AMD GPU hangs on resuming from suspend

seth · 2025-08-24 18:48:25

The OP wrote:

I'm on the Zen kernel (although the exact same thing happens on the most recent base and lts kernels)

so you're unlikely to experience the *same* problem?

@eternalflameiv
https://gitlab.freedesktop.org/drm/amd/-/issues/
nb. that your last journal snippet isn't really meaningful, the problems start w/

Aug 22 07:38:55 EternityPC kernel: amdgpu 0000:08:00.0: amdgpu: SMU: I'm not done with your previous command: SMN_C2PMSG_66:0x0000003A SMN_C2PMSG_82:0x00000000
Aug 22 07:38:55 EternityPC kernel: amdgpu 0000:08:00.0: amdgpu: RunBtc failed!
Aug 22 07:38:55 EternityPC kernel: amdgpu 0000:08:00.0: amdgpu: Failed to setup smc hw!
Aug 22 07:38:55 EternityPC kernel: amdgpu 0000:08:00.0: amdgpu: resume of IP block <smu> failed -62
Aug 22 07:38:55 EternityPC kernel: amdgpu 0000:08:00.0: amdgpu: amdgpu_device_ip_resume failed (-62).
Aug 22 07:38:55 EternityPC kernel: amdgpu 0000:08:00.0: PM: dpm_run_callback(): pci_pm_resume returns -62
Aug 22 07:38:55 EternityPC kernel: amdgpu 0000:08:00.0: PM: failed to resume async: error -62

and then the GPU keeps failing to recover from this through resets

there're a couple of maybe related bugs

Edit: https://gitlab.freedesktop.org/drm/amd/-/issues/2614 seems related to runpm…

amdgpu.runpm=0 amdgpu.bapm=0 amdgpu.aspm=0

Last edited by seth (2025-08-24 18:51:08)

eternalflameiv · 2025-08-24 19:11:17

I'm currently on 6.16.1, so the problem is probably not the same. I'll upgrade and see if it goes away, though.

Update: neither a kernel upgrade nor these command line options resolved the issue. I'm attempting to run the debug tools recommended by drm and preparing to file a bug report upstream.

Last edited by eternalflameiv (2025-08-24 20:06:44)

janp · 2025-08-25 06:34:48

eternalflameiv wrote:

I'm currently on 6.16.1, so the problem is probably not the same. I'll upgrade and see if it goes away, though.
Update: neither a kernel upgrade nor these command line options resolved the issue. I'm attempting to run the debug tools recommended by drm and preparing to file a bug report upstream.

I have a 5600 Xt and, as I said before, I got this problem after upgrading to 6.16.2. I downgraded the kernel to 6.16.1 and suspend/resume works correctly again.
I got a look to the 6.16.2 changelog and it seems that they touched more than something about suspend/resume. In particular they touched the drm/amdgpu module.
I add that I noticed that something changed about suspend/resume in the networking subsystem. After some kernel upgrade I need to always reconfigure some iptables behavior that before was correctly restored at resume.
To sum up, they may broke up something in the kernel.

seth · 2025-08-25 07:44:38

To sum up, your situation s vastly different from the OPs.

What does the journal after the failed resume actually look like for you?
(Make sure to reboot w/ the https://wiki.archlinux.org/title/Keyboa … el_(SysRq) + REISUB to preserve the journal)

janp · 2025-08-25 10:46:58

seth wrote:

To sum up, your situation s vastly different from the OPs.
What does the journal after the failed resume actually look like for you?
(Make sure to reboot w/ the https://wiki.archlinux.org/title/Keyboa … el_(SysRq) + REISUB to preserve the journal)

Are you responding to me?

I got the very same message as the OPs

Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: SMU: I'm not done with your previous c>
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: RunBtc failed!
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: Failed to setup smc hw!
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: resume of IP block <smu> failed -62
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: amdgpu_device_ip_resume failed (-62).
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: PM: dpm_run_callback(): pci_pm_resume returns >
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: PM: failed to resume async: error -62

eternalflameiv · 2025-08-25 12:18:19

janp wrote:

seth wrote:
To sum up, your situation s vastly different from the OPs.
(...)
Are you responding to me?
I got the very same message as the OPs(...)

seth seems to mean that we can't be experiencing the exact same bug, since your issues were relieved by downgrading to 6.12.1 and mine were still occurring on that same kernel.

janp · 2025-08-25 13:12:02

eternalflameiv wrote:

janp wrote:
seth wrote:
To sum up, your situation s vastly different from the OPs.
(...)
Are you responding to me?
I got the very same message as the OPs(...)
seth seems to mean that we can't be experiencing the exact same bug, since your issues were relieved by downgrading to 6.12.1 and mine were still occurring on that same kernel.

