[libc++] Add remaining benchmarks from [alg.modifying.operations] #127354
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
llvmbot
added
the
libc++
|
@llvm/pr-subscribers-libcxx Author: Louis Dionne (ldionne) ChangesThis patch adds benchmarks for all the remaining algorithms in [alg.modifying.operations] that we didn't already have a benchmark for. Patch is 93.48 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/127354.diff 28 Files Affected:
diff --git a/libcxx/include/module.modulemap b/libcxx/include/module.modulemap
index b0720703bd0de..6af6e96af7d75 100644
--- a/libcxx/include/module.modulemap
+++ b/libcxx/include/module.modulemap
@@ -700,6 +700,7 @@ module std [system] {
}
module ranges_remove {
header "__algorithm/ranges_remove.h"
+ export std.ranges.subrange // return type
}
module ranges_replace_copy_if {
header "__algorithm/ranges_replace_copy_if.h"
@@ -724,7 +725,7 @@ module std [system] {
}
module ranges_rotate_copy {
header "__algorithm/ranges_rotate_copy.h"
- export std.algorithm.in_out_result
+ export std.ranges.subrange // return type
}
module ranges_rotate { header "__algorithm/ranges_rotate.h" }
module ranges_sample { header "__algorithm/ranges_sample.h" }
diff --git a/libcxx/test/benchmarks/algorithms/fill.bench.cpp b/libcxx/test/benchmarks/algorithms/fill.bench.cpp
deleted file mode 100644
index 6a48b25b7eb63..0000000000000
--- a/libcxx/test/benchmarks/algorithms/fill.bench.cpp
+++ /dev/null
@@ -1,51 +0,0 @@
-//===----------------------------------------------------------------------===//
-//
-// 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_fill_n_vector_bool(benchmark::State& state) {
- std::vector<bool> vec1(state.range());
- for (auto _ : state) {
- benchmark::DoNotOptimize(vec1);
- benchmark::DoNotOptimize(std::fill_n(vec1.begin(), vec1.size(), false));
- }
-}
-BENCHMARK(bm_fill_n_vector_bool)->DenseRange(1, 8)->Range(16, 1 << 20);
-
-static void bm_ranges_fill_n_vector_bool(benchmark::State& state) {
- std::vector<bool> vec1(state.range());
- for (auto _ : state) {
- benchmark::DoNotOptimize(vec1);
- benchmark::DoNotOptimize(std::ranges::fill_n(vec1.begin(), vec1.size(), false));
- }
-}
-BENCHMARK(bm_ranges_fill_n_vector_bool)->DenseRange(1, 8)->Range(16, 1 << 20);
-
-static void bm_fill_vector_bool(benchmark::State& state) {
- std::vector<bool> vec1(state.range());
- for (auto _ : state) {
- benchmark::DoNotOptimize(vec1);
- std::fill(vec1.begin(), vec1.end(), false);
- }
-}
-BENCHMARK(bm_fill_vector_bool)->DenseRange(1, 8)->Range(16, 1 << 20);
-
-static void bm_ranges_fill_vector_bool(benchmark::State& state) {
- std::vector<bool> vec1(state.range());
- for (auto _ : state) {
- benchmark::DoNotOptimize(vec1);
- benchmark::DoNotOptimize(std::ranges::fill(vec1, false));
- }
-}
-BENCHMARK(bm_ranges_fill_vector_bool)->DenseRange(1, 8)->Range(16, 1 << 20);
-
-BENCHMARK_MAIN();
diff --git a/libcxx/test/benchmarks/algorithms/modifying/fill.bench.cpp b/libcxx/test/benchmarks/algorithms/modifying/fill.bench.cpp
new file mode 100644
index 0000000000000..64c7364be6549
--- /dev/null
+++ b/libcxx/test/benchmarks/algorithms/modifying/fill.bench.cpp
@@ -0,0 +1,86 @@
+//===----------------------------------------------------------------------===//
+//
+// 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
+
+#include <algorithm>
+#include <cstddef>
+#include <deque>
+#include <iterator>
+#include <list>
+#include <string>
