Just updated CentOS 9 Stream on a Lenovo T17 Gen 4 Intel and not it won't suspend with the following error:
[ 52.604998] Restarting kernel threads ... done. [ 52.605111] OOM killer enabled. [ 52.605111] Restarting tasks ... done. [ 52.606604] random: crng reseeded on system resumption [ 52.616014] thermal thermal_zone9: failed to read out thermal zone (-61) [ 52.791625] PM: suspend exit [ 52.791733] PM: suspend entry (s2idle) [ 52.797260] Filesystems sync: 0.005 seconds [ 52.797579] Freezing user space processes ... (elapsed 0.001 seconds) done. [ 52.799127] OOM killer disabled. [ 52.799128] Freezing remaining freezable tasks ... [ 72.805437] Freezing of tasks failed after 20.006 seconds (1 tasks refusing to freeze, wq_busy=0): [ 72.805450] task:NFSv4 callback state:I stack:0 pid:2191 ppid:2 flags:0x00004000 [ 72.805453] Call Trace: [ 72.805454] <TASK> [ 72.805456] __schedule+0x21b/0x550 [ 72.805463] schedule+0x2d/0x70 [ 72.805466] nfs41_callback_svc+0x186/0x190 [nfsv4] [ 72.805508] ? __pfx_autoremove_wake_function+0x10/0x10 [ 72.805512] ? __pfx_nfs41_callback_svc+0x10/0x10 [nfsv4] [ 72.805536] kthread+0xdd/0x100 [ 72.805538] ? __pfx_kthread+0x10/0x10 [ 72.805539] ret_from_fork+0x29/0x50 [ 72.805543] </TASK>
Kernel:
5.14.0-404.el9.x86_64
How do I boot the previous kernel?
Holding Shift down does nothing.
Pressing escape brings up some kind of emergency screen that has no options other than to "Press any key to exit".
Is CentOS 9 Stream considered stable enough to use as an everyday desktop?
Mike
On Wed, Jan 10, 2024 at 11:57 PM Michael B Allen ioplex@gmail.com wrote:
Just updated CentOS 9 Stream on a Lenovo T17 Gen 4 Intel and now it won't suspend with the following error:
...
[ 72.805437] Freezing of tasks failed after 20.006 seconds (1 tasks refusing to freeze, wq_busy=0): [ 72.805450] task:NFSv4 callback state:I stack:0 pid:2191 ppid:2 flags:0x00004000
FYI
After reverting to 5.14.0-388 yesterday, stability is restored. I can consistently suspend and resume without issues.
In hindsight, kernel 391 was also giving me issues. My wired network would sporadically fail to come up after suspend. No amount of fiddling would restore. Only rebooting. I have not seen that behavior with 388 either.
This would suggest an issue with networking / suspend between 388 and 391.
Q: If I update, am I going to lose kernel 388? How can I persist that one specific kernel indefinitely and still be able to update the rest of the system?
Mike
# grubby --info=ALL | grep ^kernel kernel="/boot/vmlinuz-5.14.0-404.el9.x86_64" kernel="/boot/vmlinuz-5.14.0-391.el9.x86_64" kernel="/boot/vmlinuz-5.14.0-388.el9.x86_64" kernel="/boot/vmlinuz-0-rescue-aaab5fbe787947ec94b3c7574b9d41e6" # grubby --default-kernel /boot/vmlinuz-5.14.0-388.el9.x86_64
I have two suggestions about how to keep - for a while - the 388 kernel.
First - Use DNF to remove the bad kernels. Then when a new one comes in it will take one of those slots.
Second - Increase the number of installed kernels. That change is in yum.conf.
https://linuxconfig.org/how-to-remove-old-unused-kernels-on-centos-linux
If the number is too high, you may run out of room in /boot partition. Testing required....
=============== Bill Gee
On 1/12/24 13:37, Michael B Allen wrote:
On Wed, Jan 10, 2024 at 11:57 PM Michael B Allen ioplex@gmail.com wrote:
Just updated CentOS 9 Stream on a Lenovo T17 Gen 4 Intel and now it won't suspend with the following error:
...
[ 72.805437] Freezing of tasks failed after 20.006 seconds (1 tasks refusing to freeze, wq_busy=0): [ 72.805450] task:NFSv4 callback state:I stack:0 pid:2191 ppid:2 flags:0x00004000
FYI
After reverting to 5.14.0-388 yesterday, stability is restored. I can consistently suspend and resume without issues.
In hindsight, kernel 391 was also giving me issues. My wired network would sporadically fail to come up after suspend. No amount of fiddling would restore. Only rebooting. I have not seen that behavior with 388 either.
This would suggest an issue with networking / suspend between 388 and 391.
Q: If I update, am I going to lose kernel 388? How can I persist that one specific kernel indefinitely and still be able to update the rest of the system?
