I’ve downgraded mesa on Fedora 40 for now with dnf --enablerepo updates-archive --exclude mesa-libGLU downgrade mesa*, from 24.1.4-2.fc40 to 24.1.2-7.fc40, we’ll see how it goes.
1 Like
That command doesn’t seem to work for me. In fish I’m getting an error about wildcard globbing, and in bash I get Unknown repo: 'updates-archive'.
Oops sorry, I forgot: You have to install the repo:
dnf install fedora-repos-archive
If I remember correctly that enables it by default. That should be ok for day to day operations where you’re not downgrading. If for some reason you want it disabled by default like I have then edit /etc/yum.repos.d/fedora-updates-archive.repo and set enabled=0 there.
For Fedora 40 folks, patched mesa is now on updates-testing.
dnf --enablerepo updates-testing --refresh update mesa*, make sure that pulls the -3 build with the patch.
2 Likes
Sounds like there’s some positive commentary on this, looking forward to this going stable.
Anyone else getting AMD GPU crashes/resets during HW accelerated video playback after mesa and mesa freeworld upgrade to 21.1.4-2 on Fedora?
Started getting them pretty much immediately after update, had to undo the transaction.
Edit: oops, I missed that there was actually a thread on this further down; sorry!
Hi all,
i recieved my FW 16 1,5 Weeks ago and i like it a lot.
I run Fedora 40 and until now i had no stability issues.
Device:
Framework 16
Ryzen 7840HS
64 GB DDR5 - 5600
No GPU Module
Since some update ago 2 or 3 Days i’m seeing graphical issues when using firefox.
This is provocable by hovering over youtube shorts or videos and was capturable with obs.
Other video sites work absolutely fine and until now i was unable to reproduce it anywhere else.
The short has the classic glitch effect and only gets rendered about 30% with pixel fragments at the bottom. After provoking this about 3 or 4 times the screen randomly turns black and turns back on after a few seconds.
After a few rounds of this, the DE freezes and becomes unresponsive.
When switching to the F4 terminal everything still works so the system itself is still responsive. A reboot resets everything back to start.
Dmesg is outputting something when these black sceens are happening:
[ 832.896438] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring vcn_unified_0 timeout, signaled seq=667, emitted seq=668
[ 832.896627] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process RDD Process pid 13389 thread firefox:cs0 pid 16003
[ 832.896790] amdgpu 0000:c1:00.0: amdgpu: GPU reset begin!
[ 833.167944] [drm] Register(0) [regUVD_POWER_STATUS] failed to reach value 0x00000001 != 0x00000002n
[ 833.357368] [drm] Register(0) [regUVD_RB_RPTR] failed to reach value 0x00000080 != 0x00000040n
[ 833.546944] [drm] Register(0) [regUVD_POWER_STATUS] failed to reach value 0x00000001 != 0x00000002n
[ 833.552309] amdgpu 0000:c1:00.0: amdgpu: MODE2 reset
[ 833.589334] amdgpu 0000:c1:00.0: amdgpu: GPU reset succeeded, trying to resume
[ 833.590143] [drm] PCIE GART of 512M enabled (table at 0x00000080FFD00000).
[ 833.590216] amdgpu 0000:c1:00.0: amdgpu: SMU is resuming...
[ 833.591850] amdgpu 0000:c1:00.0: amdgpu: SMU is resumed successfully!
[ 833.594342] [drm] DMUB hardware initialized: version=0x08003D00
[ 833.974983] [drm] kiq ring mec 3 pipe 1 q 0
[ 833.976761] [drm] VCN decode and encode initialized successfully(under DPG Mode).
[ 833.977085] amdgpu 0000:c1:00.0: [drm:jpeg_v4_0_hw_init [amdgpu]] JPEG decode initialized successfully.
[ 833.977628] amdgpu 0000:c1:00.0: amdgpu: ring gfx_0.0.0 uses VM inv eng 0 on hub 0
[ 833.977630] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0
[ 833.977633] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0
[ 833.977634] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 6 on hub 0
[ 833.977636] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 7 on hub 0
[ 833.977637] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 8 on hub 0
[ 833.977638] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 9 on hub 0
[ 833.977639] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 10 on hub 0
[ 833.977641] amdgpu 0000:c1:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 11 on hub 0
[ 833.977642] amdgpu 0000:c1:00.0: amdgpu: ring sdma0 uses VM inv eng 12 on hub 0
[ 833.977644] amdgpu 0000:c1:00.0: amdgpu: ring vcn_unified_0 uses VM inv eng 0 on hub 8
[ 833.977645] amdgpu 0000:c1:00.0: amdgpu: ring jpeg_dec uses VM inv eng 1 on hub 8
[ 833.977646] amdgpu 0000:c1:00.0: amdgpu: ring mes_kiq_3.1.0 uses VM inv eng 13 on hub 0
[ 833.980582] amdgpu 0000:c1:00.0: amdgpu: recover vram bo from shadow start
[ 833.980584] amdgpu 0000:c1:00.0: amdgpu: recover vram bo from shadow done
[ 833.980597] amdgpu 0000:c1:00.0: amdgpu: GPU reset(4) succeeded!
[ 834.043638] [drm:amdgpu_cs_ioctl [amdgpu]] *ERROR* Failed to initialize parser -125!
[ 834.105900] firefox:cs0[16003]: segfault at 0 ip 00005614c6914b25 sp 00007fae65dff9c0 error 6 in firefox[5614c6902000+5c000] likely on CPU 1 (core 0, socket 0)
[ 834.105910] Code: 48 8b 2d b6 24 06 00 48 8b 75 00 e8 65 e2 fe ff 48 8b 75 00 bf 0a 00 00 00 e8 e7 e2 fe ff 48 8d 05 a0 27 06 00 48 89 18 31 d2 <89> 14 25 00 00 00 00 0f 0b f3 0f 1e fa 50 58 48 8d 3d 5d c4 04 00
My first guess would be some random drm problem but i am not knowledgeable enough to search for the root cause.
Should i open a proper support case or does someone experience similar problems?
Edit: Videos on youtube seem to be affected too
1 Like
This appears to be recent for me too on an AMD FW13. I eventually get screen blanking. I started to troubleshoot last night but ran out of time. I will try to get more today.
What’s the thread?
I downgraded using the following for now:
sudo dnf downgrade libva-2 mesa-filesystem
That seems to have stabilized things for now, that doesn’t address the actual problem, but it’s a mesa problem as far as I can tell. Fedora has a bug already on it, I believe. Going to track it down later after work hours.
1 Like
It seems 24.0.5-1 is the version that works for me. updates-testing version still had a problem for me.
1 Like
In the past week or so I’ve encountered this bug often alongside another one where screen brightness behaves weirdly, doing things like
- slowly increasing (buttons don’t interrupt it), screen black, screen back on a min brightness, more quickly increasing, screen black, session crash
- go to max brightness, but updates to the correct level when a brightness button is pressed
- pop to max brightness (buttons don’t interrupt it), and just crash from there
Is this the case for anyone else? If not I’ll wait for the fix to this to come out and make a different post if it doesn’t fix that too.
I’ve also seen weird ambient light sensor changes, this FW13 laptop is brand new for me so I don’t know if this is normal behaviour or not. I know this doesn’t add much but just wanted to mention it.
I just installed the latest updates to Mesa, and based on a quick check it appears that might have fixed the issue.
1 Like
Excellent, I’ll do the update tomorrow or Friday and report back.
Same - 24.1.4-3 has a workaround for this issue, and so far after installing I’m not getting crashes in FF during video playback.
Reporting back as I stated I would, no issues after a little under 2 days of testing.