Did you try with the "plain" 6.16.1 kernel ? I got problems when I tried to use the Zen kernels a few months ago. Don't want to use them again.

seth · 2025-08-25 17:55:32

The failure message looks close enough, though.
Please post the complete journal covering the failure as well as a "good" suspend/resume w/ the 6.16.1 kernel

If this is a regression for you, maybe we can spot between the two suspends where things are falling apart (it would notably be interesting if you get the pci_pm_resume failure all the time but previously the reset succeeded)

On a formal note, everyone lease avoid bloating the thread w/ pointless full quotes, with thanks from my mouse wheel

janp · 2025-08-25 19:42:30

seth wrote:

The failure message looks close enough, though.

Please use the link below for a full journal from my machine with 6.16.2, 6.16.3 and back to 6.16.1

http://0x0.st/K-Ca.txt

I agree with you about being synthetic however at least a reference is welcome.

Thanks

fooblub · 2025-08-26 13:04:49

Same problem here with 5700 XT and 6.16.3-arch1-1

sudo journalctl -b -1 -r

Aug 26 14:48:10 home kernel: PM: suspend exit
Aug 26 14:48:10 home kernel: random: crng reseeded on system resumption
Aug 26 14:48:10 home kernel: Restarting tasks: Done
Aug 26 14:48:10 home kernel: Restarting tasks: Starting
Aug 26 14:48:10 home kernel: OOM killer enabled.
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: PM: failed to resume async: error -62
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: PM: dpm_run_callback(): pci_pm_resume returns -62
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: amdgpu_device_ip_resume failed (-62).
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: resume of IP block <smu> failed -62
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: Failed to setup smc hw!
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: RunBtc failed!
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: SMU: I'm not done with your previous command: SMN_C2PMSG_66:0x0000003A SMN_C2PMSG_82:0x00000000
Aug 26 14:48:10 home kernel: r8169 0000:04:00.0 enp4s0: Link is Up - 1Gbps/Full - flow control rx/tx
Aug 26 14:48:10 home kernel: ata2: SATA link down (SStatus 0 SControl 300)
Aug 26 14:48:10 home kernel: ata1: SATA link down (SStatus 0 SControl 300)
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: smc_dpm_info table revision(format.content): 4.5
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: use vbios provided pptable
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: SMU is resuming...
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: SECUREDISPLAY: optional securedisplay ta ucode is not available
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: RAP: optional rap ta ucode is not available
Aug 26 14:48:10 home kernel: r8169 0000:04:00.0 enp4s0: Link is Down
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: RAS: optional ras ta ucode is not available
Aug 26 14:48:10 home kernel: nvme nvme1: 24/0/0 default/read/poll queues
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: reserve 0x900000 from 0x81fd000000 for PSP TMR
Aug 26 14:48:10 home kernel: nvme nvme0: Ignoring bogus Namespace Identifiers
Aug 26 14:48:10 home kernel: nvme nvme0: 8/0/0 default/read/poll queues
Aug 26 14:48:10 home kernel: nvme nvme0: D3 entry latency set to 10 seconds
Aug 26 14:48:10 home kernel: amdgpu 0000:0b:00.0: amdgpu: PSP is resuming...
Aug 26 14:48:10 home kernel: [drm] PCIE GART of 512M enabled (table at 0x00000081FEE00000).
Aug 26 14:48:10 home kernel: pcieport 0000:00:08.1: PME: Spurious native interrupt!
Aug 26 14:48:10 home kernel: printk: Suspending console(s) (use no_console_suspend to debug)
Aug 26 14:48:10 home kernel: Freezing remaining freezable tasks completed (elapsed 0.001 seconds)
Aug 26 14:48:10 home kernel: Freezing remaining freezable tasks
Aug 26 14:48:10 home kernel: OOM killer disabled.
Aug 26 14:48:10 home kernel: Freezing user space processes completed (elapsed 0.001 seconds)
