Perf: Optimize function append in include/fmt/base.h #4541

fyrsta7 · 2025-09-14T19:49:03Z

Summary

This PR optimizes the performance of the append function in include/fmt/base.h.
This performance improvement was identified while profiling the spdlog logging library, a major downstream project that bundles fmt. The benchmarks show that this optimization provides significant gains in spdlog's throughput and latency. For context, this change was first submitted to spdlog (see gabime/spdlog#3465) where I was advised to contribute it upstream directly to fmt.
Across 47 test cases from spdlog's benchmark suite, this change achieves a maximum improvement of 16.67% while guaranteeing no regression exceeds 0.99% in any other case.

Test Plan

Correctness: All existing unit tests in the fmt project pass.
Performance: The performance impact was evaluated using the comprehensive benchmark suite from the spdlog project. This was chosen because spdlog is a key real-world use case, and its benchmarks effectively measure the performance of fmt's core formatting operations under various conditions (single-threaded, multi-threaded, different logging patterns). The commands used were ./bench/bench and ./bench/latency from the spdlog repository.

Performance Evaluation & Results

Testing Protocol:

The benchmark was run on an isolated Ubuntu 24.04 server using the spdlog benchmark suite.
The first run was discarded to account for cold-start effects.
The results below are the average of 5 subsequent runs.
The improvement for a test case is calculated as (new_value - old_value) / old_value * 100% if a higher value is better (e.g., throughput), or (old_value - new_value) / new_value * 100% if a lower value is better (e.g., latency). A positive percentage indicates a performance gain.

Results:

Test Case	Improvement
`overall_throughput_improvement`	`3.00%`
`overall_latency_improvement`	`3.19%`
`single_threaded.level_off.normal.messages_per_sec`	`0.37%`
`single_threaded.level_off.backtrace_on.messages_per_sec`	`4.33%`
`single_threaded.rotating_st.normal.messages_per_sec`	`2.82%`
`single_threaded.rotating_st.backtrace_on.messages_per_sec`	`1.95%`
`single_threaded.basic_st.normal.messages_per_sec`	`3.15%`
`single_threaded.basic_st.backtrace_on.messages_per_sec`	`3.48%`
`single_threaded.daily_st.normal.messages_per_sec`	`3.13%`
`single_threaded.daily_st.backtrace_on.messages_per_sec`	`3.12%`
`multi_threaded_1.rotating_mt.normal.messages_per_sec`	`1.41%`
`multi_threaded_1.rotating_mt.backtrace_on.messages_per_sec`	`3.45%`
`multi_threaded_1.daily_mt.normal.messages_per_sec`	`2.73%`
`multi_threaded_1.daily_mt.backtrace_on.messages_per_sec`	`12.95%`
`multi_threaded_1.basic_mt.normal.messages_per_sec`	`2.48%`
`multi_threaded_1.basic_mt.backtrace_on.messages_per_sec`	`2.73%`
`multi_threaded_1.level_off.normal.messages_per_sec`	`-0.19%`
`multi_threaded_1.level_off.backtrace_on.messages_per_sec`	`4.35%`
`multi_threaded_4.rotating_mt.normal.messages_per_sec`	`2.49%`
`multi_threaded_4.rotating_mt.backtrace_on.messages_per_sec`	`1.82%`
`multi_threaded_4.daily_mt.normal.messages_per_sec`	`6.08%`
`multi_threaded_4.daily_mt.backtrace_on.messages_per_sec`	`3.00%`
`multi_threaded_4.basic_mt.normal.messages_per_sec`	`3.94%`
`multi_threaded_4.basic_mt.backtrace_on.messages_per_sec`	`-0.69%`
`multi_threaded_4.level_off.normal.messages_per_sec`	`-0.99%`
`multi_threaded_4.level_off.backtrace_on.messages_per_sec`	`4.07%`
`single_threaded.level_off.backtrace_on.elapsed_time`	`0.00%`
`single_threaded.rotating_st.normal.elapsed_time`	`0.00%`
`single_threaded.rotating_st.backtrace_on.elapsed_time`	`0.00%`
`single_threaded.basic_st.normal.elapsed_time`	`0.00%`
`single_threaded.basic_st.backtrace_on.elapsed_time`	`16.67%`
`single_threaded.daily_st.normal.elapsed_time`	`3.33%`
`single_threaded.daily_st.backtrace_on.elapsed_time`	`12.50%`
`multi_threaded_1.rotating_mt.normal.elapsed_time`	`3.03%`
`multi_threaded_1.rotating_mt.backtrace_on.elapsed_time`	`4.76%`
`multi_threaded_1.daily_mt.normal.elapsed_time`	`0.00%`
`multi_threaded_1.daily_mt.backtrace_on.elapsed_time`	`14.89%`
`multi_threaded_1.basic_mt.normal.elapsed_time`	`0.00%`
`multi_threaded_1.basic_mt.backtrace_on.elapsed_time`	`0.00%`
`multi_threaded_1.level_off.backtrace_on.elapsed_time`	`0.00%`
`multi_threaded_4.rotating_mt.normal.elapsed_time`	`1.59%`
`multi_threaded_4.rotating_mt.backtrace_on.elapsed_time`	`0.00%`
`multi_threaded_4.daily_mt.normal.elapsed_time`	`6.06%`
`multi_threaded_4.daily_mt.backtrace_on.elapsed_time`	`2.47%`
`multi_threaded_4.basic_mt.normal.elapsed_time`	`1.67%`
`multi_threaded_4.basic_mt.backtrace_on.elapsed_time`	`0.00%`
`multi_threaded_4.level_off.backtrace_on.elapsed_time`	`0.00%`

vitaut

Thanks for the PR!

include/fmt/base.h

vitaut · 2025-09-21T16:19:59Z

include/fmt/base.h

-      if (free_cap < count) count = free_cap;
+      auto count = to_unsigned(end - begin);
+      if (free_cap < count) {
+        try_reserve(size_ + count);


Let's replace this with

grow_(*this, size_ + count);

to avoid an extra capacity check.

Thanks for the suggestion!
I've replaced try_reserve with grow_ to avoid the extra capacity check as you pointed out. The code has been updated.

vitaut · 2025-09-21T19:40:33Z

Merged, thanks!

fyrsta7 · 2025-09-21T19:42:35Z

Thank you for your quick action! 🫶

@fyrsta7

# 12.1.0 - 2025-10-29 - Optimized `buffer::append`, resulting in up to ~16% improvement on spdlog benchmarks (fmtlib/fmt#4541). Thanks @fyrsta7. - Worked around an ABI incompatibility in `std::locale_ref` between clang and gcc (fmtlib/fmt#4573). - Made `std::variant` and `std::expected` formatters work with `format_as` (fmtlib/fmt#4574, fmtlib/fmt#4575). Thanks @phprus. - Made `fmt::join<string_view>` work with C++ modules (fmtlib/fmt#4379, fmtlib/fmt#4577). Thanks @Arghnews. - Exported `fmt::is_compiled_string` and `operator""_cf` from the module (fmtlib/fmt#4544). Thanks @CrackedMatter. - Fixed a compatibility issue with C++ modules in clang (fmtlib/fmt#4548). Thanks @tsarn. - Added support for cv-qualified types to the `std::optional` formatter (fmtlib/fmt#4561, fmtlib/fmt#4562). Thanks @OleksandrKvl. - Added demangling support (used in exception and `std::type_info` formatters) for libc++ and clang-cl (fmtlib/fmt#4542, fmtlib/fmt#4560, fmtlib/fmt#4568, fmtlib/fmt#4571). Thanks @FatihBAKIR and @rohitsutreja. - Switched to global `malloc`/`free` to enable allocator customization (fmtlib/fmt#4569, fmtlib/fmt#4570). Thanks @rohitsutreja. - Made the `FMT_USE_CONSTEVAL` macro configurable by users (fmtlib/fmt#4546). Thanks @SnapperTT. - Fixed compilation with locales disabled in the header-only mode (fmtlib/fmt#4550). - Fixed compilation with clang 21 and `-std=c++20` (fmtlib/fmt#4552). - Fixed a dynamic linking issue with clang-cl (fmtlib/fmt#4576, fmtlib/fmt#4584). Thanks @FatihBAKIR. - Fixed a warning suppression leakage on gcc (fmtlib/fmt#4588). Thanks @ZedThree. - Made more internal color APIs `constexpr` (fmtlib/fmt#4581). Thanks @ishani. - Fixed compatibility with clang as a host compiler for NVCC (fmtlib/fmt#4564). Thanks @valgur. - Fixed various warnings and lint issues (fmtlib/fmt#4565, fmtlib/fmt#4572, fmtlib/fmt#4557). Thanks @LiangHuDream and @teruyamato0731. - Improved documentation (fmtlib/fmt#4549, fmtlib/fmt#4551, fmtlib/fmt#4566, fmtlib/fmt#4567, fmtlib/fmt#4578,). Thanks @teruyamato0731, @petersteneteg and @zimmerman-dev.

@fyrsta7

- Update from version 11.2.0 to 12.1.0 - Update of rootfile - so-bump so mpd requires shipping - Changelog 12.1.0 - Optimized `buffer::append`, resulting in up to ~16% improvement on spdlog benchmarks (fmtlib/fmt#4541). Thanks @fyrsta7. - Worked around an ABI incompatibility in `std::locale_ref` between clang and gcc (fmtlib/fmt#4573). - Made `std::variant` and `std::expected` formatters work with `format_as` (fmtlib/fmt#4574, fmtlib/fmt#4575). Thanks @phprus. - Made `fmt::join<string_view>` work with C++ modules (fmtlib/fmt#4379, fmtlib/fmt#4577). Thanks @Arghnews. - Exported `fmt::is_compiled_string` and `operator""_cf` from the module (fmtlib/fmt#4544). Thanks @CrackedMatter. - Fixed a compatibility issue with C++ modules in clang (fmtlib/fmt#4548). Thanks @tsarn. - Added support for cv-qualified types to the `std::optional` formatter (fmtlib/fmt#4561, fmtlib/fmt#4562). Thanks @OleksandrKvl. - Added demangling support (used in exception and `std::type_info` formatters) for libc++ and clang-cl (fmtlib/fmt#4542, fmtlib/fmt#4560, fmtlib/fmt#4568, fmtlib/fmt#4571). Thanks @FatihBAKIR and @rohitsutreja. - Switched to global `malloc`/`free` to enable allocator customization (fmtlib/fmt#4569, fmtlib/fmt#4570). Thanks @rohitsutreja. - Made the `FMT_USE_CONSTEVAL` macro configurable by users (fmtlib/fmt#4546). Thanks @SnapperTT. - Fixed compilation with locales disabled in the header-only mode (fmtlib/fmt#4550). - Fixed compilation with clang 21 and `-std=c++20` (fmtlib/fmt#4552). - Fixed a dynamic linking issue with clang-cl (fmtlib/fmt#4576, fmtlib/fmt#4584). Thanks @FatihBAKIR. - Fixed a warning suppression leakage on gcc (fmtlib/fmt#4588). Thanks @ZedThree. - Made more internal color APIs `constexpr` (fmtlib/fmt#4581). Thanks @ishani. - Fixed compatibility with clang as a host compiler for NVCC (fmtlib/fmt#4564). Thanks @valgur. - Fixed various warnings and lint issues (fmtlib/fmt#4565, fmtlib/fmt#4572, fmtlib/fmt#4557). Thanks @LiangHuDream and @teruyamato0731. - Improved documentation (fmtlib/fmt#4549, fmtlib/fmt#4551, fmtlib/fmt#4566, fmtlib/fmt#4567, fmtlib/fmt#4578,). Thanks @teruyamato0731, @petersteneteg and @zimmerman-dev. 12.0.0 - Optimized the default floating point formatting (fmtlib/fmt#3675, fmtlib/fmt#4516). In particular, formatting a `double` with format string compilation into a stack allocated buffer is more than 60% faster in version 12.0 compared to 11.2 according to [dtoa-benchmark](https://github.com/fmtlib/dtoa-benchmark): ``` Function Time (ns) Speedup fmt11 34.471 1.00x fmt12 21.000 1.64x ``` <img width="766" height="609" src="https://github.com/user-attachments/assets/d7d768ad-7543-468c-b0bb-449abf73b31b" /> - Added `constexpr` support to `fmt::format`. For example: ```c++ #include <fmt/compile.h> using namespace fmt::literals; std::string s = fmt::format(""_cf, 42); ``` now works at compile time provided that `std::string` supports `constexpr` (fmtlib/fmt#3403, fmtlib/fmt#4456). Thanks @msvetkin. - Added `FMT_STATIC_FORMAT` that allows formatting into a string of the exact required size at compile time. For example: ```c++ #include <fmt/compile.h> constexpr auto s = FMT_STATIC_FORMAT("{}", 42); ``` compiles to just ```s __ZL1s: .asciiz "42" ``` It can be accessed as a C string with `s.c_str()` or as a string view with `s.str()`. - Improved C++20 module support (fmtlib/fmt#4451, fmtlib/fmt#4459, fmtlib/fmt#4476, fmtlib/fmt#4488, fmtlib/fmt#4491, fmtlib/fmt#4495). Thanks @arBmind, @tkhyn, @Mishura4, @anonymouspc and @autoantwort. - Switched to using estimated display width in precision. For example: ```c++ fmt::print("|{:.4}|\n|1234|\n", "🐱🐱🐱"); ``` prints ![](https://github.com/user-attachments/assets/6c4446b3-13eb-43b9-b74a-b4543540ad6a) because `🐱` has an estimated width of 2 (fmtlib/fmt#4272, fmtlib/fmt#4443, fmtlib/fmt#4475). Thanks @nikhilreddydev and @localspook. - Fix interaction between debug presentation, precision, and width for strings (fmtlib/fmt#4478). Thanks @localspook. - Implemented allocator propagation on `basic_memory_buffer` move (fmtlib/fmt#4487, fmtlib/fmt#4490). Thanks @toprakmurat. - Fixed an ambiguity between `std::reference_wrapper<T>` and `format_as` formatters (fmtlib/fmt#4424, fmtlib/fmt#4434). Thanks @jeremy-rifkin. - Removed the following deprecated APIs: - `has_formatter`: use `is_formattable` instead, - `basic_format_args::parse_context_type`, `basic_format_args::formatter_type` and similar aliases in context types, - wide stream overload of `fmt::printf`, - wide stream overloads of `fmt::print` that take text styles, - `is_*char` traits, - `fmt::localtime`. - Deprecated wide overloads of `fmt::fprintf` and `fmt::sprintf`. - Improved diagnostics for the incorrect usage of `fmt::ptr` (fmtlib/fmt#4453). Thanks @TobiSchluter. - Made handling of ANSI escape sequences more efficient (fmtlib/fmt#4511, fmtlib/fmt#4528). Thanks @localspook and @Anas-Hamdane. - Fixed a buffer overflow on all emphasis flags set (fmtlib/fmt#4498). Thanks @dominicpoeschko. - Fixed an integer overflow for precision close to the max `int` value. - Fixed compatibility with WASI (fmtlib/fmt#4496, fmtlib/fmt#4497). Thanks @whitequark. - Fixed `back_insert_iterator` detection, preventing a fallback on slower path that handles arbitrary iterators (fmtlib/fmt#4454). - Fixed handling of invalid glibc `FILE` buffers (fmtlib/fmt#4469). - Added `wchar_t` support to the `std::byte` formatter (fmtlib/fmt#4479, fmtlib/fmt#4480). Thanks @phprus. - Changed component prefix from `fmt-` to `fmt_` for compatibility with NSIS/CPack on Windows, e.g. `fmt-doc` changed to `fmt_doc` (fmtlib/fmt#4441, fmtlib/fmt#4442). Thanks @n-stein. - Added the `FMT_CUSTOM_ASSERT_FAIL` macro to simplify providing a custom `fmt::assert_fail` implementation (fmtlib/fmt#4505). Thanks @HazardyKnusperkeks. - Switched to `FMT_THROW` on reporting format errors so that it can be overriden by users when exceptions are disabled (fmtlib/fmt#4521). Thanks @HazardyKnusperkeks. - Improved master project detection and disabled install targets when using {fmt} as a subproject by default (fmtlib/fmt#4536). Thanks @crueter. - Made various code improvements (fmtlib/fmt#4445, fmtlib/fmt#4448, fmtlib/fmt#4473, fmtlib/fmt#4522). Thanks @localspook, @tchaikov and @way4sahil. - Added Conan instructions to the docs (fmtlib/fmt#4537). Thanks @uilianries. - Removed Bazel files to avoid issues with downstream packaging (fmtlib/fmt#4530). Thanks @mering. - Added more entries for generated files to `.gitignore` (fmtlib/fmt#4355, fmtlib/fmt#4512). Thanks @dinomight and @localspook. - Fixed various warnings and compilation issues (fmtlib/fmt#4447, fmtlib/fmt#4470, fmtlib/fmt#4474, fmtlib/fmt#4477, fmtlib/fmt#4471, fmtlib/fmt#4483, fmtlib/fmt#4515, fmtlib/fmt#4533, fmtlib/fmt#4534). Thanks @dodomorandi, @localspook, @remyjette, @Tomek-Stolarczyk, @Mishura4, @mattiasljungstrom and @FatihBAKIR. Signed-off-by: Adolf Belka <[email protected]> Signed-off-by: Michael Tremer <[email protected]>

See github.com/fmtlib/fmt/pull/4541

fyrsta7 added 2 commits September 15, 2025 03:45

Perf: Optimize function append in include/fmt/base.h

2f344ea

fix build error

af8af1d

vitaut reviewed Sep 17, 2025

View reviewed changes

include/fmt/base.h Show resolved Hide resolved

vitaut reviewed Sep 21, 2025

View reviewed changes

fyrsta7 added 2 commits September 22, 2025 02:48

perf: Use grow_ to avoid redundant capacity check in append

9650c70

fix format

8d63f6e

vitaut merged commit 4cce5f4 into fmtlib:master Sep 21, 2025
41 checks passed

polter-rnd added a commit to polter-rnd/slimlog that referenced this pull request Dec 15, 2025

Util: backport MemoryBuffer optimization from fmtlib

64073fd

See github.com/fmtlib/fmt/pull/4541

polter-rnd added a commit to polter-rnd/slimlog that referenced this pull request Dec 30, 2025

Util: backport MemoryBuffer optimization from fmtlib

d15e180

See github.com/fmtlib/fmt/pull/4541

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

Perf: Optimize function append in include/fmt/base.h #4541

Perf: Optimize function append in include/fmt/base.h #4541

fyrsta7 commented Sep 14, 2025

Uh oh!

vitaut left a comment

Uh oh!

Uh oh!

vitaut Sep 21, 2025

Uh oh!

fyrsta7 Sep 21, 2025

Uh oh!

Uh oh!

vitaut commented Sep 21, 2025

Uh oh!

fyrsta7 commented Sep 21, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Perf: Optimize function append in include/fmt/base.h #4541

Perf: Optimize function append in include/fmt/base.h #4541

Conversation

fyrsta7 commented Sep 14, 2025

Summary

Test Plan

Performance Evaluation & Results

Uh oh!

vitaut left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

vitaut Sep 21, 2025

Choose a reason for hiding this comment

Uh oh!

fyrsta7 Sep 21, 2025

Choose a reason for hiding this comment

Uh oh!

Uh oh!

vitaut commented Sep 21, 2025

Uh oh!

fyrsta7 commented Sep 21, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants