[Lldb-commits] [PATCH] D78972: Treat hasWeakODRLinkage symbols as external in REPL computations

[Jim Ingham via Phabricator via lldb-commits]

jingham added a comment.

In D78972#2041680 <https://reviews.llvm.org/D78972#2041680>, @labath wrote:

> I don't think we need to have different behavior for repl and --top-level. I'm mainly just confused about what is the right behavior to aim for.
>
> So, assuming we want these to behave as if everything was in a single TU, my next question is: What is the purpose of the `external` flag? In this setup, it seems like everything should be "external" (i.e., made available to subsequent expressions)...

Eh, this got lost in my queue...  I think the thing you are trying to avoid is if the compiler generates some static helper function in when building one of the actual TU's (as opposed to our fictitious TU) and then we we export that we might confuse it when it goes to emit the same thing in the next TU, since it isn't expecting to find it as an external symbol.  Since I don't know all the games of this sort any of the languages we support is likely to play, I'd rather only export the things that were clearly meant to be exported.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D78972/new/

https://reviews.llvm.org/D78972