Aug 26 14:48:10 home kernel: Freezing user space processes
Aug 26 14:26:01 home kernel: Filesystems sync: 0.041 seconds
Aug 26 14:26:01 home systemd-timesyncd[811]: No network connectivity, watching for changes.
Aug 26 14:26:01 home systemd-networkd[1218]: enp4s0: DHCP lease lost
Aug 26 14:26:01 home bluetoothd[1137]: Controller resume with wake event 0x0
Aug 26 14:26:01 home systemd-networkd[1218]: enp4s0: Lost carrier
Aug 26 14:26:01 home kernel: PM: suspend entry (s2idle)
Aug 26 14:26:01 home kernel: PM: suspend exit
Aug 26 14:26:01 home kernel: random: crng reseeded on system resumption
Aug 26 14:26:01 home kernel: Restarting tasks: Done
Aug 26 14:26:01 home kernel: Restarting tasks: Starting
Aug 26 14:26:01 home kernel: OOM killer enabled.
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring jpeg_dec uses VM inv eng 5 on hub 8
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring vcn_enc1 uses VM inv eng 4 on hub 8
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring vcn_enc0 uses VM inv eng 1 on hub 8
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring vcn_dec uses VM inv eng 0 on hub 8
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring sdma1 uses VM inv eng 13 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring sdma0 uses VM inv eng 12 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring kiq_0.2.1.0 uses VM inv eng 11 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 10 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 9 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 8 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 7 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 6 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 5 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: ring gfx_0.0.0 uses VM inv eng 0 on hub 0
Aug 26 14:26:01 home kernel: ---[ end trace 0000000000000000 ]---
Aug 26 14:26:01 home kernel:  </TASK>
Aug 26 14:26:01 home kernel:  ret_from_fork_asm+0x1a/0x30
Aug 26 14:26:01 home kernel:  ? __pfx_kthread+0x10/0x10
Aug 26 14:26:01 home kernel:  ret_from_fork+0x19a/0x1d0
Aug 26 14:26:01 home kernel:  ? __pfx_kthread+0x10/0x10
Aug 26 14:26:01 home kernel:  ? __pfx_kthread+0x10/0x10
Aug 26 14:26:01 home kernel:  kthread+0xfc/0x240
Aug 26 14:26:01 home kernel:  ? __pfx_worker_thread+0x10/0x10
Aug 26 14:26:01 home kernel:  worker_thread+0x2d7/0x410
Aug 26 14:26:01 home kernel:  process_one_work+0x193/0x350
Aug 26 14:26:01 home kernel:  ? srso_alias_return_thunk+0x5/0xfbef5
Aug 26 14:26:01 home kernel:  async_run_entry_fn+0x36/0x140
Aug 26 14:26:01 home kernel:  async_resume+0x21/0x30
Aug 26 14:26:01 home kernel:  device_resume+0x183/0x280
Aug 26 14:26:01 home kernel:  dpm_run_callback+0x4a/0x150
Aug 26 14:26:01 home kernel:  ? __pfx_pci_pm_resume+0x10/0x10
Aug 26 14:26:01 home kernel:  amdgpu_pmops_resume+0x46/0x80 [amdgpu 941737443c43cca1a47ab1af779fa2cc0679b091]
Aug 26 14:26:01 home kernel:  amdgpu_device_resume+0xae/0x310 [amdgpu 941737443c43cca1a47ab1af779fa2cc0679b091]
Aug 26 14:26:01 home kernel:  amdgpu_device_ip_resume_phase3+0x5d/0x80 [amdgpu 941737443c43cca1a47ab1af779fa2cc0679b091]
Aug 26 14:26:01 home kernel:  amdgpu_ip_block_resume+0x27/0x50 [amdgpu 941737443c43cca1a47ab1af779fa2cc0679b091]
Aug 26 14:26:01 home kernel:  dm_resume+0x398/0x7a0 [amdgpu 941737443c43cca1a47ab1af779fa2cc0679b091]
Aug 26 14:26:01 home kernel:  drm_atomic_helper_resume+0xa1/0x170
Aug 26 14:26:01 home kernel:  drm_atomic_helper_commit_duplicated_state+0xde/0xf0
Aug 26 14:26:01 home kernel:  ? __pfx___drm_printfn_info+0x10/0x10
Aug 26 14:26:01 home kernel:  drm_atomic_commit+0xb1/0xe0
Aug 26 14:26:01 home kernel:  drm_atomic_helper_commit+0x13c/0x180
Aug 26 14:26:01 home kernel:  commit_tail+0xa1/0x130
Aug 26 14:26:01 home kernel:  <TASK>
Aug 26 14:26:01 home kernel: Call Trace:
Aug 26 14:26:01 home kernel: PKRU: 55555554
Aug 26 14:26:01 home kernel: CR2: 000000710158725a CR3: 000000041d7d9000 CR4: 0000000000f50ef0
