[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: How can I test the graphics card?

It seems the message, I am replying to, was rejected by the mailing list server, likely due to an attachment with a screenshot. That is why all quoted lines are preserved. 

On 06/01/2025 00:53, Serkan Kurt wrote:

Thank you Max. I managed to install the "nv84" firmware by adapting
the instructions from
"https://nouveau.freedesktop.org/VideoAcceleration.html";. The
graphical system has accelerated significantly.


Nice.


Previous errors do not
appear in dmesg. But when I run the "Blender" software, the text still
does not appear and the dmesg output is as follows;
Instead of dmesg I recommend "journalctl" as root, perhaps with "-b" (and maybe more) option. Log entries from various applications may be helpful. 


[   58.669512] nouveau 0000:01:00.0: gr: DATA_ERROR 00000004 [INVALID_VALUE]
[   58.669525] nouveau 0000:01:00.0: gr: 00100000 [] ch 4 [001f729000
Xwayland[2304]] subc 4 class 502d mthd 0838 data 000121b0


Notice the warning concerning Blender requirements in
Sven Joachim. Tue, 31 Dec 2024 20:09:09 +0100.
<https://lists.debian.org/msgid-search/87ikqzbrfu.fsf@turtle.gmx.de>
Lookup for these messages in the nouveau bug tracker and discussion groups. You may try kernel and graphics stack packages from the Debian backports repository. If issues have not fixed yet, alternatives may be: - Buy another graphics card that is better supported by Linux drivers. (Sorry, I can not suggest anything specific.) - If the developers do not mind to improve support of your device and your a ready to compile custom kernels, mesa, and some other packages then you may file a bug. 


[   58.669652] nouveau 0000:01:00.0: gr: TRAP_MP_EXEC - TP 0 MP 0:
00000010 [INVALID_OPCODE] at 06e6c8 warp 1, opcode feef4977 fddf3f37
[   58.669664] nouveau 0000:01:00.0: gr: TRAP_MP_EXEC - TP 0 MP 1:
00000010 [INVALID_OPCODE] at 06e6c8 warp 0, opcode feef4977 fddf3f37
[   58.669670] nouveau 0000:01:00.0: gr: 00200000 [] ch 4 [001f729000
Xwayland[2304]] subc 3 class 8297 mthd 0d74 data 00000000

When I restarted Blender;

[  307.480197] nouveau 0000:01:00.0: fifo: CACHE_ERROR - ch 5
[systemd-logind[502]] subc 0 mthd 0060 data beef0201
[  348.340249] perf: interrupt took too long (3960 > 3956), lowering
kernel.perf_event_max_sample_rate to 50500
[  723.319669] perf: interrupt took too long (4960 > 4950), lowering
kernel.perf_event_max_sample_rate to 40250
[  891.057444] nouveau 0000:01:00.0: gr: DATA_ERROR 00000004 [INVALID_VALUE]
[  891.057459] nouveau 0000:01:00.0: gr: 00100000 [] ch 6 [001f42f000
Xwayland[2304]] subc 4 class 502d mthd 0838 data 000121b0
[  911.824655] nouveau 0000:01:00.0: fifo: CACHE_ERROR - ch 5
[systemd-logind[502]] subc 0 mthd 0060 data beef0201
[ 1020.563973] nouveau 0000:01:00.0: fifo: CACHE_ERROR - ch 5
[systemd-logind[502]] subc 0 mthd 0060 data beef0201
[ 1020.568736] nouveau 0000:01:00.0: fifo: CACHE_ERROR - ch 5
[systemd-logind[502]] subc 0 mthd 0060 data beef0201
[ 1020.798138] nouveau 0000:01:00.0: fifo: CACHE_ERROR - ch 5
[systemd-logind[502]] subc 0 mthd 0060 data beef0201

Note: Before installing the firmware, I installed the nvidia drivers
from sid. But the graphical interface did not open and I had to
reinstall it.

Do you have any other advice?

Max Nikulin <manikulin@gmail.com>, 2 Oca 2025 Per, 05:44 tarihinde şunu yazdı:


On 31/12/2024 18:07, Serkan Kurt wrote:

nouveau 0000:01:00.0: firmware: failed to load nouveau/nv84_xuc00f (-2)
nouveau 0000:01:00.0: firmware: failed to load nouveau/nv84_xuc00f (-2)
nouveau 0000:01:00.0: Direct firmware load for nouveau/nv84_xuc00f
failed with error -2


Have you trying to find more info related to firmware files mentioned in
error messages? They may be necessary, however in some cases drivers
tries to load firmware that is available through public channels. In
some cases firmware can not be included in .deb packages due to
licensing issues, but when desperately needed, can be extracted from
proprietary drivers downloaded from vendor sites.

My experience is that freezes due to graphics card drivers might happen
for any vendor. Specifically to nvidia, several years ago I switched
back and forth between nouveau and proprietary drivers, patching the
latter for up to date kernels, due to some artefact in both cases.
Reply to: