Libreboot 20221214 released!

Leah Rowe

14 December 2022


Return to index

Article published by: Leah Rowe

Date of publication: 14 December 2022

Free as in freedom!

The last Libreboot release, version 20220710, was released on 10 July in 2022. This new release, Libreboot 20221214, is released today on December 14th, 2022. This is intended to be a testing release.

This is marked as a testing release, but most/all boards should be fairly stable.

PLEASE READ THESE INSTRUCTIONS BEFORE INSTALLING, OR YOU MAY BRICK YOUR MACHINE!! - Please click the link and follow the instructions there, before flashing. For posterity, here is the link again.

Build from source

This release was build-tested on Debian 11. Your mileage may vary, with other distros.

Work done since last release

For more detailed change logs, look at the Git repository. This is a summary of changes.

New boards, x86

This list may not be complete, but it is as follows:

GA-G41M-ES2L ROMs

GA-G41M-ES2L ROM images are back, but your mileage may vary. Only the SeaBIOS payload is enabled, on this board.

No lbmk changes were done for this, because the ROMs were simply excluded in the previous release, but this board was not deleted from lbmk.

UPDATE ON 20 December 2022: per many user reports, these ROMs work very well. GA-G41M-ES2L support is therefore stable, for all intents and purposes. This section of the release announcement previously alluding to issues, but those speculations were premature, based on limited prior testing.

New boards, ARM

NOTE: These boards use u-boot payload, instead of depthcharge.

These ARM boards are all Chromebooks.

This work on the Chromebooks and u-boot integration is courtesy of Alper Nebi Yasak (alpernebbi on libreboot IRC).

In these configurations, u-boot is a payload of coreboot.

Removed boards

ASUS KGPE-D16, KCMA-D8 and KFSN4-DRE are removed, because raminit never worked reliably on these boards to begin with and they were barely maintained anymore. Use an older release of Libreboot, if you still want to use these boards.

Known issues

Intel ME firmware missing in ROMs that need it

If you compile ROM images with lbmk directly, the build system automatically fetches ME images from the internet and neuters plus truncated them, using me_cleaner. This downloading is done to avoid distributing them directly in Libreboot, and they get scrubbed by the release build scripts.

To re-insert neutered/truncated ME into your image, look at the guide.

This applies to sandybridge, ivybridge and haswell Intel platforms, e.g. X220, T420, X230, T430, T440p. On older Intel platforms such as GM45+ICH9M (X200, T400, etc) the Intel ME image isn’t needed and Libreboot ships with ME-disabled configuration.

MRC image missing in Haswell ROMs

Ditto ME images. To re-insert, follow the guide.

Internal flashing on 16MB X230 images

The X230 has two ways of upgrading the default 12MB (8MB and 4MB chips) flash to 16MB: remove a bunch of resistors and capacitors on the board (a guide is yet to be written for this) and both flash ICs, and replace SPI1 with a 16MB chip e.g. Winbond W25Q128FVSIG.

The other method: replace just SPI2 which is a 4MB flash, with an 8MB flash e.g. Winbond W25Q64FVSIG. This latter method was tested, and it is made to boot by changing component2density in the IFD to 8MB, but this is non-standard per Intel specifications.

It boots, but then internal flashing still only lets you flash the first 12MB, even though the bootblock is at the end of that 16MB flash and does boot perfectly!

So, 16MB images on X230 are experimental. When these ROM images are built, the required IFD modification is already done in Libreboot’s IFD for this setup.

S3 suspend/resume on Haswell (T440p/W541)

Totally broken. The suspected cause is component2density setting in IFD being wrong: ifdtool reports it as being too big for what’s actually flashed.

No fix has been found yet that doesn’t brick the machine, so this bug is still present in the Libreboot 20221214 release.

The issue is that the MRC cache is in the wrong place in memory, because of that bug in the IFD, but when ifdtool is used to correct component2density, it bricks, which leads to wonder: is ifdtool’s knowledge of these bits on haswell faulty?

When X230 16MB was experimented with, Haswell was also looked at, and what seemed to be the same component density bits were set, also resulting in a brick (it’s why only 12MB images are available for haswell in libreboot).

Chromebook boards mostly untested

Alper has tested the gru_kevin ROM images produced by lbmk, but the other images are untested as of this day.

Markdown file for this page: https://libreboot.org/news/libreboot20221214.md

Subscribe to RSS for this site

Site map

This HTML page was generated by the Libreboot Static Site Generator.