Aug 26 14:26:01 home kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Aug 26 14:26:01 home kernel: FS:  0000000000000000(0000) GS:ffff8b5ba095b000(0000) knlGS:0000000000000000
Aug 26 14:26:01 home kernel: R13: 0000000000000038 R14: 0000000000000001 R15: ffff8b544fb00180
Aug 26 14:26:01 home kernel: R10: 0000000000000006 R11: ffff8b57579a5c80 R12: ffff8b5510006200
Aug 26 14:26:01 home kernel: RBP: ffffd22710e0fb88 R08: ffff8b544fb00010 R09: 0000000000000000
Aug 26 14:26:01 home kernel: RDX: 0000000000000296 RSI: 0000000000000296 RDI: ffff8b544fb00164
Aug 26 14:26:01 home kernel: RAX: 00000000ffffffea RBX: ffff8b54516b2000 RCX: 0000000000000000
Aug 26 14:26:01 home kernel: RSP: 0018:ffffd22710e0f930 EFLAGS: 00010282
Aug 26 14:26:01 home kernel: Code: ff 0f 0b e9 25 d6 ff ff 49 8b b4 24 50 01 00 00 48 8b bd 38 fe ff ff e8 10 0e 34 00 48 85 c0 0f 85 41 d3 ff ff e9 ee 9b 38 00 <0f> 0b e9 8b da ff ff 0f 0b e9 73 da ff ff 0f 0b e9 b2 e9 ff ff 48
Aug 26 14:26:01 home kernel: RIP: 0010:amdgpu_dm_atomic_commit_tail+0x33ee/0x3840 [amdgpu]
Aug 26 14:26:01 home kernel: Workqueue: async async_run_entry_fn
Aug 26 14:26:01 home kernel: Hardware name: To Be Filled By O.E.M. X570S PG Riptide/X570S PG Riptide, BIOS P5.50 10/13/2023
Aug 26 14:26:01 home kernel: CPU: 1 UID: 0 PID: 102262 Comm: kworker/u98:48 Not tainted 6.16.3-arch1-1 #1 PREEMPT(full)  1f0972552899c74baa9bd5c16d3531764dd9a1bb
Aug 26 14:26:01 home kernel:  k10temp rapl i2c_smbus mdio_bus mac_hid loop crypto_user nfnetlink ip_tables x_tables dm_crypt encrypted_keys trusted asn1_encoder tee dm_mod polyval_clmulni ghash_clmulni_intel sha512_ssse3 nvme sha1_>
Aug 26 14:26:01 home kernel: Modules linked in: vhost_net vhost vhost_iotlb tap nft_masq nft_reject_ipv4 act_csum cls_u32 sch_htb veth nf_conntrack_netlink xt_nat xt_tcpudp xt_conntrack xt_MASQUERADE bridge stp llc xt_set ip_set nf>
Aug 26 14:26:01 home kernel: WARNING: CPU: 1 PID: 102262 at drivers/gpu/drm/amd/amdgpu/../display/amdgpu_dm/amdgpu_dm.c:9085 amdgpu_dm_atomic_commit_tail+0x33ee/0x3840 [amdgpu]
Aug 26 14:26:01 home kernel: ------------[ cut here ]------------
Aug 26 14:26:01 home kernel: [drm] kiq ring mec 2 pipe 1 q 0
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: SMU is resumed successfully!
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: smc_dpm_info table revision(format.content): 4.5
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: use vbios provided pptable
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: SMU is resuming...
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: SECUREDISPLAY: optional securedisplay ta ucode is not available
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: RAP: optional rap ta ucode is not available
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: RAS: optional ras ta ucode is not available
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: reserve 0x900000 from 0x81fd000000 for PSP TMR
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: PSP is resuming...
Aug 26 14:26:01 home kernel: [drm] PCIE GART of 512M enabled (table at 0x00000081FEE00000).
Aug 26 14:26:01 home kernel: ata1: SATA link down (SStatus 0 SControl 300)
Aug 26 14:26:01 home kernel: ata2: SATA link down (SStatus 0 SControl 300)
Aug 26 14:26:01 home kernel: r8169 0000:04:00.0 enp4s0: Link is Down
Aug 26 14:26:01 home kernel: nvme nvme1: 24/0/0 default/read/poll queues
Aug 26 14:26:01 home kernel: nvme nvme0: Ignoring bogus Namespace Identifiers
Aug 26 14:26:01 home kernel: nvme nvme0: 8/0/0 default/read/poll queues
Aug 26 14:26:01 home kernel: nvme nvme0: D3 entry latency set to 10 seconds
Aug 26 14:26:01 home kernel: PM: noirq suspend of devices failed
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: PM: failed to suspend async noirq: error -22
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: PM: dpm_run_callback(): pci_pm_suspend_noirq returns -22
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: PM: pci_pm_suspend_noirq(): amdgpu_pmops_suspend_noirq [amdgpu] returns -22
