Gstreamer fails on Fedora 38 with Amd GPU

[TijZwa]

If I run python3 /usr/lib64/python3.11/site-packages/xpra/codecs/loader.py -v enc_gstreamer the command fails with amdgpu: The CS has been rejected (-125), but the context isn't robust.

Dmesg log:

[ 2767.491407] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* ring vcn_enc0 timeout, signaled seq=192, emitted seq=193
[ 2767.491661] [drm:amdgpu_job_timedout [amdgpu]] *ERROR* Process information: process python3 pid 18703 thread python3:cs0 pid 18746
[ 2767.491873] amdgpu 0000:0a:00.0: amdgpu: GPU reset begin!
[ 2767.682642] [drm] Register(0) [mmUVD_POWER_STATUS] failed to reach value 0x00000001 != 0x00000002
[ 2767.876425] [drm] Register(0) [mmUVD_RB_RPTR] failed to reach value 0x00000280 != 0x00000240
[ 2768.066497] [drm] Register(0) [mmUVD_POWER_STATUS] failed to reach value 0x00000001 != 0x00000002
[ 2768.106589] amdgpu 0000:0a:00.0: amdgpu: MODE2 reset
[ 2768.106669] amdgpu 0000:0a:00.0: amdgpu: GPU reset succeeded, trying to resume
[ 2768.106861] [drm] PCIE GART of 1024M enabled.
[ 2768.106863] [drm] PTB located at 0x000000F41FC00000
[ 2768.106881] [drm] PSP is resuming...
[ 2768.830789] [drm] reserve 0x400000 from 0xf41f800000 for PSP TMR
[ 2769.099401] amdgpu 0000:0a:00.0: amdgpu: RAS: optional ras ta ucode is not available
[ 2769.109832] amdgpu 0000:0a:00.0: amdgpu: RAP: optional rap ta ucode is not available
[ 2769.109833] amdgpu 0000:0a:00.0: amdgpu: SECUREDISPLAY: securedisplay ta ucode is not available
[ 2769.109835] amdgpu 0000:0a:00.0: amdgpu: SMU is resuming...
[ 2769.110764] amdgpu 0000:0a:00.0: amdgpu: SMU is resumed successfully!
[ 2769.111310] [drm] DMUB hardware initialized: version=0x01010024
[ 2769.305099] [drm] kiq ring mec 2 pipe 1 q 0
[ 2769.308504] [drm] VCN decode and encode initialized successfully(under DPG Mode).
[ 2769.308538] [drm] JPEG decode initialized successfully.
[ 2769.308540] amdgpu 0000:0a:00.0: amdgpu: ring gfx uses VM inv eng 0 on hub 0
[ 2769.308541] amdgpu 0000:0a:00.0: amdgpu: ring gfx_low uses VM inv eng 1 on hub 0
[ 2769.308542] amdgpu 0000:0a:00.0: amdgpu: ring gfx_high uses VM inv eng 4 on hub 0
[ 2769.308543] amdgpu 0000:0a:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 5 on hub 0
[ 2769.308543] amdgpu 0000:0a:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 6 on hub 0
[ 2769.308544] amdgpu 0000:0a:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 7 on hub 0
[ 2769.308544] amdgpu 0000:0a:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 8 on hub 0
[ 2769.308545] amdgpu 0000:0a:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 9 on hub 0
[ 2769.308545] amdgpu 0000:0a:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 10 on hub 0
[ 2769.308546] amdgpu 0000:0a:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 11 on hub 0
[ 2769.308546] amdgpu 0000:0a:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 12 on hub 0
[ 2769.308547] amdgpu 0000:0a:00.0: amdgpu: ring kiq_2.1.0 uses VM inv eng 13 on hub 0
[ 2769.308547] amdgpu 0000:0a:00.0: amdgpu: ring sdma0 uses VM inv eng 0 on hub 1
[ 2769.308548] amdgpu 0000:0a:00.0: amdgpu: ring vcn_dec uses VM inv eng 1 on hub 1
[ 2769.308548] amdgpu 0000:0a:00.0: amdgpu: ring vcn_enc0 uses VM inv eng 4 on hub 1
[ 2769.308549] amdgpu 0000:0a:00.0: amdgpu: ring vcn_enc1 uses VM inv eng 5 on hub 1
[ 2769.308550] amdgpu 0000:0a:00.0: amdgpu: ring jpeg_dec uses VM inv eng 6 on hub 1
[ 2769.310627] amdgpu 0000:0a:00.0: amdgpu: recover vram bo from shadow start
[ 2769.310629] amdgpu 0000:0a:00.0: amdgpu: recover vram bo from shadow done
[ 2769.310646] amdgpu 0000:0a:00.0: amdgpu: GPU reset(9) succeeded!
[ 2769.310906] [drm:amdgpu_cs_ioctl [amdgpu]] *ERROR* Failed to initialize parser -125!

Complete log with XPRA_ALL_DEBUG=1:
out.log

[TijZwa]

There are many reports of this issue in kernel 6.2.x
Kernel 6.2.14 (in updates-testing) does not fix the issue.

Let me downgrade to Fedora 37 and see if that helps.

[totaam]

Will follow up in #3964