0
0
Fork 0
mirror of https://github.com/matrix-construct/construct synced 2024-12-25 15:04:10 +01:00

ircd::simd: Add horizontal adder template.

This commit is contained in:
Jason Volk 2021-03-15 19:53:22 -07:00
parent 90f88cf78d
commit 05a3e505cc
3 changed files with 116 additions and 1 deletions

114
include/ircd/simd/hadd.h Normal file
View file

@ -0,0 +1,114 @@
// The Construct
//
// Copyright (C) The Construct Developers, Authors & Contributors
// Copyright (C) 2016-2020 Jason Volk <jason@zemos.net>
//
// Permission to use, copy, modify, and/or distribute this software for any
// purpose with or without fee is hereby granted, provided that the above
// copyright notice and this permission notice is present in all copies. The
// full license for this software is available in the LICENSE file.
#pragma once
#define HAVE_IRCD_SIMD_HADD_H
// Horizontal add
namespace ircd::simd
{
template<class T,
class R = T>
typename std::enable_if<simd::lanes<T>() == 2, R>::type
hadd(const T, const T) noexcept;
template<class T,
class R = T>
typename std::enable_if<simd::lanes<T>() == 4, R>::type
hadd(const T, const T) noexcept;
template<class T,
class R = T>
typename std::enable_if<simd::lanes<T>() == 8, R>::type
hadd(const T, const T) noexcept;
template<class T,
class R = T>
typename std::enable_if<simd::lanes<T>() == 16, R>::type
hadd(const T, const T) noexcept;
}
template<class T,
class R>
inline typename std::enable_if<ircd::simd::lanes<T>() == 16, R>::type
ircd::simd::hadd(const T a,
const T b)
noexcept
{
return R
{
a[0x01] + a[0x00],
a[0x03] + a[0x02],
a[0x05] + a[0x04],
a[0x07] + a[0x06],
b[0x01] + b[0x00],
b[0x03] + b[0x02],
b[0x05] + b[0x04],
b[0x07] + b[0x06],
a[0x09] + a[0x08],
a[0x0b] + a[0x0a],
a[0x0d] + a[0x0c],
a[0x0f] + a[0x0e],
b[0x09] + b[0x08],
b[0x0b] + b[0x0a],
b[0x0d] + b[0x0c],
b[0x0f] + b[0x0e],
};
}
template<class T,
class R>
inline typename std::enable_if<ircd::simd::lanes<T>() == 8, R>::type
ircd::simd::hadd(const T a,
const T b)
noexcept
{
return R
{
a[0x01] + a[0x00],
a[0x03] + a[0x02],
b[0x01] + b[0x00],
b[0x03] + b[0x02],
a[0x05] + a[0x04],
a[0x07] + a[0x06],
b[0x05] + b[0x04],
b[0x07] + b[0x06],
};
}
template<class T,
class R>
inline typename std::enable_if<ircd::simd::lanes<T>() == 4, R>::type
ircd::simd::hadd(const T a,
const T b)
noexcept
{
return R
{
a[0x01] + a[0x00],
a[0x03] + a[0x02],
b[0x01] + b[0x00],
b[0x03] + b[0x02],
};
}
template<class T,
class R>
inline typename std::enable_if<ircd::simd::lanes<T>() == 2, R>::type
ircd::simd::hadd(const T a,
const T b)
noexcept
{
return R
{
a[0x01] + a[0x00],
b[0x01] + b[0x00],
};
}

View file

@ -13,7 +13,7 @@
namespace ircd::simd
{
/// Perform a horizontal operation among lanes. The operation is specified
/// Perform a horizontal left-reduce among lanes. The operation is given
/// by the caller who supplies a functor template like `std::bit_or` or
/// `std::plus` etc. The result resides in lane[0] of the return vector
/// while all other lanes of the return vector are undefined/junk as far

View file

@ -32,6 +32,7 @@
#include "lzcnt.h"
#include "tzcnt.h"
#include "reduce.h"
#include "hadd.h"
#include "any.h"
#include "all.h"
#include "stream.h"