+#include <vector>
+
+#include "benchmark/benchmark.h"
+#include "../../GenerateInput.h"
+#include "test_macros.h"
+
+template <class Container, class Operation>
+void bm(std::string operation_name, Operation fill) {
+ auto bench = [fill](auto& st) {
+ std::size_t const size = st.range(0);
+ using ValueType = typename Container::value_type;
+ ValueType x = Generate<ValueType>::random();
+ ValueType y = Generate<ValueType>::random();
+ Container c(size, y);
+
+ for ([[maybe_unused]] auto _ : st) {
+ fill(c.begin(), c.end(), x);
+ std::swap(x, y);
+ benchmark::DoNotOptimize(c);
+ benchmark::DoNotOptimize(x);
+ benchmark::DoNotOptimize(y);
+ benchmark::ClobberMemory();
+ }
+ };
+ benchmark::RegisterBenchmark(operation_name, bench)->Arg(32)->Arg(1024)->Arg(8192);
+}
+
+template <class Operation>
+void bm_vector_bool(std::string operation_name, Operation fill) {
+ auto bench = [fill](auto& st) {
+ std::size_t const size = st.range(0);
+ bool x = true;
+ bool y = false;
+ std::vector<bool> c(size, y);
+
+ for ([[maybe_unused]] auto _ : st) {
+ fill(c.begin(), c.end(), x);
+ std::swap(x, y);
+ benchmark::DoNotOptimize(c);
+ benchmark::DoNotOptimize(x);
+ benchmark::DoNotOptimize(y);
+ benchmark::ClobberMemory();
+ }
+ };
+ benchmark::RegisterBenchmark(operation_name, bench)->Arg(32)->Arg(1024)->Arg(8192);
+}
+
+int main(int argc, char** argv) {
+ auto std_fill = [](auto first, auto last, auto const& value) { return std::fill(first, last, value); };
+ auto ranges_fill = [](auto first, auto last, auto const& value) { return std::ranges::fill(first, last, value); };
+
+ // std::fill
+ bm<std::vector<int>>("std::fill(vector<int>)", std_fill);
+ bm<std::deque<int>>("std::fill(deque<int>)", std_fill);
+ bm<std::list<int>>("std::fill(list<int>)", std_fill);
+ bm_vector_bool("std::fill(vector<bool>)", std_fill);
+
+ // ranges::fill
+ bm<std::vector<int>>("ranges::fill(vector<int>)", ranges_fill);
+ bm<std::deque<int>>("ranges::fill(deque<int>)", ranges_fill);
+ bm<std::list<int>>("ranges::fill(list<int>)", ranges_fill);
+#if TEST_STD_VER >= 23 // vector<bool>::iterator is not an output_iterator before C++23
+ bm_vector_bool("ranges::fill(vector<bool>)", ranges_fill);
+#endif
+
+ benchmark::Initialize(&argc, argv);
+ benchmark::RunSpecifiedBenchmarks();
+ benchmark::Shutdown();
+ return 0;
+}
diff --git a/libcxx/test/benchmarks/algorithms/modifying/fill_n.bench.cpp b/libcxx/test/benchmarks/algorithms/modifying/fill_n.bench.cpp
new file mode 100644
index 0000000000000..aaeb0982956f9
--- /dev/null
+++ b/libcxx/test/benchmarks/algorithms/modifying/fill_n.bench.cpp
@@ -0,0 +1,86 @@
+//===----------------------------------------------------------------------===//
+//
+// 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
+
+#include <algorithm>
+#include <cstddef>
+#include <deque>
+#include <iterator>
+#include <list>
+#include <string>
+#include <vector>
+
+#include "benchmark/benchmark.h"
+#include "../../GenerateInput.h"
+#include "test_macros.h"
+
+template <class Container, class Operation>
+void bm(std::string operation_name, Operation fill_n) {
+ auto bench = [fill_n](auto& st) {
+ std::size_t const size = st.range(0);
+ using ValueType = typename Container::value_type;
+ ValueType x = Generate<ValueType>::random();
+ ValueType y = Generate<ValueType>::random();
+ Container c(size, y);
+
+ for ([[maybe_unused]] auto _ : st) {
+ fill_n(c.begin(), size, x);
+ std::swap(x, y);
+ benchmark::DoNotOptimize(c);
+ benchmark::DoNotOptimize(x);
+ benchmark::DoNotOptimize(y);
+ benchmark::ClobberMemory();
+ }
+ };
+ benchmark::RegisterBenchmark(operation_name, bench)->Arg(32)->Arg(1024)->Arg(8192);
+}
+
+template <class Operation>
+void bm_vector_bool(std::string operation_name, Operation fill_n) {
+ auto bench = [fill_n](auto& st) {
+ std::size_t const size = st.range(0);
+ bool x = true;
+ bool y = false;
+ std::vector<bool> c(size, y);
+
+ for ([[maybe_unused]] auto _ : st) {
+ fill_n(c.begin(), size, x);
+ std::swap(x, y);
+ benchmark::DoNotOptimize(c);
+ benchmark::DoNotOptimize(x);
+ benchmark::DoNotOptimize(y);
+ benchmark::ClobberMemory();
+ }
+ };
+ benchmark::RegisterBenchmark(operation_name, bench)->Arg(32)->Arg(1024)->Arg(8192);
+}
+
+int main(int argc, char** argv) {
+ auto std_fill_n = [](auto out, auto n, auto const& value) { return std::fill_n(out, n, value); };
+ auto ranges_fill_n = [](auto out, auto n, auto const& value) { return std::ranges::fill_n(out, n, value); };
+
+ // std::fill_n
+ bm<std::vector<int>>("std::fill_n(vector<int>)", std_fill_n);
+ bm<std::deque<int>>("std::fill_n(deque<int>)", std_fill_n);
+ bm<std::list<int>>("std::fill_n(list<int>)", std_fill_n);
+ bm_vector_bool("std::fill_n(vector<bool>)", std_fill_n);
+
+ // ranges::fill_n
+ bm<std::vector<int>>("ranges::fill_n(vector<int>)", ranges_fill_n);
+ bm<std::deque<int>>("ranges::fill_n(deque<int>)", ranges_fill_n);
+ bm<std::list<int>>("ranges::fill_n(list<int>)", ranges_fill_n);
+#if TEST_STD_VER >= 23 // vector<bool>::iterator is not an output_iterator before C++23
+ bm_vector_bool("ranges::fill_n(vector<bool>)", ranges_fill_n);
+#endif
+
+ benchmark::Initialize(&argc, argv);
+ benchmark::RunSpecifiedBenchmarks();
+ benchmark::Shutdown();
+ return 0;
+}
diff --git a/libcxx/test/benchmarks/algorithms/modifying/generate.bench.cpp b/libcxx/test/benchmarks/algorithms/modifying/generate.bench.cpp
new file mode 100644
index 0000000000000..15b039a4d3009
--- /dev/null
+++ b/libcxx/test/benchmarks/algorithms/modifying/generate.bench.cpp
@@ -0,0 +1,59 @@
+//===----------------------------------------------------------------------===//
+//
+// 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
+
+#include <algorithm>
+#include <cstddef>
+#include <deque>
+#include <iterator>
+#include <list>
+#include <string>
+#include <vector>
+
+#include "benchmark/benchmark.h"
+#include "../../GenerateInput.h"
+
+template <class Container, class Operation>
+void bm(std::string operation_name, Operation generate) {
+ auto bench = [generate](auto& st) {
+ std::size_t const size = st.range(0);
+ Container c(size);
+ using ValueType = typename Container::value_type;
+ ValueType x = Generate<ValueType>::random();
+
+ for ([[maybe_unused]] auto _ : st) {
+ auto f = [&x] { return x; };
+ generate(c.begin(), c.end(), f);
+ benchmark::DoNotOptimize(c);
+ benchmark::DoNotOptimize(x);
+ benchmark::ClobberMemory();
+ }
+ };
+ benchmark::RegisterBenchmark(operation_name, bench)->Arg(32)->Arg(1024)->Arg(8192);
+}
+
+int main(int argc, char** argv) {
+ auto std_generate = [](auto first, auto last, auto f) { return std::generate(first, last, f); };
+ auto ranges_generate = [](auto first, auto last, auto f) { return std::ranges::generate(first, last, f); };
+
+ // std::generate
+ bm<std::vector<int>>("std::generate(vector<int>)", std_generate);
+ bm<std::deque<int>>("std::generate(deque<int>)", std_generate);
