Michael137 wrote: Just want to note that I think the more general issue is that we can have entities in LLDB holding onto `ModuleSP`s preventing them from being orphaned in the first place. In the case I looked at it was a rogue thread plan kept alive by `ProcessGDBRemote`. So the assertion would still trigger even if LLDB did orphan all the modules it could.
I think @jimingham's stance on this was that we shouldn't try tearing down the debugger unless all process have exited. That would work around the issue you've identified but also the fact that processes can still hold onto global modules when we tear down the debugger. I'll let others comment on the patch itself and whether to take it as a stop-gap. https://github.com/llvm/llvm-project/pull/176884 _______________________________________________ lldb-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
