amdgpu module crashes upon resuming from hibernation

pzdn · 2026-07-15 15:50:35

Hello, I'm in possession of Thinkpad P14s Gen7 laptop with AMD Ryzen AI 7 PRO 450 with Radeon 860M iGPU. I've had this laptop for a couple of days now and ever since i set up hibernation, the system often times can't resume successfully. Sometimes it does in fact resume without any problems whatsoever, but often i see just a blank screen when the DE should load - I then have to forcefully restart the device. I also once happened to see various colorful artifacts on the entire screen instead of a blank one, which would indicate some graphics recovery issue. I inspected the journal logs to find anything that would help me determine the cause of the issue and I can see logs indicating that the amdgpu module has crashed:

Jul 15 17:17:28 thinkpad kernel: ACPI: EC: interrupt blocked
Jul 15 17:17:28 thinkpad kernel: ACPI: EC: interrupt unblocked
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: [drm] PCIE GART of 512M enabled (table at 0x00000081FFB00000).
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: SMU is resuming...
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: SMU is resumed successfully!
Jul 15 17:17:28 thinkpad kernel: nvme nvme0: D3 entry latency set to 10 seconds
Jul 15 17:17:28 thinkpad kernel: nvme nvme0: 16/0/0 default/read/poll queues
Jul 15 17:17:28 thinkpad kernel: ------------[ cut here ]------------
Jul 15 17:17:28 thinkpad kernel: drm_crtc_vblank_get(new_state->base.crtc) != 0
Jul 15 17:17:28 thinkpad kernel: WARNING: drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:9844 at amdgpu_dm_atomic_commit_tail+0x34f2/0x3990 [amdgpu], CPU#4: kworker/u64:57/3125
Jul 15 17:17:28 thinkpad kernel: Modules linked in: ccm rfcomm snd_seq_dummy snd_hrtimer snd_seq snd_seq_device nft_fib_inet nft_fib_ipv4 nft_fib_ipv6 nft_fib nft_reject_inet nf_reject_ipv4 nf_reject_ipv6 nft_reje>
Jul 15 17:17:28 thinkpad kernel:  soundwire_generic_allocation mousedev mt7925e snd_hda_codec_realtek_lib soundwire_bus snd_hda_scodec_component mt7925_common snd_soc_sdca snd_hda_codec_atihdmi snd_hda_codec_gener>
Jul 15 17:17:28 thinkpad kernel:  thinkpad_acpi i2c_algo_bit drm_ttm_helper sparse_keymap aesni_intel ttm snd drm_exec soundcore drm_panel_backlight_quirks gf128mul hid_sensor_hub gpu_sched aead rfkill nvme drm_su>
Jul 15 17:17:28 thinkpad kernel: CPU: 4 UID: 0 PID: 3125 Comm: kworker/u64:57 Tainted: G        W  OE       7.1.3-arch1-3 #1 PREEMPT(full)  86fc3a2f6e9718881495ee2d1f774c90f2991f18
Jul 15 17:17:28 thinkpad kernel: Tainted: [W]=WARN, [O]=OOT_MODULE, [E]=UNSIGNED_MODULE
Jul 15 17:17:28 thinkpad kernel: Hardware name: LENOVO 21X00019PB/21X00019PB, BIOS R38ET26W (1.06 ) 06/03/2026
Jul 15 17:17:28 thinkpad kernel: Workqueue: async async_run_entry_fn
