Add capability of outputting DxDiag when running pipelines #1056
Add capability of outputting DxDiag when running pipelines #1056joaosaffran wants to merge 20 commits intollvm:mainfrom
Conversation
| run_dxdiag: | ||
| description: "Run dxdiag" | ||
| required: false | ||
| default: false | ||
| type: boolean |
There was a problem hiding this comment.
Any reason not to just run this all the time? Seems that we'd want to grab the dxdiag output when we see a test run has failed and so it'd be nice if that was always available with the rest of the results.
| - name: Check if dxdiag report exists | ||
| id: check_dxdiag | ||
| if: inputs.OS == 'windows' && inputs.run_dxdiag | ||
| shell: powershell |
There was a problem hiding this comment.
Is there a reason this has to be a separate step?
I'm wondering if you ended up doing this to work around dxdiag running asynchronously, which means that the "DxDiag report saved to $output" message gets printed before the output is actually? In which case this would only happen to work by luck and sometimes you'd not get output at all.
If you run dxdiag like this in powershell:
dxdiag /t $output | Out-NullThen that command will only complete once the file has been written, and so you can safely assume it is there.
damyanp
left a comment
There was a problem hiding this comment.
LGTM, although I wonder if we need to limit this to just failures? If we suspect that a failure may be caused by a driver update then we'd want to compare dxdiag between a successful run and the failing run.
This patch makes pipelines emit dxdiag information when failing.