[libcxx-commits] [libcxx] [libc++] Optimize ranges::copy_backward for vector<bool>::iterator (PR #121026)
Peng Liu via libcxx-commits
libcxx-commits at lists.llvm.org
Mon Dec 23 21:49:11 PST 2024
https://github.com/winner245 updated https://github.com/llvm/llvm-project/pull/121026
>From 5265f3a4b9a98231bb9b85f3dc2ab80076b5b7f3 Mon Sep 17 00:00:00 2001
From: Peng Liu <winner245 at hotmail.com>
Date: Mon, 23 Dec 2024 12:29:59 -0500
Subject: [PATCH] Optimize ranges::copy_backward for vector<bool>::iterator
---
libcxx/docs/ReleaseNotes/20.rst | 3 +
libcxx/include/__algorithm/copy_backward.h | 131 +++++++++++++++++
libcxx/include/__bit_reference | 134 +-----------------
libcxx/include/bitset | 2 +
.../algorithms/copy_backward.bench.cpp | 55 +++++++
.../alg.copy/copy_backward.pass.cpp | 91 +++++++-----
.../alg.copy/ranges.copy_backward.pass.cpp | 100 ++++++++-----
7 files changed, 318 insertions(+), 198 deletions(-)
create mode 100644 libcxx/test/benchmarks/algorithms/copy_backward.bench.cpp
diff --git a/libcxx/docs/ReleaseNotes/20.rst b/libcxx/docs/ReleaseNotes/20.rst
index c8a07fb8b73348..9c4e43899226f6 100644
--- a/libcxx/docs/ReleaseNotes/20.rst
+++ b/libcxx/docs/ReleaseNotes/20.rst
@@ -73,6 +73,9 @@ Improvements and New Features
optimized, resulting in a performance improvement of up to 2x for trivial element types (e.g., `std::vector<int>`),
and up to 3.4x for non-trivial element types (e.g., `std::vector<std::vector<int>>`).
+- The ``std::ranges::copy_backward`` algorithm has been optimized for ``std::vector<bool>::iterator``\s, resulting in
+ a performance improvement of up to 2000x.
+
Deprecations and Removals
-------------------------
diff --git a/libcxx/include/__algorithm/copy_backward.h b/libcxx/include/__algorithm/copy_backward.h
index 48a768f577f553..02ffc14361e6a9 100644
--- a/libcxx/include/__algorithm/copy_backward.h
+++ b/libcxx/include/__algorithm/copy_backward.h
@@ -10,11 +10,14 @@
#define _LIBCPP___ALGORITHM_COPY_BACKWARD_H
#include <__algorithm/copy_move_common.h>
+#include <__algorithm/copy_n.h>
#include <__algorithm/iterator_operations.h>
#include <__algorithm/min.h>
#include <__config>
+#include <__fwd/bit_reference.h>
#include <__iterator/iterator_traits.h>
#include <__iterator/segmented_iterator.h>
+#include <__memory/pointer_traits.h>
#include <__type_traits/common_type.h>
#include <__type_traits/enable_if.h>
#include <__type_traits/is_constructible.h>
@@ -34,6 +37,124 @@ template <class _AlgPolicy, class _InIter, class _Sent, class _OutIter>
_LIBCPP_HIDE_FROM_ABI _LIBCPP_CONSTEXPR_SINCE_CXX20 pair<_InIter, _OutIter>
__copy_backward(_InIter __first, _Sent __last, _OutIter __result);
+template <class _Cp, bool _IsConst>
+_LIBCPP_CONSTEXPR_SINCE_CXX20 _LIBCPP_HIDE_FROM_ABI __bit_iterator<_Cp, false> __copy_backward_aligned(
+ __bit_iterator<_Cp, _IsConst> __first, __bit_iterator<_Cp, _IsConst> __last, __bit_iterator<_Cp, false> __result) {
+ using _In = __bit_iterator<_Cp, _IsConst>;
+ using difference_type = typename _In::difference_type;
+ using __storage_type = typename _In::__storage_type;
+
+ const int __bits_per_word = _In::__bits_per_word;
+ difference_type __n = __last - __first;
+ if (__n > 0) {
+ // do first word
+ if (__last.__ctz_ != 0) {
+ difference_type __dn = std::min(static_cast<difference_type>(__last.__ctz_), __n);
+ __n -= __dn;
+ unsigned __clz = __bits_per_word - __last.__ctz_;
+ __storage_type __m = (~__storage_type(0) << (__last.__ctz_ - __dn)) & (~__storage_type(0) >> __clz);
+ __storage_type __b = *__last.__seg_ & __m;
+ *__result.__seg_ &= ~__m;
+ *__result.__seg_ |= __b;
+ __result.__ctz_ = static_cast<unsigned>(((-__dn & (__bits_per_word - 1)) + __result.__ctz_) % __bits_per_word);
+ // __last.__ctz_ = 0
+ }
+ // __last.__ctz_ == 0 || __n == 0
+ // __result.__ctz_ == 0 || __n == 0
+ // do middle words
+ __storage_type __nw = __n / __bits_per_word;
+ __result.__seg_ -= __nw;
+ __last.__seg_ -= __nw;
+ std::copy_n(std::__to_address(__last.__seg_), __nw, std::__to_address(__result.__seg_));
+ __n -= __nw * __bits_per_word;
+ // do last word
+ if (__n > 0) {
+ __storage_type __m = ~__storage_type(0) << (__bits_per_word - __n);
