Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch
See info in area-owners.md if you want to be subscribed.

@dotnet/jit-contrib

I am surprised the CI for original PR pass. @markples - do you know why?

A PR race condition - #86108 was merged after the CI run but before merge.

It looks like the test is failing in outerloop/CG2 - can we disable it everywhere against #86112 (don't want to disable only on CG2 because it'll fail on NAOT and under collectible assemblies)?

(I was about to submit a change for that)

A PR race condition - #86108 was merged after the CI run but before merge.

Not sure if I follow. #86108 CI should have ran into this, right?

disable test

done.

If two PRs are open at the same time, they each test without the other. (I.e., we don't serialize testing/merging.)

I updated the .cs file to match. I plan on strengthening the checks in various ways to avoid mistakes, but also the copy/paste practice of writing tests is vulnerable while the merging process is going on, which will pass.

If two PRs are open at the same time,

Got it. What was the other PR you are referring to?

#85850 and #86108

Mine was open for a while because I hadn't gone through the "extra" test results and probably should have checked the history before merging. I just checked and there don't seem to be any other new tests in any of the directories that I just put into merged groups.

#85850 and #86108

That makes more sense now.