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: GPU mode1 reset failed
Aug 26 14:26:01 home kernel: [drm] psp mode 1 reset failed!
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: psp reg (0x16061) wait timed out, mask: 8000ffff, read: ffffffff exp: 80000000
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: GPU psp mode1 reset
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: GPU mode1 reset
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: smc_dpm_info table revision(format.content): 4.5
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: use vbios provided pptable
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: SMU is resuming...
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: SECUREDISPLAY: optional securedisplay ta ucode is not available
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: RAP: optional rap ta ucode is not available
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: RAS: optional ras ta ucode is not available
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: reserve 0x900000 from 0x81fd000000 for PSP TMR
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: PSP is resuming...
Aug 26 14:26:01 home kernel: [drm] PCIE GART of 512M enabled (table at 0x00000081FEE00000).
Aug 26 14:26:01 home kernel: ata1: SATA link down (SStatus 0 SControl 300)
Aug 26 14:26:01 home kernel: ata2: SATA link down (SStatus 0 SControl 300)
Aug 26 14:26:01 home kernel: r8169 0000:04:00.0 enp4s0: Link is Down
Aug 26 14:26:01 home kernel: nvme nvme1: 24/0/0 default/read/poll queues
Aug 26 14:26:01 home kernel: nvme nvme0: Ignoring bogus Namespace Identifiers
Aug 26 14:26:01 home kernel: nvme nvme0: 8/0/0 default/read/poll queues
Aug 26 14:26:01 home kernel: nvme nvme0: D3 entry latency set to 10 seconds
Aug 26 14:26:01 home kernel: PM: noirq suspend of devices failed
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: PM: failed to suspend async noirq: error -22
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: PM: dpm_run_callback(): pci_pm_suspend_noirq returns -22
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: PM: pci_pm_suspend_noirq(): amdgpu_pmops_suspend_noirq [amdgpu] returns -22
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: GPU mode1 reset failed
Aug 26 14:26:01 home kernel: [drm] psp mode 1 reset failed!
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: psp reg (0x16061) wait timed out, mask: 8000ffff, read: ffffffff exp: 80000000
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: GPU psp mode1 reset
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: GPU mode1 reset
Aug 26 14:26:01 home kernel: amdgpu 0000:0b:00.0: amdgpu: MODE1 reset
Aug 26 14:26:01 home kernel: r8169 0000:04:00.0 enp4s0: Link is Down
Aug 26 14:26:01 home kernel: printk: Suspending console(s) (use no_console_suspend to debug)
Aug 26 14:26:01 home kernel: Freezing remaining freezable tasks completed (elapsed 0.000 seconds)
Aug 26 14:26:01 home kernel: Freezing remaining freezable tasks
Aug 26 14:26:01 home kernel: OOM killer disabled.
Aug 26 14:26:01 home kernel: Freezing user space processes completed (elapsed 0.001 seconds)
Aug 26 14:26:01 home kernel: Freezing user space processes
Aug 26 14:25:58 home kernel: Filesystems sync: 0.002 seconds
Aug 26 14:25:58 home kernel: PM: suspend entry (deep)
Aug 26 14:25:58 home systemd-sleep[102185]: Performing sleep operation 'suspend'...
Aug 26 14:25:58 home systemd[1]: user-1000.slice: Unit now frozen-by-parent.
Aug 26 14:25:58 home systemd-sleep[102185]: Successfully froze unit 'user.slice'.
Aug 26 14:25:58 home systemd[1]: user.slice: Unit now frozen.
Aug 26 14:25:58 home systemd[1]: session-1.scope: Unit now frozen-by-parent.
Aug 26 14:25:58 home systemd[1]: user@1000.service: Unit now frozen-by-parent.
Aug 26 14:25:58 home systemd[1]: Starting System Suspend then Hibernate...
Aug 26 14:25:58 home systemd[1]: Reached target Sleep.
Aug 26 14:25:58 home systemd-logind[1141]: The system will suspend and later hibernate now!

