[Solved] Issues with kernel 6.1.0-15-amd64/stable 6.1.66-1

[undesign]

I updated to the latest Debian 12.4 and I have the issues described here:
- no wifi connection
- extremely slow reboot
- system is not responsive
- etc

If I use an older image, then it is fine. For the moment, I "solve" it, by removing the 6.1.0-15 and using an older kernel. That one works properly (6.1.0.13).

the /boot partition is the only one with ext4:

Code: Select all

sda              8:0    0 233.8G  0 disk  
├─sda1           8:1    0   571M  0 part  /boot/efi
├─sda2           8:2    0   954M  0 part  /boot
└─sda3           8:3    0 232.3G  0 part  
  └─sda3_crypt 254:0    0 232.3G  0 crypt 
    ├─vg-swap  254:1    0   8.4G  0 lvm   [SWAP]
    └─vg-root  254:2    0 223.9G  0 lvm   /

For the moment, this is the only system where this issue manifests, the other are running properly.

Questions:
- should I wait for the next kernel package? In theory this latest version (6.1.0-15) should fix this.
- how can I check /dev/sda2 for file system inconsistencies? When I run a fsck I get an error that the file system is mounted.
I didn't dare to unmount it and to move the SSD in another machine is not an option, this is a Macbook.
- what should I do to automatically install kernel headers for the respective version?
- now, with the older kernel, the new kernel (6.1.0-15) is not displayed when I run 'apt update'. Why is that?

[alexis-burrows]

I'm new here. Same issues for me. I updated directly from image 6.1.0-13 to 15 today.
Unable to use 6.1.0-15.

WiFi / Bluetooth broken, USB devices not reading right, unable to shut system down/reboot. I got some logs from journalctl and can paste if needed.
Not sure what info to gather
So I'm stuck on 6.1.0-13.

Thanks for posting, I will be following thread and hopefully learning more.

[Aki]

Hello,

I suspect that what are you reporting is not related to the already fixed:

• Debian Bug report logs - #1057843 linux: ext4 data corruption in 6.1.64-1

You can install the newer kernel, boot the computer and then gather logs with the following commands:

Code: Select all

sudo journalctl -b 0 --no-pager > kernel.log

or

Code: Select all

su -l -c "cd $PWD; journalctl -b 0 --no-pager > kernel.log"

Then you can analyze the file named "kernel.log" and/or report it in one the following ways:

• paste them into the Debian Pastezone service (https://paste.debian.net) and report the URL/link of the paste in a follow-up message

• attaching the logs to a follow-up forum message as a compressed zip or gz file

• paste logs into the body of a follow-up message between code tags (if they fit the size of a forum message)

Hope this helps.

[konuq]

With latest 6.1.0-15 kernel, I am having same problems as op's + system hangs at reboot and shutdown. Reverting back to 6.1.0-13 fixes everything, hope this will be fixed soon.

[alexis-burrows]

I was able to run the first command to gather the kernel.log
The system then would hangup was unable to run any commands or interact with the desktop
Switching to tty was to no avail.
Commands would hang there as well and I was forced to perform hard reset/shutdown PC.

app.php/pastebin/?mode=view&s=71

[undesign]

@alexis-burrows: can you try to start in maintenance mode (from the grub menu) and try again?

This will start in terminal (no GUI) after you will provide the root password (therefore you do not need sudo).

[alexis-burrows]

I will asap. I thought I was having disk problems too recently, so I squeezed in a scan of one of my drives.

Is maintenance mode the same as "recovery mode" or is it something else that I have to access with an installation image?

[undesign]

You can enter Maintenance mode by choosing advanced options in grub (typically it is the second menu entry). Then you should see a list of available kernels. Every kernel version is displayed twice: one for normal mode, one for maintenance. Choose maintenance.

[Aki]

Hello,

@alexis-burrows:

I was able to run the first command to gather the kernel.log
The system then would hangup was unable to run any commands or interact with the desktop
Switching to tty was to no avail.
Commands would hang there as well and I was forced to perform hard reset/shutdown PC.

app.php/pastebin/?mode=view&s=71

According to your log, these are two traces that could be interesting to investigate.

