tests: treat gpg 'Bad passphrase' as transient under CI load - #2480
Conversation
Same agent-state failure as 'Corrupted protection' (issue #2457), just reported with different wording by some gpg versions - seen failing cli_tests-Compression on debian-13. Deterministic passphrase bugs still fail after all retry attempts.
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## main #2480 +/- ##
==========================================
- Coverage 85.47% 85.46% -0.01%
==========================================
Files 125 125
Lines 22962 22962
==========================================
- Hits 19626 19625 -1
- Misses 3336 3337 +1 ☔ View full report in Codecov by Harness. 🚀 New features to boost your workflow:
|
|
@ni4 one-line test-infra follow-up for your single-approval class, all checks green (only the known fuzzing exceptions): adds |
ni4
left a comment
There was a problem hiding this comment.
LGTM, thanks! Merging with single approval as CI-only.
Summary
Corrupted protectionon some gpg versions surfaces asBad passphraseon others - seen failingcli_tests-Compressionon the debian-13 leg of cli: verify signatures of side-loadable DLLs on Windows (#2411) #2478 (log:gpg signing failed: Bad passphrase, first occurrence, no code relation to that PR).GPG_TRANSIENT_ERRORSso the existing in-call retry + whole-test re-run machinery covers it. Deterministic passphrase failures (genuine regressions) still fail after all attempts - the retry only converts single-shot agent hiccups into passes.Test plan