From: "Christian König" <christian.koenig@amd.com>
To: "Dr. David Alan Gilbert" <dave@treblig.org>
Cc: alexander.deucher@amd.com, amd-gfx@lists.freedesktop.org
Subject: Re: oops/null pointer in 0010:dma_fence_is_signaled+0x12/0x60 [amdgpu]
Date: Mon, 16 Mar 2026 13:37:34 +0100 [thread overview]
Message-ID: <cf41d26e-acba-4096-abc0-277d3434c565@amd.com> (raw)
In-Reply-To: <abf48cuwzRBczbeA@gallifrey>
On 3/16/26 13:34, Dr. David Alan Gilbert wrote:
> * Christian König (christian.koenig@amd.com) wrote:
>> Hi,
>>
>> On 3/16/26 01:57, Dr. David Alan Gilbert wrote:
>>> Hi,
>>> I'm not sure if this is repeatable, but I landed with a null
>>> pointer during a GPU reset, so thought I should probably
>>> report it:
>>> 6.19.7-300.fc44.x86_64
>>> Mar 16 00:24:39 dalek kernel: BUG: kernel NULL pointer dereference, address: 0000000000000018
>>> Mar 16 00:24:39 dalek kernel: #PF: supervisor read access in kernel mode
>>> ....
>>> Mar 16 00:24:39 dalek kernel: Workqueue: amdgpu-reset-dev drm_sched_job_timedout [gpu_sched]
>>> Mar 16 00:24:39 dalek kernel: RIP: 0010:dma_fence_is_signaled+0x12/0x60 [amdgpu]
>>>
>>> see full oops below;
>>>
>>> 09:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Lexa PRO [Radeon 540/540X/550/550X / RX 540X/550/550X] (rev c7)
>>> AMD Ryzen 9 3950X
>>>
>>> I suspect the timeout was real, and caused by a runaway llama, I forgot the
>>> flag to stop it trying to use the GPU for image encoding; but the null page
>>> seems unfortunate. Impressively the audio still kept playing via it:
>>>
>>> I have the devcoredump copied if it's of interest
>>> (Note for self: ~/amd.core-2026-03-16)
>>
>>
>> Can you open up a bug report for that
>
> Sure; on bugzilla.kernel.org ?
No, https://gitlab.freedesktop.org/
Sorry I though that this would be obvious.
Regards,
Christian.
>> and please provide what line of code amdgpu_device_gpu_recover.cold+0x244 decodes to.
>
> Yeh, let me see what I can do with the Fedora symbol packages.
>
> Dave
>
>>
>> Thanks,
>> Christian.
>>
>>>
>>> Dave
>>>
>>>
>>> Mar 16 00:24:28 dalek kernel: amdgpu 0000:09:00.0: amdgpu: Dumping IP State
>>> Mar 16 00:24:28 dalek kernel: amdgpu 0000:09:00.0: amdgpu: Dumping IP State Completed
>>> Mar 16 00:24:28 dalek kernel: amdgpu 0000:09:00.0: amdgpu: [drm] AMDGPU device coredump file has been created
>>> Mar 16 00:24:28 dalek kernel: amdgpu 0000:09:00.0: amdgpu: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
>>> Mar 16 00:24:28 dalek kernel: amdgpu 0000:09:00.0: amdgpu: ring gfx timeout, signaled seq=1705630, emitted seq=1705633
>>> Mar 16 00:24:28 dalek kernel: amdgpu 0000:09:00.0: amdgpu: Process llama-mtmd-cli pid 299886 thread llama-mtmd-cli pid 299886
>>> Mar 16 00:24:28 dalek kernel: amdgpu 0000:09:00.0: amdgpu: GPU reset begin!. Source: 1
>>> Mar 16 00:24:32 dalek kernel: amdgpu 0000:09:00.0: amdgpu: failed to suspend display audio
>>> Mar 16 00:24:32 dalek kernel: amdgpu 0000:09:00.0: amdgpu: Guilty job already signaled, skipping HW reset
>>> Mar 16 00:24:32 dalek kernel: amdgpu 0000:09:00.0: amdgpu: GPU reset(1) succeeded!
>>> Mar 16 00:24:32 dalek kernel: amdgpu 0000:09:00.0: [drm] device wedged, but recovered through reset
>>> Mar 16 00:24:34 dalek lightdm[40555]: ATTENTION: default value of option mesa_glthread overridden by environment.