+ __storage_type __b = *--__last.__seg_ & __m;
+ *--__result.__seg_ &= ~__m;
+ *__result.__seg_ |= __b;
+ __result.__ctz_ = static_cast<unsigned>(-__n & (__bits_per_word - 1));
+ }
+ }
+ return __result;
+}
+
+template <class _Cp, bool _IsConst>
+_LIBCPP_CONSTEXPR_SINCE_CXX20 _LIBCPP_HIDE_FROM_ABI __bit_iterator<_Cp, false> __copy_backward_unaligned(
+ __bit_iterator<_Cp, _IsConst> __first, __bit_iterator<_Cp, _IsConst> __last, __bit_iterator<_Cp, false> __result) {
+ using _In = __bit_iterator<_Cp, _IsConst>;
+ using difference_type = typename _In::difference_type;
+ using __storage_type = typename _In::__storage_type;
+
+ const int __bits_per_word = _In::__bits_per_word;
+ difference_type __n = __last - __first;
+ if (__n > 0) {
+ // do first word
+ if (__last.__ctz_ != 0) {
+ difference_type __dn = std::min(static_cast<difference_type>(__last.__ctz_), __n);
+ __n -= __dn;
+ unsigned __clz_l = __bits_per_word - __last.__ctz_;
+ __storage_type __m = (~__storage_type(0) << (__last.__ctz_ - __dn)) & (~__storage_type(0) >> __clz_l);
+ __storage_type __b = *__last.__seg_ & __m;
+ unsigned __clz_r = __bits_per_word - __result.__ctz_;
+ __storage_type __ddn = std::min(__dn, static_cast<difference_type>(__result.__ctz_));
+ if (__ddn > 0) {
+ __m = (~__storage_type(0) << (__result.__ctz_ - __ddn)) & (~__storage_type(0) >> __clz_r);
+ *__result.__seg_ &= ~__m;
+ if (__result.__ctz_ > __last.__ctz_)
+ *__result.__seg_ |= __b << (__result.__ctz_ - __last.__ctz_);
+ else
+ *__result.__seg_ |= __b >> (__last.__ctz_ - __result.__ctz_);
+ __result.__ctz_ = static_cast<unsigned>(((-__ddn & (__bits_per_word - 1)) + __result.__ctz_) % __bits_per_word);
+ __dn -= __ddn;
+ }
+ if (__dn > 0) {
+ // __result.__ctz_ == 0
+ --__result.__seg_;
+ __result.__ctz_ = static_cast<unsigned>(-__dn & (__bits_per_word - 1));
+ __m = ~__storage_type(0) << __result.__ctz_;
+ *__result.__seg_ &= ~__m;
+ __last.__ctz_ -= __dn + __ddn;
+ *__result.__seg_ |= __b << (__result.__ctz_ - __last.__ctz_);
+ }
+ // __last.__ctz_ = 0
+ }
+ // __last.__ctz_ == 0 || __n == 0
+ // __result.__ctz_ != 0 || __n == 0
+ // do middle words
+ unsigned __clz_r = __bits_per_word - __result.__ctz_;
+ __storage_type __m = ~__storage_type(0) >> __clz_r;
+ for (; __n >= __bits_per_word; __n -= __bits_per_word) {
+ __storage_type __b = *--__last.__seg_;
+ *__result.__seg_ &= ~__m;
+ *__result.__seg_ |= __b >> __clz_r;
+ *--__result.__seg_ &= __m;
+ *__result.__seg_ |= __b << __result.__ctz_;
+ }
+ // do last word
+ if (__n > 0) {
+ __m = ~__storage_type(0) << (__bits_per_word - __n);
+ __storage_type __b = *--__last.__seg_ & __m;
+ __clz_r = __bits_per_word - __result.__ctz_;
+ __storage_type __dn = std::min(__n, static_cast<difference_type>(__result.__ctz_));
+ __m = (~__storage_type(0) << (__result.__ctz_ - __dn)) & (~__storage_type(0) >> __clz_r);
+ *__result.__seg_ &= ~__m;
+ *__result.__seg_ |= __b >> (__bits_per_word - __result.__ctz_);
+ __result.__ctz_ = static_cast<unsigned>(((-__dn & (__bits_per_word - 1)) + __result.__ctz_) % __bits_per_word);
+ __n -= __dn;
+ if (__n > 0) {
+ // __result.__ctz_ == 0
+ --__result.__seg_;
+ __result.__ctz_ = static_cast<unsigned>(-__n & (__bits_per_word - 1));
+ __m = ~__storage_type(0) << __result.__ctz_;
+ *__result.__seg_ &= ~__m;
+ *__result.__seg_ |= __b << (__result.__ctz_ - (__bits_per_word - __n - __dn));
+ }
+ }
+ }
+ return __result;
+}
+
template <class _AlgPolicy>
struct __copy_backward_impl {
template <class _InIter, class _Sent, class _OutIter>
@@ -107,6 +228,16 @@ struct __copy_backward_impl {
}
}
+ template <class _Cp, bool _IsConst>
+ _LIBCPP_HIDE_FROM_ABI _LIBCPP_CONSTEXPR_SINCE_CXX20 pair<__bit_iterator<_Cp, _IsConst>, __bit_iterator<_Cp, false> >
+ operator()(__bit_iterator<_Cp, _IsConst> __first,
+ __bit_iterator<_Cp, _IsConst> __last,
+ __bit_iterator<_Cp, false> __result) {
+ if (__last.__ctz_ == __result.__ctz_)
+ return std::make_pair(__last, std::__copy_backward_aligned(__first, __last, __result));
+ return std::make_pair(__last, std::__copy_backward_unaligned(__first, __last, __result));
+ }
+
// At this point, the iterators have been unwrapped so any `contiguous_iterator` has been unwrapped to a pointer.
