[TRACKING] Freezing Arch Linux AMD

Excellent reports everyone. Appreciate it. We have an internal ticket that is being addressed regarding the i2c errors. We have our own logs as well.

4 Likes

I’ve also been able to see this same log output in dmesg while also experiencing freezing and stuttering behavior (as I imagine existed with BIOS 3.02.) I’m running Debian Testing, Gnome, BIOS 3.03 as delivered with the laptop.

Linux framework 6.5.0-4-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.5.10-1 (2023-11-03) x86_64 GNU/Linux
2 Likes

I found some related info at [drm:dc_dmub_srv_wait_idle [amdgpu]] *ERROR* Error waiting for DMUB idle: status=3 log spam on 5.16.3 (#1887) · Issues · drm / amd · GitLab

1 Like

Hm I encounter the same errors on debian stable with backported 6.5.0-3 and latest firmware form kernel.org, with same freezing and stuttering behavior. As soon as the screen is of the machine freezes. Only a hard shutdown helps.

Still experiencing the freezing from the touchpad.

are you by any chance running 4800MTs ram?

Appreciate everyone’s reports on this.

For those just visiting:

Thanks Matt! Are you tracking two separate issues, or just one? I am experiencing the issue and symptoms at the top of this thread, but nothing related to the touchpad or i2c, as far as I can tell.

We have an internal ticket for i2c errors as it affects supported distros.

The DMUB error is tracked here at the link you provided.

1 Like

Started getting this on tumbleweed, kernel 6.6.6.

No external monitor attached, mostly happens while on YouTube or intellij idea. Hard reset only way out, very disruptive for work.

Tried Fedora 39 on the same kernel version (fedora’s take on it), and also got a freeze but here it was recoverable, system became responsive again after a few seconds.

Didn’t think of copy pasting my dmesg, but I got a bunch of amdgpu errors that resemble those mentioned here, with something about queueing

Are you seeing the issue on 6.6.7 or newer?

Haven’t tried on anything newer.

I’m jumping back and forth between tumbleweed and Fedora, but since they’re both on the same kernel now, I’m back to using TW, since both distros will crash on my eventually.

This morning I got this on fedora, user session totally borked, both screens turned white (but mouse cursor was changing from pointer to text according to where I moved it, so the DE itself was running fine, just with a coat of white paint on top.

I Ctrl Alt F7 into a cli session and journalctl -k showed this

Hope it gets fixed soon, right now I feel like a beta tester although this HW has been announced several months ago.

1 Like

Hope it gets fixed soon, right now I feel like a beta tester although this HW has been announced several months ago.

That’s how Linux is on new hardware, takes time for devs to get some stuff working.

Are you on the latest bios from from framework? v03.03
# dmidecode -s bios-version

1 Like

:frowning:

Yes, I’m on 03.03

Are you noticing any other failures in your journal or dmesg relating to amdgpu?

any progress made on the i2c errors?

I had yesterday the same issue, after a reboot all is works again but dmesg shows this:
i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration

The machine is working perfectly, but the following messages appeared on journalctl -g i2c:

Jan 13 14:38:23 mls.frame kernel: i2c_hid_acpi i2c-FRMW0004:00: failed to change power setting.
Jan 13 14:38:23 mls.frame kernel: i2c_hid_acpi i2c-FRMW0004:00: PM: dpm_run_callback(): acpi_subsys_resume+0x0/0x80 returns -121
Jan 13 14:38:23 mls.frame kernel: i2c_hid_acpi i2c-FRMW0004:00: PM: failed to resume async: error -121
Jan 15 10:33:36 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 15 11:22:57 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 12:05:05 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 13:00:20 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 13:12:56 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 13:15:47 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 13:18:37 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 13:26:46 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 13:51:20 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 15 14:30:11 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 15 15:07:15 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 15 15:07:37 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 15 17:14:19 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 17:42:25 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 17:55:55 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 15 19:00:26 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 19:18:04 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 19:47:19 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 15 19:47:22 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 20:01:58 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 20:30:16 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 21:19:40 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 21:25:16 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 22:16:14 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 22:18:14 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 22:28:22 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 15 23:07:40 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 15 23:39:58 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 15 23:55:12 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration
Jan 16 00:00:07 mls.frame kernel: i2c_hid_acpi i2c-FRMW0005:00: i2c_hid_get_input: incomplete report (7/65535)
Jan 16 16:14:19 mls.frame kernel: i2c_designware AMDI0010:00: i2c_dw_handle_tx_abort: lost arbitration

Just experienced the GPU crashing again. I’m on the kernel 6.7.1

Jan 26 09:58:15 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:15 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:15 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:16 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:16 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:16 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:16 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:17 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:17 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:17 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:17 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:18 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:18 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:18 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:19 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:19 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:19 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
Jan 26 09:58:19 smigs-space kernel: amdgpu 0000:c1:00.0: [drm] *ERROR* Error queueing DMUB command: status=2
2 Likes

Pretty certain this can be resolved with these firmware blobs. I’ve had them installed about a week now and haven’t had this specific crash happen anymore. I haven’t had any other crashes, so I can’t speak to those.

Try downloading the two files from this amdgpu commit and placing them in /lib/firmware/amdgpu/ and then running the following (probably good practice to backup the existing files before replacing them):

# Keep in mind that '-k all' will operate on all installed kernals. Skip this flag if you only want to affect the latest kernel
sudo update-initramfs -c -k all

Source: AMDGPU crash Error queuing DMUB command: status=2, Error waiting for DMUB idle: status=3 (#2862) · Issues · drm / amd · GitLab

1 Like