+ bm<std::list<int>>("std::generate(list<int>)", std_generate);
+
+ // ranges::generate
+ bm<std::vector<int>>("ranges::generate(vector<int>)", ranges_generate);
+ bm<std::deque<int>>("ranges::generate(deque<int>)", ranges_generate);
+ bm<std::list<int>>("ranges::generate(list<int>)", ranges_generate);
+
+ benchmark::Initialize(&argc, argv);
+ benchmark::RunSpecifiedBenchmarks();
+ benchmark::Shutdown();
+ return 0;
+}
diff --git a/libcxx/test/benchmarks/algorithms/modifying/generate_n.bench.cpp b/libcxx/test/benchmarks/algorithms/modifying/generate_n.bench.cpp
new file mode 100644
index 0000000000000..75b088411810f
--- /dev/null
+++ b/libcxx/test/benchmarks/algorithms/modifying/generate_n.bench.cpp
@@ -0,0 +1,59 @@
+//===----------------------------------------------------------------------===//
+//
+// 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
+
+#include <algorithm>
+#include <cstddef>
+#include <deque>
+#include <iterator>
+#include <list>
+#include <string>
+#include <vector>
+
+#include "benchmark/benchmark.h"
+#include "../../GenerateInput.h"
+
+template <class Container, class Operation>
+void bm(std::string operation_name, Operation generate_n) {
+ auto bench = [generate_n](auto& st) {
+ std::size_t const size = st.range(0);
+ Container c(size);
+ using ValueType = typename Container::value_type;
+ ValueType x = Generate<ValueType>::random();
+
+ for ([[maybe_unused]] auto _ : st) {
+ auto f = [&x] { return x; };
+ generate_n(c.begin(), size, f);
+ benchmark::DoNotOptimize(c);
+ benchmark::DoNotOptimize(x);
+ benchmark::ClobberMemory();
+ }
+ };
+ benchmark::RegisterBenchmark(operation_name, bench)->Arg(32)->Arg(1024)->Arg(8192);
+}
+
+int main(int argc, char** argv) {
+ auto std_generate_n = [](auto out, auto n, auto f) { return std::generate_n(out, n, f); };
+ auto ranges_generate_n = [](auto out, auto n, auto f) { return std::ranges::generate_n(out, n, f); };
+
+ // std::generate_n
+ bm<std::vector<int>>("std::generate_n(vector<int>)", std_generate_n);
+ bm<std::deque<int>>("std::generate_n(deque<int>)", std_generate_n);
+ bm<std::list<int>>("std::generate_n(list<int>)", std_generate_n);
+
+ // ranges::generate_n
+ bm<std::vector<int>>("ranges::generate_n(vector<int>)", ranges_generate_n);
+ bm<std::deque<int>>("ranges::generate_n(deque<int>)", ranges_generate_n);
+ bm<std::list<int>>("ranges::generate_n(list<int>)", ranges_generate_n);
+
+ benchmark::Initialize(&argc, argv);
+ benchmark::RunSpecifiedBenchmarks();
+ benchmark::Shutdown();
+ return 0;
+}
diff --git a/libcxx/test/benchmarks/algorithms/modifying/remove.bench.cpp b/libcxx/test/benchmarks/algorithms/modifying/remove.bench.cpp
new file mode 100644
index 0000000000000..764dfe73ab70c
--- /dev/null
+++ b/libcxx/test/benchmarks/algorithms/modifying/remove.bench.cpp
@@ -0,0 +1,133 @@
+//===----------------------------------------------------------------------===//
+//
+// 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
+
+#include <algorithm>
+#include <cstddef>
+#include <deque>
+#include <iterator>
+#include <list>
+#include <string>
+#include <vector>
+
+#include "benchmark/benchmark.h"
+#include "../../GenerateInput.h"
+
+// Create a sequence of the form xxxxxxxxxxyyyyyyyyyy and remove
+// the prefix of x's from it.
+//
+// We perform this benchmark in a batch because we need to restore the
+// state of the container after the operation.
