iree-org / iree

A retargetable MLIR-based machine learning compiler and runtime toolkit.
http://iree.dev/
Apache License 2.0
2.47k stars 548 forks source link

Re-enable test_amd_w7900 CI job. #17675

Closed ScottTodd closed 1 week ago

ScottTodd commented 1 week ago

This job was disabled due to runner issues and then test failures. The runners have been more stable and the tests have (hopefully) been fixed.

Depends on https://github.com/iree-org/iree/pull/17674 Fixes https://github.com/iree-org/iree/issues/17370

ci-exactly: build_all, test_amd_w7900

ScottTodd commented 1 week ago

Looks like this crashed the GPU on

Runner name: 'shark09'
Runner group name: 'Open'
Machine name: 'shark10-ci'

https://github.com/iree-org/iree/actions/runs/9551239129/job/26327378915#step:7:280

may need to change ctest to run less tests in parallel

ScottTodd commented 1 week ago

may need to change ctest to run less tests in parallel

similar to https://github.com/iree-org/iree/blob/eede9f26b2e5a593eebe4362cd0de389dc01d8af/.github/workflows/ci.yml#L320