template <class _In, class _Out, __enable_if_t<__can_lower_copy_assignment_to_memmove<_In, _Out>::value, int> = 0>
_LIBCPP_HIDE_FROM_ABI _LIBCPP_CONSTEXPR_SINCE_CXX14 pair<_In*, _Out*>
diff --git a/libcxx/include/__bit_reference b/libcxx/include/__bit_reference
index 9fa24c98d493fd..1d9294993201fd 100644
--- a/libcxx/include/__bit_reference
+++ b/libcxx/include/__bit_reference
@@ -10,6 +10,7 @@
#ifndef _LIBCPP___BIT_REFERENCE
#define _LIBCPP___BIT_REFERENCE
+#include <__algorithm/copy_backward.h>
#include <__algorithm/copy_n.h>
#include <__algorithm/min.h>
#include <__bit/countr.h>
@@ -293,134 +294,6 @@ copy(__bit_iterator<_Cp, _IsConst> __first, __bit_iterator<_Cp, _IsConst> __last
return std::__copy_unaligned(__first, __last, __result);
}
-// copy_backward
-
-template <class _Cp, bool _IsConst>
-_LIBCPP_CONSTEXPR_SINCE_CXX20 _LIBCPP_HIDE_FROM_ABI __bit_iterator<_Cp, false> __copy_backward_aligned(
- __bit_iterator<_Cp, _IsConst> __first, __bit_iterator<_Cp, _IsConst> __last, __bit_iterator<_Cp, false> __result) {
- using _In = __bit_iterator<_Cp, _IsConst>;
- using difference_type = typename _In::difference_type;
- using __storage_type = typename _In::__storage_type;
-
- const int __bits_per_word = _In::__bits_per_word;
- difference_type __n = __last - __first;
- if (__n > 0) {
- // do first word
- if (__last.__ctz_ != 0) {
- difference_type __dn = std::min(static_cast<difference_type>(__last.__ctz_), __n);
- __n -= __dn;
- unsigned __clz = __bits_per_word - __last.__ctz_;
- __storage_type __m = (~__storage_type(0) << (__last.__ctz_ - __dn)) & (~__storage_type(0) >> __clz);
- __storage_type __b = *__last.__seg_ & __m;
- *__result.__seg_ &= ~__m;
- *__result.__seg_ |= __b;
- __result.__ctz_ = static_cast<unsigned>(((-__dn & (__bits_per_word - 1)) + __result.__ctz_) % __bits_per_word);
- // __last.__ctz_ = 0
- }
- // __last.__ctz_ == 0 || __n == 0
- // __result.__ctz_ == 0 || __n == 0
- // do middle words
- __storage_type __nw = __n / __bits_per_word;
- __result.__seg_ -= __nw;
- __last.__seg_ -= __nw;
- std::copy_n(std::__to_address(__last.__seg_), __nw, std::__to_address(__result.__seg_));
- __n -= __nw * __bits_per_word;
- // do last word
- if (__n > 0) {
- __storage_type __m = ~__storage_type(0) << (__bits_per_word - __n);
- __storage_type __b = *--__last.__seg_ & __m;
- *--__result.__seg_ &= ~__m;
- *__result.__seg_ |= __b;
- __result.__ctz_ = static_cast<unsigned>(-__n & (__bits_per_word - 1));
- }
- }
- return __result;
-}
-
-template <class _Cp, bool _IsConst>
-_LIBCPP_CONSTEXPR_SINCE_CXX20 _LIBCPP_HIDE_FROM_ABI __bit_iterator<_Cp, false> __copy_backward_unaligned(
- __bit_iterator<_Cp, _IsConst> __first, __bit_iterator<_Cp, _IsConst> __last, __bit_iterator<_Cp, false> __result) {
- using _In = __bit_iterator<_Cp, _IsConst>;
- using difference_type = typename _In::difference_type;
- using __storage_type = typename _In::__storage_type;
-
- const int __bits_per_word = _In::__bits_per_word;
- difference_type __n = __last - __first;
- if (__n > 0) {
- // do first word
- if (__last.__ctz_ != 0) {
- difference_type __dn = std::min(static_cast<difference_type>(__last.__ctz_), __n);
- __n -= __dn;
- unsigned __clz_l = __bits_per_word - __last.__ctz_;
- __storage_type __m = (~__storage_type(0) << (__last.__ctz_ - __dn)) & (~__storage_type(0) >> __clz_l);
- __storage_type __b = *__last.__seg_ & __m;
- unsigned __clz_r = __bits_per_word - __result.__ctz_;
- __storage_type __ddn = std::min(__dn, static_cast<difference_type>(__result.__ctz_));
- if (__ddn > 0) {
- __m = (~__storage_type(0) << (__result.__ctz_ - __ddn)) & (~__storage_type(0) >> __clz_r);
- *__result.__seg_ &= ~__m;
- if (__result.__ctz_ > __last.__ctz_)
- *__result.__seg_ |= __b << (__result.__ctz_ - __last.__ctz_);
- else
- *__result.__seg_ |= __b >> (__last.__ctz_ - __result.__ctz_);
- __result.__ctz_ = static_cast<unsigned>(((-__ddn & (__bits_per_word - 1)) + __result.__ctz_) % __bits_per_word);
