AMD phoenix/phoenix2 platforms facing amdgpu(PHX) hangs during stress loading
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
HWE Next |
New
|
Undecided
|
Unassigned | ||
linux-firmware (Ubuntu) |
Fix Released
|
Undecided
|
You-Sheng Yang | ||
Jammy |
Fix Released
|
High
|
You-Sheng Yang | ||
Mantic |
Fix Released
|
High
|
You-Sheng Yang | ||
Noble |
Fix Released
|
Undecided
|
You-Sheng Yang |
Bug Description
[SRU Justification]
[Impact]
With stress tool like 3DMark or GravityMark, facing amdgpu(PHX) hangs within a few minutes or sometimes even quicker
[Fix]
Upstream firmware fixes for Phoenix (GC 11.0.1)/Phoenix 2 (GC 11.0.4), and other prerequisites:
* amdgpu/gc_11_0_1_* up to commit 56c0e7e ("amdgpu: update GC 11.0.1 firmware")
* amdgpu/
* amdgpu/
* amdgpu/gc_11_0_4_* up to commit 680d98c ("amdgpu: update GC 11.0.4 firmware")
* amdgpu/
[Test Case]
Run stress tool like 3DMark or GravityMark.
[Where problems could occur]
Binary firmware update recommended by chip vendor. No known issue so far.
[Other Info]
Phoenix is supported in linux-oem-
========== original bug report ==========
With stress tool like 3DMark or GravityMark, facing amdgpu(PHX) hangs within a few minutes or sometimes even quicker. Also using mantic + v6.7 hit the hang, so need to update new FWs to fix this issue.
PHX series
https:/
https:/
https:/
https:/
https:/
[ 415.782623] [drm:amdgpu_
[ 415.782833] [drm:amdgpu_
[ 415.783004] amdgpu 0000:0d:00.0: amdgpu: GPU reset begin!
[ 415.944129] [drm:mes_
[ 415.944317] [drm:amdgpu_
[ 416.074161] [drm:mes_
[ 416.074327] [drm:amdgpu_
[ 416.204184] [drm:mes_
[ 416.204356] [drm:amdgpu_
[ 416.334204] [drm:mes_
[ 416.334377] [drm:amdgpu_
[ 416.464226] [drm:mes_
[ 416.464398] [drm:amdgpu_
[ 416.594247] [drm:mes_
[ 416.594418] [drm:amdgpu_
[ 416.724265] [drm:mes_
[ 416.724432] [drm:amdgpu_
[ 416.854275] [drm:mes_
[ 416.854437] [drm:amdgpu_
[ 416.984284] [drm:mes_
[ 416.984456] [drm:amdgpu_
[ 416.996743] amdgpu 0000:0d:00.0: amdgpu: MODE2 reset
[ 417.026498] amdgpu 0000:0d:00.0: amdgpu: GPU reset succeeded, trying to resume
[ 417.026909] [drm] PCIE GART of 512M enabled (table at 0x000000801FD00
[ 417.027149] amdgpu 0000:0d:00.0: amdgpu: SMU is resuming...
[ 417.029520] amdgpu 0000:0d:00.0: amdgpu: SMU is resumed successfully!
[ 417.032154] [drm] DMUB hardware initialized: version=0x08003000
[ 417.190837] [drm] kiq ring mec 3 pipe 1 q 0
[ 417.192870] [drm] VCN decode and encode initialized successfully(under DPG Mode).
[ 417.193037] amdgpu 0000:0d:00.0: [drm:jpeg_
[ 417.193447] amdgpu 0000:0d:00.0: amdgpu: ring gfx_0.0.0 uses VM inv eng 0 on hub 0
[ 417.193449] amdgpu 0000:0d:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0
[ 417.193451] amdgpu 0000:0d:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0
[ 417.193452] amdgpu 0000:0d:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 6 on hub 0
[ 417.193453] amdgpu 0000:0d:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 7 on hub 0
[ 417.193454] amdgpu 0000:0d:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 8 on hub 0
[ 417.193455] amdgpu 0000:0d:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 9 on hub 0
[ 417.193456] amdgpu 0000:0d:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 10 on hub 0
[ 417.193458] amdgpu 0000:0d:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 11 on hub 0
[ 417.193459] amdgpu 0000:0d:00.0: amdgpu: ring sdma0 uses VM inv eng 12 on hub 0
[ 417.193460] amdgpu 0000:0d:00.0: amdgpu: ring vcn_unified_0 uses VM inv eng 0 on hub 8
[ 417.193461] amdgpu 0000:0d:00.0: amdgpu: ring jpeg_dec uses VM inv eng 1 on hub 8
[ 417.193462] amdgpu 0000:0d:00.0: amdgpu: ring mes_kiq_3.1.0 uses VM inv eng 13 on hub 0
[ 417.195893] amdgpu 0000:0d:00.0: amdgpu: recover vram bo from shadow start
[ 417.195894] amdgpu 0000:0d:00.0: amdgpu: recover vram bo from shadow done
[ 417.195904] amdgpu 0000:0d:00.0: amdgpu: GPU reset(2) succeeded!
[ 417.197048] [drm] Skip scheduling IBs!
[ 417.197057] [drm] Skip scheduling IBs!
[ 417.197063] [drm] Skip scheduling IBs!
[ 443.578688] [drm:amdgpu_
summary: |
- AMD phenix/phenix2 platforms facing amdgpu(PHX) hangs during stress + AMD phoenix/phoenix2 platforms facing amdgpu(PHX) hangs during stress loading |
Changed in linux-firmware (Ubuntu Jammy): | |
status: | New → In Progress |
Changed in linux-firmware (Ubuntu Mantic): | |
status: | New → In Progress |
Changed in linux-firmware (Ubuntu Noble): | |
status: | New → Incomplete |
status: | Incomplete → Triaged |
Changed in linux-firmware (Ubuntu Jammy): | |
importance: | Undecided → High |
Changed in linux-firmware (Ubuntu Mantic): | |
importance: | Undecided → High |
Changed in linux-firmware (Ubuntu Jammy): | |
assignee: | nobody → You-Sheng Yang (vicamo) |
Changed in linux-firmware (Ubuntu Mantic): | |
assignee: | nobody → You-Sheng Yang (vicamo) |
Changed in linux-firmware (Ubuntu Noble): | |
assignee: | nobody → You-Sheng Yang (vicamo) |
description: | updated |
tags: | added: amd oem-priority originate-from-2051539 |
tags: | added: kern-9038 |
Changed in linux-firmware (Ubuntu Noble): | |
status: | Triaged → Fix Released |
tags: | added: verification-needed-mantic |
All the fixes are in upstream repository, so there should be no work to do for Noble once it migrate to upstream HEAD.