Description of the problem: The system shows a trace during boot and hangs. Kernel version: 5.0.0-0.rc4.git0, 5.0.0-0.rc6.git0 and 5.0.0-0.rc6.git1. Kernel 5.0.0-0.rc3.git0 boots ok. Version 4.20.3-200 also worked ok back in Fedora 29. Reproducible: Always. No extra modules or proprietary drivers. journalctl --no-hostname -k --boot=-1 (or -2 or -3) only shows previous successful boots. Please instruct me on how to gather information about a failed boot attempt (other than taking a picture with my phone). My hardware is a Chuwi Lapbook SE Chinese notebook. Celeron N4100, Intel chipset and graphics. BTW: seamless boot works. If I leave rhgb and quiet params, all I see is the Chuwi logo forever. I have to remove rhgb to actually see the trace.
Created attachment 1534998 [details] Screenshot
I can guess the problem is related to the mmc driver, and I have to add that Fedora is installed on the integrated 32 GB eMMC on this notebook.
If you try the debug version of the 4.20 kernel do you see the same issue? I ask because I left debugging code on a couple of times recently and that crash is coming from debug code https://koji.fedoraproject.org/koji/packageinfo?packageID=8 you can install kernel-debug from here
Yes, I tried kernel 4.20 debug and it crashed.
Okay today's -rc7 build should have debugging enabled. Please try that. I also started a thread with the mmc maintainers upstream.
https://koji.fedoraproject.org/koji/taskinfo?taskID=32976235 can you test this scratch build when it finishes? This has debugging enabled and a potential fix from upstream.
rc7 git 2.1 doesn't work, but rc7 git 3.1 does!
So this is fixed in FC30. Should I close this bug?
Yes I'll close it, the fix should have been picked up.