- __dn -= __ddn;
- }
- if (__dn > 0) {
- // __result.__ctz_ == 0
- --__result.__seg_;
- __result.__ctz_ = static_cast<unsigned>(-__dn & (__bits_per_word - 1));
- __m = ~__storage_type(0) << __result.__ctz_;
- *__result.__seg_ &= ~__m;
- __last.__ctz_ -= __dn + __ddn;
- *__result.__seg_ |= __b << (__result.__ctz_ - __last.__ctz_);
- }
- // __last.__ctz_ = 0
- }
- // __last.__ctz_ == 0 || __n == 0
- // __result.__ctz_ != 0 || __n == 0
- // do middle words
- unsigned __clz_r = __bits_per_word - __result.__ctz_;
- __storage_type __m = ~__storage_type(0) >> __clz_r;
- for (; __n >= __bits_per_word; __n -= __bits_per_word) {
- __storage_type __b = *--__last.__seg_;
- *__result.__seg_ &= ~__m;
- *__result.__seg_ |= __b >> __clz_r;
- *--__result.__seg_ &= __m;
- *__result.__seg_ |= __b << __result.__ctz_;
- }
- // do last word
- if (__n > 0) {
- __m = ~__storage_type(0) << (__bits_per_word - __n);
- __storage_type __b = *--__last.__seg_ & __m;
- __clz_r = __bits_per_word - __result.__ctz_;
- __storage_type __dn = std::min(__n, static_cast<difference_type>(__result.__ctz_));
- __m = (~__storage_type(0) << (__result.__ctz_ - __dn)) & (~__storage_type(0) >> __clz_r);
- *__result.__seg_ &= ~__m;
- *__result.__seg_ |= __b >> (__bits_per_word - __result.__ctz_);
- __result.__ctz_ = static_cast<unsigned>(((-__dn & (__bits_per_word - 1)) + __result.__ctz_) % __bits_per_word);
- __n -= __dn;
- if (__n > 0) {
- // __result.__ctz_ == 0
- --__result.__seg_;
- __result.__ctz_ = static_cast<unsigned>(-__n & (__bits_per_word - 1));
- __m = ~__storage_type(0) << __result.__ctz_;
- *__result.__seg_ &= ~__m;
- *__result.__seg_ |= __b << (__result.__ctz_ - (__bits_per_word - __n - __dn));
- }
- }
- }
- return __result;
-}
-
-template <class _Cp, bool _IsConst>
-inline _LIBCPP_HIDE_FROM_ABI _LIBCPP_CONSTEXPR_SINCE_CXX20 __bit_iterator<_Cp, false> copy_backward(
- __bit_iterator<_Cp, _IsConst> __first, __bit_iterator<_Cp, _IsConst> __last, __bit_iterator<_Cp, false> __result) {
- if (__last.__ctz_ == __result.__ctz_)
- return std::__copy_backward_aligned(__first, __last, __result);
- return std::__copy_backward_unaligned(__first, __last, __result);
-}
-
// move
template <class _Cp, bool _IsConst>
@@ -983,9 +856,8 @@ private:
template <class _Dp, bool _IC>
_LIBCPP_CONSTEXPR_SINCE_CXX20 friend __bit_iterator<_Dp, false> __copy_backward_unaligned(
__bit_iterator<_Dp, _IC> __first, __bit_iterator<_Dp, _IC> __last, __bit_iterator<_Dp, false> __result);
- template <class _Dp, bool _IC>
- _LIBCPP_CONSTEXPR_SINCE_CXX20 friend __bit_iterator<_Dp, false>
- copy_backward(__bit_iterator<_Dp, _IC> __first, __bit_iterator<_Dp, _IC> __last, __bit_iterator<_Dp, false> __result);
+ template <class _AlgPolicy>
+ friend struct __copy_backward_impl;
template <class _Cl, class _Cr>
friend __bit_iterator<_Cr, false>
__swap_ranges_aligned(__bit_iterator<_Cl, false>, __bit_iterator<_Cl, false>, __bit_iterator<_Cr, false>);
diff --git a/libcxx/include/bitset b/libcxx/include/bitset
index 8b361824805571..bb09bc5415fb03 100644
--- a/libcxx/include/bitset
+++ b/libcxx/include/bitset
@@ -129,6 +129,8 @@ template <size_t N> struct hash<std::bitset<N>>;
#if __cplusplus < 201103L && defined(_LIBCPP_USE_FROZEN_CXX03_HEADERS)
# include <__cxx03/bitset>
#else
+# include <__algorithm/copy.h>
+# include <__algorithm/copy_backward.h>
# include <__algorithm/count.h>
# include <__algorithm/fill.h>
# include <__algorithm/fill_n.h>
diff --git a/libcxx/test/benchmarks/algorithms/copy_backward.bench.cpp b/libcxx/test/benchmarks/algorithms/copy_backward.bench.cpp
new file mode 100644
index 00000000000000..3f797b5555a138
--- /dev/null
+++ b/libcxx/test/benchmarks/algorithms/copy_backward.bench.cpp
@@ -0,0 +1,55 @@
+//===----------------------------------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+// UNSUPPORTED: c++03, c++11, c++14, c++17, c++20
+
+#include <algorithm>
+#include <benchmark/benchmark.h>
