Intel Corporation CometLake-S GT2 [UHD Graphics 630] Boot Failure
Hi,
I have an ASUS ROG STRIX Z390-E GAMING that's a few years old now. When I first got the machine I managed to install Debian 10 (buster) on it with some fussing. I remember having to build the most recent kernel (v5.4.8) at the time and install some newer X drivers though I can't remember exactly what they were. After that the machine has largely worked fine, though occasionally hangs during boot.
Since then I've periodically tried updating the kernel to something more recent but every time I try the machine will not boot. It varies how: often it hangs early but usually it hangs when the desktop is started. Sometimes I can see and move a mouse before it hangs. When it hangs I see a black screen, and the computer is dead dead. If I ssh in before starting the desktop, that session hangs too. So getting debug information is very hard. I can boot newer kernels as just fine as long as the desktop is not started.
A couple years back I successfully upgraded to Debian 11 (bullseye) hoping it would solve the issue. This mostly worked, but I had to keep that old custom built v5.4.8 kernel -- all the X packages are now stock Debian but the stock kernel would not boot. I've again periodically tried a few recent kernels builds with no success. Today I tried the current drm-tip and that kernel hangs consistently quite early in the boot process. I've also tried the latest Debian 12 (bookworm) Live USB image which hangs after showing the mouse on a black screen for a short time; as well as a Ubuntu 22.04.3 Live CD which hangs fairly early in its process.
Over the years, in attempting to solve this issue I've updated the bios, the Intel ME firmware, reset the bios defaults and played with a handful of bios options. None of these actions helped.
When using the working v5.4.8 kernel I also see some concerning errors in the dmesg, despite the machine fully booted and everything working fine.
[ 1.285824] i915 0000:00:02.0: VT-d active for gfx access
[ 1.285880] i915 0000:00:02.0: vgaarb: deactivate vga console
[ 1.298673] i915 0000:00:02.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=io+mem:owns=io+mem
[ 1.298938] [drm] Finished loading DMC firmware i915/kbl_dmc_ver1_04.bin (v1.4)
[ 1.330487] [drm:intel_dp_aux_xfer [i915]] *ERROR* dp aux hw did not signal timeout!
[ 1.802524] i915 0000:00:02.0: Failed to idle engines, declaring wedged!
[ 1.804380] [drm:gen8_reset_engines [i915]] *ERROR* rcs'0 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 1.805207] [drm:gen8_reset_engines [i915]] *ERROR* rcs'0 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
[ 1.846551] i915 0000:00:02.0: Failed to initialize GPU, declaring it wedged!
[ 1.846562] i915 0000:00:02.0: Please file a bug at https://bugs.freedesktop.org/enter_bug.cgi?product=DRI against DRM/Intel providing the dmesg log by booting with drm.debug=0xf
[ 1.876331] [drm] Initialized i915 1.6.0 20190822 for 0000:00:02.0 on minor 0
[ 1.941845] fbcon: i915drmfb (fb0) is primary device
[ 2.021476] i915 0000:00:02.0: fb0: i915drmfb frame buffer device
The full dmesg log for this boot is here: dmesg.log
At this point I'd like to find a solution to this so I can confidently update to bookworm and I'd also really like to get this machine on a modern kernel.
Thanks,
Logan