Michael137 wrote: > Yes, in the totally reduced reproducer - another debugger might be able to do > this name lookup with only a declaration, but I understand that at least > Clang's AST expression evaluator wouldn't be able to handle this case. > (though perhaps the lldb-eval work will provide some opportunity to catch > this case) > > But at least in our real-world example I think the definition is available, > in another DWO in the same executable. But I don't have a small reproducer > for that - small examples "just work" (they do find the type definition, > don't crash) - but in the larger one we still see the crash depending on > exactly where the expression is evaluated (within a fully statically linked > binary - so the definition is available, broadly speaking, no matter the > context of evaluation - not sure what makes the difference) > > If it's helpful to try to identify the more nuanced case where a crash occurs > despite the definition being available /somewhere/, I can try to figure that > out but it seems difficult to do so.
I'll submit a PR later today to get us back to previous behaviour https://github.com/llvm/llvm-project/pull/72235 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