+#include <vector>
+
+static void bm_ranges_copy_backward(benchmark::State& state, bool aligned) {
+ auto n = state.range();
+ std::vector<bool> in(n, true);
+ std::vector<bool> out(aligned ? n : n + 8);
+ benchmark::DoNotOptimize(&in);
+ auto dst = aligned ? out.end() : out.end() - 4;
+ for (auto _ : state) {
+ benchmark::DoNotOptimize(std::ranges::copy_backward(in, dst));
+ benchmark::DoNotOptimize(&out);
+ }
+}
+
+static void bm_copy_backward(benchmark::State& state, bool aligned) {
+ auto n = state.range();
+ std::vector<bool> in(n, true);
+ std::vector<bool> out(aligned ? n : n + 8);
+ benchmark::DoNotOptimize(&in);
+ auto beg = in.begin();
+ auto end = in.end();
+ auto dst = aligned ? out.end() : out.end() - 4;
+ for (auto _ : state) {
+ benchmark::DoNotOptimize(std::copy_backward(beg, end, dst));
+ benchmark::DoNotOptimize(&out);
+ }
+}
+
+static void bm_ranges_copy_backward_aligned(benchmark::State& state) { bm_ranges_copy_backward(state, true); }
+static void bm_ranges_copy_backward_unaligned(benchmark::State& state) { bm_ranges_copy_backward(state, false); }
+
+static void bm_copy_backward_aligned(benchmark::State& state) { bm_copy_backward(state, true); }
+static void bm_copy_backward_unaligned(benchmark::State& state) { bm_copy_backward(state, false); }
+
+// Test the range version of std::copy for vector<bool>::iterator
+BENCHMARK(bm_ranges_copy_backward_aligned)->Range(8, 1 << 16)->DenseRange(102400, 204800, 4096);
+BENCHMARK(bm_ranges_copy_backward_unaligned)->Range(8, 1 << 20);
+
+// Test the iterator-pair version of std::copy for vector<bool>::iterator
+BENCHMARK(bm_copy_backward_aligned)->Range(8, 1 << 20);
+BENCHMARK(bm_copy_backward_unaligned)->Range(8, 1 << 20);
+
+BENCHMARK_MAIN();
diff --git a/libcxx/test/std/algorithms/alg.modifying.operations/alg.copy/copy_backward.pass.cpp b/libcxx/test/std/algorithms/alg.modifying.operations/alg.copy/copy_backward.pass.cpp
index 928903de1ade2d..74982aaccac4bf 100644
--- a/libcxx/test/std/algorithms/alg.modifying.operations/alg.copy/copy_backward.pass.cpp
+++ b/libcxx/test/std/algorithms/alg.modifying.operations/alg.copy/copy_backward.pass.cpp
@@ -15,6 +15,7 @@
#include <algorithm>
#include <cassert>
+#include <vector>
#include "test_macros.h"
#include "test_iterators.h"
@@ -36,47 +37,66 @@ class Derived : public PaddedBase {
};
template <class InIter, class OutIter>
-TEST_CONSTEXPR_CXX20 void
-test_copy_backward()
-{
+TEST_CONSTEXPR_CXX20 void test_copy_backward() {
{
const unsigned N = 1000;
- int ia[N] = {};
+ int ia[N] = {};
for (unsigned i = 0; i < N; ++i)
- ia[i] = i;
+ ia[i] = i;
int ib[N] = {0};
- OutIter r = std::copy_backward(InIter(ia), InIter(ia+N), OutIter(ib+N));
+ OutIter r = std::copy_backward(InIter(ia), InIter(ia + N), OutIter(ib + N));
assert(base(r) == ib);
for (unsigned i = 0; i < N; ++i)
- assert(ia[i] == ib[i]);
+ assert(ia[i] == ib[i]);
}
}
-TEST_CONSTEXPR_CXX20 bool
-test()
-{
- test_copy_backward<bidirectional_iterator<const int*>, bidirectional_iterator<int*> >();
- test_copy_backward<bidirectional_iterator<const int*>, random_access_iterator<int*> >();
- test_copy_backward<bidirectional_iterator<const int*>, int*>();
+template <std::size_t N>
+struct TestBitIter {
+ std::vector<bool> in;
+ TEST_CONSTEXPR TestBitIter() : in(N, false) {
+ for (std::size_t i = 0; i < N; i += 2)
+ in[i] = true;
+ }
+ TEST_CONSTEXPR_CXX20 void operator()() {
+ { // Test copy_backward with aligned bytes
+ std::vector<bool> out(N);
+ std::copy_backward(in.begin(), in.end(), out.end());
+ assert(in == out);
+ }
+ { // Test copy_backward with unaligned bytes
+ std::vector<bool> out(N + 8);
+ std::copy_backward(in.begin(), in.end(), out.end() - 4);
+ for (std::size_t i = 0; i < N; ++i)
+ assert(out[i + 4] == in[i]);
+ }
+ }
+ TEST_CONSTEXPR std::size_t size() const { return N; }
+};
- test_copy_backward<random_access_iterator<const int*>, bidirectional_iterator<int*> >();