>>> Mar 16 00:24:34 dalek kernel: amdgpu 0000:09:00.0: amdgpu: Dumping IP State
>>> Mar 16 00:24:34 dalek kernel: amdgpu 0000:09:00.0: amdgpu: Dumping IP State Completed
>>> Mar 16 00:24:34 dalek kernel: amdgpu 0000:09:00.0: amdgpu: [drm] AMDGPU device coredump file has been created
>>> Mar 16 00:24:34 dalek kernel: amdgpu 0000:09:00.0: amdgpu: [drm] Check your /sys/class/drm/card1/device/devcoredump/data
>>> Mar 16 00:24:34 dalek kernel: amdgpu 0000:09:00.0: amdgpu: ring gfx timeout, signaled seq=1705632, emitted seq=1705637
>>> Mar 16 00:24:34 dalek kernel: amdgpu 0000:09:00.0: amdgpu: GPU reset begin!. Source: 1
>>> Mar 16 00:24:38 dalek kernel: amdgpu 0000:09:00.0: amdgpu: failed to suspend display audio
>>> Mar 16 00:24:39 dalek kernel: BUG: kernel NULL pointer dereference, address: 0000000000000018
>>> Mar 16 00:24:39 dalek kernel: #PF: supervisor read access in kernel mode
>>> Mar 16 00:24:39 dalek kernel: #PF: error_code(0x0000) - not-present page
>>> Mar 16 00:24:39 dalek kernel: PGD 849708067 P4D 849708067 PUD 15d277067 PMD 0
>>> Mar 16 00:24:39 dalek kernel: Oops: Oops: 0000 [#1] SMP NOPTI
>>> Mar 16 00:24:39 dalek kernel: CPU: 7 UID: 0 PID: 298062 Comm: kworker/u128:2 Not tainted 6.19.7-300.fc44.x86_64 #1 PREEMPT(lazy)
>>> Mar 16 00:24:39 dalek kernel: Hardware name: To Be Filled By O.E.M. To Be Filled By O.E.M./X570 Pro4, BIOS P3.10 07/13/2020
>>> Mar 16 00:24:39 dalek kernel: Workqueue: amdgpu-reset-dev drm_sched_job_timedout [gpu_sched]
>>> Mar 16 00:24:39 dalek kernel: RIP: 0010:dma_fence_is_signaled+0x12/0x60 [amdgpu]
>>> Mar 16 00:24:39 dalek kernel: Code: 0f 1f 84 00 00 00 00 00 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 48 8b 47 30 48 d1 e8 89 c2 83 e2 01 75 2c 48 8b>
>>> Mar 16 00:24:39 dalek kernel: RSP: 0018:ffffcf349c553d20 EFLAGS: 00010246
>>> Mar 16 00:24:39 dalek kernel: RAX: 0000000000000000 RBX: ffffcf349c553da0 RCX: 0000000000000000
>>> Mar 16 00:24:39 dalek kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8bce4ecd4380
>>> Mar 16 00:24:39 dalek kernel: RBP: 0000000000000000 R08: 0000000010000020 R09: ffff8bc9c0400b68
>>> Mar 16 00:24:39 dalek kernel: R10: 0000000000000080 R11: ffffffffa16760a0 R12: ffff8bc9e9100000
>>> Mar 16 00:24:39 dalek kernel: R13: ffff8bca7ccab200 R14: 0000000000000000 R15: 0000000000000000
>>> Mar 16 00:24:39 dalek kernel: FS: 0000000000000000(0000) GS:ffff8bd90c066000(0000) knlGS:0000000000000000
>>> Mar 16 00:24:39 dalek kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>> Mar 16 00:24:39 dalek kernel: CR2: 0000000000000018 CR3: 00000001c591f000 CR4: 0000000000350ef0
>>> Mar 16 00:24:39 dalek kernel: Call Trace:
>>> Mar 16 00:24:39 dalek kernel: <TASK>
>>> Mar 16 00:24:39 dalek kernel: amdgpu_device_gpu_recover.cold+0x244/0x2ec [amdgpu]
>>> Mar 16 00:24:39 dalek kernel: amdgpu_job_timedout.cold+0x218/0x258 [amdgpu]
>>> Mar 16 00:24:39 dalek kernel: ? srso_return_thunk+0x5/0x5f
>>> Mar 16 00:24:39 dalek kernel: drm_sched_job_timedout+0x8b/0x190 [gpu_sched]
>>> Mar 16 00:24:39 dalek kernel: ? srso_return_thunk+0x5/0x5f
>>> Mar 16 00:24:39 dalek kernel: process_one_work+0x190/0x350
