diff options
| author | Michele Di Giorgio <michele.digiorgio@arm.com> | 2018-04-26 10:24:30 +0100 |
|---|---|---|
| committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:52:35 +0000 |
| commit | f6f08dac6d57770c191d1bc77123f0ddd2363d3f (patch) | |
| tree | 3409794e82c069398fb6eaf74f5fbce645adc2c9 /src/core/CL/cl_kernels/gemm.cl | |
| parent | a4244190b6c7dc7d30d6adc621ca9a8b84b677ee (diff) | |
| download | ComputeLibrary-f6f08dac6d57770c191d1bc77123f0ddd2363d3f.tar.gz | |
COMPMID-1044: Optimizing GCGEMM - Support for not reshaped GEMM on GLES
Change-Id: I22fe80393ec70e4501a4f9f9cad14014029d035d
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129134
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Diffstat (limited to 'src/core/CL/cl_kernels/gemm.cl')
| -rw-r--r-- | src/core/CL/cl_kernels/gemm.cl | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/src/core/CL/cl_kernels/gemm.cl b/src/core/CL/cl_kernels/gemm.cl index 69bc09f023..9ed3af8adc 100644 --- a/src/core/CL/cl_kernels/gemm.cl +++ b/src/core/CL/cl_kernels/gemm.cl @@ -1129,7 +1129,7 @@ __kernel void gemm_mm_interleaved_transposed_qs16(IMAGE_DECLARATION(src0), #if defined(COLS_A) && defined(NUM_ELEMS_PROCESSED_PER_THREAD_X) && (NUM_ELEMS_PROCESSED_PER_THREAD_Y) #if defined(DATA_TYPE) #define VECTOR_TYPE VEC_DATA_TYPE(DATA_TYPE, NUM_ELEMS_PROCESSED_PER_THREAD_X) -/** This OpenCL kernel computes the matrix by matrix multiplication between the matrix A (src0) and matrix B (src1) in case both matrices have not beed reshaped +/** This OpenCL kernel computes the matrix by matrix multiplication between the matrix A (src0) and matrix B (src1) in case both matrices have not been reshaped * * @note This OpenCL kernel works with floating point data types (F16/F32) * @note The floating point data type must be passed at compile time using -DDATA_TYPE (e.g. -DDATA_TYPE=float) @@ -1304,7 +1304,7 @@ __kernel void gemm_mm_floating_point(IMAGE_DECLARATION(src0), } #endif // defined(DATA_TYPE) -/** This OpenCL kernel computes the matrix by matrix multiplication between the matrix A (src0) and matrix B (src1) in case both matrices have not beed reshaped +/** This OpenCL kernel computes the matrix by matrix multiplication between the matrix A (src0) and matrix B (src1) in case both matrices have not been reshaped * * @note This OpenCL kernel works with the 32-bit floating point data type (float) and uses the fma units. * @note The number of elements processed along the x and y directions must be passed at compile time using -DNUM_ELEMS_PROCESSED_PER_THREAD_X and -DNUM_ELEMS_PROCESSED_PER_THREAD_Y. |