Add Google Highway SIMD acceleration for ImageBufAlgo operations #4986
+1,388
−1,869
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Integrated Google Highway (hwy) as a required dependency and updated build scripts accordingly. Added a new SIMD-accelerated resample_hwy implementation in imagebufalgo_xform.cpp, which is used for resampling when both source and destination have local pixels and the image is not deep. The scalar fallback remains for other cases.
Introduces Highway (hwy) SIMD-accelerated implementations for add, sub, mul, and pow operations in imagebufalgo, using fast pointer-based code paths when localpixels are available. Also updates resample_hwy to support both float and double types, improving performance and type safety for SIMD image processing.
Deleted all CI, analysis, documentation, release, and related workflow YAML files from .github/workflows. This disables all automated GitHub Actions for the repository.
Introduces a new header, imagebufalgo_hwy_pvt.h, encapsulating SIMD type traits and vectorized load/store utilities using Highway. Refactors add, sub, and mul implementations in imagebufalgo_addsub.cpp and imagebufalgo_muldiv.cpp to use these utilities, improving code clarity and enabling more robust SIMD handling for various pixel types.
Refactors SIMD kernel runners and type promotion/demotion utilities in imagebufalgo_hwy_pvt.h for better extensibility and correctness, including support for int16_t and improved documentation. Updates all relevant imagebufalgo implementations to use reinterpret_cast and static_cast for type safety, and enhances pow_impl_hwy to use SIMD for scalar exponents. Also links hwy_contrib in CMake and replaces direct Highway includes with the new header where appropriate.
Collaborator
|
It's gonna take me a few days to do a thorough review -- I'm under the weather this week and mostly sacked out in bed with some strong meds, so brain isn't what it should be. But I like the direction of this. Stylistically, it's not very different from the small number of places that I've "hard coded" our simd.h approaches to certain math IBA ops, but mostly I only did it for a small number of common cases, for example over_impl_rgbafloat special case of IBA::over() when all images are 4 channel float, in imagebufalgo_composite.cpp. But you're handling many more cases, and in a more principled way using hwy. This PR seems to have completely removed the entire directory of GHA workflows. |
Benchmarking scripts for OIIO resample operations in Windows (PowerShell, BAT) and Linux (Bash), along with a C++ benchmark for HWY arithmetic operations. Includes Visual Studio project files for building and organizing the HWY benchmark. Signed-off-by: Vlad (Kuzmin) Erium <libalias@gmail.com>
Contributor
Author
|
Sure, no rush. I also will experiment with some IBA functions, that might be a good for SIMD use |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
perf(IBA): Add Google Highway SIMD fast paths for core operations
Implement SIMD acceleration using Google Highway for ImageBufAlgo add, sub,
mul, pow, and resample operations. Provides 2-8x performance improvement for
contiguous pixel layouts with portable vectorization across x86 and ARM.
New file imagebufalgo_hwy_pvt.h provides reusable SIMD infrastructure with
automatic type promotion/demotion (uint8/uint16/int16/uint32/half/float/double),
generic operation kernels, and smart fallback to scalar code for strided layouts.
Operations use runtime vector width detection (ScalableTag), FMA instructions
where applicable, and handle partial vectors correctly. Code follows OIIO style
with modern C++ casts and comprehensive documentation.
Requires: Google Highway library (MIT license, header-only)
Modified: imagebufalgo_{addsub,muldiv,pixelmath,xform}.cpp + new hwy_pvt.h
Checklist:
behavior.
testsuite.
PR, by pushing the changes to my fork and seeing that the automated CI
passed there. (Exceptions: If most tests pass and you can't figure out why
the remaining ones fail, it's ok to submit the PR and ask for help. Or if
any failures seem entirely unrelated to your change; sometimes things break
on the GitHub runners.)
fixed any problems reported by the clang-format CI test.
corresponding Python bindings. If altering ImageBufAlgo functions, I also
exposed the new functionality as oiiotool options.