This is an archive of the discontinued LLVM Phabricator instance.

Differential D139489

[mlir][sparse] Replace vector.print with printMemref for some tests.
ClosedPublic

Authored by bixia on Dec 6 2022, 5:00 PM.

Details

Reviewers
aartbik
nicolasvasilache
Commits
rGefaa78cae080: [mlir][sparse] Replace vector.print with printMemref for some tests.

Diff Detail

Event Timeline

bixia created this revision.Dec 6 2022, 5:00 PM
Herald added a reviewer: aartbik. · View Herald TranscriptDec 6 2022, 5:00 PM
Herald added a project: Restricted Project. · View Herald Transcript
Herald added subscribers: hanchung, jsetoain, Moerafaat and 23 others. · View Herald Transcript
bixia requested review of this revision.Dec 6 2022, 5:00 PM
Herald added a reviewer: nicolasvasilache. · View Herald TranscriptDec 6 2022, 5:00 PM
Herald added a project: Restricted Project. · View Herald Transcript
Herald added subscribers: stephenneuendorffer, nicolasvasilache. · View Herald Transcript
aartbik added inline comments.Dec 6 2022, 5:28 PM
mlir/test/Integration/Dialect/SparseTensor/CPU/concatenate.mlir
329

here and below, any reason the -NEXT needed to be dropped?
trying to visualize what changed here

Harbormaster completed remote builds in B201557: Diff 480702.Dec 7 2022, 2:42 AM
bixia added inline comments.Dec 7 2022, 8:28 AM
mlir/test/Integration/Dialect/SparseTensor/CPU/concatenate.mlir
329

printMemref starts with a meta data line, in this case, it is
Memref base@ = 0x35aebc0ca800 rank = 1 offset = 0 sizes = [32] strides = [1] data =

we can add a CHECK-NEXT line to connect this if that is what we want.

bixia updated this revision to Diff 481090.Dec 7 2022, 3:39 PM

Convert more tests.

bixia retitled this revision from [mlir][sparse] Replace vector.print with printMemRef for a test. to [mlir][sparse] Replace vector.print with printMemref for some tests..Dec 7 2022, 3:40 PM
Harbormaster completed remote builds in B201834: Diff 481090.Dec 8 2022, 1:02 AM
aartbik accepted this revision.Dec 9 2022, 3:58 PM
This revision is now accepted and ready to land.Dec 9 2022, 3:58 PM
Closed by commit rGefaa78cae080: [mlir][sparse] Replace vector.print with printMemref for some tests. (authored by bixia). · Explain WhyDec 12 2022, 4:54 PM
This revision was automatically updated to reflect the committed changes.
bixia added a commit: rGefaa78cae080: [mlir][sparse] Replace vector.print with printMemref for some tests..
bixia mentioned this in D139887: [mlir][sparse] Make some integration tests run with vectorization..Dec 13 2022, 11:45 AM

