As you can see in https://ews-build.webkit.org/#/builders/60/builds/10553, the layout-tests and re-run-layout-tests both failed. The bot then proceeded to the compile-webkit-without-patch step and failed to compile. Surprisingly, the bot still proceeded to the run-layout-tests-without-patch step, which failed with the following: 16:30:10.005 10889 WebKitTestRunner was not found at /Volumes/Data/worker/macOS-AppleSilicon-Big-Sur-Debug-WK2-Tests-EWS/build/WebKitBuild/Debug/WebKitTestRunner I would expect that the EWS run would be considered inconclusive or invalid, but instead it blamed the patch under test for introducing a new test failure that really should have been considered pre-existing. Other examples from around the same time: https://ews-build.webkit.org/#/builders/60/builds/10560 https://ews-build.webkit.org/#/builders/60/builds/10561 https://ews-build.webkit.org/#/builders/60/builds/10562 https://ews-build.webkit.org/#/builders/60/builds/10563
<rdar://problem/82371726>
Created attachment 436560 [details] Patch
Tested on uat instance, e.g.: https://ews-build.webkit-uat.org/#/builders/34/builds/40260
Committed r281660 (241013@main): <https://commits.webkit.org/241013@main> All reviewed patches have been landed. Closing bug and clearing flags on attachment 436560 [details].
Restarted buildbot to pick up this change.
buildbot on EWS and build.webkit.org has been restarted to pick up this change.
This change seems to be working fine. e.g.: https://ews-build.webkit.org/#/builders/51/builds/20327 https://ews-build.webkit.org/#/builders/51/builds/20385 https://ews-build.webkit.org/#/builders/51/builds/20418 https://ews-build.webkit.org/#/builders/51/builds/20448 https://ews-build.webkit.org/#/builders/10/builds/103629