Michael Buch
Michael Buch
Yup, GCC emits the template name with a space. I think that's probably partly why we decided to make it the default printing policy to be `SplitTemplateClosers == true`. Changing...
> Another way (potentially as a transitional step) to solve this would be to include the basename in the accelerator tables (in addition to the DW_AT_name and DW_AT_linkage_name that we...
> Maybe operators = should analyze result separately, like in [APFloat.h:1240](https://github.com/llvm/llvm-project/blob/7812fcf3d79ef7fe9ec6bcdfc8fd9143864956cb/llvm/include/llvm/ADT/APFloat.h#L1240). Thanks for the report and analysis. Agreed, checking for the possibility of an unordered comparison result seems like the...
> > Tried on Darwin. This is what the ast looks like: > > ``` > > (lldb) target module dump ast > > Dumping clang ast for 43 modules....
> But what I wondered was - you showed the dumped AST was correct. So /then/ I don't understand why the expression can't be evaluated, if we managed to get...
Can you repro this with a more recent lldb version? Crash doesnt repro for me with lldb-14 (and top-of-tree)