Repository navigation
tests: reference the serve-27 xfail row's tracker as rocm-cli#1058 - #1065
Conversation
The row's bug field was a bare #1058, unlike the other rows' tracker IDs and ambiguous in a report. Use the same form as the comments that name the issue. Signed-off-by: Tomas Saaristola <tsaarist@amd.com>
Security scan found no critical or high findings on lines changed by this pull requestView the security scan run for complete reports. “New” means the finding is located in a changed file on an added or modified line; it is not a base-versus-head findings delta. |
juhovainio
left a comment
There was a problem hiding this comment.
One-line metadata change in expectations.toml (serve-27 xfail row: #1058 -> rocm-cli#1058). #1058 is open and matches the row's reason text (AIM stop failure on Strix Halo Ubuntu). No code or behavior change, no scenario needed, diff leak scan clean, CI has no failures so far (GPU lanes pending, consistent with the runner outage).
Not verified by me: the PR's claim that nothing parses bug.
nowycondro
left a comment
There was a problem hiding this comment.
Review — no blocking findings
Full review of the whole change (a one-line, single-file diff).
Follow-up to #1018; brings the bug field of the serve-aim-lifecycle-parity xfail row in expectations.toml from a bare #1058 to rocm-cli#1058, matching the form already used for this same issue in nightly.yml, docs/testing.md, and model_serving.feature.
Blocking
None
Non-blocking
None
Decisions for the author
None
Positive signals
- The change is traced back to the actual consumer of the
bugfield (expectation.rs, where it's a plainStringserialized as-is into the report) rather than assumed safe, which is exactly the right level of care for a one-line data-value edit in a file that's otherwise all data.
Deployment notes
None
What this covered
The full diff (b109db12…c4434b7, the single changed file tests/e2e-cucumber/expectations.toml), the full file's header/grammar comments, every reference to the bug field in tests/e2e-cucumber/src/expectation.rs, and a repo-wide search confirming rocm-cli#1058 is already the established form elsewhere for this exact issue. Also checked commit history/blame around the changed row and the file's own prior conventions for tracker references — one apparent precedent (a comment elsewhere in the file writing out ROCm/rocm-cli#897) turned out to be prose inside a comment, not a parallel bug = field value, so it isn't a comparable inconsistency and isn't reported as a finding.
All passes ran (code, agent-instruction adherence, history, prior-changes, code-comments). CI on the PR head shows all completed checks green; a few self-hosted E2E jobs were still queued/in-progress at review time, which is unrelated to this change (test metadata only, no behavior change). PR description and the linked issue (#1058, open, describing exactly the serve-aim-lifecycle-parity stop-not-confirmed symptom this row documents) both match the change accurately. Not reconciled against any discussion on the PR itself.
|
Thanks both. On the one point juhovainio did not verify: |
Follow-up to #1018 (review point from juhovainio).
The
serve-aim-lifecycle-parityxfail row added in #1018 hadbug = "#1058". The other rows carry a tracker reference in that field, and a bare#1058is ambiguous in the E2E report, which prints it on its own. This changes it torocm-cli#1058, the same form the comments innightly.yml,docs/testing.mdandmodel_serving.featurealready use. Thebugfield is a free-form string that nothing parses, so behaviour is unchanged.CI-only/test-metadata change; no Gherkin scenario is affected, so none is added.
Verified:
cargo test -p e2e-cucumber --libpasses (343).