Revision Contents

PathSize
mlir/
test/
Integration/
Dialect/
SparseTensor/
CPU/
222 lines

Diff 480702

mlir/test/Integration/Dialect/SparseTensor/CPU/concatenate.mlir

// DEFINE: %{option} = enable-runtime-library=true // DEFINE: %{option} = enable-runtime-library=true
// DEFINE: %{command} = mlir-opt %s --sparse-compiler=%{option} | \ // DEFINE: %{command} = mlir-opt %s --sparse-compiler=%{option} | \
// DEFINE: mlir-cpu-runner \ // DEFINE: mlir-cpu-runner \
// DEFINE: -e entry -entry-point-result=void \ // DEFINE: -e entry -entry-point-result=void \
// DEFINE: -shared-libs=%mlir_lib_dir/libmlir_c_runner_utils%shlibext | \ // DEFINE: -shared-libs=%mlir_lib_dir/libmlir_c_runner_utils%shlibext,%mlir_lib_dir/libmlir_runner_utils%shlibext | \
// DEFINE: FileCheck %s // DEFINE: FileCheck %s
// //
// RUN: %{command} // RUN: %{command}
// //
// Do the same run, but now with direct IR generation. // Do the same run, but now with direct IR generation.
// REDEFINE: %{option} = enable-runtime-library=false // REDEFINE: %{option} = "enable-runtime-library=false enable-buffer-initialization=true"
// RUN: %{command} // RUN: %{command}
#MAT_C_C = #sparse_tensor.encoding<{dimLevelType = ["compressed", "compressed"]}> #MAT_C_C = #sparse_tensor.encoding<{dimLevelType = ["compressed", "compressed"]}>
#MAT_D_C = #sparse_tensor.encoding<{dimLevelType = ["dense", "compressed"]}> #MAT_D_C = #sparse_tensor.encoding<{dimLevelType = ["dense", "compressed"]}>
#MAT_C_D = #sparse_tensor.encoding<{dimLevelType = ["compressed", "dense"]}> #MAT_C_D = #sparse_tensor.encoding<{dimLevelType = ["compressed", "dense"]}>
#MAT_D_D = #sparse_tensor.encoding<{ #MAT_D_D = #sparse_tensor.encoding<{
dimLevelType = ["dense", "dense"], dimLevelType = ["dense", "dense"],
dimOrdering = affine_map<(i,j) -> (j,i)> dimOrdering = affine_map<(i,j) -> (j,i)>
Show All 10 Lines
}> }>
#MAT_D_C_P = #sparse_tensor.encoding<{ #MAT_D_C_P = #sparse_tensor.encoding<{
dimLevelType = [ "dense", "compressed" ], dimLevelType = [ "dense", "compressed" ],
dimOrdering = affine_map<(i,j) -> (j,i)> dimOrdering = affine_map<(i,j) -> (j,i)>
}> }>
module { module {
func.func private @printMemrefF64(%ptr : tensor<*xf64>)
func.func private @printMemref1dF64(%ptr : memref<?xf64>) attributes { llvm.emit_c_interface }
// //
// Tests without permutation. // Tests without permutation.
// //
// Concats all sparse matrices (with different encodings) to a sparse matrix. // Concats all sparse matrices (with different encodings) to a sparse matrix.
func.func @concat_sparse_sparse(%arg0: tensor<2x4xf64, #MAT_C_C>, %arg1: tensor<3x4xf64, #MAT_C_D>, %arg2: tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_C_C> { func.func @concat_sparse_sparse(%arg0: tensor<2x4xf64, #MAT_C_C>, %arg1: tensor<3x4xf64, #MAT_C_D>, %arg2: tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_C_C> {
%0 = sparse_tensor.concatenate %arg0, %arg1, %arg2 {dimension = 0 : index} %0 = sparse_tensor.concatenate %arg0, %arg1, %arg2 {dimension = 0 : index}
: tensor<2x4xf64, #MAT_C_C>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C> to tensor<9x4xf64, #MAT_C_C> : tensor<2x4xf64, #MAT_C_C>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C> to tensor<9x4xf64, #MAT_C_C>
