PPU/SPU LLVM: Emulate VPERM2B with a 256 bit wide VPERMB (AVX512 optimization) #10959
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Outdated suggestions cannot be applied.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.
Saves 1 uop by using 256 wide VPERMB instead of VPERM2B. (Compiles down to a vinserti128 and vpermb)
On my tigerlake laptop the AVX2 path has a throughput of one shufb per 4 cycles. The old AVX512 path has a throughput of 1 per 3 cycles, and the new AVX512 path has a throughput of 1 per 2.3 cycles. (tested in a little benchmark I wrote in asm before I rewrote it in LLVM IR.)