>>> Mar 16 00:24:39 dalek kernel: worker_thread+0x18d/0x2f0
>>> Mar 16 00:24:39 dalek kernel: ? __pfx_worker_thread+0x10/0x10
>>> Mar 16 00:24:39 dalek kernel: kthread+0xfa/0x240
>>> Mar 16 00:24:39 dalek kernel: ? finish_task_switch.isra.0+0x82/0x2a0
>>> Mar 16 00:24:39 dalek kernel: ? __pfx_kthread+0x10/0x10
>>> Mar 16 00:24:39 dalek kernel: ? __pfx_kthread+0x10/0x10
>>> Mar 16 00:24:39 dalek kernel: ret_from_fork+0x130/0x1a0
>>> Mar 16 00:24:39 dalek kernel: ? __pfx_kthread+0x10/0x10
>>> Mar 16 00:24:39 dalek kernel: ret_from_fork_asm+0x1a/0x30
>>> Mar 16 00:24:39 dalek kernel: </TASK>
>>> Mar 16 00:24:39 dalek kernel: Modules linked in: dm_crypt snd_seq_dummy snd_hrtimer nft_masq nft_reject_ipv4 act_csum cls_u32 sch_htb nf_nat_tftp nf_conntr>
>>> Mar 16 00:24:39 dalek kernel: drm_panel_backlight_quirks gpu_sched drm_suballoc_helper video drm_buddy drm_display_helper nvme nvme_core cec ghash_clmulni>
>>> Mar 16 00:24:39 dalek kernel: CR2: 0000000000000018
>>> Mar 16 00:24:39 dalek kernel: ---[ end trace 0000000000000000 ]---
>>> Mar 16 00:24:39 dalek kernel: RIP: 0010:dma_fence_is_signaled+0x12/0x60 [amdgpu]
>>> Mar 16 00:24:39 dalek kernel: Code: 0f 1f 84 00 00 00 00 00 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 90 48 8b 47 30 48 d1 e8 89 c2 83 e2 01 75 2c 48 8b>
>>> Mar 16 00:24:39 dalek kernel: RSP: 0018:ffffcf349c553d20 EFLAGS: 00010246
>>> Mar 16 00:24:39 dalek kernel: RAX: 0000000000000000 RBX: ffffcf349c553da0 RCX: 0000000000000000
>>> Mar 16 00:24:39 dalek kernel: RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8bce4ecd4380
>>> Mar 16 00:24:39 dalek kernel: RBP: 0000000000000000 R08: 0000000010000020 R09: ffff8bc9c0400b68
>>> Mar 16 00:24:39 dalek kernel: R10: 0000000000000080 R11: ffffffffa16760a0 R12: ffff8bc9e9100000
>>> Mar 16 00:24:39 dalek kernel: R13: ffff8bca7ccab200 R14: 0000000000000000 R15: 0000000000000000
>>> Mar 16 00:24:39 dalek kernel: FS: 0000000000000000(0000) GS:ffff8bd90c066000(0000) knlGS:0000000000000000
>>> Mar 16 00:24:39 dalek kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
>>> Mar 16 00:24:39 dalek kernel: CR2: 0000000000000018 CR3: 00000001c591f000 CR4: 0000000000350ef0
>>> Mar 16 00:24:39 dalek kernel: note: kworker/u128:2[298062] exited with irqs disabled
>>>
>>> --
>>> -----Open up your eyes, open up your mind, open up your code -------
>>> / Dr. David Alan Gilbert | Running GNU/Linux | Happy \
>>> \ dave @ treblig.org | | In Hex /
>>> \ _________________________|_____ http://www.treblig.org |_______/
>>
next prev parent reply other threads:[~2026-03-16 12:37 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-16 0:57 oops/null pointer in 0010:dma_fence_is_signaled+0x12/0x60 [amdgpu] Dr. David Alan Gilbert
2026-03-16 8:22 ` Christian König
2026-03-16 12:34 ` Dr. David Alan Gilbert
2026-03-16 12:37 ` Christian König [this message]
2026-03-16 14:18 ` Dr. David Alan Gilbert
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=cf41d26e-acba-4096-abc0-277d3434c565@amd.com \
--to=christian.koenig@amd.com \
--cc=alexander.deucher@amd.com \
--cc=amd-gfx@lists.freedesktop.org \
--cc=dave@treblig.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox