Integrate new pretranspose_b_array with extra fused transpose of B
This patch fuses the transposition taking place in Acl with the transformations done in arm_gemm (called pretranspose_b_array) if the underlying kernel and transform supports it. This should improve start-up time (as it's for constant Rhs matrices) and memory footprint. The transformations in arm_gemm are kernel specific. The Rhs matrix is transformed into certain layouts to improve the performance.
Resolves: COMPMID-6595
Change-Id: Id2932dd966e59f903c279417bebcea83d9a42464
Signed-off-by: Gunes Bayir <gunes.bayir@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/11144
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Viet-Hoa Do <viet-hoa.do@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index be7a6ef..c5a1721 100644
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -508,6 +508,7 @@
core/NEON/kernels/arm_gemm/gemm_quint8.cpp
core/NEON/kernels/arm_gemm/gemm_uint16.cpp
core/NEON/kernels/arm_gemm/gemm_uint8.cpp
+ core/NEON/kernels/arm_gemm/interleave-8way.cpp
core/NEON/kernels/arm_gemm/interleave_indirect.cpp
core/NEON/kernels/arm_gemm/kernels/a64_ffhybrid_bf16fp32_mmla_6x16/generic.cpp
core/NEON/kernels/arm_gemm/kernels/a64_ffhybrid_fp16_mla_6x32/generic.cpp