You are not logged in.
Hello.
I am having an issue on my Thinkpad T14 gen 5 AMD laptop where, after a reboot, the WiFi has suddenly stopped working. Network manager does not allow for WiFi connection to be made.
I believe I have found the culprit in journalctl:
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Mounted /efi.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Finished Load/Save Screen Backlight Brightness of backlight:amdgpu_bl1.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Reached target Local File Systems.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Listening on Boot Entries Service Socket.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Listening on System Extension Image Management.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Starting Rebuild Dynamic Linker Cache...
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Starting Set Up Additional Binary Formats...
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Finished Load/Save OS Random Seed.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Update Boot Loader Random Seed was skipped because no trigger condition checks were met.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: proc-sys-fs-binfmt_misc.automount: Got automount request for /proc/sys/fs/binfmt_misc, triggered by 871 (systemd-bi>
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Mounting Arbitrary Executable File Formats File System...
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Mounted Arbitrary Executable File Formats File System.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Finished Set Up Additional Binary Formats.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Finished Flush Journal to Persistent Storage.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Starting Create System Files and Directories...
Sep 10 12:34:01 gdb-arch-t14-5 kernel: ath12k_pci 0000:02:00.0: chip_id 0x2 chip_family 0x4 board_id 0xff soc_id 0x40170200
Sep 10 12:34:01 gdb-arch-t14-5 kernel: ath12k_pci 0000:02:00.0: fw_version 0x100301e1 fw_build_timestamp 2023-12-06 04:05 fw_build_id QC_IMAGE_VERSION_STRING=>
Sep 10 12:34:01 gdb-arch-t14-5 kernel: ath12k_pci 0000:02:00.0: failed to fetch board data for bus=pci,vendor=17cb,device=1107,subsystem-vendor=17aa,subsystem>
Sep 10 12:34:01 gdb-arch-t14-5 kernel: ath12k_pci 0000:02:00.0: failed to fetch board.bin from WCN7850/hw2.0
Sep 10 12:34:01 gdb-arch-t14-5 kernel: ath12k_pci 0000:02:00.0: qmi failed to load bdf:
Sep 10 12:34:01 gdb-arch-t14-5 kernel: ath12k_pci 0000:02:00.0: qmi failed to load board data file:-2
Sep 10 12:34:01 gdb-arch-t14-5 systemd-tmpfiles[880]: Opening file "/tmp/.X11-unix/X0" failed, proceeding without lock: No such device or address
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Finished Create System Files and Directories.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: First Boot Wizard was skipped because of an unmet condition check (ConditionFirstBoot=yes).
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: First Boot Complete was skipped because of an unmet condition check (ConditionFirstBoot=yes).
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Starting Rebuild Journal Catalog...
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Save Transient machine-id to Disk was skipped because of an unmet condition check (ConditionPathIsMountPoint=/etc/m>
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Starting Record System Boot/Shutdown in UTMP...
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Finished Record System Boot/Shutdown in UTMP.
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Finished Rebuild Journal Catalog.
Sep 10 12:34:01 gdb-arch-t14-5 (udev-worker)[451]: ucsi-source-psy-USBC000:001: Process '/usr/share/slimbookamdcontroller/src/applyconfig.py' failed with exit>
Sep 10 12:34:01 gdb-arch-t14-5 systemd[1]: Finished Rebuild Dynamic Linker Cache.
The error appears to be at:
Sep 10 12:34:01 gdb-arch-t14-5 kernel: ath12k_pci 0000:02:00.0: failed to fetch board data for bus=pci,vendor=17cb,device=1107,subsystem-vendor=17aa,subsystem>
Sep 10 12:34:01 gdb-arch-t14-5 kernel: ath12k_pci 0000:02:00.0: failed to fetch board.bin from WCN7850/hw2.0
Sep 10 12:34:01 gdb-arch-t14-5 kernel: ath12k_pci 0000:02:00.0: qmi failed to load bdf:
Sep 10 12:34:01 gdb-arch-t14-5 kernel: ath12k_pci 0000:02:00.0: qmi failed to load board data file:-2
I have tried using both wpa_supplicant and iwd as a backend on Network Manager, to no avail. This would make sense, as it seems to be a kernel issue.
The only thing I can think of that changed in-between this boot and the last boot is that I ran:
sudo ethtool -s enp1s0f0 wol g
In an attempt to solve some issues I was having (and am still having) with s2idle.
I am on kernel 6.10.9.
Thanks in advance for any help.
Offline
This basically has two possible causes, a kernel regression and/or a linux-firmware regression. Did you do an update of either before rebooting? If there was only a kernel update, try downgrading the kernel, if it was firmware, try downgrading the linux-firmware package (and potentially the kernel, though you can generally regard this distinctly and handle them in isolation, so downgrade just linux-firmware, retest, maybe bring in the kernel, retest)
Offline
Hmm I think a linux firmware update was issued. I'll try and downgrade it. Thanks.
EDIT: It was linux-firmware and linux-firmwase-whence that were upgraded. I don't have earlier versions in the cache.
Last edited by Mezzy12 (2024-09-10 13:13:39)
Offline
Online
Just did that there. It's a regression alright. Works fine with the 8 of Septembers firmware.
Offline
Looks like the issue was triggered by https://gitlab.com/kernel-firmware/linu … 1c76f70927
Does linux 6.11-rc7 which contains https://github.com/torvalds/linux/commi … 55d80f72be work as expected?
sudo pacman -U https://pkgbuild.com/\~gromit/linux-bisection-kernels/linux-mainline-6.11rc7-1-x86_64.pkg.tar.zst
Last edited by loqs (2024-09-10 23:10:09)
Offline
Will try that when I get the opportunity. Should I also upgrade linux-firmware and linux-firmware-whence?
