[PATCH] [X86][SSE] Add v16i8/v32i8 multiplication support
llvm-dev at redking.me.uk
Mon Apr 20 11:11:16 PDT 2015
Hi qcolombet, andreadb, spatel, delena,
Patch to allow int8 vectors to be multiplied on the SSE unit instead of being scalarized.
The patch sign extends the i8 lanes to i16, uses the SSE2 pmullw multiplication instruction, then packs the lower byte from each result.
Once vpackuswb zmm support is present this should also work for v64i8 multiplication on AVX512BW targets.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 8970 bytes
Desc: not available
More information about the llvm-commits