seth · 2025-08-26 13:30:05

From janp's journal

Aug 23 23:13:05 bilbo systemd[1]: Reached target Sleep.
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: MODE1 reset
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: GPU mode1 reset
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: GPU psp mode1 reset
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: psp reg (0x16061) wait timed out, mask: 8000ffff, read: ffffffff exp: 80000000
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: GPU mode1 reset failed
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: PM: pci_pm_suspend_noirq(): amdgpu_pmops_suspend_noirq [amdgpu] returns -22
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: PM: dpm_run_callback(): pci_pm_suspend_noirq returns -22
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: PM: failed to suspend async noirq: error -22
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: PSP is resuming...
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: reserve 0x900000 from 0x817d000000 for PSP TMR
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: RAS: optional ras ta ucode is not available
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: RAP: optional rap ta ucode is not available
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: SECUREDISPLAY: optional securedisplay ta ucode is not available
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: SMU is resuming...
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: use vbios provided pptable
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: smc_dpm_info table revision(format.content): 4.5
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: SMU is resumed successfully!
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring gfx_0.0.0 uses VM inv eng 0 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring comp_1.0.0 uses VM inv eng 1 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring comp_1.1.0 uses VM inv eng 4 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring comp_1.2.0 uses VM inv eng 5 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring comp_1.3.0 uses VM inv eng 6 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring comp_1.0.1 uses VM inv eng 7 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring comp_1.1.1 uses VM inv eng 8 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring comp_1.2.1 uses VM inv eng 9 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring comp_1.3.1 uses VM inv eng 10 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring kiq_0.2.1.0 uses VM inv eng 11 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring sdma0 uses VM inv eng 12 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring sdma1 uses VM inv eng 13 on hub 0
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring vcn_dec uses VM inv eng 0 on hub 8
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring vcn_enc0 uses VM inv eng 1 on hub 8
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring vcn_enc1 uses VM inv eng 4 on hub 8
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: ring jpeg_dec uses VM inv eng 5 on hub 8
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: PSP is resuming...
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: reserve 0x900000 from 0x817d000000 for PSP TMR
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: RAS: optional ras ta ucode is not available
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: RAP: optional rap ta ucode is not available
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: SECUREDISPLAY: optional securedisplay ta ucode is not available
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: SMU is resuming...
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: use vbios provided pptable
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: smc_dpm_info table revision(format.content): 4.5
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: SMU: I'm not done with your previous command: SMN_C2PMSG_66:0x0000003A SMN_C2PMSG_82:0x00000000
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: RunBtc failed!
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: Failed to setup smc hw!
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: resume of IP block <smu> failed -62
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: amdgpu_device_ip_resume failed (-62).
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: PM: dpm_run_callback(): pci_pm_resume returns -62
Aug 24 09:16:00 bilbo kernel: amdgpu 0000:2f:00.0: PM: failed to resume async: error -62
Aug 24 09:16:00 bilbo systemd[1]: Stopped target Sleep.
Aug 24 09:16:10 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: Dumping IP State

@fooblub has the same (please don't use -r) - is this also a 6.16.2 regression?

Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: MODE1 reset
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: GPU mode1 reset
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: GPU psp mode1 reset
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: psp reg (0x16061) wait timed out, mask: 8000ffff, read: ffffffff exp: 80000000
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: GPU mode1 reset failed
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: PM: pci_pm_suspend_noirq(): amdgpu_pmops_suspend_noirq [amdgpu] returns -22
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: PM: dpm_run_callback(): pci_pm_suspend_noirq returns -22
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: PM: failed to suspend async noirq: error -22
Aug 23 23:13:08 bilbo kernel: amdgpu 0000:2f:00.0: amdgpu: PSP is resuming...

https://cdn.kernel.org/pub/linux/kernel … Log-6.16.2
https://github.com/gregkh/linux/commit/ … 621b0b1c9f ??
Both of you are btw. suspending S3 ("deep") where eternalflameiv is limited to s2idle.

fooblub · 2025-08-27 06:22:10

@seth thanks for looking into this ❤️ sorry for the -r, won't use it for my future reports. I don't know if this is a 6.16.2 regression, currently don't have much time for testing.

For the last week or so I'm having a really bad time with amdgpu: besides this suspend issue I can also reliably introduce a crash using webgl in chromium and firefox

Last edited by fooblub (2025-08-27 06:22:20)

Camilo · 2025-08-28 09:17:55

Hi I may be having a similar issue. GPU is

Device: AMD Radeon RX 5700 XT (radeonsi, navi10, LLVM 20.1.8, DRM 3.64, 6.16.3-arch1-1) (0x731f)

Yesterday morning I installed updates (just showing ones that might be relevant):

[2025-08-27T08:01:34+0100] [ALPM] upgraded linux (6.16.2.arch1-1 -> 6.16.3.arch1-1)
[2025-08-27T08:01:36+0100] [ALPM] upgraded linux-headers (6.16.2.arch1-1 -> 6.16.3.arch1-1)
[2025-08-27T08:01:38+0100] [ALPM] upgraded vulkan-radeon (1:25.2.1-1 -> 1:25.2.1-2)
[2025-08-27T08:01:32+0100] [ALPM] upgraded mesa (1:25.2.1-1 -> 1:25.2.1-2)
...

