ci(gfx1250): Add public gfx1250 hipblaslt + tensilelite HW CI workflow by archana-ramalingam · Pull Request #12164 · ROCm/rocm-libraries

archana-ramalingam · 2026-09-16T00:03:48Z

JIRA ID : AIHPBLAS-4510

Motivation

Add per-PR gfx1250 HW test coverage for hipBLASLt and TensileLite

Technical Details

Add public gfx1250 hipblaslt + tensilelite HW test workflow
Scope guard: org-member label + path filter; forks run only when vetted.
Only requires the gfx1250 build to succeed
Enabled Per-path scoping
Triggers: workflow_run (after TheRock Multi-Arch CI) and workflow_dispatch.
Fetches the prebuilt gfx1250 ROCm artifact
Graceful-skip when a build reused a baseline (no artifact); real fetch errors fail.
Pin the gfx1250v0 target, pass--gpu-targets to pytest, and use a 90-min per-test cap.

Test Plan

gfx1250 public CI passes

Test Result

gfx1250 public CI passed: 106216395195

Submission Checklist

Look over the contributing guidelines at https://github.com/ROCm/TheRock/blob/main/GOVERNANCE.md#pull-requests.

Risk level

Low — new opt-in workflow, no change to existing CI; dedicated runner; read-only permissions; fork code gated by an org-member label.

…ite-test-workflow

therock-pr-bot · 2026-09-16T00:10:55Z

✅ All Checks Passed — Ready for Review

Check	Status	Details
📝 PR Description	✅ Pass	—
⛔ Forbidden Files	✅ Pass	—
🧪 Unit Test	✅ Pass	PR does not contain code files — Unit Test auto-passed
🔎 pre-commit	✅ Pass	—
🚫 Draft PR	🔜 To Be Enabled	—
🚩 Feature Flag	🔜 To Be Enabled	—
📊 Code Coverage	🔜 To Be Enabled	—
🤖 therock-pr-bot	✅ Pass	—

🎉 All checks passed! This PR is ready for review.

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

🙋 Wish to Override Policy?

therock-pr-bot · 2026-09-16T00:10:59Z

🎉 All checks passed! This PR is ready for review.

…uites

…se-checkout

… suffix

… success gate

…name GIDs; explicit pipefail

geomin12

thanks, lgtm! we can land it and iterate and see how it goes

…marker, ccache

…ite-test-workflow

…O NOT MERGE)

JIRA ID : AIHPBLAS-4510 ## Motivation - 4 gfx1250 TensileLite `common` configs fail on the current gfx1250 HW. - xfail them so the per-PR and nightly gfx1250 lanes stay green while the gaps are tracked. - The separate gfx942 `test_compile` failure is fixed in #12086 ## Technical Details - Mark the currently-failing gfx1250 configs `xfail-gfx1250` - Fix `findAvailableArchs` (`config_helpers.py`) to strip a trailing version suffix arch — without this the marks never fire on a versioned runner and the configs stay red. - FFM path uses HSA_MODEL_MEMFILE to skip any HW specific `xfail-<arch>`, as these tests failing on HW pass on FFM. - Add 3 tests in `Tensile/Tests/common/test_config_helpers.py` covering the arch-suffix strip and the FFM-gate suppression. ## Test Plan - 3 new common tests in `test_config_helpers.py` - FFM tests pass as expected - Full gfx1250 HW xfail behavior is exercised once #12164 's workflow runs post-merge against a rebuilt artifact. ## Test Result - [x] `pytest Tensile/Tests/common/test_config_helpers.py` → 7 passed (4 existing + 3 new). - [x] FFM tests pass as expected - https://github.com/ROCm/rocm-libraries/pull/12178/checks?check_run_id=105181618286 - gfx1250 HW result pending the post-merge #12164 run. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/TheRock/blob/main/GOVERNANCE.md#pull-requests. ## Risk level - Low — only adds xfail marks to already-failing gfx1250 configs plus a one-line arch-normalization with regression tests; no change to any passing test. `xfail_strict` turns an unexpected pass red, so a fixed config can't stay silently masked. --------- Co-authored-by: Tony Davis <tony.davis@amd.com>

…bug probe

…ite-test-workflow

…(PR#12355 HW test) (DO NOT MERGE)

