test(chat): deflake Copilot exec retry attempt-count assertions#4895
Open
devantler wants to merge 1 commit into
Open
test(chat): deflake Copilot exec retry attempt-count assertions#4895devantler wants to merge 1 commit into
devantler wants to merge 1 commit into
Conversation
TestRunCopilotCmdWithRetry's "runs once on success" and "does not retry a non-ETXTBSY failure" subtests write a shell script with os.WriteFile and immediately exec it, asserting attempts == 1. Under heavy parallel load on Linux CI, a concurrent fork can momentarily inherit the still-open write fd to the freshly written file, so the first execve returns a genuine transient ETXTBSY. runCopilotCmdWithRetry correctly retries it, inflating attempts to 2 and intermittently failing the required Test check on unrelated PRs. Add writeSettledExecutable, which writes the script and then execs it until a launch returns a non-ETXTBSY result. The file's only write fd comes from the single os.WriteFile, so once any exec succeeds no process holds a write fd and every subsequent exec is deterministic — keeping the exact attempts == 1 assertion intact rather than loosening it. On macOS (no ETXTBSY enforcement on exec) the loop returns after one exec, so there is no added cost there. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Contributor
There was a problem hiding this comment.
Pull request overview
Deflakes TestRunCopilotCmdWithRetry by pre-settling freshly written test scripts before timing the attempts == 1 exec assertions, avoiding transient ETXTBSY races on Linux CI.
Changes:
- Add
writeSettledExecutablehelper that retries exec until non-ETXTBSYresult. - Use the new helper in the two subtests asserting an exact attempt count.
Contributor
❌MegaLinter analysis: Error❌ COPYPASTE / jscpd - 7 errors❌ REPOSITORY / osv-scanner - 1 error❌ ACTION / zizmor - 1 error✅ Linters with no issuesactionlint, bash-exec, git_diff, hadolint, jsonlint, lychee, markdown-table-formatter, markdownlint, prettier, prettier, shellcheck, shfmt, stylelint, syft, trivy-sbom, trufflehog, v8r, v8r, yamllint Notices📣 MegaLinter 9.5.0 is out! Discover the new features and security recommendations in the release announcement. (Skip this info by defining See detailed reports in MegaLinter artifacts
|
Contributor
Code Coverage OverviewLanguages: Go Go / code-coverage/goThe overall coverage in the branch is 55%. Coverage data for the branch is not yet available. Show a code coverage summary of the most covered files.
Code Coverage is in Public Preview. Learn more and provide us with your feedback. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What
Makes
TestRunCopilotCmdWithRetrydeterministic by settling the freshly written test script before the timed/counted exec. Touches onlypkg/cli/cmd/chat/chat_exec_retry_test.go.Why
Two subtests —
returns nil and runs once on successanddoes not retry a non-ETXTBSY failure— write a shell script withos.WriteFileand immediately exec it, assertingattempts == 1.Under heavy parallel load on Linux CI runners, a concurrent fork in another test can momentarily inherit the still-open write fd to the freshly written file, so the first
execvereturns a genuine transientETXTBSY("text file busy").runCopilotCmdWithRetrycorrectly retriesETXTBSY, soattemptsbecomes2and the exact-count assertion fails. This intermittently blocks the required 🧪 Test check on unrelated PRs.This is a pre-existing flake, unrelated to any feature work.
How
Adds a
writeSettledExecutablehelper that writes the script and then execs it until a launch returns a non-ETXTBSYresult. The file's only write fd comes from the singleos.WriteFile, so once any exec succeeds no process holds a write fd and every subsequent exec is deterministic. This keeps the strong, intent-revealingattempts == 1assertion intact rather than loosening it.The same flaky pattern affected both
attempts == 1subtests, so both now use the settled helper. On macOS (ETXTBSYon exec of a write-open file isn't enforced) the loop returns after a single exec, so there's no added cost there; it only absorbs the race on Linux, which is where CI both hits the flake and measures coverage.Verification
go build+go vet ./pkg/cli/cmd/chat/— passgo test -run TestRunCopilotCmdWithRetry -count=50 ./pkg/cli/cmd/chat/— pass (run twice)go test -race ./pkg/cli/cmd/chat/— passgolangci-lint run— the changed file is clean (remaininggoconstfindings are pre-existing inchat_test.go/chat_additional_test.go, untouched here)🤖 Generated with Claude Code