XPost: linux.kernel

[To: +debian-kernel@lists.debian.org]

Dear Jeff, dear Debian Linux kernel team,


Am 14.11.23 um 19:19 schrieb Jeff Johnson:

On 11/14/2023 9:32 AM, Jeff Johnson wrote:

On 11/12/2023 4:13 AM, Paul Menzel wrote:

On the Dell XPS 13 9360 with Debian sid/unstable, Linux 6.5.10 logs the
error below:

$ sudo dmesg | grep -e "DMI:" -e "Linux version" -e microcode
[ 0.000000] microcode: updated early: 0xf0 -> 0xf4, date = 2023-02-22
[ 0.000000] Linux version 6.5.0-4-amd64 (debian-kernel@lists.debian.org) (gcc-13 (Debian 13.2.0-6) 13.2.0, GNU ld (GNU Binutils for Debian) 2.41) #1 SMP PREEMPT_DYNAMIC Debian 6.5.10-1 (2023-11-03)
[ 0.000000] DMI: Dell Inc. XPS 13 9360/0596KF, BIOS 2.21.0 06/02/2022
[ 0.580689] microcode: Microcode Update Driver: v2.2.

$ sudo dmesg --level alert,crit,err
[ 0.053566] x86/cpu: SGX disabled by BIOS.
[ 1.942183] psmouse serio1: synaptics: Unable to query device: -5 >>> [ 14.401112] ath10k_pci 0000:3a:00.0: firmware: failed to load ath10k/pre-cal-pci-0000:3a:00.0.bin (-2)
[ 14.401143] firmware_class: See https://wiki.debian.org/Firmware for information about missing firmware
[ 14.401185] ath10k_pci 0000:3a:00.0: firmware: failed to load ath10k/pre-cal-pci-0000:3a:00.0.bin (-2)
[ 14.401233] ath10k_pci 0000:3a:00.0: firmware: failed to load ath10k/cal-pci-0000:3a:00.0.bin (-2)
[ 14.401273] ath10k_pci 0000:3a:00.0: firmware: failed to load ath10k/cal-pci-0000:3a:00.0.bin (-2)

As it’s logged with error level, I’d like to address the ath10k_pci
errors, but I am unable to find the firmware in the Linux firmware
archive [1].

What can I do about this?

[1]: https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/

Adding the ath10k list so this isn't lost in the noise

The calibration files are optional and you should not see any warnings
if they are not present.

Note ath10k explicitly calls firmware_request_nowarn() when it loads the files since it doesn't want any warnings from the firmware loader if the files aren't present. <https://elixir.bootlin.com/linux/v6.5.10/source/drivers/net/wireless/ath/ath10k/core.c#L929>

And ath10k itself explicitly doesn't warn: <https://elixir.bootlin.com/linux/v6.5.10/source/drivers/net/wireless/ath/ath10k/core.c#L1259>

So I'm confused where these warnings originate.

Based upon the message:
[ 14.401143] firmware_class: See https://wiki.debian.org/Firmware
for information about missing firmware

it seems you are not running a stock kernel. So perhaps Debian has
modified the firmware loading such that it ignores the FW_OPT_NO_WARN
flag and warns even when told not to do so? This does not appear to be
an upstream kernel issue.

Thank you very much for the analysis. It seems to be indeed a Debian
specific patch [1].

Dear Debian Linux kernel team, is my observation about the error log a
result of the patch and intended?


Kind regards,

Paul


[1]: debian/patches/debian/firmware_class-refer-to-debian-wiki-firmware-page.patch

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

XPost: linux.kernel

FTR: I do NOT speak on behalf of the Debian kernel team.

On Tuesday, 14 November 2023 20:58:58 CET Paul Menzel wrote:

Based upon the message:
[ 14.401143] firmware_class: See https://wiki.debian.org/Firmware
for information about missing firmware

it seems you are not running a stock kernel. So perhaps Debian has
modified the firmware loading such that it ignores the FW_OPT_NO_WARN
flag and warns even when told not to do so? This does not appear to be
an upstream kernel issue.

Thank you very much for the analysis. It seems to be indeed a Debian
specific patch [1].

Dear Debian Linux kernel team, is my observation about the error log a
result of the patch and intended?

[1]: debian/patches/debian/firmware_class-refer-to-debian-wiki-firmware-page.patch

There are a number of Debian patches wrt firmware and *I* think they should get a (serious) review. In https://bugs.debian.org/1040738 I requested a review of the firmware related patches and described one of the issues I encountered myself. I've also been involved in triaging Debian kernel bug issues and there I've encountered several more such cases.

AFAIK quite a bit of work has happened upstream to make the firmware messages more appropriate and I think the Debian patches haven't been (properly) adjusted for those changes.

So they are (very) likely caused by the Debian patches and thus expected, but I'm hesitant to call them intended ;-)
-----BEGIN PGP SIGNATURE-----

iHUEABYIAB0WIQT1sUPBYsyGmi4usy/XblvOeH7bbgUCZVPrfgAKCRDXblvOeH7b boiXAQDA1uG7ZfQM5AYj9rUiuH0ukanafGQ2IiEA/RJJ4CiqrQD+LQCWZ+QZJF2n ZI0Wzz8jlYnQUxCwRfbYW9pRT6Oo9go=
=ZsyL
-----END PGP SIGNATURE-----

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)