dwblaikie wrote:
> Just noticed that this breaks
> `llvm/test/tools/dsymutil/X86/DWARFLinkerParallel/odr-string.test` because
> for `DW_TAG_typedef`, llvm-dwarfdump will print the fully qualified name
> instead of the base name in DW_AT_type that refers to
> it([here](https://github.com/llvm/llvm-project/blob/33124910c9f4877c069c0358159763c03e10ca31/llvm/lib/DebugInfo/DWARF/DWARFDie.cpp#L199-L205)).
> Do you think we should keep the original bahavior by using the base name of
> typedef? Or update the test to accept fully qualified name?
@JDevlieghere can probably speak more to that, but sounds very golden-file-y &
can probably be updated.
I take it it's this part of the test?
```
CHECK:DW_TAG_reference_type
CHECK: 0x[[CONST_STR_REF:[0-9a-f]*]]:
DW_TAG_reference_type{{.*[[:space:]].*}}DW_AT_type{{.*}}0x[[CONST_STRING:[0-9a-f]*]]
"const string"
CHECK:DW_TAG_const_type
CHECK: 0x[[CONST_STRING]]:
DW_TAG_const_type{{.*[[:space:]].*}}DW_AT_type{{.*}}0x[[STRING]] "string"
```
Where those names go from "const string" to "const std::__1::string" (and from
"string" to "std::__1::string")? or something like that.
https://github.com/llvm/llvm-project/pull/117239
_______________________________________________
lldb-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
