[bm] improve charconv benchmarks

biojppm · Mar 16, 2022 · 2fd5583 · 2fd5583
1 parent f31fc2d
commit 2fd5583
Show file tree

Hide file tree

Showing 6 changed files with 580 additions and 239 deletions.
diff --git a/.github/workflows/benchmarks.yml b/.github/workflows/benchmarks.yml
@@ -24,8 +24,8 @@ env:
 
 
 jobs:
-  benchmarks:
-    name: bm/c++${{matrix.std}}/${{matrix.cxx}}/${{matrix.bt}}
+  bm_x86_64:
+    name: bm/x86_64/c++${{matrix.std}}/${{matrix.cxx}}/${{matrix.bt}}
     if: |
       (!contains(github.event.head_commit.message, 'skip all')) ||
       (!contains(github.event.head_commit.message, 'skip benchmarks')) ||
@@ -36,23 +36,15 @@ jobs:
       fail-fast: false
       matrix:
         include:
-          - {std: 11, cxx: g++-10, bt: Debug  , os: ubuntu-18.04, bitlinks: static64 static32}
           - {std: 11, cxx: g++-10, bt: Release, os: ubuntu-18.04, bitlinks: static64 static32}
-          - {std: 17, cxx: g++-10, bt: Debug  , os: ubuntu-18.04, bitlinks: static64 static32}
           - {std: 17, cxx: g++-10, bt: Release, os: ubuntu-18.04, bitlinks: static64 static32}
-          - {std: 20, cxx: g++-10, bt: Debug  , os: ubuntu-18.04, bitlinks: static64 static32}
           - {std: 20, cxx: g++-10, bt: Release, os: ubuntu-18.04, bitlinks: static64 static32}
           #
-          - {std: 11, cxx: vs2019, bt: Debug  , os: windows-2019, bitlinks: static64 static32}
           - {std: 11, cxx: vs2019, bt: Release, os: windows-2019, bitlinks: static64 static32}
-          - {std: 17, cxx: vs2019, bt: Debug  , os: windows-2019, bitlinks: static64 static32}
           - {std: 17, cxx: vs2019, bt: Release, os: windows-2019, bitlinks: static64 static32}
-          - {std: 20, cxx: vs2019, bt: Debug  , os: windows-2019, bitlinks: static64 static32}
           - {std: 20, cxx: vs2019, bt: Release, os: windows-2019, bitlinks: static64 static32}
           #
-          - {std: 11, cxx: xcode, xcver: 13, bt: Debug  , os: macos-11, bitlinks: static64}
           - {std: 11, cxx: xcode, xcver: 13, bt: Release, os: macos-11, bitlinks: static64}
-          - {std: 17, cxx: xcode, xcver: 13, bt: Debug  , os: macos-11, bitlinks: static64}
           - {std: 17, cxx: xcode, xcver: 13, bt: Release, os: macos-11, bitlinks: static64}
     env: {BM: ON, STD: "${{matrix.std}}", CXX_: "${{matrix.cxx}}", BT: "${{matrix.bt}}", BITLINKS: "${{matrix.bitlinks}}", VG: "${{matrix.vg}}", SAN: "${{matrix.san}}", LINT: "${{matrix.lint}}", OS: "${{matrix.os}}"}
     steps:
@@ -81,7 +73,7 @@ jobs:
           set -x
           desc=$(git describe || git rev-parse --short HEAD)
           for bl in ${{matrix.bitlinks}} ; do
-            dst=$(echo benchmark_results/$desc/${{matrix.cxx}}-${{matrix.bt}}-c++${{matrix.std}}-$bl | sed 's:++-:xx:g' | sed 's:+:x:g')
+            dst=$(echo benchmark_results/$desc/x86_64/${{matrix.cxx}}-${{matrix.bt}}-c++${{matrix.std}}-$bl | sed 's:++-:xx:g' | sed 's:+:x:g')
             mkdir -p $dst
             find build -name bm-results
             mv -vf build/$bl/bm/bm-results/* $dst/.
@@ -91,3 +83,92 @@ jobs:
         with:
           name: benchmark_results
           path: benchmark_results/
+
+  #--------------------------------------------------------------------------------------------------
+  bm_rarearch:
+    name: bm/${{matrix.arch}}/c++${{matrix.std}}/${{matrix.bt}}
+    if: |
+      (!contains(github.event.head_commit.message, 'skip all')) ||
+      (!contains(github.event.head_commit.message, 'skip benchmarks')) ||
+      contains(github.event.head_commit.message, 'only benchmarks')
+    continue-on-error: true
+    runs-on: ubuntu-20.04
+    strategy:
+      fail-fast: false
+      matrix:
+        include:
+          - {std: 11, bt: Release, arch: aarch64, distro: ubuntu20.04}
+          - {std: 17, bt: Release, arch: aarch64, distro: ubuntu20.04}
+          #
+          - {std: 11, bt: Release, arch: ppc64le, distro: ubuntu20.04}
+          - {std: 17, bt: Release, arch: ppc64le, distro: ubuntu20.04}
+          #
+          # the github runners are failing for the following:
+          #- {std: 11, bt: Release, arch: s390x  , distro: ubuntu20.04}
+          #- {std: 17, bt: Release, arch: s390x  , distro: ubuntu20.04}
+          ##
+          #- {std: 11, bt: Release, arch: armv6  , distro: ubuntu18.04}
+          #- {std: 17, bt: Release, arch: armv6  , distro: ubuntu18.04}
+          ##
+          #- {std: 11, bt: Release, arch: armv7  , distro: ubuntu18.04}
+          #- {std: 17, bt: Release, arch: armv7  , distro: ubuntu18.04}
+    steps:
+      - {name: checkout, uses: actions/checkout@v2, with: {submodules: recursive}}
+      - name: test
+        uses: uraimo/run-on-arch-action@v2.0.5
+        with:
+          arch: ${{matrix.arch}}
+          distro: ${{matrix.distro}}
+          install: |
+            set -x
+            apt-get update -y
+            apt-get install -y \
+              git \
+              build-essential
+            # arm platforms need an up-to-date cmake:
+            # https://gitlab.kitware.com/cmake/cmake/-/issues/20568
+            if [ "${{matrix.arch}}" == "armv6" ] || [ "${{matrix.arch}}" == "armv7" ] ; then
+              apt-get install -y \
+                gpg \
+                wget \
+                apt-transport-https
+              wget --no-check-certificate -O - https://apt.kitware.com/keys/kitware-archive-latest.asc 2>/dev/null | gpg --dearmor - | tee /usr/share/keyrings/kitware-archive-keyring.gpg >/dev/null
+              echo 'deb [signed-by=/usr/share/keyrings/kitware-archive-keyring.gpg] https://apt.kitware.com/ubuntu/ focal main' | tee /etc/apt/sources.list.d/kitware.list >/dev/null
+              apt-get update -y
+              rm /usr/share/keyrings/kitware-archive-keyring.gpg
+              apt-get install kitware-archive-keyring
+              apt-get update -y
+            fi
+            apt-get install -y cmake cmake-data
+            cmake --version
+          run: |
+            set -x
+            uname -a
+            pwd
+            ls -lFhp .
+            #
+            bdir=build_${{matrix.arch}}_${{matrix.bt}}_${{matrix.std}}
+            idir=install_${{matrix.arch}}_${{matrix.bt}}_${{matrix.std}}
+            mkdir -p $bdir
+            #
+            cmake -S . -B $bdir \
+              -DCMAKE_INSTALL_PREFIX=$idir \
+              -DCMAKE_BUILD_TYPE=${{matrix.bt}} \
+              -DC4_CXX_STANDARD=${{matrix.std}} \
+              -DCXX_STANDARD=${{matrix.std}} \
+              -DC4CORE_DEV=ON \
+              -DC4CORE_BUILD_TESTS=OFF \
+              -DC4CORE_BUILD_BENCHMARKS=ON \
+              -DC4CORE_SANITIZE=OFF \
+              -DC4CORE_LINT=OFF \
+              -DC4CORE_VALGRIND=OFF
+            #
+            cmake --build $bdir -j --target c4core-bm-build
+            #
+            cmake --build $bdir -j 1 --target c4core-bm-run
+            #
+            desc=$(git describe || git rev-parse --short HEAD)
+            dst=$(echo benchmark_results/$desc/${{matrix.arch}}/${{matrix.bt}}-c++${{matrix.std}} | sed 's:++-:xx:g' | sed 's:+:x:g')
+            mkdir -p $dst
+            find $bdir -name bm-results
+            mv -vf $bdir/bm/bm-results/* $dst/.
diff --git a/README.md b/README.md
@@ -92,73 +92,6 @@ are transitively used.
 All of the utilities in this library are under the namespace `c4`; any
 exposed macros use the prefix `C4_`: eg `C4_ASSERT()`.
 
