On Sun, May 26, 2024 at 02:34:05AM -0400, Philip wrote: > Part of the problem seems to be that the Github actions runner has a bug > on OSX: https://github.com/actions/runner/issues/884 I think this may be a red herring. The issue described there is that processes in the runner environment are spawned with SIGPIPE ignored. But one of the things we do in common-main.c is make sure SIGPIPE is set back to the default. So unless the problem is in a non-git program (say, an issue with the shell, or prove, or something), I don't think this would affect us. That issue also goes back to 2020, and anecdotally, I would say that the hangs we are seeing started in the last 6 months or so. -Peff