Comment 4 for bug 1973167

After grepping dmesg for all lines about the "00:15" devices, I found two lines starting with "DMAR:", which made me think of another machine running debian which had issues with DMAR: devices that were related to the Intel IOMMU.

So I tried the workaround I knew from there:

Edit /etc/default/grub and add "intel_iommu=off" to the GRUB_CMDLINE_LINUX_DEFAULT value, then run "sudo update-grub" to update the grub configuration and reboot.

and voila, intel-lpss initializes within a few milliseconds again:

[ 6.518592] intel-lpss 0000:00:15.0: enabling device (0000 -> 0002)
[ 6.560693] idma64 idma64.0: Found Intel integrated DMA 64-bit
[ 6.569509] intel-lpss 0000:00:15.1: enabling device (0000 -> 0002)
[ 6.569691] idma64 idma64.1: Found Intel integrated DMA 64-bit
[ 6.574164] mei_me 0000:00:16.0: enabling device (0000 -> 0002)

$ cat /proc/cmdline
BOOT_IMAGE=/boot/vmlinuz-4.15.0-177-generic root=UUID=<blabla> ro intel_iommu=off

So some change between 4.15.0-176 and 4.15.0-177 seemingly broke the Intel IOMMU. Question is whether it was a formerly "dormant" BIOS bug that was only unveiled by some change, or whether it is a newly introduced Linux bug that broke Intel IOMMU support...