Jul 15 17:17:28 thinkpad kernel: RIP: 0010:amdgpu_dm_atomic_commit_tail+0x34f2/0x3990 [amdgpu]
Jul 15 17:17:28 thinkpad kernel: Code: ff 49 8b b4 24 58 01 00 00 48 8b bd 30 fe ff ff e8 53 59 3a 00 48 85 c0 0f 85 60 d2 ff ff e9 81 d0 40 00 31 f6 e9 91 d3 ff ff <0f> 0b e9 82 da ff ff 0f 0b e9 6a da ff ff 0f 0>
Jul 15 17:17:28 thinkpad kernel: RSP: 0018:ffffd4d389347908 EFLAGS: 00010282
Jul 15 17:17:28 thinkpad kernel: RAX: 00000000ffffffea RBX: ffff8e7798149000 RCX: 0000000000000000
Jul 15 17:17:28 thinkpad kernel: RDX: 0000000000000286 RSI: 0000000000000286 RDI: ffff8e7797700144
Jul 15 17:17:28 thinkpad kernel: RBP: ffffd4d389347b70 R08: 0000000000000000 R09: ffffd4d3893479e0
Jul 15 17:17:28 thinkpad kernel: R10: 0000000000000000 R11: ffff8e77a9151d00 R12: ffff8e7806268800
Jul 15 17:17:28 thinkpad kernel: R13: 0000000000000000 R14: 0000000000000000 R15: ffff8e7797700160
Jul 15 17:17:28 thinkpad kernel: FS:  0000000000000000(0000) GS:ffff8e811ee51000(0000) knlGS:0000000000000000
Jul 15 17:17:28 thinkpad kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Jul 15 17:17:28 thinkpad kernel: CR2: 00007f6f56c8c410 CR3: 00000006ce022000 CR4: 0000000000f50ef0
Jul 15 17:17:28 thinkpad kernel: PKRU: 55555554
Jul 15 17:17:28 thinkpad kernel: Call Trace:
Jul 15 17:17:28 thinkpad kernel:  <TASK>
Jul 15 17:17:28 thinkpad kernel:  ? drm_atomic_helper_check_planes+0x157/0x220
Jul 15 17:17:28 thinkpad kernel:  ? amdgpu_device_rreg+0x1c/0x40 [amdgpu 6a36ef633a5e3f82fe6400810fbdff3004cbc814]
Jul 15 17:17:28 thinkpad kernel:  commit_tail+0xc1/0x160
Jul 15 17:17:28 thinkpad kernel:  drm_atomic_helper_commit+0x158/0x1a0
Jul 15 17:17:28 thinkpad kernel:  drm_atomic_commit+0xb1/0xe0
Jul 15 17:17:28 thinkpad kernel:  ? __pfx___drm_printfn_info+0x10/0x10
Jul 15 17:17:28 thinkpad kernel:  drm_atomic_helper_commit_duplicated_state+0x14e/0x160
Jul 15 17:17:28 thinkpad kernel:  drm_atomic_helper_resume+0xd5/0x1a0
Jul 15 17:17:28 thinkpad kernel:  dm_resume+0x76f/0x9b0 [amdgpu 6a36ef633a5e3f82fe6400810fbdff3004cbc814]
Jul 15 17:17:28 thinkpad kernel:  amdgpu_ip_block_resume+0x27/0x50 [amdgpu 6a36ef633a5e3f82fe6400810fbdff3004cbc814]
Jul 15 17:17:28 thinkpad kernel:  amdgpu_device_ip_resume_phase3+0x5d/0x80 [amdgpu 6a36ef633a5e3f82fe6400810fbdff3004cbc814]
Jul 15 17:17:28 thinkpad kernel:  amdgpu_device_resume+0xc5/0x370 [amdgpu 6a36ef633a5e3f82fe6400810fbdff3004cbc814]
Jul 15 17:17:28 thinkpad kernel:  ? __pfx_pci_pm_resume+0x10/0x10
Jul 15 17:17:28 thinkpad kernel:  ? pci_device_is_present+0x52/0x70
Jul 15 17:17:28 thinkpad kernel:  amdgpu_pmops_resume+0x4f/0xa0 [amdgpu 6a36ef633a5e3f82fe6400810fbdff3004cbc814]
Jul 15 17:17:28 thinkpad kernel:  ? __pfx_pci_pm_resume+0x10/0x10
Jul 15 17:17:28 thinkpad kernel:  dpm_run_callback+0x62/0x1b0
Jul 15 17:17:28 thinkpad kernel:  ? dpm_wait_for_superior+0x117/0x180