▲ Show 20 Lines • Show All 129 Lines • ▼ Show 20 Linesmodule {
// //
func.func @concat_mix_sparse_dyn(%arg0: tensor<4x2xf64>, %arg1: tensor<4x3xf64, #MAT_C_D>, %arg2: tensor<4x4xf64, #MAT_D_C>) -> tensor<?x?xf64, #MAT_C_C> { func.func @concat_mix_sparse_dyn(%arg0: tensor<4x2xf64>, %arg1: tensor<4x3xf64, #MAT_C_D>, %arg2: tensor<4x4xf64, #MAT_D_C>) -> tensor<?x?xf64, #MAT_C_C> {
%0 = sparse_tensor.concatenate %arg0, %arg1, %arg2 {dimension = 1 : index} %0 = sparse_tensor.concatenate %arg0, %arg1, %arg2 {dimension = 1 : index}
: tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C> to tensor<?x?xf64, #MAT_C_C> : tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C> to tensor<?x?xf64, #MAT_C_C>
return %0 : tensor<?x?xf64, #MAT_C_C> return %0 : tensor<?x?xf64, #MAT_C_C>
} }
func.func @dump_mat_9x4(%A: tensor<9x4xf64, #MAT_C_C>) { func.func @dump_mat_9x4(%A: tensor<9x4xf64, #MAT_C_C>) {
%c0 = arith.constant 0 : index
%du = arith.constant -1.0 : f64
%c = sparse_tensor.convert %A : tensor<9x4xf64, #MAT_C_C> to tensor<9x4xf64> %c = sparse_tensor.convert %A : tensor<9x4xf64, #MAT_C_C> to tensor<9x4xf64>
%v = vector.transfer_read %c[%c0, %c0], %du: tensor<9x4xf64>, vector<9x4xf64> %cu = tensor.cast %c : tensor<9x4xf64> to tensor<*xf64>
vector.print %v : vector<9x4xf64> call @printMemrefF64(%cu) : (tensor<*xf64>) -> ()
%n = sparse_tensor.number_of_entries %A : tensor<9x4xf64, #MAT_C_C> %n = sparse_tensor.number_of_entries %A : tensor<9x4xf64, #MAT_C_C>
vector.print %n : index vector.print %n : index
%1 = sparse_tensor.values %A : tensor<9x4xf64, #MAT_C_C> to memref<?xf64> %1 = sparse_tensor.values %A : tensor<9x4xf64, #MAT_C_C> to memref<?xf64>
%2 = vector.transfer_read %1[%c0], %du: memref<?xf64>, vector<18xf64> call @printMemref1dF64(%1) : (memref<?xf64>) -> ()
vector.print %2 : vector<18xf64>
return return
} }
func.func @dump_mat_perm_9x4(%A: tensor<9x4xf64, #MAT_C_C_P>) { func.func @dump_mat_perm_9x4(%A: tensor<9x4xf64, #MAT_C_C_P>) {
%c0 = arith.constant 0 : index
%du = arith.constant -1.0 : f64
%c = sparse_tensor.convert %A : tensor<9x4xf64, #MAT_C_C_P> to tensor<9x4xf64> %c = sparse_tensor.convert %A : tensor<9x4xf64, #MAT_C_C_P> to tensor<9x4xf64>
%v = vector.transfer_read %c[%c0, %c0], %du: tensor<9x4xf64>, vector<9x4xf64> %cu = tensor.cast %c : tensor<9x4xf64> to tensor<*xf64>
vector.print %v : vector<9x4xf64> call @printMemrefF64(%cu) : (tensor<*xf64>) -> ()
%n = sparse_tensor.number_of_entries %A : tensor<9x4xf64, #MAT_C_C_P> %n = sparse_tensor.number_of_entries %A : tensor<9x4xf64, #MAT_C_C_P>
vector.print %n : index vector.print %n : index
%1 = sparse_tensor.values %A : tensor<9x4xf64, #MAT_C_C_P> to memref<?xf64> %1 = sparse_tensor.values %A : tensor<9x4xf64, #MAT_C_C_P> to memref<?xf64>
%2 = vector.transfer_read %1[%c0], %du: memref<?xf64>, vector<18xf64> call @printMemref1dF64(%1) : (memref<?xf64>) -> ()
vector.print %2 : vector<18xf64>
return return
} }
func.func @dump_mat_dense_9x4(%A: tensor<9x4xf64>) { func.func @dump_mat_dense_9x4(%A: tensor<9x4xf64>) {
%c0 = arith.constant 0 : index %u = tensor.cast %A : tensor<9x4xf64> to tensor<*xf64>
%du = arith.constant -1.0 : f64 call @printMemrefF64(%u) : (tensor<*xf64>) -> ()