-### Multi-platform / multi-compiler utilities
-
-```c++
-// TODO: elaborate on the topics:
-#include <c4/error.hpp>
-
-C4_LIKELY()/C4_UNLIKELY()
-
-C4_RESTRICT, $, c$, $$, c$$
-#include <c4/restrict.hpp>
-#include <c4/unrestrict.hpp>
-
-#include <c4/windows_push.hpp>
-#include <c4/windows_pop.hpp>
-
-C4_UNREACHABLE()
-
-c4::type_name()
-```
-
-### Runtime assertions and error handling
-
-```c++
-// TODO: elaborate on the topics:
-
-error callback
-
-C4_ASSERT()
-C4_XASSERT()
-C4_CHECK()
-
-C4_ERROR()
-C4_NOT_IMPLEMENTED()
-```
-
-### Memory allocation
-
-```c++
-// TODO: elaborate on the topics:
-
-c4::aalloc(), c4::afree() // aligned allocation
-
-c4::MemoryResource // global and scope
-
-c4::Allocator
-```
-
-### Mass initialization/construction/destruction
-
-```c++
-// TODO: elaborate on the topics:
-
-c4::construct()/c4::construct_n()
-
-c4::destroy()/c4::destroy_n()
-
-c4::copy_construct()/c4::copy_construct_n()
-
-c4::copy_assign()/c4::copy_assign_n()
-
-c4::move_construct()/c4::move_construct_n()
-
-c4::move_assign()/c4::move_assign_n()
-
-c4::make_room()/c4::destroy_room()
-```
-
 
 ### Writeable string views: c4::substr and c4::csubstr
 
@@ -172,16 +105,78 @@ Here: [`#include <c4/charconv.hpp>`](src/c4/charconv.hpp)
 ```c++
 // TODO: elaborate on the topics:
 
+c4::read_dec(), c4::write_dec()
+c4::read_hex(), c4::write_hex()
+c4::read_oct(), c4::write_oct()
+c4::read_bin(), c4::write_bin()
+
 c4::utoa(), c4::atou()
 c4::itoa(), c4::atoi()
 c4::ftoa(), c4::atof()
 c4::dtoa(), c4::atod()
+c4::xtoa(), c4::atox()
 
 c4::to_chars(), c4::from_chars()
 c4::to_chars_sub()
 c4::to_chars_first()
 ```
 