Jul 15 17:17:28 thinkpad kernel:  device_resume+0x19c/0x2a0
Jul 15 17:17:28 thinkpad kernel:  async_resume+0x21/0x30
Jul 15 17:17:28 thinkpad kernel:  async_run_entry_fn+0x36/0x160
Jul 15 17:17:28 thinkpad kernel:  process_one_work+0x19f/0x390
Jul 15 17:17:28 thinkpad kernel:  worker_thread+0x1b1/0x310
Jul 15 17:17:28 thinkpad kernel:  ? __pfx_worker_thread+0x10/0x10
Jul 15 17:17:28 thinkpad kernel:  kthread+0xe4/0x120
Jul 15 17:17:28 thinkpad kernel:  ? __pfx_kthread+0x10/0x10
Jul 15 17:17:28 thinkpad kernel:  ret_from_fork+0x2a7/0x330
Jul 15 17:17:28 thinkpad kernel:  ? __pfx_kthread+0x10/0x10
Jul 15 17:17:28 thinkpad kernel:  ret_from_fork_asm+0x1a/0x30
Jul 15 17:17:28 thinkpad kernel:  </TASK>
Jul 15 17:17:28 thinkpad kernel: ---[ end trace 0000000000000000 ]---
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring gfx_0.0.0 uses VM inv eng 0 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring comp_1.0.0 uses VM inv eng 1 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring comp_1.1.0 uses VM inv eng 4 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring comp_1.2.0 uses VM inv eng 6 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring comp_1.3.0 uses VM inv eng 7 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring comp_1.0.1 uses VM inv eng 8 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring comp_1.1.1 uses VM inv eng 9 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring comp_1.2.1 uses VM inv eng 10 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring comp_1.3.1 uses VM inv eng 11 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring sdma0 uses VM inv eng 12 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring vcn_unified_0 uses VM inv eng 0 on hub 8
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring jpeg_dec_0 uses VM inv eng 1 on hub 8
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring mes_kiq_3.1.0 uses VM inv eng 13 on hub 0
Jul 15 17:17:28 thinkpad kernel: amdgpu 0000:c4:00.0: ring vpe uses VM inv eng 4 on hub 8
Jul 15 17:17:28 thinkpad kernel: OOM killer enabled.
Jul 15 17:17:28 thinkpad kernel: Restarting tasks: Starting
Jul 15 17:17:28 thinkpad kernel: Restarting tasks: Done
Jul 15 17:17:28 thinkpad kernel: efivarfs: resyncing variable state
Jul 15 17:17:28 thinkpad kernel: efivarfs: finished resyncing variable state
Jul 15 17:17:28 thinkpad kernel: random: crng reseeded on system resumption
Jul 15 17:17:28 thinkpad systemd-sleep[3614]: System returned from sleep operation 'suspend-then-hibernate'.
Jul 15 17:17:28 thinkpad kernel: PM: suspend exit
Jul 15 17:17:29 thinkpad systemd-sleep[3614]: Performing sleep operation 'hibernate'...
Jul 15 17:17:29 thinkpad kernel: PM: hibernation: hibernation entry
-- Boot 1ead0a80bf7247009d5b49983027913d --