- test_copy_backward<random_access_iterator<const int*>, random_access_iterator<int*> >();
- test_copy_backward<random_access_iterator<const int*>, int*>();
+TEST_CONSTEXPR_CXX20 bool test() {
+ test_copy_backward<bidirectional_iterator<const int*>, bidirectional_iterator<int*> >();
+ test_copy_backward<bidirectional_iterator<const int*>, random_access_iterator<int*> >();
+ test_copy_backward<bidirectional_iterator<const int*>, int*>();
- test_copy_backward<const int*, bidirectional_iterator<int*> >();
- test_copy_backward<const int*, random_access_iterator<int*> >();
- test_copy_backward<const int*, int*>();
+ test_copy_backward<random_access_iterator<const int*>, bidirectional_iterator<int*> >();
+ test_copy_backward<random_access_iterator<const int*>, random_access_iterator<int*> >();
+ test_copy_backward<random_access_iterator<const int*>, int*>();
+
+ test_copy_backward<const int*, bidirectional_iterator<int*> >();
+ test_copy_backward<const int*, random_access_iterator<int*> >();
+ test_copy_backward<const int*, int*>();
#if TEST_STD_VER > 17
- test_copy_backward<contiguous_iterator<const int*>, bidirectional_iterator<int*>>();
- test_copy_backward<contiguous_iterator<const int*>, random_access_iterator<int*>>();
- test_copy_backward<contiguous_iterator<const int*>, int*>();
-
- test_copy_backward<bidirectional_iterator<const int*>, contiguous_iterator<int*>>();
- test_copy_backward<random_access_iterator<const int*>, contiguous_iterator<int*>>();
- test_copy_backward<contiguous_iterator<const int*>, contiguous_iterator<int*>>();
- test_copy_backward<const int*, contiguous_iterator<int*>>();
+ test_copy_backward<contiguous_iterator<const int*>, bidirectional_iterator<int*>>();
+ test_copy_backward<contiguous_iterator<const int*>, random_access_iterator<int*>>();
+ test_copy_backward<contiguous_iterator<const int*>, int*>();
+
+ test_copy_backward<bidirectional_iterator<const int*>, contiguous_iterator<int*>>();
+ test_copy_backward<random_access_iterator<const int*>, contiguous_iterator<int*>>();
+ test_copy_backward<contiguous_iterator<const int*>, contiguous_iterator<int*>>();
+ test_copy_backward<const int*, contiguous_iterator<int*>>();
#endif
{ // Make sure that padding bits aren't copied
@@ -96,15 +116,22 @@ test()
assert(std::equal(a, a + 10, expected));
}
- return true;
+ { // Test vector<bool>::iterator optimization
+ TestBitIter<8>()();
+ TestBitIter<16>()();
+ TestBitIter<32>()();
+ TestBitIter<64>()();
+ TestBitIter<1024>()();
+ }
+
+ return true;
}
-int main(int, char**)
-{
- test();
+int main(int, char**) {
+ test();
#if TEST_STD_VER > 17
- static_assert(test());
+ static_assert(test());
#endif
return 0;
diff --git a/libcxx/test/std/algorithms/alg.modifying.operations/alg.copy/ranges.copy_backward.pass.cpp b/libcxx/test/std/algorithms/alg.modifying.operations/alg.copy/ranges.copy_backward.pass.cpp
index 343447446ab2d7..ec4b7a858e21fd 100644
--- a/libcxx/test/std/algorithms/alg.modifying.operations/alg.copy/ranges.copy_backward.pass.cpp
+++ b/libcxx/test/std/algorithms/alg.modifying.operations/alg.copy/ranges.copy_backward.pass.cpp
@@ -61,16 +61,16 @@ template <class In, class Out, class Sent>
constexpr void test_iterators() {
{ // simple test
{
- std::array in {1, 2, 3, 4};
+ std::array in{1, 2, 3, 4};
std::array<int, 4> out;
std::same_as<std::ranges::in_out_result<In, Out>> auto ret =
- std::ranges::copy_backward(In(in.data()), Sent(In(in.data() + in.size())), Out(out.data() + out.size()));
+ std::ranges::copy_backward(In(in.data()), Sent(In(in.data() + in.size())), Out(out.data() + out.size()));
assert(in == out);
assert(base(ret.in) == in.data() + in.size());
assert(base(ret.out) == out.data());
}
{
- std::array in {1, 2, 3, 4};
+ std::array in{1, 2, 3, 4};
std::array<int, 4> out;
auto range = std::ranges::subrange(In(in.data()), Sent(In(in.data() + in.size())));
std::same_as<std::ranges::in_out_result<In, Out>> auto ret =
@@ -94,7 +94,7 @@ constexpr void test_iterators() {
