nico wrote:

(To be clear, I'm fine with landing this.)

> non conforming two phase lookup by default would also be a concern.

It is! There were quite a few discussions around this. Aha, see #42377.

We usually say the goal is to have a very standards compliant impl, while also 
being able to parse system headers. This can involve things like making things 
that should be errors instead be warnings that are mapped to errors by default 
(…but warnings are implicitly ignored in system headers), etc. 
https://docs.google.com/presentation/d/1oxNHaVjA9Gn_rTzX6HIpJHP7nXRua_0URXxxJ3oYRq0/edit#slide=id.p
 / https://www.youtube.com/watch?v=rXi065XC6zY talks about this some, in case 
you haven't seen it :)

Re cross builds: That's something we do too. For case insensitivity, other than 
symlinks, one possible approach is to use FUSE with ciopfs 
(https://source.chromium.org/chromium/chromium/src/+/main:build/vs_toolchain.py;l=504?q=ciopfs).
 Another approach is -ivfsoverlay. The git log of 
llvm/cmake/platforms/WinMsvc.cmake has some information on that approach.

https://github.com/llvm/llvm-project/pull/107509
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

Reply via email to