…250 markers to excluded test (#12355) JIRA ID : AIHPBLAS-4510 ## Motivation - On gfx1250 hardware, spmm_gate now passes, so its xfail-gfx1250 marker is stale (XPASS-strict fails). Remove it. - sgemm_streamk_asm.yaml (multi-arch gfx11/gfx12) isn't tagged gfx1250, so the `-m gfx1250` CI suite never collects it — add the marker. - test_config_helpers.py: tag its framework tests into the gfx1250 suite. - Once spmm_gate is un-xfailed, the xfail-fixture tests #12282 pointed at it no longer hold — remove them. ## Technical Details - Remove the xfail-gfx1250 mark from: - Tensile/Tests/common/sparse/gfx1250/spmm_gate.yaml - Add gfx1250 marker to: - sgemm_streamk_asm.yaml - test_config_helpers.py - Drop the 2 tests tied to spmm_gate xfail fixture ## Test Plan - gfx1250 TensileLite common run — the test should pass, not XPASS. ## Test Result - Will be tested via #12164 only after merge. ## Submission Checklist - [x] Look over the contributing guidelines at https://github.com/ROCm/TheRock/blob/main/GOVERNANCE.md#pull-requests. ## Risk level Low — test-only: removes a stale xfail marker + fragile fixture-coupled tests, adds collection markers. No runtime/library code touched.

…h validation (DO NOT MERGE)

…ite-test-workflow

tony-davis · 2026-09-23T17:40:30Z

Bypassing CI for merge. New opt-in workflow only.

assistant-librarian · 2026-09-25T12:04:13Z

TheRock Submodule Bump Activity

Newest first

2026-09-25 — Included in TheRock via ROCm/TheRock#8472.

Brings the branch up to date with develop and ports what landed there under the gfx1250v0 name onto gfx1250-strict. Tuning logic. ROCm#12336 and ROCm#12271 replaced 4 gfx1250v0 logic files and added 7 (including a new Origami pair). The 4 are dropped here too. The 7 move to gfx1250-strict/, with ScheduleName and ArchitectureName rewritten the same way d05bf20 did for the rest. Every file in the tree now matches its develop counterpart apart from those two names. Test marks. ROCm#12282 skipped three streamk configs on gfx1250v0, and it and ROCm#12355 dropped xfail-gfx1250 from those and spmm_gate now that they pass. They keep skip-gfx1250-strict and lose the xfail. fused_a2a gained skip-gfx1250 on develop, so it also gets skip-gfx1250-strict, per this branch's convention. findAvailableArchs keeps this branch's version. The develop version made gfx1250v0 inherit base gfx1250 marks. That is the opposite of the mirrored marks here, where a strict config says skip-gfx1250. The four gfx1250v0 tests in test_config_helpers.py are ported to pin the strict behaviour instead, including that a base skip-gfx1250 does not fire on a strict target. ClientWriter. ROCm#12569 added clientLibraryFiles (msgpack master lookup) and routed a stepping through computeOutputArchNames, which this branch removes. It auto-merged into TensileCreateLibrary/__init__.py as an import of a function that no longer exists, so Tensile failed to import. The export is dropped. clientLibraryFiles keeps the msgpack fix but loses the remap, since a stepping is now its own subdir and master suffix. Its stepping test now checks that gfx1250-strict is found by name and that the gfx1250 files next to it are not picked up. Not changed: .github/workflows/gfx1250_hipblaslt_tensilelite_test.yml (ROCm#12164, manual dispatch) still passes --gpu-targets gfx1250v0 and selects with -m "gfx1250 and not skip-gfx1250". It needs porting to gfx1250-strict as a follow-up. Co-authored-by: Cursor <cursoragent@cursor.com>

ci(hipblaslt): gfx1250 hipblaslt + tensilelite HW test workflow

a9341d1

archana-ramalingam requested a review from a team as a code owner September 16, 2026 00:03

archana-ramalingam requested review from geomin12 and talumbau September 16, 2026 00:03

github-actions Bot added github actions project: none Does not target any component labels Sep 16, 2026

archana-ramalingam changed the title ~~ci(gfx1250): Add public gfx1250 hipblaslt + tensilelite HW test workflow~~ ci(gfx1250): Add public gfx1250 hipblaslt + tensilelite HW CI workflow Sep 16, 2026