std::array<int, 0> in;
std::array<int, 0> out;
auto range = std::ranges::subrange(In(in.data()), Sent(In(in.data() + in.size())));
- auto ret = std::ranges::copy_backward(range, Out(out.data()));
+ auto ret = std::ranges::copy_backward(range, Out(out.data()));
assert(base(ret.in) == in.data() + in.size());
assert(base(ret.out) == out.data());
}
@@ -104,16 +104,16 @@ constexpr void test_iterators() {
template <class InContainer, class OutContainer, class In, class Out, class Sent = In>
constexpr void test_containers() {
{
- InContainer in {1, 2, 3, 4};
+ InContainer in{1, 2, 3, 4};
OutContainer out(4);
std::same_as<std::ranges::in_out_result<In, Out>> auto ret =
- std::ranges::copy_backward(In(in.begin()), Sent(In(in.end())), Out(out.end()));
+ std::ranges::copy_backward(In(in.begin()), Sent(In(in.end())), Out(out.end()));
assert(std::ranges::equal(in, out));
assert(base(ret.in) == in.end());
assert(base(ret.out) == out.begin());
}
{
- InContainer in {1, 2, 3, 4};
+ InContainer in{1, 2, 3, 4};
OutContainer out(4);
auto range = std::ranges::subrange(In(in.begin()), Sent(In(in.end())));
std::same_as<std::ranges::in_out_result<In, Out>> auto ret = std::ranges::copy_backward(range, Out(out.end()));
@@ -125,13 +125,12 @@ constexpr void test_containers() {
template <class Iter, class Sent>
constexpr void test_join_view() {
- auto to_subranges = std::views::transform([](auto& vec) {
- return std::ranges::subrange(Iter(vec.begin()), Sent(Iter(vec.end())));
- });
+ auto to_subranges =
+ std::views::transform([](auto& vec) { return std::ranges::subrange(Iter(vec.begin()), Sent(Iter(vec.end()))); });
{ // segmented -> contiguous
std::vector<std::vector<int>> vectors = {};
- auto range = vectors | to_subranges;
+ auto range = vectors | to_subranges;
std::vector<std::ranges::subrange<Iter, Sent>> subrange_vector(range.begin(), range.end());
std::array<int, 0> arr;
@@ -140,7 +139,7 @@ constexpr void test_join_view() {
}
{ // segmented -> contiguous
std::vector<std::vector<int>> vectors = {{1, 2, 3, 4}, {5, 6, 7, 8}, {9, 10}, {}};
- auto range = vectors | to_subranges;
+ auto range = vectors | to_subranges;
std::vector<std::ranges::subrange<Iter, Sent>> subrange_vector(range.begin(), range.end());
std::array<int, 10> arr;
@@ -149,7 +148,7 @@ constexpr void test_join_view() {
}
{ // contiguous -> segmented
std::vector<std::vector<int>> vectors = {{0, 0, 0, 0}, {0, 0}, {0, 0, 0, 0}, {}};
- auto range = vectors | to_subranges;
+ auto range = vectors | to_subranges;
std::vector<std::ranges::subrange<Iter, Sent>> subrange_vector(range.begin(), range.end());
std::array arr = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10};
@@ -158,10 +157,10 @@ constexpr void test_join_view() {
}
{ // segmented -> segmented
std::vector<std::vector<int>> vectors = {{1, 2, 3, 4}, {5, 6, 7, 8}, {9, 10}, {}};
- auto range1 = vectors | to_subranges;
+ auto range1 = vectors | to_subranges;
std::vector<std::ranges::subrange<Iter, Sent>> subrange_vector(range1.begin(), range1.end());
std::vector<std::vector<int>> to_vectors = {{0, 0, 0, 0}, {0, 0, 0, 0}, {}, {0, 0}};
- auto range2 = to_vectors | to_subranges;
+ auto range2 = to_vectors | to_subranges;
std::vector<std::ranges::subrange<Iter, Sent>> to_subrange_vector(range2.begin(), range2.end());
std::ranges::copy_backward(subrange_vector | std::views::join, (to_subrange_vector | std::views::join).end());
@@ -224,6 +223,29 @@ constexpr void test_proxy_in_iterators() {
test_sentinels<ProxyIterator, Out>();
}
+template <std::size_t N>
+struct TestBitIter {
+ std::vector<bool> in;
+ TEST_CONSTEXPR TestBitIter() : in(N, false) {
+ for (std::size_t i = 0; i < N; i += 2)
+ in[i] = true;
+ }
+ TEST_CONSTEXPR_CXX20 void operator()() {
+ { // Test copy_backward with aligned bytes
+ std::vector<bool> out(N);
+ std::ranges::copy_backward(in, out.end());
+ assert(in == out);
+ }
+ { // Test copy_backward with unaligned bytes
