mclow.lists added a comment.
A few comments; more to come after the earlier patches land.
================
Comment at: libcxx/include/experimental/simd:1429
+struct deduce<_Tp, _Np, __simd_abi<__kind, __old_size>...> {
+ using type = __simd_abi<__kind, _Np>;
+};
----------------
Is this deliberately incomplete?
Comparing to N4755 - I see:
The member type shall be present if and only if:
* T is a vectorizable type, and
* simd_abi::fixed_size<N> is supported (see 9.2.1), and
* every type in the Abis pack is an ABI tag.
Will a later patch fix this up?
================
Comment at: libcxx/include/experimental/simd:1725
// NOTE: P0820 extension
template <size_t __array_size, class _Tp, class _Abi>
----------------
We don't need the comments any more that P0820 has been adopted.
================
Comment at: libcxx/include/experimental/simd:1732
__array_size>>::type
split_by(const simd<_Tp, _Abi>& __v) {
+ array<simd<_Tp, simd_abi::deduce_t<
----------------
Isn't this now named `split`?
https://reviews.llvm.org/D44663
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
