[PATCH] D48712: [X86] Lowering integer truncation intrinsics to native IR

[Simon Pilgrim via Phabricator via cfe-commits]

RKSimon added inline comments.


================
Comment at: clang/lib/Headers/avx512vlbwintrin.h:1501
+      (__v8qi){0, 0, 0, 0, 0, 0, 0, 0}, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11,
+      12, 13, 14, 15);
 }
----------------
Are we happy with using illegal types like this? What about flipping the shuffle and convert?

```
  return (__m128i)__builtin_convertvector(
                    __builtin_shufflevector((__v8hi)__A,
                                            (__v8hi){0, 0, 0, 0, 0, 0, 0, 0},
                                            0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15), __v16qi);
```


https://reviews.llvm.org/D48712