MirrorHub/construct
0
0
Fork 0
mirror of https://github.com/matrix-construct/construct synced 2024-06-16 17:08:20 +02:00
Code Issues Releases Wiki Activity

ircd::simd: Add ymmx shifter specializations.

Browse source
This commit is contained in:
Jason Volk 2020-06-25 11:51:35 -07:00
parent 1ad2a1546a
commit 6a33be4494
1 changed files with 34 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

34
include/ircd/simd.h
View file

@ -163,6 +163,8 @@ namespace ircd::simd
template<int bits, class T> T shr(const T &a) noexcept;
template<int bits> u128x1 shl(const u128x1 &a) noexcept;
template<int bits> u128x1 shr(const u128x1 &a) noexcept;
template<int bits> u256x1 shl(const u256x1 &a) noexcept;
template<int bits> u256x1 shr(const u256x1 &a) noexcept;
// readable output and debug
template<class T> string_view str_reg(const mutable_buffer &buf, const T &, const uint &fmt = 0) noexcept;
@ -212,3 +214,35 @@ noexcept
return _mm_bslli_si128(a, b / 8);
}
#ifdef __AVX2__
template<int b>
[[using gnu: always_inline, gnu_inline, artificial]]
extern inline ircd::u256x1
ircd::simd::shr(const u256x1 &a)
noexcept
{
static_assert
(
b % 8 == 0, "ymmx register only shifts right at bytewise resolution."
);
return _mm256_srli_si256(a, b / 8);
}
#endif
#ifdef __AVX2__
template<int b>
[[using gnu: always_inline, gnu_inline, artificial]]
extern inline ircd::u256x1
ircd::simd::shl(const u256x1 &a)
noexcept
{
static_assert
(
b % 8 == 0, "ymmx register only shifts right at bytewise resolution."
);
return _mm256_slli_si256(a, b / 8);
}
#endif