Just updated four machines with the latest -current updates, including the 6.9.0 kernel. Three work fine, but one now has broken graphics. X refuses to start. SDDM gets stuck in a crash loop. Both startx and startkwayland produce an instant server crash. startxfce4 works fine!

The computer is a locally re-badged Clevo N141ZU. lspci reports the graphics card to be Intel WhiskeyLake-U GT2 [UHD Graphics 620]

Interestingly, this is the same laptop that caused me a lot of grief trying to install Slackware a while back: https://www.linuxquestions.org/quest...re-4175671086/

I'm not sure if this is related though.

Luckily, I still have kernel 6.6.30 installed as a backup, and it still works perfectly.

I'm also attaching my Xorg.0.log - not that its very helpful! Everything looks OK until: [ 174.860] (II) Server zapped. Shutting down.

Any suggestions gratefully received!

--
Pete

Attached Files

Xorg.0.log (34.7 KB, 12 views)

upon updating last current updates with kernel 6.9.0 Xorg is in some sort of crashing loop. this happens on multiple machines... as reported by the OP 6.6.30 works fine. this is my Xorg.0.log from my desktop running current

https://termbin.com/zkmn

1 members found this post helpful.

It may be useful to peak at dmesg and /var/log/messages and syslog to see any kernel gotchas with amdgpu and drm. Xorg doesn't log kernel errors

I'm also affected. In /var/log/messages I can see the following output for each loop:

This matches the output of dmesg:

There is a known problem with the built 6.9.0 kernel on some intel CPUs pinpointed here:

https://www.linuxquestions.org/quest...ml#post6501916

regards Henrik

1 members found this post helpful.

this is dmesg

https://termbin.com/0i70

this is /var/log/messages

https://termbin.com/gpyz

this is /var/log/syslog

https://termbin.com/9nu1

sorry.... these are running kernel 6.9.0

dmesg

https://termbin.com/uvii

/var/log/messages

https://termbin.com/rrig

/var/log/syslog

https://termbin.com/hulo

I also had the issue with Slackware64-current after the 6.9 kernel upgrade. I pressed e when grub was loading and added mitigations=off as mentioned in https://www.linuxquestions.org/quest...ml#post6501920 and was able to use my machine again. I hope this helps someone else

the mitigations=off worked for me too

Seems to be the current solution.

Thanks to Henrik for pointing out the thread with the solution, and to all others for confirming it works!

This morning's updates contain a revised kernel, which has solved the problem.

Apologies for not responding earlier, but I'm in the UK, and the posts from our American and Canadian friends arrived after I was in bed!

--
Pete

what are the implications of running mitigations=off?
most of that stuff needs physical access anyway dont they?

Theoretically a malicious process can find out what's in the memory of another process on the same processor, but IMO you shouldn't be running executables you don't trust anyway.

I can remember that Firefox was patched to reduce the granularity of its timers so this sort of things couldn't be exploited through javascript from remote websites, so that shouldn't be a concern any more.

I believe it's still a concern for shared hosting companies who use VMs or containerisation for isolation between different customers on the same hardware. Probably not a big deal for your average desktop user or on dedicated servers.

4 members found this post helpful.