-
Notifications
You must be signed in to change notification settings - Fork 2.6k
Reenable CoreFX tests that were marked as fixed. #22358
Reenable CoreFX tests that were marked as fixed. #22358
Conversation
|
@dotnet-bot test Ubuntu16.04 arm64 Cross Checked corefx_baseline Build and Test |
|
The ARM tests are actually failing on floating point formatting, re-run won't help. /cc @tannergooding |
|
@filipnavara I thought dotnet/corefx#35016 fixed that, did not it? |
|
I didn't immediately pinpoint the cause when looking at it. Hopefully @tannergooding will have better idea what is going on. |
|
The CoreFX tests, AFAIK, are not running against a live copy of CoreFX, but an older copy. CC. @jkotas in case that has changed. |
|
@tannergooding it did not change, we need to wait another sync like #18557 to have these tests passed. |
That's still the case (but I don't know who updates it and when). |
|
There are two different systems to run CoreFX tests in CoreCLR CI: (1) A system that uses the fixed test snapshot. It is what the CoreFX legs that are on by default in CoreCLR CI use. (2) A system that builds a live copy of CoreFX and then runs the tests. This system suffers from multiple sources of flakiness (e.g. https://github.com/dotnet/coreclr/issues/22344#issuecomment-459913137). This system is used for optional JIT stress mode legs. This is the system that is being updated in this PR. The plan how to move this forward and unify these systems is being discussed in https://github.com/dotnet/corefx/issues/34957#issuecomment-459415999 |
af7d8d7 to
112f1a0
Compare
|
is ci dead again? |
|
@dotnet-bot test this please |
nah (except for the AzurePipelines bot trigger words) |
|
@dotnet-bot test this please |
112f1a0 to
5158caf
Compare
|
"CentOS7.1 x64 Debug Innerloop Build" looks like build infrastructure failure "Ubuntu16.04 arm64 Cross Checked corefx_jitstress2 Build and Test" fails on |
Created #22449 and excluded this test for this PR. |
|
ci doesn't want to test it again: @dotnet-bot test this please |
|
@dotnet-bot test Ubuntu16.04 arm64 Cross Checked corefx_baseline Build and Test |
|
test Ubuntu16.04 arm64 Cross Checked corefx_jitstress2 Build and Test |
|
@dotnet-bot test Windows_NT arm64 Cross Checked Innerloop Build and Test |
|
PTAL @BruceForstall @dotnet/jit-contrib |
|
@dotnet-bot test Ubuntu16.04 arm64 Cross Checked corefx_baseline Build and Test @dotnet-bot test OSX10.12 x64 Checked Innerloop Build and Test |
Yes we do. I hope we can get this done by tomorrow. |
|
The other idea is to get this in and disable the failing test in corefx.issues.json. we can do that if we think we need to rush getting the new drop from corefx |
This PR has been waiting for 5 days, it can wait another, don't worry. |
|
@dotnet-bot test Ubuntu16.04 arm64 Cross Checked corefx_baseline Build and Test |
|
@dotnet-bot test Ubuntu16.04 arm64 Cross Checked corefx_baseline Build and Test |
|
@dotnet-bot test Ubuntu16.04 arm64 Cross Checked corefx_jitstress2 Build and Test |
|
@sandreenko to let you know, the changes for fixing the Range/Index in corefx is delayed because we have been not getting the coreclr new build to corefx. |
|
I thought CoreFX part was merged with dotnet/corefx@b0b1617, was not it? |
|
@sandreenko no, this PR was just the mirroring of the shared files between coreclr and corefx. The needed changes in corefx is in the PR dotnet/corefx#35003. As I mentioned before we have the option to disable the failing tests in https://github.com/dotnet/coreclr/blob/master/tests/CoreFX/CoreFX.issues.json if you don't want to wait. |
|
@dotnet-bot test Ubuntu16.04 arm64 Cross Checked corefx_baseline Build and Test |
No description provided.