On Sat, Nov 22, 2014 at 6:21 PM, Dylan
On 22/11/14 15:34, Mark Goldstein wrote: ...
Yes, I understand that. Note that 173.xxxx branch was dropped after the "incompatible" changes in one of opensource components: kernel and/or x server. I do not have enough knowledge to say whether these changes were necessary and why it was not possible to keep backward compatibility. These things happen from time to time.
AIUI it is the result of legacy kernel features being removed and NVidia not being prepared to update their code. There was plenty of time (years) between implementation of the replacement kernel features and removal of the legacy code. This sort of code housekeeping has to be done to prevent the kernel balooning; remove potentially compromised code; improve efficiency and development processes among many other reasons...
OK, sounds logical.
Unfortunately, nouveau never worked well on my HW (two main issues: fickering, and shifting of the view area of the screen. I use the same screen on my main machine and on old test machine, with KVM), so I'm normally switching to proprietary driver.
This (especially "shifting of the view area") sounds more like an issue with the KVM (in-)correctly reporting or transferring monitor metrics, or otherwise confusing the display adaptor/monitor communication than an issue with the nouveau driver per se. I have often seen similar problems when there is a sync rate or resolution difference between machines sharing a monitor on KVM.
Yep, I also suspect that KVM has a part in it, though with NVidia driver everything works. I can see that nouveau driver detects display type correctly (from EDID info). Maybe I'll play a bi with configurations (copying from xorg.conf generated by nvidia). Thanks, -- Mark Goldstein -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse+owner@opensuse.org