Repository navigation
Conversation
xaionaro
force-pushed
the
github-actions
branch
2 times, most recently
from
October 15, 2026 10:44
74c75b6 to
c79a963
Compare
Nothing runs the checks of the Makefile on a change before it lands. Add a workflow that runs them on ubuntu-24.04 for every pull request and every push to main: - lint-and-test: make setup, then make -k -o qemu, which runs every check but the QEMU boots. - qemu: the packages of tests/qemu/ubuntu-packages, the udev rule that GitHub documents to open /dev/kvm to the runner user, then make qemu. The kernel RPMs and DriverSample.efi are cached under a key of the files that pin them and saved only when the job passes, as run.sh reuses DriverSample.efi unchecked. The logs of failed boots are uploaded. The actions are pinned to commit SHAs, and the token can only read the repository. Signed-off-by: Dmitrii Okunev <xaionaro@meta.com>
xaionaro
force-pushed
the
github-actions
branch
from
October 15, 2026 11:12
c79a963 to
4f34ce0
Compare
On some GitHub runners every boot failed its health checks although all tests passed: run.sh boots the guest with -cpu host, and on Intel hosts that hide IBRS from the guest, Linux 6.17 logs a RETBleed error while it boots (arch/x86/kernel/cpu/bugs.c, retbleed_update_mitigation). init failed on any kernel error since boot, so what the host's CPU makes the kernel log decided the result. init now writes a start record to /dev/kmsg and snapshots the taint and the warning count right before the test, then fails on a kernel error after the start record, on a changed taint or warning count, or when the start record is gone. The warning count catches a WARN whose taint bit the boot already set. The values are read with cat, as busybox read gets only the first digit of a sysctl. run.sh prints the host's CPU model with the tool versions, which the logs of the failing runner lacked. Signed-off-by: Dmitrii Okunev <xaionaro@meta.com>
make lint test did not work, as the Makefile had no lint target, and the QEMU tests were called qemu, so CI ran make -k -o qemu to run every check but the boots. Name the checks after what they do: - lint runs fmt, clippy-x86_64, clippy-aarch64, clippy-uefi and doc. qemu-lint goes: the root cargo fmt --check already covers the QEMU test crates, as they are workspace members; doc now documents the whole workspace, publish-hiidb included; and the UEFI clippy of publish-hiidb becomes clippy-uefi. - test-e2e-qemu, formerly qemu, boots the scenarios; it no longer runs the lints first, as test does not either. CI runs make -k lint test and make test-e2e-qemu, in jobs of those names. Signed-off-by: Dmitrii Okunev <xaionaro@meta.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Just enabling the existing auto-tests in the CI via GitHub Actions.