The first is the following (the missing rtl8192cufw_TMSC.bin is included in the package firmware-realtek; see https://packages.debian.org/search?sear ... e&arch=any):

Code: Select all

Dec 12 07:34:31 pleiades kernel: usb 1-8: firmware: failed to load rtlwifi/rtl8192cufw_TMSC.bin (-2)
Dec 12 07:34:31 pleiades kernel: usb 1-8: Direct firmware load for rtlwifi/rtl8192cufw_TMSC.bin failed with error -2
Dec 12 07:34:31 pleiades kernel: usb 1-8: firmware: failed to load rtlwifi/rtl8192cufw.bin (-2)
Dec 12 07:34:31 pleiades kernel: usb 1-8: firmware: failed to load rtlwifi/rtl8192cufw.bin (-2)
Dec 12 07:34:31 pleiades kernel: usb 1-8: Direct firmware load for rtlwifi/rtl8192cufw.bin failed with error -2
Dec 12 07:34:31 pleiades kernel: rtlwifi: Loading alternative firmware rtlwifi/rtl8192cufw.bin

and the following error about disk /dev/hdd (the disk is not accessible):

Code: Select all

Dec 12 07:34:30 pleiades kernel: scsi 6:0:0:0: Direct-Access     WD       My Book 1235     1005 PQ: 0 ANSI: 6
Dec 12 07:34:30 pleiades kernel: scsi 6:0:0:1: Enclosure         WD       SES Device       1005 PQ: 0 ANSI: 6
Dec 12 07:34:30 pleiades kernel: sd 6:0:0:0: [sdd] Very big device. Trying to use READ CAPACITY(16).
Dec 12 07:34:30 pleiades kernel: sd 6:0:0:0: [sdd] 7813969920 512-byte logical blocks: (4.00 TB/3.64 TiB)
Dec 12 07:34:30 pleiades kernel: sd 6:0:0:0: [sdd] 4096-byte physical blocks
Dec 12 07:34:30 pleiades kernel: sd 6:0:0:0: [sdd] Write Protect is off
Dec 12 07:34:30 pleiades kernel: sd 6:0:0:0: [sdd] Mode Sense: 53 00 10 08
Dec 12 07:34:30 pleiades kernel: sd 6:0:0:0: [sdd] No Caching mode page found
Dec 12 07:34:30 pleiades kernel: sd 6:0:0:0: [sdd] Assuming drive cache: write through
[..]
Dec 12 07:34:31 pleiades kernel:  sdd: sdd1
Dec 12 07:34:31 pleiades kernel: sd 6:0:0:0: [sdd] Attached SCSI disk
[..]
Dec 12 07:34:31 pleiades udisksd[911]: Error probing device: Error sending ATA command IDENTIFY DEVICE to '/dev/sdd': Unexpected sense data returned:
                                       0000: f0 00 01 00  50 00 00 0a  80 00 00 00  00 1d 00 00    ....P...........
                                       0010: 00 00 00 00  00 00 00 00  00 00 00 00  00 00 00 00    ................
                                        (g-io-error-quark, 0)

Does you computer become unresponsive due to /dev/sdd disk not responding ?

[Swevenham]

With latest 6.1.0-15 kernel, I am having same problems as op's + system hangs at reboot and shutdown. Reverting back to 6.1.0-13 fixes everything, hope this will be fixed soon.

Yeah, my wi-fi seems to work, but shutdown/reboot doesn't. I'm wary of removing kernels, so guess i'm configuring grub to boot with 6.1.0-13 for now.

[alexis-burrows]

@Aki

According to your log, these are two traces that could be interesting to investigate.

The first is the following (the missing rtl8192cufw_TMSC.bin is included in the package firmware-realtek; see https://packages.debian.org/search?sear ... e&arch=any):

I installed the driver for the USB dongle from the company's website.
I'll uninstall it and try to use the firmware-realtek package from the repos instead.

Does you computer become unresponsive due to /dev/sdd disk not responding ?

It never has before. I'm not sure if that's what is causing the hangups or not.
There isn't an issue with the disk that I'm aware of. I just did a SMART test on it.
It's and encrypted external drive and I have it set to standby after a few minutes of inactivity, using hdparm utility.

Thank you for your help

[stevepusser]

It's a known issue with kernels based on 6.1.66 and 6.6.5. Whoops!!

https://www.phoronix.com/news/Linux-6.6.6-Released

6.1.67 reverts that one change to fix that, but I don't see it in the Debian pool yet. Let's see if I can just make a patch that for 6.1.66...

Just rebuilt Liquorix 6.6-7, based on 6.6.6.

[stevepusser]

Ahhh, here's the incremental 6.1.66 to 67 patch. https://cdn.kernel.org/pub/linux/kernel ... 1.66-67.xz

[stevepusser]

And 6.1.67 was just committed to the Bookworm repo. Should fix the issue if the wi-fi works well enough to upgrade.

[sunrat]

Just rebuilt Liquorix 6.6-7, based on 6.6.6.

The neighbour of the beast.

