[PATCH] D60882: [WebAssembly] Error on relocations against undefined data symbols.

[Sam Clegg via Phabricator via llvm-commits]

sbc100 added a comment.

In D60882#1472591 <https://reviews.llvm.org/D60882#1472591>, @ruiu wrote:

> Although I understand that reporting an error is sometimes useful, are you sure that that's the behavior that users want? I wonder if we would end up adding `--really-allow-undefined` in the future to override this new behavior.


The problem is that this would actually mean "treat all data symbols as weak undefined with no way to provide them".      Which I guess somebody might want but its hard to imagine.   the current users of `--allow-undefined` kind of expect to be able to provide the undefined symbols at runtime.

In reality what we will probably end up doing is linker relaxation to allow "i32.load <foo>" to re-written to "global.get <foo>", although I'm not sure when we will end up going there.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D60882/new/

https://reviews.llvm.org/D60882