+The charconv funcions above are very fast; even faster than C++'s fastest facility `std::from_chars()`, `std::to_chars()`. For continuous benchmark results, browse through c4core's [github CI benchmark runs](https://github.com/biojppm/c4core/actions/workflows/benchmarks.yml). For example, a benchmark run on Linux/g++11.2 shows that:
+- `c4::to_chars()` can be expected to be roughly...
+  - ~40% to 2x faster than `std::to_chars()`
+  - ~10x-30x faster than `sprintf()`
+  - ~50x-100x faster than a naive `stringstream::operator<<()` followed by `stringstream::str()`
+- `c4::from_chars()` can be expected to be roughly...
+  - ~10%-30% faster than `std::from_chars()`
+  - ~10x faster than `scanf()`
+  - ~30x-50x faster than a naive `stringstream::str()` followed by `stringstream::operator>>()`
+
+Here are the results:
+
+| Write throughput         |         | Read throughput          |          |
+|:-------------------------|--------:|:-------------------------|---------:|
+|  **write `uint8_t`**     | **MB/s**| **read `uint8_t`**       |  **MB/s**|
+| `c4::to_chars<u8>`       |  526.86 |  `c4::from_chars<u8>`    |   163.06 |
+| `std::to_chars<u8>`      |  379.03 |  `std::from_chars<u8>`   |   154.85 |
+| `std::sprintf<u8>`       |   20.49 |  `std::scanf<u8>`        |    15.75 |
+| `std::stringstream<u8>`  |    3.82 |  `std::stringstream<u8>` |     3.83 |
+|  **write `int8_t`**      | **MB/s**| **read `int8_t`**        |  **MB/s**|
+| `c4::to_chars<i8>`       |  599.98 |  `c4::from_chars<i8>`    |   184.20 |
+| `std::to_chars<i8>`      |  246.32 |  `std::from_chars<i8>`   |   156.40 |
+| `std::sprintf<i8>`       |   19.15 |  `std::scanf<i8>`        |    16.44 |
+| `std::stringstream<i8>`  |    3.83 |  `std::stringstream<i8>` |     3.89 |
+|  **write `uint16_t`**    | **MB/s**| **read `uint16_t`**      |  **MB/s**|
+| `c4::to_chars<u16>`      |  486.40 |  `c4::from_chars<u16>`   |   349.48 |
+| `std::to_chars<u16>`     |  454.24 |  `std::from_chars<u16>`  |   319.13 |
+| `std::sprintf<u16>`      |   38.74 |  `std::scanf<u16>`       |    28.12 |
+| `std::stringstream<u16>` |    7.08 |  `std::stringstream<u16>`|     6.73 |
+|  **write `int16_t`**     | **MB/s**| **read `int16_t`**       |  **MB/s**|
+| `c4::to_chars<i16>`      |  507.44 |  `c4::from_chars<i16>`   |   282.95 |
+| `std::to_chars<i16>`     |  297.49 |  `std::from_chars<i16>`  |   186.18 |
+| `std::sprintf<i16>`      |   39.03 |  `std::scanf<i16>`       |    28.45 |
+| `std::stringstream<i16>` |    6.98 |  `std::stringstream<i16>`|     6.49 |
+|  **write `uint32_t`**    | **MB/s**| **read `uint32_t`**      |  **MB/s**|
+| `c4::to_chars<u32>`      |  730.12 |  `c4::from_chars<u32>`   |   463.95 |
+| `std::to_chars<u32>`     |  514.76 |  `std::from_chars<u32>`  |   329.42 |
+| `std::sprintf<u32>`      |   71.19 |  `std::scanf<u32>`       |    44.97 |
+| `std::stringstream<u32>` |   14.05 |  `std::stringstream<u32>`|    12.57 |
+|  **write `int32_t`**     | **MB/s**| **read `int32_t`**       |  **MB/s**|
+| `c4::to_chars<i32>`      |  618.76 |  `c4::from_chars<i32>`   |   345.53 |
+| `std::to_chars<i32>`     |  394.72 |  `std::from_chars<i32>`  |   224.46 |
+| `std::sprintf<i32>`      |   71.14 |  `std::scanf<i32>`       |    43.49 |
+| `std::stringstream<i32>` |   13.91 |  `std::stringstream<i32>`|    12.03 |
+|  **write `uint64_t`**    | **MB/s**| **read `uint64_t`**      |  **MB/s**|
+| `c4::to_chars<u64>`      | 1118.87 |  `c4::from_chars<u64>`   |   928.49 |
+| `std::to_chars<u64>`     |  886.58 |  `std::from_chars<u64>`  |   759.03 |
+| `std::sprintf<u64>`      |  140.96 |  `std::scanf<u64>`       |    91.60 |
+| `std::stringstream<u64>` |   28.01 |  `std::stringstream<u64>`|    25.00 |
+|  **write `int64_t`**     | **MB/s**| **read `int64_t`**       |  **MB/s**|
+| `c4::to_chars<i64>`      | 1198.78 |  `c4::from_chars<i64>`   |   713.76 |
+| `std::to_chars<i64>`     |  882.17 |  `std::from_chars<i64>`  |   646.18 |
+| `std::sprintf<i64>`      |  138.79 |  `std::scanf<i64>`       |    90.07 |
+| `std::stringstream<i64>` |   27.62 |  `std::stringstream<i64>`|    25.12 |
+
+
 ### String formatting and parsing
 