Mike
# grubby --info=ALL | grep ^kernel kernel="/boot/vmlinuz-5.14.0-404.el9.x86_64" kernel="/boot/vmlinuz-5.14.0-391.el9.x86_64" kernel="/boot/vmlinuz-5.14.0-388.el9.x86_64" kernel="/boot/vmlinuz-0-rescue-aaab5fbe787947ec94b3c7574b9d41e6" # grubby --default-kernel /boot/vmlinuz-5.14.0-388.el9.x86_64 _______________________________________________ CentOS mailing list CentOS@centos.org https://lists.centos.org/mailman/listinfo/centos
On Fri, Jan 12, 2024 at 4:53 PM Bill Gee bgee@campercaver.net wrote:
I have two suggestions about how to keep - for a while - the 388 kernel.
First - Use DNF to remove the bad kernels. Then when a new one comes in it will take one of those slots.
Second - Increase the number of installed kernels. That change is in yum.conf.
Done. Thanks. Hopefully whatever the bug is will get worked out at some point ...
Mike
On Fri, Jan 12, 2024 at 10:57 PM Michael B Allen ioplex@gmail.com wrote:
Done. Thanks. Hopefully whatever the bug is will get worked out at some point ...
Just FYI
I updated to 5.14.0-410. The login screen appeared for 2 seconds and then it dropped to console with:
Failed to start: Crash recovery kernel arming
Tried to reboot but it hang on the splash screen. The console shows "Failed to start ..." for all services like ModemManager and so on.
Rebooted back into 388 and it's running fine again.
So it looks like my stability is getting worse for me with each update ...
I'm getting worried about hanging onto 388.
Mike
On Tue, Feb 06, 2024 at 05:59:15PM -0500, Michael B Allen wrote:
On Fri, Jan 12, 2024 at 10:57 PM Michael B Allen ioplex@gmail.com wrote:
Done. Thanks. Hopefully whatever the bug is will get worked out at some point ...
Just FYI
I updated to 5.14.0-410. The login screen appeared for 2 seconds and then it dropped to console with:
Failed to start: Crash recovery kernel arming
Tried to reboot but it hang on the splash screen. The console shows "Failed to start ..." for all services like ModemManager and so on.
Rebooted back into 388 and it's running fine again.
So it looks like my stability is getting worse for me with each update ...
I'm getting worried about hanging onto 388.
The currently booted kernel will never get swapped out when you are updating, so this should not be a concern. Have you reported your issues on JIRA though?
On Wed, Feb 21, 2024 at 10:40 PM Michel Lind salimma@centosproject.org wrote:
Rebooted back into 388 and it's running fine again.
So it looks like my stability is getting worse for me with each update ...
I'm getting worried about hanging onto 388.
The currently booted kernel will never get swapped out when you are updating, so this should not be a concern. Have you reported your issues on JIRA though?
Well it just happened with 388 too. Coming out of suspend (power button), the network didn't come up. Tried gracefully closing terminals and logging out but it hanged up. After about 5 minutes Wayland did log out but then it got really hard stuck (no mouse or keyboard) on some kind of half-baked gdm screen. Had to hold the power button down for a while. Come up ok back into 388 ok.
Here's /var/log/messages starting from previous suspend. Trouble starts around first 15:50:05 timestamp.
https://www.ioplex.com/~miallen/CentOSSuspendFailLenovoT14-03-15-2024.txt
Names have been changed to protect the guilty. I'm going to remove this after a day or so so download it now if you're interested.
Hopefully it helps someone figure out what's going on.
I'm not doing anything exotic. I don't have lots of crap installed. I'm not watching videos or listening to tunes. Just coding java over NFS in terminators.
Do you have a link to instructions for posting to Jira? I've been writing C and Java since the 1900's so I might actually be able to get the diagnostics you really want.
Mike
On Fri, Mar 15, 2024 at 4:31 PM Michael B Allen ioplex@gmail.com wrote:
On Wed, Feb 21, 2024 at 10:40 PM Michel Lind salimma@centosproject.org wrote:
Rebooted back into 388 and it's running fine again.
So it looks like my stability is getting worse for me with each update
...
I'm getting worried about hanging onto 388.
The currently booted kernel will never get swapped out when you are updating, so this should not be a concern. Have you reported your issues on JIRA though?
Well it just happened with 388 too. Coming out of suspend (power button), the network didn't come up. Tried gracefully closing ...
Hopefully it helps someone figure out what's going on.
Fixed.
Updating to kernel 432 fixed it. At that time I did also switch to X11 thinking that maybe it was a factor. But I switched back to Wayland briefly before updating to kernel 437 and it worked fine. Now I've been on kernel 437 + Wayland for probably a month or so and I haven't noticed anything out of the ordinary. Very solid now.
So whatever it was, just updating fixed it.
Just FYI. Carry on ...
Mike
-
Bill Gee -
Michael B Allen -
Michel Lind