diff options
| author | Pablo Marquez Tello <pablo.tello@arm.com> | 2024-05-14 07:54:19 +0100 |
|---|---|---|
| committer | Pablo Marquez Tello <pablo.tello@arm.com> | 2024-05-14 11:58:27 +0000 |
| commit | 2217f1e60964fe586cae7ef996af7ef1c0bef2ab (patch) | |
| tree | 593dad484fe922efeb655335d37b0066d2af4bcd /src/cpu/operators/internal/CpuGemmAssemblyDispatch.cpp | |
| parent | 21fb2ad16a30a5ff29929515abe28c14b2c6b5a1 (diff) | |
| download | ComputeLibrary-2217f1e60964fe586cae7ef996af7ef1c0bef2ab.tar.gz | |
Refactor arm_gemm to enable FP16 in all multi_isa builds
* Resolves MLCE-1285
Change-Id: I22a37972aefe1c0f04accbc798baa18358ed8959
Signed-off-by: Pablo Marquez Tello <pablo.tello@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/11552
Reviewed-by: Viet-Hoa Do <viet-hoa.do@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/cpu/operators/internal/CpuGemmAssemblyDispatch.cpp')
| -rw-r--r-- | src/cpu/operators/internal/CpuGemmAssemblyDispatch.cpp | 10 |
1 files changed, 6 insertions, 4 deletions
diff --git a/src/cpu/operators/internal/CpuGemmAssemblyDispatch.cpp b/src/cpu/operators/internal/CpuGemmAssemblyDispatch.cpp index 7d85885654..cc21ccbaa1 100644 --- a/src/cpu/operators/internal/CpuGemmAssemblyDispatch.cpp +++ b/src/cpu/operators/internal/CpuGemmAssemblyDispatch.cpp @@ -945,6 +945,7 @@ Status CpuGemmAssemblyDispatch::has_opt_impl(arm_compute::WeightFormat &expected } break; #endif /* __aarch64__ */ + #if defined(ARM_COMPUTE_ENABLE_BF16) case DataType::BFLOAT16: { @@ -963,13 +964,14 @@ Status CpuGemmAssemblyDispatch::has_opt_impl(arm_compute::WeightFormat &expected break; } #endif /* defined(ARM_COMPUTE_ENABLE_BF16) */ -#ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC + +#if defined(ARM_COMPUTE_ENABLE_FP16) case DataType::F16: ARM_COMPUTE_RETURN_ERROR_ON_MSG( !(arm_gemm::has_opt_gemm<float16_t, float16_t, arm_gemm::Nothing>(arm_gemm_expected_wf, args, {})), "We could not find an optimized kernel for F16 input and F16 output"); break; -#endif /* __ARM_FEATURE_FP16_VECTOR_ARITHMETIC */ +#endif /* ARM_COMPUTE_ENABLE_FP16 */ default: ARM_COMPUTE_RETURN_ERROR_ON_MSG(true, "Usupported type. Could not find a kernel"); break; @@ -1102,11 +1104,11 @@ void CpuGemmAssemblyDispatch::configure( } break; #endif /* defined(ARM_COMPUTE_ENABLE_BF16) */ -#ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC +#ifdef ARM_COMPUTE_ENABLE_FP16 case DataType::F16: create_arm_gemm<float16_t, float16_t>(_arm_gemm, a, b, c, d, act, info); break; -#endif /* __ARM_FEATURE_FP16_VECTOR_ARITHMETIC */ +#endif /* ARM_COMPUTE_ENABLE_FP16 */ default: break; } |