Offline
Will try that when I get the opportunity. Should I also upgrade linux-firmware and linux-firmware-whence?
Yes to test if the issue triggered by linux-firmware 20240909.552ed9b8-1 is already been fixed in linux-mainline 6.11rc7 but the fix has not been back-ported to 6.10.Y stable yet.
Offline
Thanks. I'll try that now. What's the best way to explicitly tell pacman to grab the latest linux-firmware and linux-firmware-whence packages? Right now my system is using Sunday's repo archive.
Offline
Thanks. I'll try that now. What's the best way to explicitly tell pacman to grab the latest linux-firmware and linux-firmware-whence packages? Right now my system is using Sunday's repo archive.
If the packages are still in pacman's cache without changing the repo you would use:
# pacman -U /var/cache/pacman/pkg/linux-firmware-20240909.552ed9b8-1-any.pkg.tar.zst /var/cache/pacman/pkg/linux-firmware-whence-20240909.552ed9b8-1-any.pkg.tar.zst
Offline
Okay I tried that, still doesn't work unfortunately. Had to downgrade again even with the mainline kernel.
Offline
Does it work with the linux-lts kernel?
Offline
I will check now.
Offline
Does it work with the linux-lts kernel?
Nope, it very much does not hahah.
Offline
If you use linux-firmware-20240909.552ed9b8-1.1 which has https://gitlab.com/kernel-firmware/linu … 1c76f70927 reverted can you still reproduce the issue?
Offline
What is the source of that firmware? Nothing against you, but I am wary of installing something like that from google drive with pacman
Offline
Since the source of the firmware above seems to be a concern I have built & uploaded a version that is signed with my packager key (which you can manually verify like so "pacman-key --verify linux-firmware-20240909.552ed9b8-1.1-any.pkg.tar.zst{.sig,}"):
sudo pacman -U https://pkgbuild.com/\~gromit/linux-bisection-kernels/linux-firmware-20240909.552ed9b8-1.1-any.pkg.tar.zst
If you want to build it yourself instead you can clone the official PKGBUILD with "pkgctl repo clone --protocol=https linux-firmware" and apply the following patch:
+++ b/PKGBUILD
@@ -7,7 +7,7 @@ pkgname=(linux-firmware-whence linux-firmware amd-ucode
)
_tag=20240909
pkgver=20240909.552ed9b8
-pkgrel=1
+pkgrel=1.1
pkgdesc="Firmware files for Linux"
url="https://gitlab.com/kernel-firmware/linux-firmware"
license=(
@@ -40,6 +40,8 @@ prepare() {
git log --oneline -1 "${_c}"
git cherry-pick -n "${_c}"
done
+
+ git revert -n 82318c966fd1af87044299d34611751c76f70927
}
pkgver() {
Offline
Yep, that works! Thanks gromit and loqs.
Offline
@gromit any thoughts how/where to report the issue upstream?
Offline
No not really, I asked in the packaging and a kernel channel about that, maybe someone has a good idea. If not I will send a MR reverting the commit in question to https://gitlab.com/kernel-firmware/linux-firmware (which has issues disabled).
Offline
Posted a bugreport: https://lore.kernel.org/all/44ae68d2-24 … heusel.eu/
Offline
There was a separate bug report and the issue is now being tracked & discussed here: https://bugzilla.kernel.org/show_bug.cgi?id=219282
Offline
Apparently there is a potential fix incoming: https://bugzilla.kernel.org/show_bug.cgi?id=219282#c8
Offline
You can rebuild the linux-firmware package with the proposed fix applied with the following diff:
diff --git a/PKGBUILD b/PKGBUILD
index a9968cf..24f181e 100644
--- a/PKGBUILD
+++ b/PKGBUILD
@@ -25,8 +25,10 @@ options=(
!strip
!debug
)
-source=("git+$url.git?signed#tag=${_tag}")
-b2sums=('1578c2a2fc4696a21d463f03745dc5bf2fe1266001e3ceac698f52c4b891c1d045c788c6ac57555182f8cf78b6dd973e84fed34884092dc36850c63643d067f9')
+source=("git+$url.git?signed#tag=${_tag}"
+ https://git.codelinaro.org/clo/ath-firmware/ath12k-firmware/-/commit/6de0474a14f8eee790b9f1166346b01c3556a11b.patch)
+b2sums=('1578c2a2fc4696a21d463f03745dc5bf2fe1266001e3ceac698f52c4b891c1d045c788c6ac57555182f8cf78b6dd973e84fed34884092dc36850c63643d067f9'
+ '95a6028a6f65672f061b5680bb6e5754bd6a4559800bfb678f4cf33c07bd32e21e3d113adb71d8d1da703409a6e52b45874690c7873b724da4996baaa06b8304')
validpgpkeys=('4CDE8575E547BF835FE15807A31B6BD72486CFD6') # Josh Boyer <jwboyer@fedoraproject.org>
_backports=(
@@ -35,6 +37,8 @@ _backports=(
prepare() {
cd ${pkgbase}
+ git apply --directory ath12k/ -v ../6de0474a14f8eee790b9f1166346b01c3556a11b.patch
+
local _c
for _c in "${_backports[@]}"; do
git log --oneline -1 "${_c}"
Offline
Could somebody in this thread test the proposed fix as described above and report back if it works or not?
Offline