%v = vector.transfer_read %A[%c0, %c0], %du: tensor<9x4xf64>, vector<9x4xf64>
vector.print %v : vector<9x4xf64>
return return
} }
func.func @dump_mat_annotated_dense_9x4(%A: tensor<9x4xf64, #MAT_D_D>) { func.func @dump_mat_annotated_dense_9x4(%A: tensor<9x4xf64, #MAT_D_D>) {
%c0 = arith.constant 0 : index
%du = arith.constant -1.0 : f64
%n = sparse_tensor.number_of_entries %A : tensor<9x4xf64, #MAT_D_D> %n = sparse_tensor.number_of_entries %A : tensor<9x4xf64, #MAT_D_D>
vector.print %n : index vector.print %n : index
%1 = sparse_tensor.values %A : tensor<9x4xf64, #MAT_D_D> to memref<?xf64> %1 = sparse_tensor.values %A : tensor<9x4xf64, #MAT_D_D> to memref<?xf64>
%2 = vector.transfer_read %1[%c0], %du: memref<?xf64>, vector<36xf64> call @printMemref1dF64(%1) : (memref<?xf64>) -> ()
vector.print %2 : vector<36xf64>
return return
} }
func.func @dump_mat_4x9(%A: tensor<4x9xf64, #MAT_C_C>) { func.func @dump_mat_4x9(%A: tensor<4x9xf64, #MAT_C_C>) {
%c0 = arith.constant 0 : index
%du = arith.constant -1.0 : f64
%c = sparse_tensor.convert %A : tensor<4x9xf64, #MAT_C_C> to tensor<4x9xf64> %c = sparse_tensor.convert %A : tensor<4x9xf64, #MAT_C_C> to tensor<4x9xf64>
%v = vector.transfer_read %c[%c0, %c0], %du: tensor<4x9xf64>, vector<4x9xf64> %cu = tensor.cast %c : tensor<4x9xf64> to tensor<*xf64>
vector.print %v : vector<4x9xf64> call @printMemrefF64(%cu) : (tensor<*xf64>) -> ()
%n = sparse_tensor.number_of_entries %A : tensor<4x9xf64, #MAT_C_C> %n = sparse_tensor.number_of_entries %A : tensor<4x9xf64, #MAT_C_C>
vector.print %n : index vector.print %n : index
%1 = sparse_tensor.values %A : tensor<4x9xf64, #MAT_C_C> to memref<?xf64> %1 = sparse_tensor.values %A : tensor<4x9xf64, #MAT_C_C> to memref<?xf64>
%2 = vector.transfer_read %1[%c0], %du: memref<?xf64>, vector<18xf64> call @printMemref1dF64(%1) : (memref<?xf64>) -> ()
vector.print %2 : vector<18xf64>
return return
} }
func.func @dump_mat_dyn(%A: tensor<?x?xf64, #MAT_C_C>) { func.func @dump_mat_dyn(%A: tensor<?x?xf64, #MAT_C_C>) {
%c0 = arith.constant 0 : index
%du = arith.constant -1.0 : f64
%c = sparse_tensor.convert %A : tensor<?x?xf64, #MAT_C_C> to tensor<?x?xf64> %c = sparse_tensor.convert %A : tensor<?x?xf64, #MAT_C_C> to tensor<?x?xf64>
%v = vector.transfer_read %c[%c0, %c0], %du: tensor<?x?xf64>, vector<4x9xf64> %cu = tensor.cast %c : tensor<?x?xf64> to tensor<*xf64>
vector.print %v : vector<4x9xf64> call @printMemrefF64(%cu) : (tensor<*xf64>) -> ()
%n = sparse_tensor.number_of_entries %A : tensor<?x?xf64, #MAT_C_C> %n = sparse_tensor.number_of_entries %A : tensor<?x?xf64, #MAT_C_C>
vector.print %n : index vector.print %n : index
%1 = sparse_tensor.values %A : tensor<?x?xf64, #MAT_C_C> to memref<?xf64> %1 = sparse_tensor.values %A : tensor<?x?xf64, #MAT_C_C> to memref<?xf64>
%2 = vector.transfer_read %1[%c0], %du: memref<?xf64>, vector<18xf64> call @printMemref1dF64(%1) : (memref<?xf64>) -> ()
vector.print %2 : vector<18xf64>
return return
} }
func.func @dump_mat_perm_4x9(%A: tensor<4x9xf64, #MAT_C_C_P>) { func.func @dump_mat_perm_4x9(%A: tensor<4x9xf64, #MAT_C_C_P>) {
%c0 = arith.constant 0 : index
%du = arith.constant -1.0 : f64
%c = sparse_tensor.convert %A : tensor<4x9xf64, #MAT_C_C_P> to tensor<4x9xf64> %c = sparse_tensor.convert %A : tensor<4x9xf64, #MAT_C_C_P> to tensor<4x9xf64>
%v = vector.transfer_read %c[%c0, %c0], %du: tensor<4x9xf64>, vector<4x9xf64> %cu = tensor.cast %c : tensor<4x9xf64> to tensor<*xf64>
vector.print %v : vector<4x9xf64> call @printMemrefF64(%cu) : (tensor<*xf64>) -> ()
%n = sparse_tensor.number_of_entries %A : tensor<4x9xf64, #MAT_C_C_P> %n = sparse_tensor.number_of_entries %A : tensor<4x9xf64, #MAT_C_C_P>
vector.print %n : index vector.print %n : index
%1 = sparse_tensor.values %A : tensor<4x9xf64, #MAT_C_C_P> to memref<?xf64> %1 = sparse_tensor.values %A : tensor<4x9xf64, #MAT_C_C_P> to memref<?xf64>
%2 = vector.transfer_read %1[%c0], %du: memref<?xf64>, vector<18xf64> call @printMemref1dF64(%1) : (memref<?xf64>) -> ()
vector.print %2 : vector<18xf64>
return return
} }
func.func @dump_mat_dense_4x9(%A: tensor<4x9xf64>) { func.func @dump_mat_dense_4x9(%A: tensor<4x9xf64>) {
%c0 = arith.constant 0 : index %1 = tensor.cast %A : tensor<4x9xf64> to tensor<*xf64>
%du = arith.constant -1.0 : f64 call @printMemrefF64(%1) : (tensor<*xf64>) -> ()
%v = vector.transfer_read %A[%c0, %c0], %du: tensor<4x9xf64>, vector<4x9xf64>
vector.print %v : vector<4x9xf64>
return return
} }
// Driver method to call and verify kernels. // Driver method to call and verify kernels.
func.func @entry() { func.func @entry() {
%m42 = arith.constant dense< %m42 = arith.constant dense<
[ [ 1.0, 0.0 ], [ [ 1.0, 0.0 ],
Show All 28 Lines func.func @entry() {
%sm34cdp = sparse_tensor.convert %m34 : tensor<3x4xf64> to tensor<3x4xf64, #MAT_C_D_P> %sm34cdp = sparse_tensor.convert %m34 : tensor<3x4xf64> to tensor<3x4xf64, #MAT_C_D_P>
%sm42ccp = sparse_tensor.convert %m42 : tensor<4x2xf64> to tensor<4x2xf64, #MAT_C_C_P> %sm42ccp = sparse_tensor.convert %m42 : tensor<4x2xf64> to tensor<4x2xf64, #MAT_C_C_P>
%sm43cdp = sparse_tensor.convert %m43 : tensor<4x3xf64> to tensor<4x3xf64, #MAT_C_D_P> %sm43cdp = sparse_tensor.convert %m43 : tensor<4x3xf64> to tensor<4x3xf64, #MAT_C_D_P>
%sm44dcp = sparse_tensor.convert %m44 : tensor<4x4xf64> to tensor<4x4xf64, #MAT_D_C_P> %sm44dcp = sparse_tensor.convert %m44 : tensor<4x4xf64> to tensor<4x4xf64, #MAT_D_C_P>
%sm43cd_dyn = sparse_tensor.convert %m43 : tensor<4x3xf64> to tensor<?x?xf64, #MAT_C_D> %sm43cd_dyn = sparse_tensor.convert %m43 : tensor<4x3xf64> to tensor<?x?xf64, #MAT_C_D>
%sm44dc_dyn = sparse_tensor.convert %m44 : tensor<4x4xf64> to tensor<?x?xf64, #MAT_D_C> %sm44dc_dyn = sparse_tensor.convert %m44 : tensor<4x4xf64> to tensor<?x?xf64, #MAT_D_C>
// CHECK: ( ( 1, 0, 3, 0 ), ( 0, 2, 0, 0 ), ( 1, 0, 1, 1 ), ( 0, 0.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 0, 0, 1.5, 1 ), ( 0, 3.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 3, 0],
// CHECK-NEXT: [0, 2, 0, 0],
// CHECK-NEXT: [1, 0, 1, 1],
// CHECK-NEXT: [0, 0.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 1.5, 1],
// CHECK-NEXT: [0, 3.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [1, 0.5, 0, 0]]
// CHECK-NEXT: 18 // CHECK-NEXT: 18
// CHECK-NEXT: ( 1, 3, 2, 1, 1, 1, 0.5, 1, 5, 2, 1.5, 1, 3.5, 1, 5, 2, 1, 0.5 ) // CHECK: [1, 3, 2, 1, 1, 1, 0.5, 1, 5, 2, 1.5, 1, 3.5, 1, 5, 2, 1, 0.5
aartbikUnsubmitted
Not Done
ReplyInline Actions

