rupprecht wrote: I got one buildbot email: https://lab.llvm.org/buildbot/#/builders/68/builds/68776
Looks like lldb-dap crashes occasionally, and so the `disconnect` event doesn't get the response it was expecting. (The buildbot does not enable the dap request/response logging, but I have it enabled internally, so I have the stack trace -- I'll file a bug later). I saw that TestThreadLocal.py was previously ignoring errors during tear down if the test is annotated xfail, but maybe that's a more general difference. So this patch probably isn't the culprit for that crash flakiness, but may cause it to be more visible. The arm/aarch64 lldb bots have had an `ninja: error: empty path` error, they seem to be misconfigured to run `ninja "" ""`. For Mac, [Green Dragon](https://green.lab.llvm.org/green/view/LLDB/) seems to be down. So I don't have a great signal if this really broke anything. I'll look at making TestRequireHWBreakpoints.py use skip instead of xfail, I think that makes sense anyway. https://github.com/llvm/llvm-project/pull/79945 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