Since then I noticed sleep not working (monitor does not receive any signal, but pc fans and lights still on, tried blindly typing in reboot commands but nothing happened). I trigger sleep by using `/usr/bin/systemctl suspend`
journalctl output looks like this:

Aug 27 09:13:51 fusion3 kernel: amdgpu 0000:2d:00.0: amdgpu: MODE1 reset
Aug 27 09:13:51 fusion3 kernel: amdgpu 0000:2d:00.0: amdgpu: GPU mode1 reset
Aug 27 09:13:51 fusion3 kernel: amdgpu 0000:2d:00.0: amdgpu: GPU psp mode1 reset
Aug 27 09:13:51 fusion3 kernel: amdgpu 0000:2d:00.0: amdgpu: psp reg (0x16061) wait timed out, mask: 8000ffff, read: ffffffff exp: 80000000
Aug 27 09:13:51 fusion3 kernel: [drm] psp mode 1 reset failed!
Aug 27 09:13:51 fusion3 kernel: amdgpu 0000:2d:00.0: amdgpu: GPU mode1 reset failed
Aug 27 09:13:51 fusion3 kernel: amdgpu 0000:2d:00.0: PM: pci_pm_suspend_noirq(): amdgpu_pmops_suspend_noirq [amdgpu] returns -22
Aug 27 09:13:51 fusion3 kernel: amdgpu 0000:2d:00.0: PM: dpm_run_callback(): pci_pm_suspend_noirq returns -22
Aug 27 09:13:51 fusion3 kernel: amdgpu 0000:2d:00.0: PM: failed to suspend async noirq: error -22
Aug 27 09:13:51 fusion3 kernel: PM: noirq suspend of devices failed
...
Aug 27 09:52:57 fusion3 kernel: amdgpu 0000:2d:00.0: amdgpu: SMU: I'm not done with your previous command: SMN_C2PMSG_66:0x0000003A SMN_C2PMSG_82:0x00000000
Aug 27 09:52:57 fusion3 kernel: amdgpu 0000:2d:00.0: amdgpu: RunBtc failed!
Aug 27 09:52:57 fusion3 kernel: amdgpu 0000:2d:00.0: amdgpu: Failed to setup smc hw!
Aug 27 09:52:57 fusion3 kernel: amdgpu 0000:2d:00.0: amdgpu: resume of IP block <smu> failed -62
Aug 27 09:52:57 fusion3 kernel: amdgpu 0000:2d:00.0: amdgpu: amdgpu_device_ip_resume failed (-62).
Aug 27 09:52:57 fusion3 kernel: amdgpu 0000:2d:00.0: PM: dpm_run_callback(): pci_pm_resume returns -62
Aug 27 09:52:57 fusion3 kernel: amdgpu 0000:2d:00.0: PM: failed to resume async: error -62

I think that gap in time stamps was when I was away from my pc, not realising it had not slept properly, then maybe coming back and pressing the power button?

Last edited by Camilo (2025-08-28 09:24:11)

ledbettj · 2025-08-28 14:53:51

I'm also seeing this; https://gitlab.freedesktop.org/drm/amd/-/issues/4531 seems to have identified the issue.

emgemg · 2025-08-30 23:03:39

I've had the same issue for about a week now, also with AMD Radeon RX 5700 XT. I'm seeing the same errors as in issue 4531, and for me downgrading the kernel to 6.16.1 fixes it for now.

ledbettj · 2025-09-20 14:27:32

The offending commit was reverted in 6.16.8 (and an actual fix was merged for 6.17), so it should be safe to upgrade now I believe.

fooblub · 2025-09-20 17:12:01

Thanks for the info. 6.16.8 fixed the issue for me.

