[PATCH] D79248: [lld][WebAssembly] Implement --unresolved-symbols

Sam Clegg via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Tue Nov 17 15:25:11 PST 2020


sbc100 added a comment.

In D79248#2400768 <https://reviews.llvm.org/D79248#2400768>, @MaskRay wrote:

> In D79248#2400523 <https://reviews.llvm.org/D79248#2400523>, @sbc100 wrote:
>
>> @MaskRay any objections to landing this?
>>
>> BTW, how to you feel about being a reviewer on wasm-ld change in general.. if you are OK with it I could use a reviewer for my work here now that @Rui is less active.
>
> I don't mind but you'll throw out some learning resources on me.. (Persoanlly I do want to study more binary formats) For now I guess the best I can provide is how some options are used in ELF..

That sounds good.  I can include wasm-specific people too in all my changes, but its good to have somebody with deep knowledge on the other linker take a look at changes like this for sure.

> I just committed D91510 <https://reviews.llvm.org/D91510> for ELF - personally I think the option is a bit awkward to use as it changes two bits simultaneously. There are more values in the wasm-lld version: do you think these values should all be mutually exclusive. If yes, this looks fine to me.

That an interesting point.  Yes I am certainly imagining the various modes being mutually exclusive.    Unlike ELF we have a couple of different ways that we want to be able to handle undefined symbols accepting various value here makes sense to me.  Just doing `-z undefs` would not be enough in general as I would also want to know *how* to handle those undefined symbols.   Once the followup to this change lands there will be 3 different ways to deal with them once we know we want to allow them.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D79248/new/

https://reviews.llvm.org/D79248



More information about the llvm-commits mailing list