+template <class Container, class Operation>
+void bm_prefix(std::string operation_name, Operation remove) {
+ auto bench = [remove](auto& st) {
+ std::size_t const size = st.range(0);
+ constexpr std::size_t BatchSize = 10;
+ using ValueType = typename Container::value_type;
+ Container c[BatchSize];
+ ValueType x = Generate<ValueType>::random();
+ ValueType y = Generate<ValueType>::random();
+ for (std::size_t i = 0; i != BatchSize; ++i) {
+ c[i] = Container(size);
+ auto half = size / 2;
+ std::fill_n(std::fill_n(c[i].begin(), half, x), half, y);
+ }
+
+ while (st.KeepRunningBatch(BatchSize)) {
+ for (std::size_t i = 0; i != BatchSize; ++i) {
+ auto result = remove(c[i].begin(), c[i].end(), x);
+ benchmark::DoNotOptimize(result);
+ benchmark::DoNotOptimize(c[i]);
+ benchmark::DoNotOptimize(x);
+ benchmark::ClobberMemory();
+ }
+
+ st.PauseTiming();
+ for (std::size_t i = 0; i != BatchSize; ++i) {
+ auto half = size / 2;
+ std::fill_n(std::fill_n(c[i].begin(), half, x), half, y);
+ }
+ st.ResumeTiming();
+ }
+ };
+ benchmark::RegisterBenchmark(operation_name, bench)->Arg(32)->Arg(1024)->Arg(8192);
+}
+
+// Create a sequence of the form xyxyxyxyxyxyxyxyxyxy and remove
+// the x's from it.
+//
+// We perform this benchmark in a batch because we need to restore the
+// state of the container after the operation.
+template <class Container, class Operation>
+void bm_sprinkled(std::string operation_name, Operation remove) {
+ auto bench = [remove](auto& st) {
+ std::size_t const size = st.range(0);
+ constexpr std::size_t BatchSize = 10;
+ using ValueType = typename Container::value_type;
+ Container c[BatchSize];
+ ValueType x = Generate<ValueType>::random();
+ ValueType y = Generate<ValueType>::random();
+ auto alternate = [&](auto out, auto n) {
+ for (std::size_t i = 0; i != n; ++i) {
+ *out++ = (i % 2 == 0 ? x : y);
+ }
+ };
+ for (std::size_t i = 0; i != BatchSize; ++i) {
+ c[i] = Container(size);
+ alternate(c[i].begin(), size);
+ }
+
+ while (st.KeepRunningBatch(BatchSize)) {
+ for (std::size_t i = 0; i != BatchSize; ++i) {
+ auto result = remove(c[i].begin(), c[i].end(), x);
+ benchmark::DoNotOptimize(result);
+ benchmark::DoNotOptimize(c[i]);
+ benchmark::DoNotOptimize(x);
+ benchmark::ClobberMemory();
+ }
+
+ st.PauseTiming();
+ for (std::size_t i = 0; i != BatchSize; ++i) {
+ alternate(c[i].begin(), size);
+ }
+ st.ResumeTiming();
+ }
+ };
+ benchmark::RegisterBenchmark(operation_name, bench)->Arg(32)->Arg(1024)->Arg(8192);
+}
+
+int main(int argc, char** argv) {
+ auto std_remove = [](auto first, auto last, auto const& value) { return std::remove(first, last, value); };
+ auto ranges_remove = [](auto first, auto last, auto const& value) { return std::ranges::remove(first, last, value); };
+
+ // std::remove
+ bm_prefix<std::vector<int>>("std::remove(vector<int>) (prefix)", std_remove);
+ bm_sprinkled<std::vector<int>>("std::remove(vector<int>) (sprinkled)", std_remove);
+
+ bm_prefix<std::deque<int>>("std::remove(deque<int>) (prefix)", std_remove);
+ bm_sprinkled<std::deque<int>>("std::remove(deque<int>) (sprinkled)", std_remove);
+
+ bm_prefix<std::list<int>>("std::remove(list<int>) (prefix)", std_remove);
+ bm_sprinkled<std::list<int>>("std::remove(list<int>) (sprinkled)", std_remove);
+
+ // ranges::remove
+ bm_prefix<std::vector<int>>("ranges::remove(vector<int>) (prefix)", ranges_remove);
+ bm_sprinkled<std::vector<int>>("ranges::remove(vector<int>) (sprinkled)", ranges_remove);
+
+ bm_prefix<std::deque<int>>("ranges::remove(deque<int>) (prefix)", ranges_remove);
+ bm_sprinkled<std::deque<int>>("ranges::remove(deque<int>) (sprinkled)", ranges_remove);