My hibernation setup looks as following: I've set up a 48GB swapfile (my machine has 48GB RAM) on my Btrfs filesystem in /swap/swapfile and I followed the Arch Wiki guide on how to set up the swapfile under Btrfs (https://wiki.archlinux.org/title/Power_ … ibernation); I used

btrfs inspect-internal map-swapfile -r /swap/swapfile

to determine the file offset and

findmnt -no UUID -T /swap/swapfile

for the file UUID; then I simply put it inside /etc/cmdline.d/resume.conf like so:

resume=UUID=29f3726a-d71c-4705-adcb-be3d4899f162 resume_offset=4252582

I use the hibernate=nocompress parameter, but I've also experimented with hibernate.compression=lz4 without any positive results. My /etc/fstab for the swapfile is pretty basic:

/swap/swapfile  none    swap    defaults        0       0

I don't know what to do essentially at this point and I desperately need some help. Please let me know what else can I try or how to debug the issue further, as it might be the kernel driver bug, but I'm not sure. I'd be vary thankful for any help. Also, I apologize for any missing debugging data in case I forgot to mention something; I'm relatively new to the community.

Last edited by pzdn (2026-07-16 11:30:27)

seth · 2026-07-15 19:50:26

Please use [code][/code] tags, not "quote" tags. Edit your post in this regard.

That's a warning, not a "crash", happens

Jul 15 17:17:28 thinkpad kernel: drm_crtc_vblank_get(new_state->base.crtc) != 0

when checking the monitors vsync counter and

Jul 15 17:17:28 thinkpad systemd-sleep[3614]: System returned from sleep operation 'suspend-then-hibernate'.
Jul 15 17:17:28 thinkpad kernel: PM: suspend exit
Jul 15 17:17:29 thinkpad systemd-sleep[3614]: Performing sleep operation 'hibernate'...
Jul 15 17:17:29 thinkpad kernel: PM: hibernation: hibernation entry

*before* the actual hibernation.

=> Is this strictly caused by hibernation (S4) or also (longer) S3 or s2idle cycles?

I then have to forcefully restart the device

Avoid that at all costs.
Can you
1. switch the VT (ctrl+alt+f3)
2. ssh into the system
3. reboot using https://wiki.archlinux.org/title/Keyboa … el_(SysRq) + REISUB (nb. you'll have to enable that before the "crash")

Are you dual booting (with windows)?

pzdn · 2026-07-16 14:08:13

Hello seth,

thank you for your remark; I've changed the "quote" tags into "code".

The issue itself only appears after resuming from hibernation (S4); my laptop does not support the deep sleep mode (S3), and I haven't noticed any behaviour like that ever when resuming from s2idle.
1. I cannot switch the VT, nothing happens when i do so.
2. I cannot determine whether ssh works reliably, as once I could connect and I had a working session, while the other time the machine didn't respond ("No route to host")
3. Yes, I can reliably restart the machine using REISUB

I do not have dual boot on this machine; Arch is the one and only installed OS.

I was digging more into this issue and I found out that removing the kms hook from /etc/mkinitcpio.conf and rebuilding the initramfs fixes the issue. I think there is some race condition when it comes to resuming the iGPU, leaving the system in a frozen state and delaying this process helps. However I presume this is more of a workaround rather than an actual fix of the issue. Nevertheless, without this hook the machine has been resuming without any issues whatsoever.

seth · 2026-07-16 16:09:43

I was digging more into this issue and I found out that removing the kms hook from /etc/mkinitcpio.conf and rebuilding the initramfs fixes the issue. I think there is some race condition when it comes to resuming the iGPU, leaving the system in a frozen state and delaying this process helps. However I presume this is more of a workaround rather than an actual fix of the issue. Nevertheless, without this hook the machine has been resuming without any issues whatsoever.

This is actually a known issue w/ nvidia's VRAM preservation and AMD might run into a similar problem, https://bbs.archlinux.org/viewtopic.php?id=285508

Does the journal have more useful information (notably from after the hibernation) when rebooting w/ the sysrq (in doubt give each stage a couple of seconds, S and U will sync and umount all filesystems what should™ preserve the journal)

Arch Linux

#1 2026-07-15 15:50:35

amdgpu module crashes upon resuming from hibernation

#2 2026-07-15 19:50:26

Re: amdgpu module crashes upon resuming from hibernation

#3 2026-07-16 14:08:13

Re: amdgpu module crashes upon resuming from hibernation

#4 2026-07-16 16:09:43

Re: amdgpu module crashes upon resuming from hibernation

Board footer