Summary: | Highly unstable default Intel HD driver xorg-video-intel-modesetting | ||
---|---|---|---|
Product: | [KDE Neon] neon | Reporter: | Mahendra Tallur <mahen> |
Component: | Packages User Edition | Assignee: | Neon Bugs <neon-bugs> |
Status: | RESOLVED FIXED | ||
Severity: | crash | CC: | jr, neon-bugs, sitter |
Priority: | NOR | ||
Version First Reported In: | unspecified | ||
Target Milestone: | --- | ||
Platform: | Neon | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: | ||
Sentry Crash Report: |
Description
Mahendra Tallur
2018-01-09 09:09:28 UTC
(My NOTE 2 is related to Kaby Lake HD Graphics performance ; some games ran much better on my 5+ y.o HD 3000 than my 1 y.o HD 630. But this is just a remark, my bug is about stability.) My initial bugreport contains a big mistake. The xserver-xorg-video-intel-modesetting driver is a dummy package. So when this one is selected, X actually makes use of the i915 driver and forces kernel mode-setting driver, so, theorically, this should be the driver built into the Ubuntu HWE kernel. However I get the crashes whatever kernel version I use : 4.10 HWE or 4.13 HWE. It would be logical the "xserver-xorg-video-intel-hwe" packages also made use of the aforementioned driver but without the kernel mode-setting, but it's apparently a different one (the package does contain some .so files) To sum up I don't know if the crashes are due to driver versions or the use of KMS. The instability now constantly occurs whatever package I use. It seems related to the fact Ubuntu LTS jumped from kernel 4.10 to 4.13 to apply the Meltdown fix... (?) Maybe modesetting is now always on ? To work around this I temporarily put back my Nvidia card. MEANWHILE I figured out my wife's laptop, which also is a Kaby Lake one (HD 520), doesn't suffer at all from the instability issue (very same software stack). Well, I cannot reproduce it anymore with newer kernels / Intel drivers (HWE in KDE Neon 16.04 or current Neon 18.04). I guess it can be considered as resolved :) |