[PATCH] D144293: [PowerPC] Fix the implicit casting for the emulated intrinsics

[Amy Kwan via Phabricator via cfe-commits]

amyk added inline comments.


================
Comment at: clang/lib/Headers/ppc_wrappers/emmintrin.h:57
 typedef __vector unsigned char __v16qu;
+typedef __vector float __v2f;
 
----------------
nemanjai wrote:
> The name `__v2f` seems strange since `__vector float` is a vector of 4 `float` values. Should this be `__v4f`?
nit: Could we also put this to where we put the `double`?


================
Comment at: clang/lib/Headers/ppc_wrappers/smmintrin.h:310
   __charmask = vec_gb(__charmask);
-  __v8hu __shortmask = (__v8hu)vec_unpackh(__charmask);
+  __v8hu __shortmask = (__v8hu)vec_unpackh((__v16qi)__charmask);
 #ifdef __BIG_ENDIAN__
----------------
Potentially silly question, but is it intentionally for that mask for be signed?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D144293/new/

https://reviews.llvm.org/D144293