[Aki]

@alexis-burrows:

@Aki

According to your log, these are two traces that could be interesting to investigate.
The first is the following (the missing rtl8192cufw_TMSC.bin is included in the package firmware-realtek; see https://packages.debian.org/search?sear ... e&arch=any):

I installed the driver for the USB dongle from the company's website.
I'll uninstall it and try to use the firmware-realtek package from the repos instead.

According to the log, your WiFi USB device uses a Realtek chipset (rtl8821cu):

Code: Select all

[..]
$ grep -i rtl "2023-12-12T07 34 30-kernel.log.txt"
Dec 12 07:34:31 pleiades kernel: usbcore: registered new interface driver rtl8821cu
Dec 12 07:34:31 pleiades kernel: rtl8821cu 3-2:1.0 wlx347de4401b77: renamed from wlan0
Dec 12 07:34:31 pleiades systemd[1]: Listening on virtlockd.socket - Virtual machine lock manager socket.
Dec 12 07:34:31 pleiades systemd[1]: Listening on virtlockd-admin.socket - Virtual machine lock manager admin socket.
Dec 12 07:34:31 pleiades systemd[1]: Listening on virtlogd.socket - Virtual machine log manager socket.
Dec 12 07:34:31 pleiades systemd[1]: Listening on virtlogd-admin.socket - Virtual machine log manager socket.
Dec 12 07:34:31 pleiades kernel: rtl8192cu: Chip version 0x10
Dec 12 07:34:31 pleiades kernel: rtl8192cu: Board Type 0
Dec 12 07:34:31 pleiades kernel: rtl_usb: rx_max_size 15360, rx_urb_num 8, in_ep 1
Dec 12 07:34:31 pleiades kernel: rtl8192cu: Loading firmware rtlwifi/rtl8192cufw_TMSC.bin
Dec 12 07:34:31 pleiades kernel: ieee80211 phy1: Selected rate control algorithm 'rtl_rc'
Dec 12 07:34:31 pleiades kernel: usbcore: registered new interface driver rtl8192cu
Dec 12 07:34:31 pleiades kernel: usb 1-8: firmware: failed to load rtlwifi/rtl8192cufw_TMSC.bin (-2)
Dec 12 07:34:31 pleiades kernel: usb 1-8: firmware: failed to load rtlwifi/rtl8192cufw_TMSC.bin (-2)
Dec 12 07:34:31 pleiades kernel: usb 1-8: Direct firmware load for rtlwifi/rtl8192cufw_TMSC.bin failed with error -2
Dec 12 07:34:31 pleiades kernel: usb 1-8: firmware: failed to load rtlwifi/rtl8192cufw.bin (-2)
Dec 12 07:34:31 pleiades kernel: usb 1-8: firmware: failed to load rtlwifi/rtl8192cufw.bin (-2)
Dec 12 07:34:31 pleiades kernel: usb 1-8: Direct firmware load for rtlwifi/rtl8192cufw.bin failed with error -2
Dec 12 07:34:31 pleiades kernel: rtlwifi: Loading alternative firmware rtlwifi/rtl8192cufw.bin
Dec 12 07:34:31 pleiades kernel: rtl8192cu 1-8:1.0 wlx74da385c7495: renamed from wlan0
Dec 12 07:34:31 pleiades NetworkManager[1055]: <info>  [1702388071.9978] rfkill1: found Wi-Fi radio killswitch (at /sys/devices/pci0000:00/0000:00:14.0/usb1/1-8/1-8:1.0/ieee80211/phy1/rfkill1) (driver rtl8192cu)
Dec 12 07:34:31 pleiades NetworkManager[1055]: <info>  [1702388071.9981] rfkill0: found Wi-Fi radio killswitch (at /sys/devices/pci0000:00/0000:00:1b.2/0000:05:00.0/usb3/3-2/3-2:1.0/ieee80211/phy0/rfkill0) (driver rtl8821cu)
Dec 12 07:34:32 pleiades kernel: rtl8192cu: MAC auto ON okay!
Dec 12 07:34:32 pleiades kernel: rtl8192cu: Tx queue select: 0x05
Dec 12 07:34:33 pleiades kernel: rtl8192c_common: Polling FW ready fail! REG_MCUFWDL:0x00000006.
Dec 12 07:34:33 pleiades kernel: rtl8192c_common: Firmware is not ready to run!
Dec 12 07:34:33 pleiades kernel: rtl8192cu: MAC auto ON okay!
Dec 12 07:34:33 pleiades kernel: rtl8192cu: Tx queue select: 0x05
Dec 12 07:34:34 pleiades kernel: rtl8192c_common: Polling FW ready fail! REG_MCUFWDL:0x00000006.
Dec 12 07:34:34 pleiades kernel: rtl8192c_common: Firmware is not ready to run!

