http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c1 Stefan Dirsch changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |patrik.jakobsson@suse.com, | |tiwai@suse.com, | |tzimmermann@suse.com Component|X.Org |Kernel Assignee|gfx-bugs@suse.de |kernel-bugs@opensuse.org Summary|HDMI monitor doesn't works |i915: HDMI monitor no | |longer detected after | |update QA Contact|gfx-bugs@suse.de |qa-bugs@suse.de --- Comment #1 from Stefan Dirsch --- Hmm. Looks like a kernel regression to me. [ 6.916] (II) modeset(0): Output HDMI-1 disconnected -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c3 --- Comment #3 from M --- I've tried to boot from other kernel but nothing, the problem still here. I've tried with 5.14.6-2 and 5.13.12-2, the kernels I can see going in "Advanced boot options for OpenSuse Tumbleweed" in the grub boot menu. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c5 --- Comment #5 from M --- As you asked: - the Xorg.0.log from the system where the monitor is detected: https://susepaste.org/74758339 - ouput of `rpm -qa --last | head -n 1` from the system where the monitor is detected: rpm -qa --last | head -n 1 - output of `rpm -qa --last` from the new system, where the monitor is NOT detected: https://susepaste.org/92924404 -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c7 Stefan Dirsch changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(quiquoqua48@gmail | |.com) | --- Comment #7 from Stefan Dirsch --- Ok. Obviously EDID cannot be read out any longer with updated kernel. You've updated > 6000 packages. Even without all the texlive packages we have still > 3000 packages. I won't check these. Let's assume it's an issue with the kernel update. Or maybe also kernel-firmware. kernel-firmware-intel-20210928-1.1.noarch ven 8 ott 2021, 14:16:14 [...] kernel-default-devel-5.14.6-2.1.x86_64 ven 8 ott 2021, 14:13:05 -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c9 --- Comment #9 from Patrik Jakobsson --- Is this a multi-gpu (Intel + Nvidia) system? Perhaps the output is not connected to the i915 card? Can you please provide the output from: cat /sys/kernel/debug/dri/0/i915_display_info of this file doesn't exists, also try: cat /sys/kernel/debug/dri/1/i915_display_info -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c11 --- Comment #11 from Stefan Dirsch --- Please also uninstall bbswitch bbswitch-kmp-default -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c13 --- Comment #13 from Stefan Dirsch --- Ok. So HDMI-0 appears to be connected to Intel. So with all these packages uninstalled could you repost X logfile. And then also try again with older 5.13.12-2 kernel. Also please run xrandr -q in the desktop on your internal screen. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c15 Stefan Dirsch changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(quiquoqua48@gmail | |.com) | --- Comment #15 from Stefan Dirsch --- --- working --- black screen on HDMI monitor [...] - (II) modeset(G0): glamor X acceleration enabled on NV167 - (II) modeset(G0): glamor initialized - (II) modeset(G0): Output HDMI-1-1 has no monitor section - (II) modeset(G0): EDID for output HDMI-1-1 + (II) modeset(G0): Refusing to try glamor on llvmpipe + (EE) modeset(G0): glamor initialization failed + (II) modeset(G0): ShadowFB: preferred YES, enabled YES + (II) modeset(G0): Double-buffered shadow updates: off + (II) modeset(G0): Output HDMI-1-2 has no monitor section + (II) modeset(G0): EDID for output HDMI-1-2 [...] + (II) Loading sub module "shadow" + (II) LoadModule: "shadow" + (II) Loading /usr/lib64/xorg/modules/libshadow.so + (II) Module shadow: vendor="X.Org Foundation" + compiled for 1.20.13, module version = 1.1.0 + ABI class: X.Org ANSI C Emulation, version 0.4 [...] - Setup complete - DRI driver: nouveau - VDPAU driver: nouveau [...] - randr: falling back to unsynchronized pixmap sharing - (II) modeset(G0): Allocate new frame buffer 1920x1080 stride That would mean that before nouveau driver was being used for 2D acceleration for some reason. This makes no sense to me. Your xrandr output looks good. No idea, why the screen stays black. -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 Stefan Dirsch changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |IN_PROGRESS -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c18 Stefan Dirsch changed: What |Removed |Added ---------------------------------------------------------------------------- Flags| |needinfo?(quiquoqua48@gmail | |.com) --- Comment #18 from Stefan Dirsch --- Hmm. I suggest to attach again the output of rpm -qa --last and the new X logfile. Maybe this time we see only a few new packages ... -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c20 --- Comment #20 from Stefan Dirsch --- Thanks. Seems a Mesa update fixed the issue now. nvdia and suse-prime packages haven't been installed again. --- xlog-old.processed 2021-10-15 14:36:08.678183328 +0200 +++ xlog-new.processed 2021-10-15 14:35:11.821941693 +0200 [...] - (II) modeset(G0): Refusing to try glamor on llvmpipe - (EE) modeset(G0): glamor initialization failed - (II) modeset(G0): ShadowFB: preferred YES, enabled YES - (II) modeset(G0): Double-buffered shadow updates: off + (II) modeset(G0): glamor X acceleration enabled on NV167 + (II) modeset(G0): glamor initialized [...] + Setup complete + DRI driver: nouveau + VDPAU driver: nouveau I would feel better if things would also work without packages xf86-video-nouveau Mesa-dri-nouveau libvdpau_nouveau been installed though. I mean just plain intel drivers. If you could give this a try by uninstalling these 3 packages, that would be great! -- You are receiving this mail because: You are on the CC list for the bug.

http://bugzilla.opensuse.org/show_bug.cgi?id=1191515 http://bugzilla.opensuse.org/show_bug.cgi?id=1191515#c22 Stefan Dirsch changed: What |Removed |Added ---------------------------------------------------------------------------- Flags|needinfo?(quiquoqua48@gmail | |.com) | --- Comment #22 from Stefan Dirsch --- Ok. So we had some temporary user space regression in Mesa's nouveau driver it seems (no idea why this driver is needed at all; we have Intel here and nvidia is just nice to have for faster rendering) and probably a kernel regression for suspend, which affectes at least HDMI connections. :-( Thanks for your patience. Unfortunately I can't provide a solution here. :-( Since you don't intend to use the nvidia capacities and outputs apparently are only connected to Intel GPU, it might be an option to disable NVIDIA GPU by BIOS settings - if possible. That would make things much easier and I could imaging things then just working out-of-the-box. -- You are receiving this mail because: You are on the CC list for the bug.