Tumbleweed 5.7.1 error loop within boot, can't invoke Ctrl-D (or anything else).

English Install/Boot/Login
1

Unable to interrupt. It is able to read /boot (grub2) initrd, vmlinuz and map. Video comes up, but initialization does not proceed. On switch to text console, I see more than 100 errors repeated each time:

“memcg_kmem_cache_create_func” (Many of these per line, about 120 such messaqes in total.)

each “batch” of errors followed by unique message “A start job is running for Udev Wait for Complete Device Initialization” with incrementing time. After several minutes, I give up and power off.

Today’s 5.7.1 Recover CD is able to inmitialize and run OK. And, when running the recovery CD, Within the NVME media/run/device/boot directory, the previous 5.6 boot entries have all been erased! (No vmlinux, no map, no initrd). So there seems a separate and disastrous bug within the ‘erase old kernels’ process, it left previous to 5.7.1. The grub2 choice remains, but can’t find a ‘vmlinuz’ to load if I select the 5.6-xx entry.

I will try to edit and simplify boot parameters for 5…7.1 (there are intel-orienented and multi-VM and other CPU bug security workarounds specified as ‘disabled’ on the command line. 3600 Zen, non-x CPU. Over here (my wife’s), 5.6.14 is still running OK, older AMD architecture.

2

There a factory mailing list thread about this:

Bootloop with Tumbleweed snapshot 20200614

3

YES! I went back to BOIS and disabled WIFI (which I do not even use), now boots to runlevel 5 quickly and correctly.

4

I guess that’s a pretty good confirmation that it is the iwlwifi module causing the problems.

Thanks for reporting back.

5

I’m closing this thread as SOLVED.