 * [`#include <c4/format.hpp>`](src/c4/format.hpp)
@@ -190,13 +185,26 @@ c4::to_chars_first()
 // TODO: elaborate on the topics:
 
 c4::cat(), c4::uncat()
-
 c4::catsep(), c4::uncatsep()
-
 c4::format(), c4::unformat()
 
+c4::catrs()
+c4::catseprs()
+c4::formatrs()
+
 // formatting:
-c4::raw, c4::craw
+c4::fmt::overflow_checked
+c4::fmt::real
+c4::fmt::boolalpha
+c4::fmt::dec
+c4::fmt::hex
+c4::fmt::oct
+c4::fmt::bin
+c4::fmt::zpad
+c4::fmt::right
+c4::fmt::left
+c4::fmt::raw, c4::fmt::craw
+c4::fmt::base64, c4::fmt::cbase64
 ```
 
 ### `c4::span` and `c4::blob`
@@ -230,3 +238,71 @@ c4::bm2str(), c4::str2bm()
 [`#include <c4/base64.hpp>`](src/c4/base64.hpp)
 
 ### Fuzzy float comparison
+
+
+### Multi-platform / multi-compiler utilities
+
+```c++
+// TODO: elaborate on the topics:
+#include <c4/error.hpp>
+
+C4_LIKELY()/C4_UNLIKELY()
+
+C4_RESTRICT, $, c$, $$, c$$
+#include <c4/restrict.hpp>
+#include <c4/unrestrict.hpp>
+
+#include <c4/windows_push.hpp>
+#include <c4/windows_pop.hpp>
+
+C4_UNREACHABLE()
+
+c4::type_name()
+```
+
+### Runtime assertions and error handling
+
+```c++
+// TODO: elaborate on the topics:
+
+error callback
+
+C4_ASSERT()
+C4_XASSERT()
+C4_CHECK()
+
+C4_ERROR()
+C4_NOT_IMPLEMENTED()
+```
+
+### Memory allocation
+
+```c++
+// TODO: elaborate on the topics:
+
+c4::aalloc(), c4::afree() // aligned allocation
+
+c4::MemoryResource // global and scope
+
+c4::Allocator
+```
+
+### Mass initialization/construction/destruction
+
+```c++
+// TODO: elaborate on the topics:
+
+c4::construct()/c4::construct_n()
+
+c4::destroy()/c4::destroy_n()
+
+c4::copy_construct()/c4::copy_construct_n()
+
+c4::copy_assign()/c4::copy_assign_n()
+
+c4::move_construct()/c4::move_construct_n()
+
+c4::move_assign()/c4::move_assign_n()
+
+c4::make_room()/c4::destroy_room()
+```