Device: Lenovo Thinkpad 11e Chromebook Gen 3 (board name Ultima)
OS: Currently none, attempting Linux
Firmware Type: UEFI Full ROM
Firmware Version: MrChromebox-2603.0 (03/29/2026)
Internal storage type: 16gb eMMC
Summary of the Issue
I have two of these devices that simply refuse to install Linux. I have tried Mint, Debian 13, WattOS (Debian 12 w/LXDE) and KolibriOS (not Linux) no avail. On Debian/WattOS, the installer hangs in a certain place and refuses to go any further. Toggling the log shows nothing. On Mint, it throws a crapton of buffer I/O errors and then the installer gives a “Failed to install” error and closes. It’s definitely something with either Coreboot or Linux itself not liking the eMMC chip for whatever reason.
Steps to Reproduce
Remove the WP screw and install full UEFI firmware with Coreboot.
Attempt to install Linux to the internal eMMC chip.
Expected Behavior
It should install the OS. I’ve done this successfully on Gen 1 Chromebooks and other Chromebooks numerous times.
EDIT: The only way I can get them working is by setting sdhci debug_quirks=0x40 and 0x60. It unfortunately makes your system rather slow. You can read the thread regarding this fix here.
someone was supposed to send me an ULTIMA to test/debug but it never happened. I’ve reached the limit of what I can do with remote debugging on a board without proper serial debug output
This seems to be fixed as of Linux Kernel 7. If for some reason you want to use older kernels, you’ll need to disable DMA and ADMA. It might also be possible to port some headers or the kernel itself from the system’s native ChromeOS, but I can’t try as my system’s eMMC is corrupted. I also don’t know if those things are public (probably not) but it might be possible to copy it to a USB drive via Developer Mode and then install the packages once Linux is installed (with DMA/ADMA disabled) and then re-enabling it. Tiny11 or some other debloated Windows might be an option too. I feel like other Braswell devices should be tested as Chromebook motherboards are usually pretty standardized and whatever causes this issue could very well be present on contemporary devices. I’ll try to get a hold of one whenever I can. Huge thanks to MrChromebox for offering assistance and getting to the bottom of this, and of course, the project that this is all based on as well.