[linux-firmware] Raspberry pi 4 wlan0 is missing

[jgcrunden]

Hi ughhug9

It's a known issue with the firmware. I had the same thing with my Pi when I updated it this morning

There's a topic for the bug here along with a workaround from graysky, deleting 4 files from /usr/lib/firmware
viewtopic.php?f=15&t=15120&start=20#p65802
I've tried doing that and it fixes it for me.

You could try that, or wait until the developers to release a fix.

Thanks

[ughhug9]

Thanks for your reply, jgcrunden

Is there a mailing list that I can assign to where info like this is sent out so the user base don't need to be caught off-guard by these types of errors?

In this particular case, It will cause me quite a bit of hassle as two of the three devices are hosted on a remote location and used in a production setting.

[jgcrunden]

You're welcome.
I'm not sure to be honest, I've only just signed up as a user to the the Arch Linux ARM forum so still getting to grips with their site and community...

[ughhug9]

Looks like the proposed solution worked for me on the device I have at home. Thanks for the suggestion!

As for the other two devices that remotely located, I will have to find a different solution. I'm thinking they should be accessible if an ethernet cable is plugged in. That should give me access and enable my to deploy the fix.

[cyberpunkrocker]

Thanks guys, this community is awesome! I just updated my RPi4 and lost wlan0, and here's an immediate fix!

I agree with mdevaev that this firmware package as it is now is extremely dangerous. Fortunately I haven't updated my two Zero W's yet, and because I have only SSH access to them, firmware update would have been disastrous.

[ughhug9]

I wasn't as lucky as cyberpunkrocker - I mindlessly updated two remote devices and both of them were unable to come back online after reboot.

Question: Is there a mailing list (similar to [arch-announce]) that can provide users with information about potentially breaking changes? Would have saved me some trouble

[robg]

Dear mdevaev,

Thanks for pointing out the related discussion on the mailing list (https://www.spinics.net/lists/arm-kernel/msg876183.html). It appears that there is misunderstanding on the kernel dev's side: The issue doesn't just apply to the RPI kernel (linux-raspberrypi4) but also to the vanilla aarch64 kernel (linux-aarch64), as tested by graysky: viewtopic.php?f=15&t=15120#p65769

Is this indeed the case? If someone can confirm, I will report it on the mailing list.

[dpj]

similar looking issue on rpi 3 b,

- working system with brcmfmac43430-sdio.bin 7.45.98.38
- updated to linux-firmware-20210208.b79d239-1-any.pkg.tar.xz
- brcmfmac43430-sdio.bin 7.45.98.97
- no wlan interface present after boot

$this->bbcode_second_pass_code('', '
uname -a
Linux alarm 5.10.16-1-ARCH #1 SMP Sat Feb 13 19:41:14 UTC 2021 aarch64 GNU/Linux
')
$this->bbcode_second_pass_code('', '
dmesg
-->%--
[ 12.669479] brcmfmac: brcmf_fw_alloc_request: using brcm/brcmfmac43430-sdio for chip BCM43430/1
[ 12.769539] audit: type=1127 audit(1612390934.499:8): pid=333 uid=0 auid=4294967295 ses=4294967295 msg=' comm="systemd-update-utmp" exe="/usr/lib/systemd/systemd-update-utmp" hostname=? addr=? terminal=? res=success'
[ 12.813685] ieee80211 phy0: brcmf_fw_crashed: Firmware has halted or crashed
[ 12.828459] ------------[ cut here ]------------
[ 12.828500] WARNING: CPU: 2 PID: 78 at kernel/workqueue.c:1477 __queue_work+0x2e0/0x364
[ 12.855171] Modules linked in: brcmfmac brcmutil rc_cec smsc vc4 cec cfg80211 rc_core smsc95xx bcm2835_v4l2(C) drm_kms_helper usbnet mii bcm2835_mmal_vchiq(C) hci_uart syscopyarea sysfillrect sysimgblt videobuf2_vmalloc fb_sys_fops btqca videobuf2_memops btbcm videobuf2_v4l2 btintel videobuf2_common raspberrypi_hwmon bluetooth videodev mc pwm_bcm2835 bcm2835_thermal snd_bcm2835(C) i2c_bcm2835 ecdh_generic ecc rfkill bcm2835_wdt fuse drm
[ 12.855331] CPU: 2 PID: 78 Comm: kworker/u8:1 Tainted: G C 5.10.16-1-ARCH #1
[ 12.855335] Hardware name: Raspberry Pi 3 Model B (DT)
[ 12.855418] Workqueue: brcmf_wq/mmc1:0001:1 brcmf_sdio_dataworker [brcmfmac]
[ 12.855431] pstate: 00000085 (nzcv daIf -PAN -UAO -TCO BTYPE=--)
[ 12.855445] pc : __queue_work+0x2e0/0x364
[ 12.855454] lr : __queue_work+0xa0/0x364
[ 12.855459] sp : ffff8000122b3c40
[ 12.855463] x29: ffff8000122b3c40 x28: 0000000000000008
[ 12.855474] x27: ffff80001188b000 x26: ffff800011cbecb8
[ 12.855484] x25: ffff80001188b000 x24: 0000000000000002
[ 12.855493] x23: 000000000000000c x22: ffff000002409c00
[ 12.855504] x21: ffff0000371b8f00 x20: ffff0000371eac00
[ 12.855514] x19: ffff0000042fb6c8 x18: 0000000000000030
[ 12.855523] x17: 0000000000000000 x16: 0000000000000000
[ 12.855533] x15: 00004270974ad038 x14: ffff800011cd91f0
[ 12.855543] x13: 0000000000000000 x12: 0000000000000040
[ 12.855553] x11: ffff000002803908 x10: ffff00000280390a
[ 12.855563] x9 : ffff800011cd6268 x8 : ffff000002803930
[ 12.855573] x7 : 0000000000000000 x6 : ffff000002803928
[ 12.855583] x5 : ffff000002803908 x4 : 0000000000000000
[ 12.855592] x3 : ffff0000371e6f00 x2 : 0000000000000001
[ 12.855601] x1 : 0000000000000000 x0 : ffff0000042fb6d0
[ 12.855612] Call trace:
[ 12.855623] __queue_work+0x2e0/0x364
[ 12.855632] queue_work_on+0x54/0x7c
[ 12.855686] brcmf_fw_crashed+0x38/0x70 [brcmfmac]
[ 12.855740] brcmf_sdio_hostmail+0x15c/0x1d4 [brcmfmac]
[ 12.855796] brcmf_sdio_dpc+0x2d4/0x590 [brcmfmac]
[ 12.855852] brcmf_sdio_dataworker+0x3c/0xa0 [brcmfmac]
[ 12.855863] process_one_work+0x1c4/0x3a0
[ 12.855874] worker_thread+0x178/0x400
[ 12.855882] kthread+0x10c/0x110
[ 12.855891] ret_from_fork+0x10/0x3c
[ 12.855897] ---[ end trace 194d5fd6f415a1d1 ]---
')
tried to delete
$this->bbcode_second_pass_code('', '
sudo rm /usr/lib/firmware/brcm/brcmfmac43430-sdio.raspberrypi,3-model-b.txt
')
but this has not helped.

[mdevaev]

$this->bbcode_second_pass_quote('robg', 'I')s this indeed the case? If someone can confirm, I will report it on the mailing list.

[robg]

I have written a reply to the thread https://www.spinics.net/lists/arm-kernel/msg876183.html pointing out that the issue also appears with the vanilla kernel. My reply is currently awaiting moderation.