Files
gcc/libstdc++-v3/include/bits/simd_iterator.h
Matthias Kretz 8be0893fd9 libstdc++: Implement [simd] for C++26
This implementation differs significantly from the
std::experimental::simd implementation. One goal was a reduction in
template instantiations wrt. what std::experimental::simd did.

Design notes:

- bits/vec_ops.h contains concepts, traits, and functions for working
  with GNU vector builtins that are mostly independent from std::simd.
  These could move from std::simd:: to std::__vec (or similar). However,
  we would then need to revisit naming. For now we kept everything in
  the std::simd namespace with __vec_ prefix in the names. The __vec_*
  functions can be called unqualified because they can never be called
  on user-defined types (no ADL). If we ever get simd<UDT> support this
  will be implemented via bit_cast to/from integral vector
  builtins/intrinsics.

- bits/simd_x86.h extends vec_ops.h with calls to __builtin_ia32_* that
  can only be used after uttering the right GCC target pragma.

- basic_vec and basic_mask are built on top of register-size GNU vector
  builtins (for now / x86). Any larger vec/mask is a tree of power-of-2
  #elements on the "first" branch. Anything non-power-of-2 that is
  smaller than register size uses padding elements that participate in
  element-wise operations. The library ensures that padding elements
  lead to no side effects. The implementation makes no assumption on the
  values of these padding elements since the user can bit_cast to
  basic_vec/basic_mask.

Implementation status:

- The implementation is prepared for more than x86 but is x86-only for
  now.

- Parts of [simd] *not* implemented in this patch:

  - std::complex<floating-point> as vectorizable types
  - [simd.permute.dynamic]
  - [simd.permute.mask]
  - [simd.permute.memory]
  - [simd.bit]
  - [simd.math]
  - mixed operations with vec-mask and bit-mask types
  - some conversion optimizations (open questions wrt. missed
    optimizations in the compiler)

- This patch implements P3844R3 "Restore simd::vec broadcast from int",
  which is not part of the C++26 WD draft yet. If the paper does not get
  accepted the feature will be reverted.

- This patch implements D4042R0 "incorrect cast between simd::vec and
simd::mask via conversion to and from impl-defined vector types" (to be
published once the reported LWG issue gets a number).

- The standard feature test macro __cpp_lib_simd is not defined yet.

Tests:

- Full coverage requires testing
  1. constexpr,
  2. constant-propagating inputs, and
  3. unknown (to the optimizer) inputs
  - for all vectorizable types
  * for every supported width (1–64 and higher)
  + for all possible ISA extensions (combinations)
  = with different fast-math flags
  ... leading to a test matrix that's far out of reach for regular
  testsuite builds.

- The tests in testsuite/std/simd/ try to cover all of the API. The
  tests can be build in every combination listed above. Per default only
  a small subset is built and tested.

- Use GCC_TEST_RUN_EXPENSIVE=something to compile the more expensive
  tests (constexpr and const-prop testing) and to enable more /
  different widths for the test type.

- Tests can still emit bogus -Wpsabi warnings (see PR98734) which are
  filtered out via dg-prune-output.

Benchmarks:

- The current implementation has been benchmarked in some aspects on
  x86_64 hardware. There is more optimization potential. However, it is
  not always clear whether optimizations should be part of the library
  if they can be implemented in the compiler.

- No benchmark code is included in this patch.

libstdc++-v3/ChangeLog:

	* include/Makefile.am: Add simd headers.
	* include/Makefile.in: Regenerate.
	* include/bits/version.def (simd): New.
	* include/bits/version.h: Regenerate.
	* include/bits/simd_alg.h: New file.
	* include/bits/simd_details.h: New file.
	* include/bits/simd_flags.h: New file.
	* include/bits/simd_iterator.h: New file.
	* include/bits/simd_loadstore.h: New file.
	* include/bits/simd_mask.h: New file.
	* include/bits/simd_mask_reductions.h: New file.
	* include/bits/simd_reductions.h: New file.
	* include/bits/simd_vec.h: New file.
	* include/bits/simd_x86.h: New file.
	* include/bits/vec_ops.h: New file.
	* include/std/simd: New file.
	* testsuite/std/simd/arithmetic.cc: New test.
	* testsuite/std/simd/arithmetic_expensive.cc: New test.
	* testsuite/std/simd/create_tests.h: New file.
	* testsuite/std/simd/creation.cc: New test.
	* testsuite/std/simd/creation_expensive.cc: New test.
	* testsuite/std/simd/loads.cc: New test.
	* testsuite/std/simd/loads_expensive.cc: New test.
	* testsuite/std/simd/mask2.cc: New test.
	* testsuite/std/simd/mask2_expensive.cc: New test.
	* testsuite/std/simd/mask.cc: New test.
	* testsuite/std/simd/mask_expensive.cc: New test.
	* testsuite/std/simd/reductions.cc: New test.
	* testsuite/std/simd/reductions_expensive.cc: New test.
	* testsuite/std/simd/shift_left.cc: New test.
	* testsuite/std/simd/shift_left_expensive.cc: New test.
	* testsuite/std/simd/shift_right.cc: New test.
	* testsuite/std/simd/shift_right_expensive.cc: New test.
	* testsuite/std/simd/simd_alg.cc: New test.
	* testsuite/std/simd/simd_alg_expensive.cc: New test.
	* testsuite/std/simd/sse_intrin.cc: New test.
	* testsuite/std/simd/stores.cc: New test.
	* testsuite/std/simd/stores_expensive.cc: New test.
	* testsuite/std/simd/test_setup.h: New file.
	* testsuite/std/simd/traits_common.cc: New test.
	* testsuite/std/simd/traits_impl.cc: New test.
	* testsuite/std/simd/traits_math.cc: New test.

