2021-05-15 12:02:31 +08:00
|
|
|
// RUN: mlir-opt %s -sparsification | FileCheck %s --check-prefix=CHECK-HIR
|
|
|
|
|
//
|
|
|
|
|
// RUN: mlir-opt %s -sparsification --sparse-tensor-conversion \
|
|
|
|
|
// RUN: --convert-linalg-to-loops | FileCheck %s --check-prefix=CHECK-MIR
|
|
|
|
|
//
|
|
|
|
|
// RUN: mlir-opt %s -sparsification --sparse-tensor-conversion \
|
|
|
|
|
// RUN: --convert-linalg-to-loops --func-bufferize --tensor-constant-bufferize \
|
|
|
|
|
// RUN: --tensor-bufferize --finalizing-bufferize | \
|
|
|
|
|
// RUN: FileCheck %s --check-prefix=CHECK-LIR
|
|
|
|
|
|
|
|
|
|
#CSR = #sparse_tensor.encoding<{dimLevelType = [ "dense", "compressed" ]}>
|
|
|
|
|
|
|
|
|
|
#trait_matvec = {
|
|
|
|
|
indexing_maps = [
|
|
|
|
|
affine_map<(i,j) -> (i,j)>, // A
|
|
|
|
|
affine_map<(i,j) -> (j)>, // b
|
|
|
|
|
affine_map<(i,j) -> (i)> // x (out)
|
|
|
|
|
],
|
|
|
|
|
iterator_types = ["parallel","reduction"],
|
|
|
|
|
doc = "x(i) += A(i,j) * b(j)"
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
// CHECK-HIR-LABEL: func @matvec(
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-HIR-SAME: %[[VAL_0:.*]]: tensor<32x64xf64, #sparse_tensor.encoding<{ dimLevelType = [ "dense", "compressed" ], pointerBitWidth = 0, indexBitWidth = 0 }>>,
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-HIR-SAME: %[[VAL_1:.*]]: tensor<64xf64>,
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-HIR-SAME: %[[VAL_2:.*]]: tensor<32xf64> {linalg.inplaceable = true}) -> tensor<32xf64> {
|
|
|
|
|
// CHECK-HIR: %[[VAL_3:.*]] = constant 32 : index
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-HIR: %[[VAL_4:.*]] = constant 0 : index
|
|
|
|
|
// CHECK-HIR: %[[VAL_5:.*]] = constant 1 : index
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-HIR: %[[VAL_6:.*]] = sparse_tensor.pointers %[[VAL_0]], %[[VAL_5]] : tensor<32x64xf64, #sparse_tensor.encoding<{ dimLevelType = [ "dense", "compressed" ], pointerBitWidth = 0, indexBitWidth = 0 }>> to memref<?xindex>
|
|
|
|
|
// CHECK-HIR: %[[VAL_7:.*]] = sparse_tensor.indices %[[VAL_0]], %[[VAL_5]] : tensor<32x64xf64, #sparse_tensor.encoding<{ dimLevelType = [ "dense", "compressed" ], pointerBitWidth = 0, indexBitWidth = 0 }>> to memref<?xindex>
|
|
|
|
|
// CHECK-HIR: %[[VAL_8:.*]] = sparse_tensor.values %[[VAL_0]] : tensor<32x64xf64, #sparse_tensor.encoding<{ dimLevelType = [ "dense", "compressed" ], pointerBitWidth = 0, indexBitWidth = 0 }>> to memref<?xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-HIR: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_1]] : memref<64xf64>
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-HIR: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-HIR: scf.for %[[VAL_11:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
|
|
|
|
|
// CHECK-HIR: %[[VAL_12:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xindex>
|
|
|
|
|
// CHECK-HIR: %[[VAL_13:.*]] = addi %[[VAL_11]], %[[VAL_5]] : index
|
|
|
|
|
// CHECK-HIR: %[[VAL_14:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_13]]] : memref<?xindex>
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-HIR: %[[VAL_15:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_11]]] : memref<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-HIR: %[[VAL_16:.*]] = scf.for %[[VAL_17:.*]] = %[[VAL_12]] to %[[VAL_14]] step %[[VAL_5]] iter_args(%[[VAL_18:.*]] = %[[VAL_15]]) -> (f64) {
|
|
|
|
|
// CHECK-HIR: %[[VAL_19:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_17]]] : memref<?xindex>
|
|
|
|
|
// CHECK-HIR: %[[VAL_20:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_17]]] : memref<?xf64>
|
|
|
|
|
// CHECK-HIR: %[[VAL_21:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_19]]] : memref<64xf64>
|
|
|
|
|
// CHECK-HIR: %[[VAL_22:.*]] = mulf %[[VAL_20]], %[[VAL_21]] : f64
|
|
|
|
|
// CHECK-HIR: %[[VAL_23:.*]] = addf %[[VAL_18]], %[[VAL_22]] : f64
|
|
|
|
|
// CHECK-HIR: scf.yield %[[VAL_23]] : f64
|
|
|
|
|
// CHECK-HIR: }
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-HIR: memref.store %[[VAL_24:.*]], %[[VAL_10]]{{\[}}%[[VAL_11]]] : memref<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-HIR: }
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-HIR: %[[VAL_25:.*]] = memref.tensor_load %[[VAL_10]] : memref<32xf64>
|
|
|
|
|
// CHECK-HIR: return %[[VAL_25]] : tensor<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-HIR: }
|
|
|
|
|
|
|
|
|
|
// CHECK-MIR-LABEL: func @matvec(
|
|
|
|
|
// CHECK-MIR-SAME: %[[VAL_0:.*]]: !llvm.ptr<i8>,
|
|
|
|
|
// CHECK-MIR-SAME: %[[VAL_1:.*]]: tensor<64xf64>,
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-MIR-SAME: %[[VAL_2:.*]]: tensor<32xf64> {linalg.inplaceable = true}) -> tensor<32xf64> {
|
|
|
|
|
// CHECK-MIR: %[[VAL_3:.*]] = constant 32 : index
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-MIR: %[[VAL_4:.*]] = constant 0 : index
|
|
|
|
|
// CHECK-MIR: %[[VAL_5:.*]] = constant 1 : index
|
|
|
|
|
// CHECK-MIR: %[[VAL_6:.*]] = call @sparsePointers(%[[VAL_0]], %[[VAL_5]]) : (!llvm.ptr<i8>, index) -> memref<?xindex>
|
|
|
|
|
// CHECK-MIR: %[[VAL_7:.*]] = call @sparseIndices(%[[VAL_0]], %[[VAL_5]]) : (!llvm.ptr<i8>, index) -> memref<?xindex>
|
|
|
|
|
// CHECK-MIR: %[[VAL_8:.*]] = call @sparseValuesF64(%[[VAL_0]]) : (!llvm.ptr<i8>) -> memref<?xf64>
|
|
|
|
|
// CHECK-MIR: %[[VAL_9:.*]] = memref.buffer_cast %[[VAL_1]] : memref<64xf64>
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-MIR: %[[VAL_10:.*]] = memref.buffer_cast %[[VAL_2]] : memref<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-MIR: scf.for %[[VAL_11:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
|
|
|
|
|
// CHECK-MIR: %[[VAL_12:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xindex>
|
|
|
|
|
// CHECK-MIR: %[[VAL_13:.*]] = addi %[[VAL_11]], %[[VAL_5]] : index
|
|
|
|
|
// CHECK-MIR: %[[VAL_14:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_13]]] : memref<?xindex>
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-MIR: %[[VAL_15:.*]] = memref.load %[[VAL_10]]{{\[}}%[[VAL_11]]] : memref<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-MIR: %[[VAL_16:.*]] = scf.for %[[VAL_17:.*]] = %[[VAL_12]] to %[[VAL_14]] step %[[VAL_5]] iter_args(%[[VAL_18:.*]] = %[[VAL_15]]) -> (f64) {
|
|
|
|
|
// CHECK-MIR: %[[VAL_19:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_17]]] : memref<?xindex>
|
|
|
|
|
// CHECK-MIR: %[[VAL_20:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_17]]] : memref<?xf64>
|
|
|
|
|
// CHECK-MIR: %[[VAL_21:.*]] = memref.load %[[VAL_9]]{{\[}}%[[VAL_19]]] : memref<64xf64>
|
|
|
|
|
// CHECK-MIR: %[[VAL_22:.*]] = mulf %[[VAL_20]], %[[VAL_21]] : f64
|
|
|
|
|
// CHECK-MIR: %[[VAL_23:.*]] = addf %[[VAL_18]], %[[VAL_22]] : f64
|
|
|
|
|
// CHECK-MIR: scf.yield %[[VAL_23]] : f64
|
|
|
|
|
// CHECK-MIR: }
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-MIR: memref.store %[[VAL_24:.*]], %[[VAL_10]]{{\[}}%[[VAL_11]]] : memref<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-MIR: }
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-MIR: %[[VAL_25:.*]] = memref.tensor_load %[[VAL_10]] : memref<32xf64>
|
|
|
|
|
// CHECK-MIR: return %[[VAL_25]] : tensor<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-MIR: }
|
|
|
|
|
|
|
|
|
|
// CHECK-LIR-LABEL: func @matvec(
|
|
|
|
|
// CHECK-LIR-SAME: %[[VAL_0:.*]]: !llvm.ptr<i8>,
|
|
|
|
|
// CHECK-LIR-SAME: %[[VAL_1:.*]]: memref<64xf64>,
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-LIR-SAME: %[[VAL_2:.*]]: memref<32xf64> {linalg.inplaceable = true}) -> memref<32xf64> {
|
|
|
|
|
// CHECK-LIR: %[[VAL_3:.*]] = constant 32 : index
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-LIR: %[[VAL_4:.*]] = constant 0 : index
|
|
|
|
|
// CHECK-LIR: %[[VAL_5:.*]] = constant 1 : index
|
|
|
|
|
// CHECK-LIR: %[[VAL_6:.*]] = call @sparsePointers(%[[VAL_0]], %[[VAL_5]]) : (!llvm.ptr<i8>, index) -> memref<?xindex>
|
|
|
|
|
// CHECK-LIR: %[[VAL_7:.*]] = call @sparseIndices(%[[VAL_0]], %[[VAL_5]]) : (!llvm.ptr<i8>, index) -> memref<?xindex>
|
|
|
|
|
// CHECK-LIR: %[[VAL_8:.*]] = call @sparseValuesF64(%[[VAL_0]]) : (!llvm.ptr<i8>) -> memref<?xf64>
|
|
|
|
|
// CHECK-LIR: scf.for %[[VAL_9:.*]] = %[[VAL_4]] to %[[VAL_3]] step %[[VAL_5]] {
|
|
|
|
|
// CHECK-LIR: %[[VAL_10:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_9]]] : memref<?xindex>
|
|
|
|
|
// CHECK-LIR: %[[VAL_11:.*]] = addi %[[VAL_9]], %[[VAL_5]] : index
|
|
|
|
|
// CHECK-LIR: %[[VAL_12:.*]] = memref.load %[[VAL_6]]{{\[}}%[[VAL_11]]] : memref<?xindex>
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-LIR: %[[VAL_13:.*]] = memref.load %[[VAL_2]]{{\[}}%[[VAL_9]]] : memref<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-LIR: %[[VAL_14:.*]] = scf.for %[[VAL_15:.*]] = %[[VAL_10]] to %[[VAL_12]] step %[[VAL_5]] iter_args(%[[VAL_16:.*]] = %[[VAL_13]]) -> (f64) {
|
|
|
|
|
// CHECK-LIR: %[[VAL_17:.*]] = memref.load %[[VAL_7]]{{\[}}%[[VAL_15]]] : memref<?xindex>
|
|
|
|
|
// CHECK-LIR: %[[VAL_18:.*]] = memref.load %[[VAL_8]]{{\[}}%[[VAL_15]]] : memref<?xf64>
|
|
|
|
|
// CHECK-LIR: %[[VAL_19:.*]] = memref.load %[[VAL_1]]{{\[}}%[[VAL_17]]] : memref<64xf64>
|
|
|
|
|
// CHECK-LIR: %[[VAL_20:.*]] = mulf %[[VAL_18]], %[[VAL_19]] : f64
|
|
|
|
|
// CHECK-LIR: %[[VAL_21:.*]] = addf %[[VAL_16]], %[[VAL_20]] : f64
|
|
|
|
|
// CHECK-LIR: scf.yield %[[VAL_21]] : f64
|
|
|
|
|
// CHECK-LIR: }
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-LIR: memref.store %[[VAL_22:.*]], %[[VAL_2]]{{\[}}%[[VAL_9]]] : memref<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-LIR: }
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
// CHECK-LIR: return %[[VAL_2]] : memref<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
// CHECK-LIR: }
|
|
|
|
|
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
func @matvec(%arga: tensor<32x64xf64, #CSR>,
|
2021-05-15 12:02:31 +08:00
|
|
|
%argb: tensor<64xf64>,
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
%argx: tensor<32xf64> {linalg.inplaceable = true}) -> tensor<32xf64> {
|
2021-05-15 12:02:31 +08:00
|
|
|
%0 = linalg.generic #trait_matvec
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
ins(%arga, %argb : tensor<32x64xf64, #CSR>, tensor<64xf64>)
|
|
|
|
|
outs(%argx: tensor<32xf64>) {
|
2021-05-15 12:02:31 +08:00
|
|
|
^bb(%A: f64, %b: f64, %x: f64):
|
|
|
|
|
%0 = mulf %A, %b : f64
|
|
|
|
|
%1 = addf %x, %0 : f64
|
|
|
|
|
linalg.yield %1 : f64
|
[mlir][sparse] add full dimension ordering support
This revision completes the "dimension ordering" feature
of sparse tensor types that enables the programmer to
define a preferred order on dimension access (other than
the default left-to-right order). This enables e.g. selection
of column-major over row-major storage for sparse matrices,
but generalized to any rank, as in:
dimOrdering = affine_map<(i,j,k,l,m,n,o,p) -> (p,o,j,k,i,l,m,n)>
Reviewed By: bixia
Differential Revision: https://reviews.llvm.org/D102856
2021-05-22 02:52:34 +08:00
|
|
|
} -> tensor<32xf64>
|
|
|
|
|
return %0 : tensor<32xf64>
|
2021-05-15 12:02:31 +08:00
|
|
|
}
|
