Skip to content

feat(ci): Enable MI455 tests as non-blocking - #8407

Merged
geomin12 merged 11 commits into
developfrom
users/geomin12/mi455-test-enablement
Jul 15, 2026
Merged

feat(ci): Enable MI455 tests as non-blocking#8407
geomin12 merged 11 commits into
developfrom
users/geomin12/mi455-test-enablement

Conversation

@geomin12

@geomin12 geomin12 commented Jul 10, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Enable MI455 tests for gfx125X-dcgpu builds by removing the hardcoded disable flag.

Added a custom build and test step for mi455, so builds are blocking but tests are not (due to limited capacity)

@github-actions github-actions Bot added the github actions Pull requests that update GitHub Actions code label Jul 10, 2026
@therock-pr-bot

therock-pr-bot Bot commented Jul 10, 2026 •
edited
Loading

Copy link
Copy Markdown

❌ PR Check — Action Required

Check Status Details
🌿 Branch Name ✅ Pass —
📝 PR Title/Description ❌ Fail Error: PR description must reference a JIRA ID, ISSUE ID, or a GitHub closing keyword.
Expected: include a JIRA ID / ISSUE ID line (separator : or -, or omitted; value may be a JIRA key, a number with/without #, or a link), OR a closing keyword + issue reference. Accepted examples:
• JIRA ID : TESTAUTO-6039
• JIRA ID - #330
• JIRA ID #330
• ISSUE ID : TESTUTO-3334
• ISSUE ID #3334
• ISSUE ID - TESTAUTO-3433
• ISSUE ID : https://github.com/<org_name>/<repo_name>/issues/1234
• Closes #10
• Fixes octo-org/octo-repo#100
• Resolves: #123
• #123
• https://github.com/<org_name>/<repo_name>/issues/123
Current: no valid JIRA/ISSUE/closing-keyword reference found
⛔ Forbidden Files ✅ Pass —
🧪 Unit Test ❌ Fail Error: Source/code files changed without an accompanying unit test.
Expected: add at least one test file named like test_<name>.py / test_<name>.cpp (or <name>_test.*).
Current: code file(s) changed: .github/scripts/therock_configure_ci.py; no test file found
🔎 pre-commit ⏳ Pending ⏳ Still running…
🚫 Draft PR 🔜 To Be Enabled —
🚩 Feature Flag 🔜 To Be Enabled —
📊 Code Coverage 🔜 To Be Enabled —

⚠️ 2 policy check(s) failed. Please address the issues above before this PR can be Reviewed.

🚫 Please fix the failed policies

  • ❌ PR Title/Description
  • ❌ Unit Test

The Not ready to Review label was added to this PR. Once all policies pass, the label is removed automatically.

📖 Need help? See the Policy FAQ for details on every check and how to fix failures.

@therock-pr-bot

therock-pr-bot Bot commented Jul 10, 2026 •
edited
Loading

Copy link
Copy Markdown

🚫 Please fix the failed policies before requesting reviews.

The following policy checks failed:

  • ❌ PR Title/Description
  • ❌ Unit Test

The Not ready to Review label has been added to this PR.
Once all policies pass, the label will be removed automatically.

@geomin12
geomin12 force-pushed the users/geomin12/mi455-test-enablement branch 2 times, most recently from 84994ba to 76391c4 Compare July 10, 2026 16:01
@geomin12
geomin12 force-pushed the users/geomin12/mi455-test-enablement branch 4 times, most recently from c6b7a44 to 08aead2 Compare July 10, 2026 19:24
dan-garvey
dan-garvey approved these changes Jul 10, 2026
View reviewed changes

@dan-garvey dan-garvey left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM as long as tests pass to expectation

@geomin12
geomin12 force-pushed the users/geomin12/mi455-test-enablement branch 3 times, most recently from 9a6cd97 to dbc1566 Compare July 13, 2026 18:34
@geomin12 @claude
ci(therock): Enable MI455 tests with expect_failure flag
ad1cfce 
- Enable MI455 tests for gfx125X-dcgpu builds
- Point therock-ci-config to users/geomin12/expect-fail-mi455 branch
  to test the expect_failure flag for non-blocking MI455 tests
- Add CI workflow change detection to trigger MI455 tests
- Update TheRock commit to 9165e6e864a1643d4804e24935e8eb663633f4a3