Merge branch 'develop' into users/aramalin/gfx1250-hipblaslt-tensilel…

4db512b

…ite-test-workflow

ci(hipblaslt): per-step 120m timeouts + 45m per-test cap on gfx1250 s…

ec5856c

…uites

assistant-librarian Bot added the organization: ROCm label Sep 16, 2026

geomin12 reviewed Sep 16, 2026

View reviewed changes

ci(hipblaslt): containerize gfx1250 test workflow; scope-first + spar…

f646616

…se-checkout

archana-ramalingam requested a review from geomin12 September 16, 2026 09:51

archana-ramalingam mentioned this pull request Sep 16, 2026

test(tensilelite): xfail failing gfx1250 on HW #12178

Merged

3 tasks done

archana-ramalingam added 2 commits September 16, 2026 10:27

ci(hipblaslt): don't gate on whole multi-arch success; drop step-name…

497281c

… suffix

ci(hipblaslt): scope hipblaslt vs tensilelite by path; drop whole-run…

2539f3a

… success gate

archana-ramalingam commented Sep 16, 2026

View reviewed changes

Comment thread .github/workflows/gfx1250_hipblaslt_tensilelite_test.yml Resolved

ci(hipblaslt): fetch fails closed (red) on any error; scope by path; …

52e5896

…name GIDs; explicit pipefail

geomin12 approved these changes Sep 16, 2026

View reviewed changes

archana-ramalingam requested a review from yash-amd September 17, 2026 05:57

yash-amd approved these changes Sep 17, 2026

View reviewed changes

archana-ramalingam and others added 2 commits September 17, 2026 18:41

ci(gfx1250): HW-test workflow fixes — fetch, libomp ldconfig, target/…

45d5d0f

…marker, ccache

Merge branch 'develop' into users/aramalin/gfx1250-hipblaslt-tensilel…

3d1795f

…ite-test-workflow

archana-ramalingam added a commit that referenced this pull request Sep 17, 2026

test(ci): sync scratch to cleaned #12164 workflow + runid hardcode (D…

fc666d5

…O NOT MERGE)

archana-ramalingam mentioned this pull request Sep 21, 2026

test(tensilelite): remove xfail-gfx1250 from passing tests & add gfx1250 markers to excluded test #12355

Merged

1 task done

archana-ramalingam and others added 2 commits September 21, 2026 10:30

ci(gfx1250): pin gfx1250v0 target, raise per-test cap to 90m, drop de…

b948a78

…bug probe

Merge branch 'develop' into users/aramalin/gfx1250-hipblaslt-tensilel…

689a966

…ite-test-workflow

ci(gfx1250): add -rxX so xfailed/xpassed test names print in the CI log

2855575

archana-ramalingam added a commit that referenced this pull request Sep 22, 2026

test(ci): sync scratch to #12164 workflow + build_run_id 35659355728 …

636d85d

…(PR#12355 HW test) (DO NOT MERGE)

archana-ramalingam added 2 commits September 22, 2026 11:24

stream per-test names with -v; drop junit XML

65b4942

cleanup

18f0ae3

talumbau approved these changes Sep 22, 2026

View reviewed changes

archana-ramalingam added a commit that referenced this pull request Sep 23, 2026

test(ci): port #12164 on-demand restructure; hardcode Case A for pus…

958ffcd

…h validation (DO NOT MERGE)

archana-ramalingam and others added 2 commits September 23, 2026 10:15

ci(gfx1250): make the HW test workflow on-demand only

9664450

Merge branch 'develop' into users/aramalin/gfx1250-hipblaslt-tensilel…

6565a99

…ite-test-workflow

tony-davis merged commit 1f16a3b into develop Sep 23, 2026
212 of 222 checks passed

tony-davis deleted the users/aramalin/gfx1250-hipblaslt-tensilelite-test-workflow branch September 23, 2026 17:40

Navigation Menu

Repository navigation

ci(gfx1250): Add public gfx1250 hipblaslt + tensilelite HW CI workflow - #12164

ci(gfx1250): Add public gfx1250 hipblaslt + tensilelite HW CI workflow#12164
tony-davis merged 16 commits into
developfrom
users/aramalin/gfx1250-hipblaslt-tensilelite-test-workflow

archana-ramalingam commented Sep 16, 2026 •

edited

Loading

Uh oh!

therock-pr-bot Bot commented Sep 16, 2026 •

edited

Loading

Uh oh!

therock-pr-bot Bot commented Sep 16, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

geomin12 left a comment

Uh oh!

tony-davis commented Sep 23, 2026

Uh oh!

Uh oh!

assistant-librarian Bot commented Sep 25, 2026

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Conversation

archana-ramalingam commented Sep 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Motivation

Technical Details

Test Plan

Test Result

Submission Checklist

Risk level

Uh oh!

therock-pr-bot Bot commented Sep 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

✅ All Checks Passed — Ready for Review

Uh oh!

therock-pr-bot Bot commented Sep 16, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

geomin12 left a comment

Choose a reason for hiding this comment

Uh oh!

tony-davis commented Sep 23, 2026

Uh oh!

Uh oh!

assistant-librarian Bot commented Sep 25, 2026

TheRock Submodule Bump Activity

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

archana-ramalingam commented Sep 16, 2026 •

edited

Loading

therock-pr-bot Bot commented Sep 16, 2026 •

edited

Loading