Arbaaz123676 opened a new pull request, #49180: URL: https://github.com/apache/arrow/pull/49180
### Rationale for this change arrow::util::Float16 currently relies on a software implementation for float↔Float16 conversions. However, GCC and Clang provide native "_Float16" support on some architectures. This PR adds an optional, compile-time-guarded fast-path that uses native "_Float16" where available, while preserving the existing software implementation as a deterministic fallback (including MSVC). ### What changes are included in this PR? - Add native "_Float16" helpers guarded by ARROW_HAVE_NATIVE_FLOAT16. - Use the native fast-path for Float16::FromFloat and Float16::ToFloat when supported. - Preserve the existing software conversion path as a fallback. - No changes to public APIs, ABI, layout, or semantics. ### Are these changes tested? The code builds successfully locally. C++ tests were not enabled in the local build configuration; CI will run the full test suite. ### Are there any user-facing changes? No. This is an internal optimisation only. This PR does not introduce any breaking changes to public APIs. Fixes #49179 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