+ std::vector<bool> out(N + 8);
+ std::ranges::copy_backward(in, out.end() - 4);
+ for (std::size_t i = 0; i < N; ++i)
+ assert(out[i + 4] == in[i]);
+ }
+ }
+ TEST_CONSTEXPR std::size_t size() const { return N; }
+};
+
constexpr bool test() {
test_in_iterators<bidirectional_iterator>();
test_in_iterators<random_access_iterator>();
@@ -237,13 +259,13 @@ constexpr bool test() {
{ // check that ranges::dangling is returned
std::array<int, 4> out;
std::same_as<std::ranges::in_out_result<std::ranges::dangling, int*>> auto ret =
- std::ranges::copy_backward(std::array {1, 2, 3, 4}, out.data() + out.size());
+ std::ranges::copy_backward(std::array{1, 2, 3, 4}, out.data() + out.size());
assert(ret.out == out.data());
assert((out == std::array{1, 2, 3, 4}));
}
{ // check that an iterator is returned with a borrowing range
- std::array in {1, 2, 3, 4};
+ std::array in{1, 2, 3, 4};
std::array<int, 4> out;
std::same_as<std::ranges::in_out_result<std::array<int, 4>::iterator, int*>> auto ret =
std::ranges::copy_backward(std::views::all(in), out.data() + out.size());
@@ -254,8 +276,8 @@ constexpr bool test() {
{ // check that every element is copied exactly once
struct CopyOnce {
- bool copied = false;
- constexpr CopyOnce() = default;
+ bool copied = false;
+ constexpr CopyOnce() = default;
constexpr CopyOnce(const CopyOnce& other) = delete;
constexpr CopyOnce& operator=(const CopyOnce& other) {
assert(!other.copied);
@@ -264,16 +286,16 @@ constexpr bool test() {
}
};
{
- std::array<CopyOnce, 4> in {};
- std::array<CopyOnce, 4> out {};
+ std::array<CopyOnce, 4> in{};
+ std::array<CopyOnce, 4> out{};
auto ret = std::ranges::copy_backward(in.begin(), in.end(), out.end());
assert(ret.in == in.end());
assert(ret.out == out.begin());
assert(std::all_of(out.begin(), out.end(), [](const auto& e) { return e.copied; }));
}
{
- std::array<CopyOnce, 4> in {};
- std::array<CopyOnce, 4> out {};
+ std::array<CopyOnce, 4> in{};
+ std::array<CopyOnce, 4> out{};
auto ret = std::ranges::copy_backward(in, out.end());
assert(ret.in == in.end());
assert(ret.out == out.begin());
@@ -284,8 +306,8 @@ constexpr bool test() {
{ // check that the range is copied backwards
struct OnlyBackwardsCopyable {
OnlyBackwardsCopyable* next = nullptr;
- bool canCopy = false;
- OnlyBackwardsCopyable() = default;
+ bool canCopy = false;
+ OnlyBackwardsCopyable() = default;
constexpr OnlyBackwardsCopyable& operator=(const OnlyBackwardsCopyable&) {
assert(canCopy);
if (next != nullptr)
@@ -294,12 +316,12 @@ constexpr bool test() {
}
};
{
- std::array<OnlyBackwardsCopyable, 3> in {};
- std::array<OnlyBackwardsCopyable, 3> out {};
- out[1].next = &out[0];
- out[2].next = &out[1];
+ std::array<OnlyBackwardsCopyable, 3> in{};
+ std::array<OnlyBackwardsCopyable, 3> out{};
+ out[1].next = &out[0];
+ out[2].next = &out[1];
out[2].canCopy = true;
- auto ret = std::ranges::copy_backward(in, out.end());
+ auto ret = std::ranges::copy_backward(in, out.end());
assert(ret.in == in.end());
assert(ret.out == out.begin());
assert(out[0].canCopy);
@@ -307,12 +329,12 @@ constexpr bool test() {
assert(out[2].canCopy);
}
{
- std::array<OnlyBackwardsCopyable, 3> in {};
- std::array<OnlyBackwardsCopyable, 3> out {};
- out[1].next = &out[0];
- out[2].next = &out[1];
+ std::array<OnlyBackwardsCopyable, 3> in{};
+ std::array<OnlyBackwardsCopyable, 3> out{};
+ out[1].next = &out[0];
+ out[2].next = &out[1];
out[2].canCopy = true;
- auto ret = std::ranges::copy_backward(in.begin(), in.end(), out.end());
+ auto ret = std::ranges::copy_backward(in.begin(), in.end(), out.end());
assert(ret.in == in.end());
assert(ret.out == out.begin());
assert(out[0].canCopy);
@@ -321,6 +343,14 @@ constexpr bool test() {
}
}
+ { // Test vector<bool>::iterator optimization
+ TestBitIter<8>()();
+ TestBitIter<16>()();
+ TestBitIter<32>()();
+ TestBitIter<64>()();
+ TestBitIter<1024>()();
+ }
+
return true;
}
More information about the libcxx-commits
mailing list