[PATCH] D61175: [MinGW] Don't let template instantiation declarations cover nested classes

[Martin Storsjö via Phabricator via cfe-commits]

mstorsjo marked an inline comment as done.
mstorsjo added inline comments.


================
Comment at: lib/Sema/SemaTemplateInstantiate.cpp:2687-2689
       if ((Context.getTargetInfo().getCXXABI().isMicrosoft() ||
-           Context.getTargetInfo().getTriple().isWindowsItaniumEnvironment()) &&
+           Context.getTargetInfo().getTriple().isWindowsItaniumEnvironment() ||
+           Context.getTargetInfo().getTriple().isWindowsGNUEnvironment()) &&
----------------
rnk wrote:
> I think this can be simplified to "if Windows" at this point. But, I'm confused why we need this change to the general template instantiation machinery... Anyway, I'll send the simplification as a code review.
Yes, I believe so. Not sure about why this is needed though.

In case GCC actually does make a move on this matter, and chooses a different path for fixing it, we might want to follow that so in that case we might need to reintroduce some condition like this. But until that hypothetical case, let's simplify it.


Repository:
  rC Clang

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D61175/new/

https://reviews.llvm.org/D61175