Fisiu · 2025-10-07 17:00:52

I'm facing this issue on 6.17.1-2 with RX6700 XT:

Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: amdgpu: GPU mode1 reset failed
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: PM: pci_pm_suspend_noirq(): amdgpu_pmops_suspend_noirq [amdgpu] returns -62
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: PM: dpm_run_callback(): pci_pm_suspend_noirq.llvm.16234474454299812026 returns -62
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: PM: failed to suspend async noirq: error -62
Oct 07 18:40:46 pendrak kernel: PM: noirq suspend of devices failed
<--other not related messages -->
Oct 07 18:40:46 pendrak kernel: [drm] PCIE GART of 512M enabled (table at 0x0000008000F00000).
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: amdgpu: PSP is resuming...
Oct 07 18:40:46 pendrak kernel: [drm:psp_v11_0_memory_training.llvm.12783815074455855840 [amdgpu]] *ERROR* send training msg failed.
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: amdgpu: Failed to process memory training!
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: amdgpu: resume of IP block <psp> failed -62
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: amdgpu: amdgpu_device_ip_resume failed (-62).
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: PM: dpm_run_callback(): pci_pm_resume.llvm.16234474454299812026 returns -62
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: PM: failed to resume async: error -62

However, with lts kernel suspend/resume works as expected.

seth · 2025-10-07 19:26:43

6.16.8 works as well but 6.17.0 breaks?

HarryGeez · 2025-10-07 20:35:18

Fisiu wrote:

I'm facing this issue on 6.17.1-2 with RX6700 XT:

Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: amdgpu: GPU mode1 reset failed
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: PM: pci_pm_suspend_noirq(): amdgpu_pmops_suspend_noirq [amdgpu] returns -62
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: PM: dpm_run_callback(): pci_pm_suspend_noirq.llvm.16234474454299812026 returns -62
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: PM: failed to suspend async noirq: error -62
Oct 07 18:40:46 pendrak kernel: PM: noirq suspend of devices failed
<--other not related messages -->
Oct 07 18:40:46 pendrak kernel: [drm] PCIE GART of 512M enabled (table at 0x0000008000F00000).
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: amdgpu: PSP is resuming...
Oct 07 18:40:46 pendrak kernel: [drm:psp_v11_0_memory_training.llvm.12783815074455855840 [amdgpu]] *ERROR* send training msg failed.
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: amdgpu: Failed to process memory training!
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: amdgpu: resume of IP block <psp> failed -62
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: amdgpu: amdgpu_device_ip_resume failed (-62).
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: PM: dpm_run_callback(): pci_pm_resume.llvm.16234474454299812026 returns -62
Oct 07 18:40:46 pendrak kernel: amdgpu 0000:03:00.0: PM: failed to resume async: error -62

However, with lts kernel suspend/resume works as expected.

I'm experiencing the exact same issue! Same card same error messages. Ironically this release is supposed to improve AMD GPU hibernation. Downgraded to 6.18 for now

Last edited by HarryGeez (2025-10-07 20:35:45)

Arch Linux

#26 2025-08-24 18:48:25

Re: AMD GPU hangs on resuming from suspend

#27 2025-08-24 19:11:17

Re: AMD GPU hangs on resuming from suspend

#28 2025-08-25 06:34:48

Re: AMD GPU hangs on resuming from suspend

#29 2025-08-25 07:44:38

Re: AMD GPU hangs on resuming from suspend

#30 2025-08-25 10:46:58

Re: AMD GPU hangs on resuming from suspend

#31 2025-08-25 12:18:19

Re: AMD GPU hangs on resuming from suspend

#32 2025-08-25 13:12:02

Re: AMD GPU hangs on resuming from suspend

#33 2025-08-25 17:55:32

Re: AMD GPU hangs on resuming from suspend

#34 2025-08-25 19:42:30

Re: AMD GPU hangs on resuming from suspend

#35 2025-08-26 13:04:49

Re: AMD GPU hangs on resuming from suspend

#36 2025-08-26 13:30:05

Re: AMD GPU hangs on resuming from suspend

#37 2025-08-27 06:22:10

Re: AMD GPU hangs on resuming from suspend

#38 2025-08-28 09:17:55

Re: AMD GPU hangs on resuming from suspend

#39 2025-08-28 14:53:51

Re: AMD GPU hangs on resuming from suspend

#40 2025-08-30 23:03:39

Re: AMD GPU hangs on resuming from suspend

#41 2025-09-20 14:27:32

Re: AMD GPU hangs on resuming from suspend

#42 2025-09-20 17:12:01

Re: AMD GPU hangs on resuming from suspend

#43 2025-10-07 17:00:52

Re: AMD GPU hangs on resuming from suspend

#44 2025-10-07 19:26:43

Re: AMD GPU hangs on resuming from suspend

#45 2025-10-07 20:35:18

Re: AMD GPU hangs on resuming from suspend

Board footer