Signed-off-by: Matthias Kretz <m.kretz@gsi.de>
2026-03-21 12:44:15 +01:00

178 lines
4.7 KiB
C++

// Implementation of <simd> -*- C++ -*-
// Copyright The GNU Toolchain Authors.
//
// This file is part of the GNU ISO C++ Library. This library is free
// software; you can redistribute it and/or modify it under the
// terms of the GNU General Public License as published by the
// Free Software Foundation; either version 3, or (at your option)
// any later version.
// This library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.
// Under Section 7 of GPL version 3, you are granted additional
// permissions described in the GCC Runtime Library Exception, version
// 3.1, as published by the Free Software Foundation.
// You should have received a copy of the GNU General Public License and
// a copy of the GCC Runtime Library Exception along with this program;
// see the files COPYING3 and COPYING.RUNTIME respectively. If not, see
// <http://www.gnu.org/licenses/>.
#ifndef _GLIBCXX_SIMD_ITERATOR_H
#define _GLIBCXX_SIMD_ITERATOR_H 1
#ifdef _GLIBCXX_SYSHDR
#pragma GCC system_header
#endif
#if __cplusplus >= 202400L
#include "simd_details.h"
namespace std _GLIBCXX_VISIBILITY(default)
{
_GLIBCXX_BEGIN_NAMESPACE_VERSION
namespace simd
{
/** @internal
* Iterator type for basic_vec and basic_mask.
*
* C++26 [simd.iterator]
*/
template <typename _Vp>
class __iterator
{
friend class __iterator<const _Vp>;
template <typename, typename>
friend class _VecBase;
template <size_t, typename>
friend class _MaskBase;
_Vp* _M_data = nullptr;
__simd_size_type _M_offset = 0;
constexpr
__iterator(_Vp& __d, __simd_size_type __off)
: _M_data(&__d), _M_offset(__off)
{}
public:
using value_type = typename _Vp::value_type;
using iterator_category = input_iterator_tag;
using iterator_concept = random_access_iterator_tag;
using difference_type = __simd_size_type;
constexpr __iterator() = default;
constexpr
__iterator(const __iterator &) = default;
constexpr __iterator&
operator=(const __iterator &) = default;
constexpr
__iterator(const __iterator<remove_const_t<_Vp>> &__i) requires is_const_v<_Vp>
: _M_data(__i._M_data), _M_offset(__i._M_offset)
{}
constexpr value_type
operator*() const
{ return (*_M_data)[_M_offset]; } // checked in operator[]
constexpr __iterator&
operator++()
{
++_M_offset;
return *this;
}
constexpr __iterator
operator++(int)
{
__iterator r = *this;
++_M_offset;
return r;
}
constexpr __iterator&
operator--()
{
--_M_offset;
return *this;
}
constexpr __iterator
operator--(int)
{
__iterator r = *this;
--_M_offset;
return r;
}
constexpr __iterator&
operator+=(difference_type __x)
{
_M_offset += __x;
return *this;
}
constexpr __iterator&
operator-=(difference_type __x)
{
_M_offset -= __x;
return *this;
}
constexpr value_type
operator[](difference_type __i) const
{ return (*_M_data)[_M_offset + __i]; } // checked in operator[]
constexpr friend bool operator==(__iterator __a, __iterator __b) = default;
constexpr friend bool operator==(__iterator __a, std::default_sentinel_t) noexcept
{ return __a._M_offset == _Vp::size.value; }
constexpr friend auto operator<=>(__iterator __a, __iterator __b)
{ return __a._M_offset <=> __b._M_offset; }
constexpr friend __iterator
operator+(const __iterator& __it, difference_type __x)
{ return __iterator(*__it._M_data, __it._M_offset + __x); }
constexpr friend __iterator
operator+(difference_type __x, const __iterator& __it)
{ return __iterator(*__it._M_data, __it._M_offset + __x); }
constexpr friend __iterator
operator-(const __iterator& __it, difference_type __x)
{ return __iterator(*__it._M_data, __it._M_offset - __x); }
constexpr friend difference_type
operator-(__iterator __a, __iterator __b)
{ return __a._M_offset - __b._M_offset; }
constexpr friend difference_type
operator-(__iterator __it, std::default_sentinel_t) noexcept
{ return __it._M_offset - difference_type(_Vp::size.value); }
constexpr friend difference_type
operator-(std::default_sentinel_t, __iterator __it) noexcept
{ return difference_type(_Vp::size.value) - __it._M_offset; }
};
} // namespace simd
_GLIBCXX_END_NAMESPACE_VERSION
} // namespace std
#endif // C++26
#endif // _GLIBCXX_SIMD_ITERATOR_H