Conversation
|
@imlk0 ，您好，您的请求已接收，请耐心等待结果。 |
|
@imlk0 ，您好，未检测到有镜像需要构建，如需重新检测请评论 /start 。 |
|
@imlk0 ，您好，您的请求已接收，请耐心等待结果。 |
|
@imlk0 ，您好，未检测到有镜像需要构建，如需重新检测请评论 /start 。 |
The 24-job test-convert matrix uses the OTP key provider, which recreates the delta every boot, so it never exercises real disk-persist persistence or the dm-integrity AEAD path. As a result issue #140 (offline resize2fs runs before fsck; surfaces on the SECOND boot of a persistent disk) and #141 (disk-persist + integrity=true first-boot dm-snapshot creation fails: only 4 KiB of the COW header is initialized while the persistent snapshot reads a 16-sector/8 KiB chunk, so sector 8 fails the integrity AEAD check) are both invisible to CI. Add a dedicated test-convert-persist job (UKI; +4 jobs = 2 distros x 2 scenarios: fsck-140 / fsck-141) and extend test-convert.sh with --delta-key <otp|stable>, --integrity <true|false>, --second-boot and --skip-enhance, plus a direct-qemu two-boot path. The qemus entrypoint recreates a fresh throwaway overlay per boot and cannot reproduce a second-boot failure, so the new path boots the SAME persistent qcow2 overlay (relative backing filename so qemu resolves it inside the container) twice. The stable exec key makes the delta persist across boots so the second boot sees the first boot's updated s_mtime. Both new tests reproduce the reported failures on unfixed v0.9.2 (INTEGRITY AEAD ERROR sector 8; resize2fs "Please run e2fsck -f") and are expected to go green once the fix for #140/#141 lands; merge alongside that fix. Assisted-by: Claude:glm-5.2 Signed-off-by: Kun Lai <laikun@linux.alibaba.com>
|
@imlk0 ，您好，您的请求已接收，请耐心等待结果。 |
|
@imlk0 ，您好，未检测到有镜像需要构建，如需重新检测请评论 /start 。 |
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
2 participants
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.
What & why
Adds CI regression coverage for #140 and #141 — the two disk-persist failure modes the existing 24-job
test-convertmatrix cannot see, because it uses the OTP key provider, which recreates the delta every boot and so never exercises real persistence or the dm-integrity AEAD path.resize2fsruns beforee2fsck; surfaces only on the second boot of a persistent disk. The first boot mounts the rootfs read-write (updatings_mtimepasts_lastcheckvia the persistent delta); the next boot'sresize2fs 1.46.0then rejects withPlease run 'e2fsck -f /dev/mapper/rootfs' first.integrity=truefails on the first boot — only 4 KiB of the COW header is initialized while the persistent snapshot reads a 16-sector / 8 KiB chunk, so sector 8 (the first sector beyond the initialized 4 KiB) fails the integrity AEAD check.Changes
tests/test-convert.sh: new--delta-key <otp|stable>,--integrity <true|false>,--second-boot,--skip-enhanceoptions;create_test_configemits a stable exec-key delta provider (same passphrase as rootfs) so the delta persists across boots; newtest_qemu_boot_direct+test_qemu_boot_persist_second. The qemus entrypoint recreates a fresh throwaway overlay per boot and cannot reproduce a second-boot failure, so the new path boots the same persistent qcow2 overlay (relative backing filename so qemu resolves it inside the container) twice.Makefile:run-convert-test-casepasses throughDELTA_KEY / INTEGRITY / SECOND_BOOT / SKIP_ENHANCE..github/workflows/build-rpm.yml: newtest-convert-persistjob — UKI, +4 jobs (2 distros × 2 scenarios:fsck-140,fsck-141). Zero changes to the existing 24-job matrix.Verification (on unfixed v0.9.2)
Both new tests reproduce the reported failures:
fsck-141device-mapper: crypt: dm-6: INTEGRITY AEAD ERROR, sector 8+0: Failed to create dm-snapshot device→ emergency modefsck-1400: Failed to resize ext4 filesystem+resize2fs 1.46.0+Please run 'e2fsck -f /dev/mapper/rootfs' first.→ emergency modecargo fmt --checkOK, shellcheck clean, YAML valid, +4 jobs confirmed via matrix parse.Merge ordering
These regression tests are expected to fail on this branch until the fix for #140/#141 lands (the repro is encoded as the failing test — TDD red). Merge alongside that fix; the fix flips these two scenarios from red to green.
Refs: #140, #141.