-
Notifications
You must be signed in to change notification settings - Fork 17.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
crypto/x509: TestPlatformVerifier failures on macOS due to mismatched error text #65461
Comments
Found new dashboard test flakes for:
2024-02-01 21:42 gotip-darwin-amd64_14 go@c9d88ea2 crypto/x509.TestPlatformVerifier (log)
2024-02-01 22:57 gotip-darwin-amd64-nocgo go@6d3c1ce8 crypto/x509.TestPlatformVerifier (log)
2024-02-01 22:59 gotip-darwin-amd64-nocgo go@117164f9 crypto/x509.TestPlatformVerifier (log)
|
Found new dashboard test flakes for:
2024-02-02 18:20 gotip-darwin-amd64-nocgo go@6b0309ce crypto/x509.TestPlatformVerifier (log)
|
(attn @rolandshoemaker) |
I took a quick look at this, and I have zero clue how this could be flakey. The pattern of failures don't provide any particularly useful insight, the error returned is completely useless and I'm unable to repro locally. Will think about this a little more, but for now I'm stumped. |
Huh. I wonder if someone was changing the LUCI builder configuration during these runs? (Maybe someone on @golang/release would have some insight?) |
@prattmic might know if there've been changes to macOS builders on LUCI that would be relevant to this. Something to check might be whether the builders running into this all have the root from #52108 added. But given the test is running (and not skipped with "test root is not in trust store"), I think it must be. |
Every failing instance linked above was on the same bot, darwin-amd64-14--c1f58526-d1c4-4792-bef1-06d6c16b49e4.golang.cc.macservice.goog. I imagine that is related, but it is still odd because all of the bots use identical macOS images. |
Bit flip somewhere in the root store? 🤷 |
I pulled up the test history on this bot. The bot has been running since 2024-01-09, and the test passed 100% of the time until a run starting 2024-02-01 11:30:06.033 America/Los_Angeles (https://cr-buildbucket.appspot.com/build/8757267625331049905). It has been failing 100% of the time since then. |
I've also verified that this is the only bot that has had any failures. |
Found new dashboard test flakes for:
2024-02-05 14:52 gotip-darwin-amd64-nocgo go@e8d87728 crypto/x509.TestPlatformVerifier (log)
2024-02-05 15:30 gotip-darwin-amd64-nocgo go@76ff0caa crypto/x509.TestPlatformVerifier (log)
2024-02-05 20:59 gotip-darwin-amd64-nocgo go@6076edc5 crypto/x509.TestPlatformVerifier (log)
|
Found new dashboard test flakes for:
2024-02-02 16:26 go1.21-darwin-amd64-nocgo release-branch.go1.21@2fdad8af crypto/x509.TestPlatformVerifier (log)
2024-02-06 12:43 gotip-darwin-amd64_14 go@6f44cc88 crypto/x509.TestPlatformVerifier (log)
|
The cluster this instance was running in had an infrastructure issue that seems to have corrupted it. I have killed this instance. |
Found new dashboard test flakes for:
2024-02-02 16:17 go1.22-darwin-amd64_14 release-branch.go1.22@b0957cfc crypto/x509.TestPlatformVerifier (log)
|
Found new dashboard test flakes for:
2024-02-20 16:10 gotip-darwin-amd64_13 go@ff4e45fb crypto/x509.TestPlatformVerifier/non-nested_KU (log)
2024-02-20 17:18 gotip-darwin-amd64_13 go@02785362 crypto/x509.TestPlatformVerifier/non-nested_KU (log)
|
Found new dashboard test flakes for:
2024-02-20 22:29 gotip-darwin-amd64_13 go@5428cc4f crypto/x509.TestPlatformVerifier/non-nested_KU (log)
|
Found new dashboard test flakes for:
2024-02-21 18:47 gotip-darwin-amd64_13 go@c07b9b00 crypto/x509.TestPlatformVerifier/non-nested_KU (log)
2024-02-21 20:35 gotip-darwin-amd64_13 go@b27d02c0 crypto/x509.TestPlatformVerifier/non-nested_KU (log)
2024-02-22 22:21 gotip-darwin-amd64_13 go@d892cb49 crypto/x509.TestPlatformVerifier/non-nested_KU (log)
|
@prattmic it is possible we have another broken instance? |
Quite possible, I was doing a sweep of potentially messed up instances (https://go.dev/cl/562399 has helped a lot, but some are still slipping through), and as part of that I've killed the instance that these tests failed on. |
Issue created automatically to collect these failures.
Example (log):
— watchflakes
The text was updated successfully, but these errors were encountered: