COMPMID-3776: Indirect GEMM
Signed-off-by: Georgios Pinitas <georgios.pinitas@arm.com>
Change-Id: I51a1b0f098bc3a8c408c50c92221e4df3061e12c
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/4343
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Sang-Hoon Park <sang-hoon.park@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
diff --git a/src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_12way_16bit.hpp b/src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_12way_16bit.hpp
index 5ab5774..f6233ef 100644
--- a/src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_12way_16bit.hpp
+++ b/src/core/NEON/kernels/arm_gemm/transforms/a64_transpose_interleave_12way_16bit.hpp
@@ -30,12 +30,12 @@
// Generic unblocked transposed 6x32-bit sized specialisation
template <>
template <typename T>
-inline void TransformImpl<6, 1, true, 4, 4, false>::Transform(
+inline void TransformImpl<6, 1, true, 4, 4, VLType::None>::Transform(
T* out, const T* const in, const int stride,
const int x0, const int xmax, const int k0, const int kmax
) {
// Redirect to a 12 x uint16_t specialisation
- TransformImpl<12, 1, true, 2, 2, false>::Transform(
+ TransformImpl<12, 1, true, 2, 2, VLType::None>::Transform(
reinterpret_cast<uint16_t *>(out),
reinterpret_cast<const uint16_t *>(in),
stride*2, x0*2, xmax*2, k0, kmax
@@ -45,7 +45,7 @@
// Generic 12x16-bit sized specialisation
template <>
template <typename T>
-inline void TransformImpl<12, 1, true, 2, 2, false>::Transform(
+inline void TransformImpl<12, 1, true, 2, 2, VLType::None>::Transform(
T* out, const T* const in, const int stride,
const int x0, const int xmax, const int k0, const int kmax
) {
@@ -135,7 +135,7 @@
template <>
template <>
-inline void TransformImpl<12, 1, true, 2, 2, false>::Transform(
+inline void TransformImpl<12, 1, true, 2, 2, VLType::None>::Transform(
uint16_t* out, const uint16_t* const in, const int stride,
const int x0, const int xmax, const int k0, const int kmax
) {