Skip to content

Conversation

@AlexGuteniev
Copy link
Contributor

Fix #5908

The auto vectorization recipe applied in #4431 did not work in 17.14.
Another recipe would have worked, but it does not work in the latest Insiders.
So chasing the compiler auto vectorization here appears to be unproductive.

Do not help the compiler ™️

Benchmark results

Benchmark Before After
rc<std::uint8_t> 917 ns 58.4 ns
rc<std::uint16_t> 2074 ns 109 ns
rc<std::uint32_t> 1299 ns 207 ns
rc<std::uint64_t> 930 ns 548 ns

@StephanTLavavej StephanTLavavej added the performance Must go faster label Dec 31, 2025
@StephanTLavavej StephanTLavavej self-assigned this Jan 6, 2026
@StephanTLavavej StephanTLavavej removed their assignment Jan 6, 2026
@StephanTLavavej StephanTLavavej moved this from Initial Review to Ready To Merge in STL Code Reviews Jan 6, 2026
@StephanTLavavej
Copy link
Member

Thanks! 😻 I pushed significant changes, please double-check.

@AlexGuteniev
Copy link
Contributor Author

All good

@StephanTLavavej StephanTLavavej moved this from Ready To Merge to Merging in STL Code Reviews Jan 7, 2026
@StephanTLavavej
Copy link
Member

I'm mirroring this to the MSVC-internal repo - please notify me if any further changes are pushed.

@StephanTLavavej StephanTLavavej merged commit b3f2c67 into microsoft:main Jan 8, 2026
45 checks passed
@github-project-automation github-project-automation bot moved this from Merging to Done in STL Code Reviews Jan 8, 2026
@StephanTLavavej
Copy link
Member

🐱 😻 😸

@AlexGuteniev AlexGuteniev deleted the replacement-copy branch January 8, 2026 17:35
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

performance Must go faster

Projects

Archived in project

Development

Successfully merging this pull request may close these issues.

Vectorize replace_copy for x64 / x86

2 participants