Combine various patchFailed*Tests properties. We don't need separate property for each type of test failure.
These properties were initially added so that somes build-step can be statically added in build factory and would be executed/skipped based on these flags. This was mostly for build-steps which should be run only after failure, e.g.: UnApplyPatchIfRequired, CompileWebKitToT.
For most part, we don't require these flags anymore, since we started using addStepsAfterCurrentStep() method inside steps.py, which add such build-step dynamically as required (only after failure) (e.g.: https://trac.webkit.org/browser/webkit/trunk/Tools/BuildSlaveSupport/ews-build/steps.py#L638). But maybe it's still worth to keep them, as a safeguard against someone accidentally adding such steps in a wrong order in the factory.
Created attachment 372841 [details]
I would have expected tests to change with this sort of property modification.
Created attachment 372883 [details]
(In reply to Jonathan Bedard from comment #2)
> I would have expected tests to change with this sort of property modification.
Unit tests are modified for 'patchFailedJSCTests' property.
I just updated TestCompileWebKitToT.test_failure() and TestUnApplyPatchIfRequired.test_failure() so that these unit-test test 'patchFailedTests' property as well.
Comment on attachment 372883 [details]
Clearing flags on attachment: 372883
Committed r246822: <https://trac.webkit.org/changeset/246822>
All reviewed patches have been landed. Closing bug.