NixOS doesn't have a traditional FHS where echo is available at
`/bin/echo`. Instead, we must rely on the PATH for any
distribution-managed utilities. Reverses
https://github.com/osresearch/heads/issues/106.
Before, the configure script sourced these from the system FHS
(/usr/include/libusb-1.0). The build failed on my NixOS build machine,
which doesn't store dependencies in a traditional FHS. And this is the
correct approach for reproducible builds.
patches/coreboot-4.15: remove patches for coreboot 4.15
No boards depend on it and is affected by CVE-2022-29264
Signed-off-by: Daniel Pineda <daniel.pineda@puri.sm>
This makes configs much less dependent on directory layout.
As of this commit the following variables are supported:
* @BOARD_BUILD_DIR@ - absolute path under build/
* @BLOB_DIR@ - absolute path to blobs/
Signed-off-by: Sergii Dmytruk <sergii.dmytruk@3mdeb.com>
-Makefile: remove local gawk and make version compare and local build
-modules: remove gawk and make
-patches: remove make
local make was added to build 4.2.1 on OSes that were having older version. It was then patched to be built on OSes having newer buildstack.
local gawk was added when GPG toolstack was older then libgpg-error 1.37. GPG toolstack was then upgraded, but local gawk stayed.
Removing those permits better parallelization and of builds and reduces CircleCI (and higher cores systems) to have race conditions and stalled builds
Testing points:
- None here. Board who exported "CONFIG_USB_KEYBOARD=y" have it packed under their initrd, but there is no logic loading the module yet.
- zlib 1.2.12 release is not respecting cross compiling. 1.2.11 disappeared from servers: taking another archive link, same hash.
- busybox 1.32.0 was not patched with 1.28.0 patch. Renaming patch so that its applied in fresh builds.
Older coreboot versions (pre-4.11) required IASL to be built separate
from the main toolchain (crossgcc), but that is no longer the case,
and doing so causes random failures from trying to build IASL as
part of the toolchain and separately, in parallel, each using
-j$(CPUS) threads.
Test: build any board using coreboot 4.15 under Debian 11, observe
no random failures from building the toolchain due to false positives
for a missing depencency .
Signed-off-by: Matt DeVillier <matt.devillier@puri.sm>
Update hash for coreboot module, coreboot-blobs.
Adjust extra flags to address SNB/IVB build issue.
Signed-off-by: Matt DeVillier <matt.devillier@puri.sm>
CircleCI: We currently drop coreboot 4.11 builds.
- There is a file missing in the builds. Not sure why/how this is happening
src/soc/intel/fsp_broadwell_de/romstage/romstage.c:41:10: fatal error: build.h: No such file or directory
Example:https://app.circleci.com/pipelines/github/tlaurion/heads/877/workflows/7d0248d2-459c-42ad-b741-8fd56a75d527/jobs/2487
- kgpe-d16_workstation building for all GPUs is unfortunately taking too much time to build (40 minutes).
- Not sure why, but it seems that the kernel build paralellization is not working for 4.11 while it works for 4.13
Makefile: Uncomment MAKE_JOBS which passes the number of jobs to numbers cores by default and --max-load of 16
CircleCI: Remove CPUS statement to use Makefile default
modules/newt: force build with one make job, otherwise there is a race condition in module which fails randomly expecting build modules. (TODO: FIX)
Interestingly, building all coreboot 4.13 boards is happening on a clean commit just above 1h limit.
More details:
- CircleCI changed job build time to a maximum of 1h each.
- CircleCI now permits parallelization of 30 jobs
- 6000 build minutes a month.
- Still waiting for osresearch/heads CircleCI project to be unlocked (currently not recognized as open source project?!)
Readd https://github.com/osresearch/heads/pull/984 without cache
Add kgpe-d16 musl-cross target prior of having kgpe-d16 depend on musl-cross target (To try to have musl-cross step successfull under 1h CircleCI new limit)
CircleCI: add a subcommand that can follow a target (to build musl-cross-make now and coreboot version specific musl-cross later)
Output of hashes is now optional
29/11/2021 CircleCI public information available states parallelization of up to 30 jobs at a time. Let's play
- We first build heads musl-cross-make and persist (passing musl-cross-make into next job)
- We then build per coreboot version board with coreboot make statement only and persist (passing musl-cross-make + coreboot's musl-cross buildstack)
- We then build per coreboot version board (reusing past build musl-cross-make and coreboot's version musl-cross buildstack)
Remove 4.11 boards for the moment to test only build time and parallelization
Update version, download hash, patch filename.
Fixes some IOMMU-related issues on Librem Mini v1/v2, L14
Signed-off-by: Matt DeVillier <matt.devillier@puri.sm>
* Flashrom was being fetched with git and was always using `master`
* No patches were being applied (i.e. `0100-enable-kgpe-d16.patch` was being ignored).
- update module hash and blobs hash
- drop patches no longer needed; migrate those that remain
- adjust Librem Mini/Mini v2 board configs
Signed-off-by: Matt DeVillier <matt.devillier@puri.sm>
