Skip to content

Add --blame-crash support to DotNetTestSettings #5

Description

@BrewingCoder

Tamp.NetCli.V10.DotNetTestSettings exposes BlameHang and BlameHangTimeout,
which map to --blame-hang / --blame-hang-timeout. There is no corresponding
property for --blame-crash.

These cover different failures. --blame-hang catches a single test exceeding a
time budget. --blame-crash is what you need when the test host process dies:
it writes a sequence file naming the test that was executing at the time, plus a
crash dump.

This mattered concretely. We had a bug where an unhandled exception on a
background thread killed the test host at a different test count on every run
(249, 541, 635). The run reported Test Run Aborted with a bare stack and no
indication of which test was in flight, and the TRX written was truthful but
partial, so it read as a clean pass of a subset. --blame-crash is the
diagnostic that answers "which test was running when the host died", and it was
not reachable through the typed API.

Suggested fix: add BlameCrash (bool) and ideally BlameCrashDumpType
(mini / full), matching the existing BlameHang shape.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions