| /llvm-project/mlir/test/Dialect/SPIRV/IR/ |
| H A D | matrix-ops.mlir | 96 // expected-error @+1 {{input and output matrices must have the same component type}}
104 // expected-error @+1 {{right and result matrices must have equal columns' count}}
112 // expected-error @+1 {{left and result matrices must have equal rows' count}}
128 // expected-error @+1 {{right and result matrices' component type must be the same}}
136 // expected-error @+1 {{left and result matrices' component type must be the same}}
|
| /llvm-project/mlir/test/Integration/Dialect/Linalg/CPU/ArmSVE/ |
| H A D | matmul.mlir | 23 // Allocate the matrices
28 // Initialise the matrices
60 // Allocate the matrices
65 // Initialise the matrices
|
| /llvm-project/mlir/test/Integration/Dialect/Vector/CPU/ |
| H A D | transpose.mlir | 17 // Construct test vectors and matrices.
45 // test matrices:
62 // transposed matrices:
75 // back to original after transpose matrices:
|
| H A D | matrix-multiply-row.mlir | 36 // flattened row-major 2-D matrices.
|
| H A D | matrix-multiply-col.mlir | 36 // flattened column-major 2-D matrices.
|
| /llvm-project/mlir/test/Examples/NVGPU/ |
| H A D | Ch4.py | 32 # 2.1 Execute TMA Load for two input matrices for each stage
81 …TMA loads two input matrices from global memory to shared memory. It performs the following operat…
137 Prologue of the GEMM kernel. It loads 2 input matrices for each stage in loop like below:
174 # Initialize A and B (input matrices) and C (accumulator)
197 # Iterate input matrices, update accumulator
|
| H A D | Ch3.py | 11 # 1. Execute TMA Load for two input matrices
32 …TMA loads two input matrices from global memory to shared memory. It performs the following operat…
94 # 1. TMA Load for two input matrices
|
| H A D | Ch5.py | 86 …TMA loads two input matrices from global memory to shared memory. It performs the following operat…
202 # Iterate input matrices, update accumulator
|
| /llvm-project/mlir/include/mlir/Analysis/Presburger/ |
| H A D | Matrix.h | 203 /// Split the rows of a matrix into two matrices according to which bits are
251 // An inherited class for integer matrices, with no new data attributes.
265 /// Given the current matrix M, returns the matrices H, U such that H is the
294 // An inherited class for rational matrices, with no new data attributes.
295 // This class is for functionality that only applies to matrices of fractions.
|
| /llvm-project/mlir/test/Integration/Dialect/SparseTensor/CPU/ |
| H A D | concatenate_dim_1.mlir | 58 // Concats all sparse matrices (with different encodings) to a sparse matrix.
65 // Concats all sparse matrices (with different encodings) to a dense matrix.
72 // Concats mix sparse and dense matrices to a sparse matrix.
79 // Concats mix sparse and dense matrices to a dense matrix.
|
| H A D | concatenate_dim_0.mlir | 56 // Concats all sparse matrices (with different encodings) to a sparse matrix.
63 // Concats all sparse matrices (with different encodings) to a dense matrix.
70 // Concats mix sparse and dense matrices to a sparse matrix.
77 // Concats mix sparse and dense matrices to a dense matrix.
|
| H A D | concatenate_dim_1_permute.mlir | 61 // Concats all sparse matrices (with different encodings) to a sparse matrix.
68 // Concats all sparse matrices (with different encodings) to a dense matrix.
75 // Concats mix sparse and dense matrices to a sparse matrix.
82 // Concats mix sparse and dense matrices to a dense matrix.
|
| H A D | concatenate_dim_0_permute.mlir | 61 // Concats all sparse matrices (with different encodings) to a sparse matrix.
68 // Concats all sparse matrices (with different encodings) to a dense matrix.
75 // Concats mix sparse and dense matrices to a sparse matrix.
82 // Concats mix sparse and dense matrices to a dense matrix.
|
| H A D | sparse_expand.mlir | 60 // Initialize various dense matrices for stress testing.
76 // Convert all these matrices to sparse format.
|
| H A D | sparse_matmul.mlir | 57 // Computes C = A x B with all matrices dense.
68 // Computes C = A x B with all matrices sparse (SpMSpM) in CSR.
80 // Computes C = A x B with all matrices sparse (SpMSpM) in DCSR.
97 // Initialize various matrices, dense for stress testing,
133 // Convert all these matrices to sparse format.
|
| H A D | sparse_matrix_ops.mlir | 95 // Adds two sparse matrices element-wise into a new sparse matrix.
113 // Multiplies two sparse matrices element-wise into a new sparse matrix.
136 // Setup sparse matrices.
|
| /llvm-project/mlir/unittests/Analysis/Presburger/ |
| H A D | LinearTransformTest.cpp | 45 // m1, m2, m3 are rank 1 matrices -- the first and second rows are identical. in TEST()
67 // m4, m5, m6 are rank 2 matrices -- the first and second rows are different. in TEST()
|
| /llvm-project/mlir/test/Integration/Dialect/Linalg/CPU/ |
| H A D | mmt4d.mlir | 13 // Allocate the matrices
32 // Initialise the matrices
|
| /llvm-project/clang/docs/ |
| H A D | MatrixTypes.rst | 11 fixed-size 2-dimensional matrices as language values and perform arithmetic on
140 ``M2`` are of a real type, they are broadcast to matrices here. — end note ]
182 Operations on floating-point matrices have the same rounding and floating-point
|
| /llvm-project/mlir/include/mlir/Dialect/GPU/IR/ |
| H A D | GPUDialect.h | 103 /// operations and are also produced as results. These matrices are meant to
105 /// performed on these matrices, i.e., operations which operate uniformly on
|
| /llvm-project/mlir/lib/ExecutionEngine/ |
| H A D | SparseTensorRuntime.cpp | 184 // Double matrices with all combinations of overhead storage. in _mlir_ciface_newSparseTensor()
218 // Float matrices with all combinations of overhead storage. in _mlir_ciface_newSparseTensor()
262 // Integral matrices with both overheads of the same type. in _mlir_ciface_newSparseTensor()
280 // Complex matrices with wide overhead. in _mlir_ciface_newSparseTensor()
|
| /llvm-project/mlir/lib/Dialect/SPIRV/IR/ |
| H A D | SPIRVOps.cpp | 369 // 1. Cooperative Matrices (1 constituent) in verify()
380 // Case 1. -- matrices. in verify()
1684 // Verify that the input and output matrices have correct shapes.
1693 // Verify that the input and output matrices have the same component type in verify()
1695 return emitError("input and output matrices must have the same " in verify()
1764 // right and result matrices columns' count must be the same in verify()
1767 "right and result matrices must have equal columns' count");
1769 // right and result matrices component type must be the same
1771 return emitError("right and result matrices' component type must" in parse()
1774 // left and result matrices componen in parse()
[all...] |
| /llvm-project/polly/docs/ |
| H A D | Performance.rst | 10 …e computation C |larr| α ⊗ C ⊕ β ⊗ A ⊗ B, where A, B, and C are three appropriately sized matrices,
|
| /llvm-project/llvm/test/CodeGen/MLRegAlloc/ |
| H A D | dev-mode-extra-features-logging.ll | 39 ; Ensure that we can still go through the mapping matrices for the rest of the
|
| /llvm-project/polly/include/polly/ |
| H A D | MatmulOptimizer.h | 41 /// In our case matrices are stored in row-major order instead of
|