones doing most of the testing.

I can confirm that this is the case.