<__msvc_int128.hpp>: optimize _Signed128::operator* with a branchless multiply

[KKoishi_ (Koishi-Satori)]

Resolves issue #6405.

std::_Signed128::operator* in <__msvc_int128.hpp> perform sign-normalizes before multiplying, but it is unnecessary and will cause performance issue.

The low 128 bits of a two's-complement product equal the unsigned product, so the sign handling is unnecessary. I replaced it with the same computation _Base128::_Multiply already performs.

I've tried to implement a version without sign handling, the new version has same behavior and will not break any ABI. The new version without sign-normalizing is more faster than current version in <__msvc_int128.hpp>.

This PR also adds a signed128_mul benchmark comparing the new implementation against a faithful replica of the old one.

Benchmark

On MSVC x64, Release (/O2), measured on a GitHub Actions hosted runner (shared 4-vCPU VM; absolute numbers vary between runs):

Benchmark	Old (sign-normalizing)	New (branchless)	Speedup
bm_signed128_mul (independent multiplies)	9.66 ns	6.94 ns	~1.39x
bm_signed128_mul_horner (5 dependent multiplies)	80.9 ns	9.15 ns	~8.8x

The first case measures multiplication throughput. The second evaluates a degree-5 polynomial with a Horner chain, where every multiply depends on the previous result; this is the same computation shape used by fixed-point math functions such as log2 and sin. Removing the sign normalization shortens the serial dependency chain of each multiply (and eliminates its branches), so latency-bound code benefits far more than the throughput microbenchmarksuggests.

[azure-pipelines]

Azure Pipelines:
There may be pipelines that require an authorized user to comment /azp run to run.

[Copilot]

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Optimizes _Signed128 multiplication by avoiding sign normalization and adds a microbenchmark to compare new vs. old behavior/performance.

Changes:

• Replaced _Signed128::operator* implementation with a direct low-128-bit product using _UMul128 + cross terms.
• Added a new Google Benchmark program to compare the new multiplication vs. the previous implementation.
• Registered the new benchmark target in the benchmarks CMake list.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 5 comments.

File	Description
stl/inc/__msvc_int128.hpp	Reworks _Signed128 multiplication implementation to compute the low 128 bits directly.
benchmarks/src/signed128_mul.cpp	Adds a benchmark comparing new vs. old signed-128 multiplication.
benchmarks/CMakeLists.txt	Adds the new signed128_mul benchmark target.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[KKoishi_ (Koishi-Satori)]

@microsoft-github-policy-service agree

[Copilot]

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated no new comments.

Suppressed comments (1)

stl/inc/__msvc_int128.hpp:1377

• Optional: this reimplements _Base128::_Multiply verbatim (defined at lines 422–427), creating a second multiplication core that can drift when the shared implementation changes. The unsigned operator already delegates to that helper at lines 1025–1026; returning its result here removes the sign-normalization branches while preserving a single implementation.

        _Signed128 _Result;
        _Result._Word[0] = _UMul128(_Left._Word[0], _Right._Word[0], _Result._Word[1]);
        _Result._Word[1] += _Left._Word[1] * _Right._Word[0];
        _Result._Word[1] += _Left._Word[0] * _Right._Word[1];

[Copilot]

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated no new comments.

[Alex Guteniev (AlexGuteniev)]

Please provide your benchmark numbers.

See Benchmarking the STL on the Wiki or follow other PR precedents.

[Copilot]

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated no new comments.

[KKoishi_ (Koishi-Satori)]

Please provide your benchmark numbers.

See Benchmarking the STL on the Wiki or follow other PR precedents.

I've added benchmark results in PR description.

[Stephan T. Lavavej (StephanTLavavej)]

/azp run STL-CI

[azure-pipelines]

Azure Pipelines:
Successfully started running 1 pipeline(s).