5000 series APU fails to resume from S3 (dcn10_init_hw) when not connected to power
Intermittently (but not rarely) upon waking from deep
(S3?) suspend, the GPU does not successfully resume.
Booting with amdgpu.gpu_recovery=1
or drm.vblankoffdelay=0
doesn't seem to make any difference.
I have compared with issues #2040 and #2092, but my logs seem different, so I do not believe (from my naive perspective) that this bug is the same.
Some dmesg extracts follow:
Feb 09 16:02:54 stelmaria kernel: [drm] PSP is resuming...
Feb 09 16:02:54 stelmaria kernel: [drm] reserve 0x400000 from 0xf41f800000 for PSP TMR
Feb 09 16:02:54 stelmaria kernel: [drm] psp gfx command SETUP_TMR(0x5) failed and response status is (0x80000306)
Feb 09 16:02:54 stelmaria kernel: [drm] failed to load ucode CP_MEC1(0x19)
Feb 09 16:02:54 stelmaria kernel: [drm] psp gfx command LOAD_IP_FW(0x6) failed and response status is (0x80000203)
Feb 09 16:02:54 stelmaria kernel: [drm] failed to load ucode VCN(0x36)
Feb 09 16:02:54 stelmaria kernel: [drm] psp gfx command LOAD_IP_FW(0x6) failed and response status is (0x80000203)
Feb 09 16:02:54 stelmaria kernel: [drm] failed to load ucode DMCUB(0x3C)
Feb 09 16:02:54 stelmaria kernel: [drm] psp gfx command LOAD_IP_FW(0x6) failed and response status is (0x80000203)
Feb 09 16:02:54 stelmaria kernel: [drm] Wait for DMUB auto-load failed: 3
Feb 09 16:02:54 stelmaria kernel: [drm] DMUB hardware initialized: version=0x01010024
Feb 09 16:02:54 stelmaria kernel: [drm:dc_dmub_srv_wait_idle [amdgpu]] *ERROR* Error waiting for DMUB idle: status=3
Feb 09 16:02:54 stelmaria kernel: Call Trace:
Feb 09 16:02:54 stelmaria kernel: <TASK>
Feb 09 16:02:54 stelmaria kernel: dcn10_init_hw+0x3a8/0x5a0 [amdgpu]
Feb 09 16:02:54 stelmaria kernel: dc_set_power_state+0x105/0x160 [amdgpu]
Feb 09 16:02:54 stelmaria kernel: dm_resume+0xe1/0x600 [amdgpu]
Feb 09 16:02:54 stelmaria kernel: amdgpu_device_ip_resume_phase2+0x4f/0x100 [amdgpu]
Feb 09 16:02:54 stelmaria kernel: amdgpu_device_resume+0xbf/0x300 [amdgpu]
Feb 09 16:02:54 stelmaria kernel: ? pci_legacy_resume+0x80/0x80
Feb 09 16:02:54 stelmaria kernel: amdgpu_pmops_resume+0x32/0x70 [amdgpu]
Feb 09 16:02:54 stelmaria kernel: ? pci_legacy_resume+0x80/0x80
Feb 09 16:02:54 stelmaria kernel: dpm_run_callback+0x47/0x150
Feb 09 16:02:54 stelmaria kernel: device_resume+0x88/0x190
Feb 09 16:02:54 stelmaria kernel: async_resume+0x19/0x30
Feb 09 16:02:54 stelmaria kernel: async_run_entry_fn+0x2d/0x130
Feb 09 16:02:54 stelmaria kernel: process_one_work+0x1c4/0x380
Feb 09 16:02:54 stelmaria kernel: worker_thread+0x4d/0x380
Feb 09 16:02:54 stelmaria kernel: ? rescuer_thread+0x3a0/0x3a0
Feb 09 16:02:54 stelmaria kernel: kthread+0xe6/0x110
Feb 09 16:02:54 stelmaria kernel: ? kthread_complete_and_exit+0x20/0x20
Feb 09 16:02:54 stelmaria kernel: ret_from_fork+0x1f/0x30
Feb 09 16:02:54 stelmaria kernel: </TASK>
Feb 09 16:02:54 stelmaria kernel: [drm] perform_link_training_with_retries: Link(0) training attempt 1 of 4 failed @ rate(10) x lane(2) : fail reason:(1)
Hardware description:
- Device: Star Labs StarBook MkVI (BIOS 1.0.4)
- CPU: AMD Ryzen 7 5800U with Radeon Graphics
- GPU: Cezanne [Radeon Vega Series / Radeon Vega Mobile Series] [1002:1638] (rev c1)
- System Memory: 64 GiB
- Display(s): Internal laptop panel
- Type of Display Connection: eDP
System information:
- Distro name and Version: Debian bookworm/testing (current as of 23-02-09)
- Kernel version:
6.1.0-3-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.8-1 (2023-01-29) x86_64 GNU/Linux
How to reproduce the issue:
- Suspend laptop
- Attempt to resume from suspend
dmesg logs:
Edited by Toby Smithe