The manufacturer installer scripts install a foreign unknown debian package named rtl8821cu-dkms.deb from https://linux.brostrend.com/rtl8821cu-dkms.deb , that's the installed one that is not working correctly:

Code: Select all

[..]
install_driver() {
    local module reinstall ver ikd

    case "$_DRIVER" in
    rtl*) module=${_DRIVER#rtl} ;;
    *) module=aic8800_fdrv ;;
    esac
    bold "Downloading the $_DRIVER driver"
    re cd "$(mktemp -d)"
    # Avoid the "Download is performed unsandboxed" apt warning
    re chmod 755 .
    download "https://linux.brostrend.com/$_DRIVER-dkms.deb"
    bold "Installing and compiling the driver"
[..]

Searching the internet, I found this third part repository:

• 8821cu ( 8821cu.ko )

I successfully recompiled the aforementioned kernel module with the latest Debian Bookworm, but I did not install or test it because I'm missing the device itself.

Anyway, according to the upstream (official) Linux source kernel repository:

• kernel/git/stable/linux.git - wifi: rtw88: Add rtw8821cu chipset support

upstream suppor in Linux kernel for the rtl8122cu chipset ("Realtek 8821CU USB wireless network adapter") was introduced in December 2022. According to the following command (run after downloading a kernel git repository), it should be available since kernel version 6.1:

Code: Select all

$ git describe --tag aff5ffd718de23cb8603f2e229204670e2644334
v6.1-rc7-1789-gaff5ffd718de

Does you computer become unresponsive due to /dev/sdd disk not responding ?

It never has before. I'm not sure if that's what is causing the hangups or not.
There isn't an issue with the disk that I'm aware of. I just did a SMART test on it.
It's and encrypted external drive and I have it set to standby after a few minutes of inactivity, using hdparm utility.

You could consider disconnecting it to exclude as a possible (contributing ?) cause.

EDIT: anyway, as @stevepusser pointed out in previous posts, a new kernel update (with some regression fixes) for Debian Bookworm is in progress (in proposed-updates repository) that it's worth trying:

• Debian Bug report logs - #1057967 linux-image-6.1.0-15-amd64 renders my physical bookworm/gnome computer largely unusable

• linux-signed-amd64_6.1.67+1_source.changes is NEW

[undesign]

According to your log, these are two traces that could be interesting to investigate.

The first is the following (the missing rtl8192cufw_TMSC.bin is included in the package firmware-realtek; see https://packages.debian.org

Yes, but how does the system works properly when a different kernel is booted?
I also have wi-fi issues, yet the firmware (from Intel in my case) is installed. Once I boot an older kernel, everything is fine.

Later edit: yesterday I updated another system of mine to this 6.1.0-15 kernel and everything was fine.

[undesign]

• Debian Bug report logs - #1057967 linux-image-6.1.0-15-amd64 renders my physical bookworm/gnome computer largely unusable

Yes, that's the bug described in this topic. It even has a report from a Macbook user.

[alirezaimi]

Hi
I have same problem after upgrade to 6.1.66-1, slow boot process and stuck in sddm kde desktop and nothing work at all, I had to switch to older kernel 6.1.64-1 from grub and its work perfectly .

[scaalp]

Bonjour,
Just to say that I also have this problem. It disappear when I boot with the previous kernel (6.1.0-14).
I'll try to read later all the post to see if a diagnostic fits with my situation.
Cordialement,
Pascal


Hardware : MacBook mid-2013 (6.2)
OS : Debian bookworm, up-to-date

- startup ok, maybe a bit longer
- wifi tries to connect
- may succeed after dozen of seconds
- but KDE Plasma hangs, and after few minutes, a lot of pop-up appears to tell that timeout and dbus problems occured.
- power off may takes 10 minutes

As the first post, I also have a LUKS partition.

Code: Select all

Sys. de fichiers       blocs de 1K   Utilisé Disponible Uti% Monté sur
udev                       1945080         0    1945080   0% /dev
tmpfs                       394808      1456     393352   1% /run
/dev/sda5                100663296  29389084   68790180  30% /
tmpfs                      1974028         0    1974028   0% /dev/shm
tmpfs                         5120        12       5108   1% /run/lock
/dev/sda4                   510984      5972     505012   2% /boot/efi
/dev/mapper/home_crypt   136787236 129126504     679520 100% /home
tmpfs                       394804        68     394736   1% /run/user/1000

Note : I also had to see if my Debian VM guest at work does the same

**EDIT** my work VM guest uses Debian 12 kernel 6.1.0-15, and has a normal bevavior (as usual).

I'll try the proposed newer kernel at home this evening.