From cea7060684ae6c33fc8e16affc1c7998d17815ae Mon Sep 17 00:00:00 2001 From: Pablo Marquez Tello Date: Thu, 17 Aug 2023 16:18:17 +0100 Subject: NEFuseBatchNormalizationKernel rework * Enable fp16 in armv8a multi_isa builds * Code guarded with __ARM_FEATURE_FP16_VECTOR_ARITHMETIC needs to be moved to an fp16.cpp file to allow compilation with -march=armv8.2-a+fp16 * fp16.cpp needs to use the template add_same_neon() so it had to be moved from impl.cpp to impl.h * Partially resolves MLCE-1102 Change-Id: Ia51007f5e663b708071958bb94bfab4535e4b2f8 Signed-off-by: Pablo Marquez Tello Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/10191 Benchmark: Arm Jenkins Tested-by: Arm Jenkins Reviewed-by: Viet-Hoa Do Comments-Addressed: Arm Jenkins --- src/CMakeLists.txt | 1 - 1 file changed, 1 deletion(-) (limited to 'src/CMakeLists.txt') diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt index 4ce3d3c2df..2a592090ba 100644 --- a/src/CMakeLists.txt +++ b/src/CMakeLists.txt @@ -755,7 +755,6 @@ target_sources( cpu/kernels/floor/neon/fp32.cpp cpu/kernels/fuse_batch_normalization/generic/fp16.cpp cpu/kernels/fuse_batch_normalization/generic/fp32.cpp - cpu/kernels/fuse_batch_normalization/generic/impl.cpp cpu/kernels/fuse_batch_normalization/nchw/all.cpp cpu/kernels/fuse_batch_normalization/nhwc/neon/fp16.cpp cpu/kernels/fuse_batch_normalization/nhwc/neon/fp32.cpp -- cgit v1.2.1