[SOLVED] Debian 10 amd-64 UEFI hangs on shutdown

[debbiethekiwi]

First attempt at installation hung waiting for a network, shouldn't we be able to install an OS offline? The installer couldn't see my intel iwlwifi, so I downloaded the debian-live-10.0.0-amd64-kde+nonfree.iso. This time it was happy, and I pottered around exploring the system, then at shutdown time, the cursor vanishes, the screen goes to a blank blue-grey Debian background, and sits there forever. OK for more than 5 minutes, when I force shutdown holding the power key.

There are unanswered question on this here and here. I've looked at
some instructions, but the shutdown/debug script produces no output, and I'm not sure how to boot in debug mode.

Sometimes a few lines of debug log appear on screen, sometimes not.
I've checked the d/l iso with sha256sum, its OK. I made the bootable USB stick with

Code: Select all

dd status=progress bs=1M if=/path/to/debian-live-10.0.0-amd64-kde+nonfree.iso of=/dev/sda

then checked its integrity is OK with

Code: Select all

dd if=/dev/sda bs=4096 count=$(($(stat -c '%s' debian-live-10.0.0-amd64-kde+nonfree.iso) / 4096)) | sha256sum

My system is a Huawei Matebook13, and I've more detail on hardware and messages seen on screen, but this is long enough for a first post...

[sunrat]

After rebooting from a hung shutdown, run

Code: Select all

journalctl -b -1 -p 3

It may show useful error messages.

[debbiethekiwi]

Result:

Code: Select all

Specifying boot ID or boot offset has no effect, no persistent journal was found.

Confession: this is a multiboot system, as I'm looking for the best distro after ending a 30 year love affair with Apple/Macintosh. I've been about four days struggling with Debian. The last time it hung I had installed inxi with apper, then ran $ inxi -Fxz. The result appeared in the konsole window and the cursor froze. I stared in disbelief for abt 25 seconds then lines slowly appeared on the bottom of the konsole referring to hard lockup on CPU2, soft lockup on CPU1. Similar messages had appeared in some previous hangs, for CPUs 4&5 or 6&7.
I rebooted into Linux Mint and had a good cleanout of MX-Linux (including its boot folder from the ESP), and lo behold Debian behaved itself for about half a dozen shutdowns and reboots. Then just as randomly it hung again leaving this on screen:

Code: Select all

[    3.787815] nouveau 0000:01:00.0: DRM: Pointer to TMDS table invalid

Debian GNU/Linux 10 debbie tty1

debbie login: [ 3371.568217] rcu: INFO: rcu_sched detected stalls on CPUs/tasks:
[ 3371.568257] rcu: o7-...0: (1 GPs behind) idle=d6a/1/0x4000000000000000 softirq=183153/183154 fqs=2591
[ 3371.568295] rcu: o(detected by 0, t=5252 jiffies, g=781737, q=2467)
[ 3371.569320] NMI watchdog: Watchdog detected hard LOCKUP on cpu 7
[ 3434.583059] rcu: INFO: rcu_sched detected stalls on CPUs/tasks
[ 3434.583097] rcu: o7-...0: (1GPs behind) idle=d6a/1/0x4000000000000000 softirq=183153/183154 fqs-10430
[ 3434.583135] rcu: o(detected by 5, t=21007 jiffies, g=781737, q=3003)

Suspecting a faulty install, I booted from the live USB stick. On shutdown it too barfed:

Code: Select all

Debian GNU/Linux 10 debian tty1

debian login: user (automatic login)

Linux debian 4.19.0-5-amd64 #1 SMP Debian 4.19.37-5 (2019-06-19) x86_64

The programs included with the Debian GNU/Linux system are free software;
the exact distribution terms for each program are described in the
individual files in /usr/share/doc/*/copyright.

Debian GNU/Linux comes with ABSOLUTELY NO WARRANTY, to the extent
permitted by applicable law.
[   13.955468] iwlwifi 0000:00:14.3: BIOS contains WGDS but no WRDS
[   user@debian:~$ [   14.445492] iwlwifi 0000:00:14.3: BIOS contains WGDS but no WRDS
[  168.163948] rcu: INFO: rcu_sched detected stalls on CPUs/tasks:
[  168.167636] rcu: o4-...0: (13 ticks this GP) idle=fa6/1/0x4000000000000000 softirq=9689/9691 fqs=<truncated off photo>
[  168.170257] rcu: o
[  168.170258] rcu: INFO: rcu_sched detected expedited stalls on CPUs/tasks: ( 4-...
[  168.173446] rcu: blocking rcu_node structures:
[  168.174914] MM1 watchdog: Watchdog detected hard LOCKUP on cpu 4
[  168.180727] rcu: rcu_sched kthread starved for 866 jiffies! g25829 f0x0 RCU_GP_WAIT_FQS(5) ->stat <truncated>
[  168.185530] rcu: RCU grace-period kthread stack dump:

Just an observation: MX-Linux had this behaviour when I tried it with no network, it plainly stated on hanging that it couldn't shutdown NetworkManager, so now I keep my network enabled when testing sytems. Debian is the only hangout now...

[debbiethekiwi]

I'm putting [Solved} on the header, but the problem remains unknown. I've successfully installed Debian 10.1 stable, with wifi drivers from the non-free firmware collection, and now I can logout, reboot or shutdown repeatedly without hangs or crashes. This is the newest computer I've ever had, and it seems strange to have to wait for the software to catch up...

[dilberts_left_nut]

This is the newest computer I've ever had

^^ This.
Got the processor microcode installed?
Later kernel than stable may help.

[debbiethekiwi]

microcode update didn't make any noticeable difference, Debian 10.0 or 10.1.
I was looking for a newer kernel, but didn't know how new I could go:I'm currently running 4.15.0.20 on Linux-Mint on this machine, and Mint has available from the series 4.15.x, 4.19.x, and 5.0.x. I've done a little bit of hacking, but I don't think I'm yet up to building a kernel