Created attachment 1922428 [details] journalctl --no-hostname -k > dmesg.txt 1. Please describe the problem: My linux workstation (AMD Ryzen 7 5700G, mainboard with 2 dp and one hdmi sockets) failed to identify and run my threee 4K monitors correctly and run them in their max resolution. When i am patient and play with the display settings a while, the problem seem to disappear until the next reboot. I am running gnome+Wayland and had no problems before with kernel5- 2. What is the Version-Release number of the kernel: Linux version 6.0.5-200.fc36.x86_64 The problem occured firstly with the very first kernel6 update. 3. Did it work previously in Fedora? If so, what kernel version did the issue *first* appear? Old kernels are available for download at https://koji.fedoraproject.org/koji/packageinfo?packageID=8 : yes in 5'er versions. 4. Can you reproduce this issue? If so, please provide the steps to reproduce the issue below: In one of two reboots is see this problems. 5. Does this problem occur with the latest Rawhide kernel? To install the Rawhide kernel, run ``sudo dnf install fedora-repos-rawhide`` followed by ``sudo dnf update --enablerepo=rawhide kernel``: nope 6. Are you running any modules that not shipped with directly Fedora's kernel?: no! 7. Please attach the kernel logs. You can get the complete kernel log for a boot with ``journalctl --no-hostname -k > dmesg.txt``. If the issue occurred on a previous boot, use the journalctl ``-b`` flag.
This seems to be a gnome configuration issuse. Plasma does not have this problem. And when i start the computer, all 3 displays are beeing displayed the rigtht way. When i login in into gnome-wayland, the primary desktop is wrong and some seetings (size of the pointer) has beeen changed. When i try to do some changes, the desktop and the displays go crazy while the control panel shows the configuration with the change primary display but some display disconnects-
Sorry again for the switch from kernel to gnome-desktop to wayland. There seems to be a problem with wayland and the display-management/detection. I tried also Plasma this morning and got a perfect freeze of the whole systeme. Switching to XFree brought back a stable environement.
I recognized several problems with the wayland/gnome. Screen switch to black or become inactive. I crash may occur, i a enlage a firefox window with maximize. In this case one or two of the 3 connected screens become inactive (sometime no signal is reporet). Some indications: - From my point of view there migh be a bug, that cause memory corruption in your code. - There may be also a problem with a selected audio output over DP to the monitor with speakers. (there are 3 monitors, 2 with DP and one with HDMI) If the monitor handling is messed up, i have to reboot in order to get my displays back.
System is running ith a AMD Ryzen 7 5700G (amd code) and there is no attached additional videocard.
I got again a problem, this is what Nov 11 00:12:19 buster firefox[4229]: Couldn't map window 0x7f4d3feec8c0 as subsurface because its parent is not mapped. Nov 11 00:12:19 buster gnome-shell[2403]: gvc_mixer_card_get_index: assertion 'GVC_IS_MIXER_CARD (card)' failed Nov 11 00:12:20 buster gnome-shell[2403]: Failed to post KMS update: drmModeAtomicCommit: Nicht genügend Hauptspeicher verfügbar Nov 11 00:12:20 buster gnome-shell[2403]: Page flip discarded: drmModeAtomicCommit: Nicht genügend Hauptspeicher verfügbar Nov 11 00:12:20 buster kernel: amdgpu 0000:08:00.0: amdgpu: 0000000052fc8dea pin failed Nov 11 00:12:20 buster kernel: [drm:dm_plane_helper_prepare_fb [amdgpu]] *ERROR* Failed to pin framebuffer with error -12 Nov 11 00:12:32 buster gnome-shell[2403]: Failed to post KMS update: drmModeAtomicCommit: Das Argument ist ungültig Nov 11 00:12:32 buster gnome-shell[2403]: Page flip discarded: drmModeAtomicCommit: Das Argument ist ungültig Nov 11 00:12:35 buster cupsd[1453]: REQUEST localhost - - "POST / HTTP/1.1" 200 186 Renew-Subscription successful-ok Nov 11 00:12:56 buster gnome-shell[2403]: Failed to post KMS update: drmModeAtomicCommit: Das Argument ist ungültig Nov 11 00:12:56 buster gnome-shell[2403]: Page flip discarded: drmModeAtomicCommit: Das Argument ist ungültig Nov 11 00:12:56 buster gnome-shell[2403]: Failed to post KMS update: drmModeAtomicCommit: Das Argument ist ungültig Nov 11 00:12:56 buster gnome-shell[2403]: Page flip discarded: drmModeAtomicCommit: Das Argument ist ungültig Nov 11 00:12:56 buster gnome-shell[2403]: Failed to post KMS update: drmModeAtomicCommit: Das Argument ist ungültig Nov 11 00:12:56 buster gnome-shell[2403]: Page flip discarded: drmModeAtomicCommit: Das Argument ist ungültig Nov 11 00:12:56 buster gnome-shell[2403]: Failed to post KMS update: drmModeAtomicCommit: Das Argument ist ungültig Nov 11 00:12:56 buster gnome-shell[2403]: Page flip discarded: drmModeAtomicCommit: Das Argument ist ungültig Das Argument ist ungültig = invalid argument
Nov 11 00:12:20 buster gnome-shell[2403]: Failed to post KMS update: drmModeAtomicCommit: Nicht genügend Hauptspeicher verfügbar Nicht genügend Hauptspeicher verfügbar=not enough main memory available Weired, that system has 64Gb main mem
https://bugzilla.redhat.com/show_bug.cgi?id=1885457 similar problem, also with a AMD sysem
I think i was able to identify the reason for this problem -> BIOS setting. Well my PC system is a AMD Ryzen 7 5700G, there is no dedicated graphic card with a separated video-ram. The GPU resised on the same die as the CPU and the video-ram is located in the normal RAM. As a consequence of this, a part of the normal memory will be used as a video ram in oder to create a video-signal. The manufacturer BIOS setting has a "dynamic" setting set, that means, that the videoram is not fixed reserverd that in turn may be a reason for the messages out-of-memory. After i had set the RAM to 4Gb, the problem disappeared .