[PATCH] D122394: [C++20][Modules] Correct an assert for modules-ts.

Iain Sandoe via Phabricator via cfe-commits cfe-commits at lists.llvm.org
Thu Mar 24 07:05:13 PDT 2022


iains added a comment.

In D122394#3405338 <https://reviews.llvm.org/D122394#3405338>, @urnathan wrote:

> Is this because of history that ModulesTS option != p1103 modules?  I thought we wanted to make the former become the latter (i.e. ModuleTS is the same as CPlusPlusModules)  This seems to be moving in the wrong direction.

This is because we are not removing the fmodules-ts flag (nor. at this point, the code that it changes) - as I understand things.  So I have now to fix (next patch) the problem that we currently generate wrong code for static vars and lambdas in modules since the linkage is defaulting to the modules-ts "module internal" model.

So, unless we shift priorities such that we take a detour to remove the modules-ts support, I need to make this change to avoid the assert firing for -fmodules-ts (it seems we have limited testing - since nothing fired until I tried to make other changes form my patchset).


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D122394/new/

https://reviews.llvm.org/D122394



More information about the cfe-commits mailing list