I had an attempted shutdown yesterday that didn't shutdown. It had more than the usual message posted, here it is: localhost login: [10495.617 342] [ T1] reboot:Powerdown [10495.6205841] [t1785] asix 1-5:1.0 enp 0s20f0u5: failed to write reg irdex 0x0000: -19 [10495.623670] [ t1785] asix 1-5:1.0 enp0s20f0u5: failed to enable software MIIaccess Hopefully this will provide more information regarding the source of the failure to shutdown problem. Mark On Sun, Oct 15, 2023 at 8:21 AM Rodney Baker <rodney.baker@outlook.com.au> wrote:
On Sunday, 15 October 2023 2:30:54 AM ACDT David C. Rankin wrote:
On 10/13/23 14:45, Andrei Borzenkov wrote:
On 11.10.2023 03:38, Mark Misulich wrote:
I had a shutdown attempt of the computer this afternoon where it didn't shut down, and hung on the shutdown with the line after the localhost login displayed and remained displayed so I could write it down. Here is what I saw: localhost login: [ 9211, 806671] [ T1] reboot; Power down
That is really the last message before kernel turns off the system. It sounds like kernel issue.
I'm curious. I plymouth is installed, even with plymouth.enable=0, is there any part of a plymouth library loaded by the boot-loader or kernel?
The reason I ask is I've run into the shutdown hang ever since Leap 42.X on this HP laptop, but uninstalling plymouth completely has cured the issue 100% of the time.
My thought, may be wrong, but if the boot-loader or loads part of plymouth, even though the kernel parameter plymouth.enable=0 is set, could whatever address holds the part of plymouth be responsible for the hang?
That's the only thing I can think of that would make any difference between using "plymouth.enable=0", or "rpm -e plymouth-dracut ... ..."
Thankfully this is a laptop, because unreliable shutdown/reboot kills the ability to remote admin a box quicker than anything else...
I see the same thing on my desktop running an ASUS Prime X299-A motherboard with Intel Core i7-7820X. On shutdown it sometimes stops showing 05 on the q- code display indicating that it's stopped in S5 sleep state instead of shutting down.
Initially I thought it had something to do with a device on the USB bus but I've now seen it happen even without that suspected problem device plugged in.
The annoying thing is that it **sometimes** shuts down correctly, but other times requires me to push and hold the power button to shut it down once it stops in S5 Sleep.
Regards, Rodney.
-- ========================================================================================================== Rodney Baker rodney.baker@outlook.com.au ==========================================================================================================