[Lldb-commits] [PATCH] D90598: [lldb/DWARF] Fix implementation of DW_OP_convert

Mon Nov 2 13:59:40 PST 2020

JDevlieghere added a comment.

In D90598#2368826 <https://reviews.llvm.org/D90598#2368826>, @dblaikie wrote:

> In D90598#2368020 <https://reviews.llvm.org/D90598#2368020>, @labath wrote:
>
>> Judging by http://lists.llvm.org/pipermail/llvm-dev/2020-October/145990.html, dsymutil has a similar bug...
>
> That's my understanding. It means I'm not sure (though I'm not the expert here - certainly leave final decisions up to @aprantl and @JDevlieghere ) how this will be fixed, since lldb will presumably still need to be able to parse/understand old dsyms using the incorrect absolute offsets. (we had a similar problem once with DW_OP_bit_piece, where it was implemented incorrectly in both LLVM and lldb and I'm not sure how the backwards compatibility story was addressed there).

I think @aprantl has proposed this in the past, but we should make `dsymutil` convert `DW_OP_convert` to `DW_OP_LLVM_convert` in its output. I also don't have a good plan for backward compatibility. If there was a way we could "know" that the DWARF came from a dSYM we could continue to support the old behavior knowing that it has always been wrong, but I'm not aware of anything like that in LLDB and I'm also not sure that's something I'd like to have in the first place...

Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D90598/new/

https://reviews.llvm.org/D90598