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@geomin12
geomin12 force-pushed the users/geomin12/mi455-test-enablement branch from dbc1566 to ad1cfce Compare July 13, 2026 18:35
geomin12 and others added 2 commits July 13, 2026 11:36
@geomin12
Merge branch 'develop' into users/geomin12/mi455-test-enablement
db7c914
@geomin12 @claude
ci: mark gfx125X tests as xfail (expected failure)
96c3dbc 
Add conditional to mark all gfx125X (MI455) tests as expected failures
using continue-on-error. This makes MI455 test failures non-blocking
while we stabilize the test infrastructure.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@geomin12
geomin12 force-pushed the users/geomin12/mi455-test-enablement branch from b75e431 to 96c3dbc Compare July 13, 2026 19:49
geomin12 and others added 2 commits July 13, 2026 20:08
@geomin12 @claude
ci: update TheRock ref to d8e27dff1d (2026-07-13)
4d866ae 
Update TheRock repository reference to latest from develop.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@geomin12 @claude
ci: separate MI455 (gfx125X) build and tests for non-blocking CI
6046436 
Separate MI455 build and test into dedicated jobs to allow PRs to merge
without waiting for MI455 tests to complete or even start running.

Architecture:
- therock-ci-linux: builds and tests gfx94X, gfx950 (blocking)
- therock-build-linux-mi455: builds gfx125X only (blocking, in summary)
- therock-test-linux-mi455: tests gfx125X, depends ONLY on the MI455
  build job (non-blocking, NOT in summary)

This allows MI455 tests to start as soon as the MI455 build completes,
without waiting for other families' tests. PRs can merge once all builds
pass, while MI455 tests continue running or wait in queue.

TEMPORARY: This separation is needed while MI455 machine capacity is
limited. Once we have sufficient MI455 machines and stable test
infrastructure, we can merge this back into the main matrix.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@geomin12
geomin12 force-pushed the users/geomin12/mi455-test-enablement branch from 92f4187 to 6046436 Compare July 14, 2026 16:11
geomin12 and others added 6 commits July 14, 2026 09:17
@geomin12 @claude
ci: run configure_test_matrix on standard GitHub runner
47b8a4b 
The configure_test_matrix job doesn't need GPU access - it just parses
config files. Run it on ubuntu-24.04 instead of the test runner to
avoid GLIBC compatibility issues on specialized GPU runners.

Pass --platform to fetch_test_configurations.py to ensure correct
platform-specific configuration is generated.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@geomin12
Merge branch 'develop' into users/geomin12/mi455-test-enablement
c570a57
@geomin12 @claude
ci: use main branch for therock-ci-config
4219a8e 
Switch therock-ci-config ref from test branch back to main.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@geomin12 @claude
ci: remove gfx125X xfail since MI455 tests are non-blocking
b856c43 
The gfx125X xfail condition is no longer needed since MI455 tests now
run in a separate non-blocking job (therock-test-linux-mi455) that is
excluded from therock_ci_summary.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
@geomin12
Merge branch 'develop' into users/geomin12/mi455-test-enablement
07cecaa
@geomin12
Merge branch 'develop' into users/geomin12/mi455-test-enablement
0282dfe
@geomin12
geomin12 marked this pull request as ready for review July 14, 2026 22:49
@geomin12
geomin12 requested a review from a team as a code owner July 14, 2026 22:49
@geomin12
geomin12 merged commit 0522418 into develop Jul 15, 2026
31 of 38 checks passed
@geomin12
geomin12 deleted the users/geomin12/mi455-test-enablement branch July 15, 2026 00:54
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@dan-garvey dan-garvey dan-garvey approved these changes

Assignees

No one assigned

Labels

github actions Pull requests that update GitHub Actions code Not ready to Review organization: ROCm

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

2 participants

@geomin12 @dan-garvey