+
+ bm_prefix<std::list<int>>("ranges::remove(list<int>) (prefix)", ranges_remove);
+ bm_sprinkled<std::list<int>>("ranges::remove(list<int>) (sprinkled)", ranges_remove);
+
+ benchmark::Initialize(&argc, argv);
+ benchmark::RunSpecifiedBenchmarks();
+ benchmark::Shutdown();
+ return 0;
+}
diff --git a/libcxx/test/benchmarks/algorithms/modifying/remove_copy.bench.cpp b/libcxx/test/benchmarks/algorithms/modifying/remove_copy.bench.cpp
new file mode 100644
index 0000000000000..88c05384eb7bd
--- /dev/null
+++ b/libcxx/test/benchmarks/algorithms/modifying/remove_copy.bench.cpp
@@ -0,0 +1,109 @@
+//===----------------------------------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-...
[truncated]
|
|
can i work on this issue |
ldionne
force-pushed
the
review/benchmark-add-all-modifying
branch
from
7d00280 to
55c1ebe
Compare
ldionne
commented
Feb 19, 2025
libcxx/test/benchmarks/algorithms/modifying/remove_if.bench.cpp
Outdated
Show resolved
Hide resolved
philnik777
reviewed
Feb 24, 2025
| } | ||
| })->Range(8, 1 << 20); | ||
| }; | ||
| bm.operator()<std::vector<int>>("std::move(vector<int>)", std_move); |
There was a problem hiding this comment.
For the copy family of algorithms I think it would be interesting to test some non-trivial type. Specifically, that the compiler would not be able to see through the operation. That could be achieved by an out-of-line definition or a [[gnu::noinline]] on the function. Maybe just an optimization barrier would work to.
There was a problem hiding this comment.
Here too, I would like to defer to a separate patch where I can focus on doing just that for the copy and the move algorithms.
libcxx/test/benchmarks/algorithms/modifying/remove_if.bench.cpp
Outdated
Show resolved
Hide resolved
libcxx/test/benchmarks/algorithms/modifying/remove_if.bench.cpp
Outdated
Show resolved
Hide resolved
libcxx/test/benchmarks/algorithms/modifying/remove_if.bench.cpp
Outdated
Show resolved
Hide resolved
libcxx/test/benchmarks/algorithms/modifying/shift_left.bench.cpp
Outdated
Show resolved
Hide resolved
libcxx/test/benchmarks/algorithms/modifying/shift_left.bench.cpp
Outdated
Show resolved
Hide resolved
| ->Arg(1024) | ||
| ->Arg(8192); | ||
| }; | ||
| bm.operator()<std::vector<int>>("std::transform(vector<int>, vector<int>)", std_transform); |
There was a problem hiding this comment.
I think we should also add benchmarks for cases where the output is a segmented iterator (i.e. output to a deque) for the _copy versions of algorithms.
There was a problem hiding this comment.
I agree, but after consideration I would like to tackle that as a separate patch. Indeed, adding this variation of the benchmark requires writing a new batched benchmark since I can't just alternate between the source and the destination range. I'd like to do that in a separate patch that also tackles the benchmarks for std::copy (which have already landed).
This patch adds benchmarks for all the remaining algorithms in [alg.modifying.operations] that we didn't already have a benchmark for.
ldionne
force-pushed
the
review/benchmark-add-all-modifying
branch
from
5c563b5 to
fd313e5
Compare
|
I'm going to land this now since I applied all of @philnik777 's comments except #128733 and #128737. We can do another pass once he's back from vacation, but landing this now will at least ensure we don't have merge conflicts. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This patch adds benchmarks for all the remaining algorithms in [alg.modifying.operations] that we didn't already have a benchmark for.