http://bugzilla.opensuse.org/show_bug.cgi?id=952781 Stefan Dirsch <sndirsch@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Priority|P5 - None |P3 - Medium Status|NEW |CONFIRMED -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=952781 http://bugzilla.opensuse.org/show_bug.cgi?id=952781#c3 --- Comment #3 from Felix Miata <mrmazda@earthlink.net> --- FWIW, these messages: Uhhuh. NMI received for unknown reason 31 on CPU 0. Do you have a strange power saving mode enabled? Dazed and confused, but trying to continue occur whether booted to 13.1, 13.2, 42.1, TW, Fedora, Mageia or Jessie. They are most reliably triggered by opening an HTML5 Youtube video. Neither nohpet nor nmi_watchdog=0 nor BIOS changes have any perceptible effect on their occurrence. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=952781 http://bugzilla.opensuse.org/show_bug.cgi?id=952781#c5 --- Comment #5 from Felix Miata <mrmazda@earthlink.net> --- AFAICT, it's not a SiS chipset. It only depends on the SiS driver for access to non-VESA modes native to virtually all displays made in the past decade: # lspci | grep VGA 0a:03.0 VGA compatible controller: XGI Technology Inc. (eXtreme Graphics Innovation) Z7/Z9 (XG20 core) (prog-if 00 [VGA controller]) It's on a 64-bit server motherboard: http://b2b.gigabyte.com/products/product-page.aspx?pid=2350 Thus, it wouldn't have been a big seller, like those server motherboards with prior century Matrox gfxchips sold around a decade ago and with its users complaining in the not too distant past about mga driver foibles. Couple that with the fact that pre-release bug reporters and pre-release testers are a tiny fraction of total user population, and nobody would expect to hear much from users of this particular gfxchip. That's something I can do, give people a chance to find out about bugs before it's too late for them to get fixed (hopefully before an upcoming GA), so they don't blindly upgrade into an unacceptable condition. Whether it would ever get fixed is a whole different matter. *I* don't need it to work, because it's nothing but a test machine. Some among the multitude of others might not be so lucky. I would like to have reported this upstream, but wasn't able to reproduce on this machine for lack of the sis driver in other installed distros, which apparently force many users needing to stay secure into either using distorted non-native display modes, changing their hardware, or changing to a more complete distro. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=952781 http://bugzilla.opensuse.org/show_bug.cgi?id=952781#c7 --- Comment #7 from Stefan Dirsch <sndirsch@suse.com> --- I don't thinks XGI's Volari Z7/Z9 GPUs ever worked with xf86-video-sis driver. They have been supported by xf86-video-xgi driver (which is basically a fork of xf86-video-sis driver). IIRC we support it still on sle11, but removed support for sle12 and current openSUSE versions. Feel free to clone the xgi git repo and build it yourself. Maybe you're lucky. git://anongit.freedesktop.org/git/xorg/driver/xf86-video-xgi -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=952781 http://bugzilla.opensuse.org/show_bug.cgi?id=952781#c9 Max Staudt <mstaudt@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |mstaudt@suse.com --- Comment #9 from Max Staudt <mstaudt@suse.com> --- Felix, thanks a lot for taking the time to test this, and in general making sure that openSUSE runs well even on older hardware. I'm an enthusiast of those things myself, and really appreciate your help. This is part of what keeps the Linux community thriving in the way it does. Unfortunately, I am unable to find a SiS graphics card of any but the oldest generation, so I am unable to replicate and debug your setup with an XGI card or any other card of that generation. What's more - going forward, User Mode Setting drivers such as xf86-video-sis will stop working in future releases, as we transition into an environment where the X server is no longer allowed to poke the memory directly. This is already the case on UEFI Secure Boot systems, as well as on current Tumbleweed when logging in via gdm (it launches the X server as a non-root process). Seeing as according to "man sis", the Z7 support (which is probably spotty, see Comment 7) does not have XVideo acceleration anyway, may I suggest looking at the Linux kernel's "sis" fbdev driver? It looks very complete, so combining it with xf86-video-fbdev probably makes a good team. This should also future-proof your setup. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=952781 http://bugzilla.opensuse.org/show_bug.cgi?id=952781#c11 --- Comment #11 from Stefan Dirsch <sndirsch@suse.com> --- Max, since Felix has a XGI graphics card card (Volari Z7/Z9), he really needs the xgi instead of the sis driver. I was able to build the xf86-video-xgi driver from git master on openSUSE Leap 42.2, so chances aren, that it even loads with the current Xserver. I wouldn't recommend using any native kernel framebuffer. They are ancient and poorly tested - if tested at all since years. I doubt that Felix's machine has already UEFI so UMS driver should still work. Dislaymanagers still starting Xserver as root are also still available. -- You are receiving this mail because: You are on the CC list for the bug.