here and below, any reason the -NEXT needed to be dropped?
trying to visualize what changed here

aartbik: here and below, any reason the -NEXT needed to be dropped? trying to visualize what changed here
bixiaAuthorUnsubmitted
Done
ReplyInline Actions

printMemref starts with a meta data line, in this case, it is
Memref base@ = 0x35aebc0ca800 rank = 1 offset = 0 sizes = [32] strides = [1] data =

we can add a CHECK-NEXT line to connect this if that is what we want.

bixia: printMemref starts with a meta data line, in this case, it is Memref base@ = 0x35aebc0ca800…
%0 = call @concat_sparse_sparse(%sm24cc, %sm34cd, %sm44dc) %0 = call @concat_sparse_sparse(%sm24cc, %sm34cd, %sm44dc)
: (tensor<2x4xf64, #MAT_C_C>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_C_C> : (tensor<2x4xf64, #MAT_C_C>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_C_C>
call @dump_mat_9x4(%0) : (tensor<9x4xf64, #MAT_C_C>) -> () call @dump_mat_9x4(%0) : (tensor<9x4xf64, #MAT_C_C>) -> ()
// CHECK-NEXT: ( ( 1, 0, 3, 0 ), ( 0, 2, 0, 0 ), ( 1, 0, 1, 1 ), ( 0, 0.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 0, 0, 1.5, 1 ), ( 0, 3.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 3, 0],
// CHECK-NEXT: [0, 2, 0, 0],
// CHECK-NEXT: [1, 0, 1, 1],
// CHECK-NEXT: [0, 0.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 1.5, 1],
// CHECK-NEXT: [0, 3.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [1, 0.5, 0, 0]]
%1 = call @concat_sparse_dense(%sm24cc, %sm34cd, %sm44dc) %1 = call @concat_sparse_dense(%sm24cc, %sm34cd, %sm44dc)
: (tensor<2x4xf64, #MAT_C_C>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64> : (tensor<2x4xf64, #MAT_C_C>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64>
call @dump_mat_dense_9x4(%1) : (tensor<9x4xf64>) -> () call @dump_mat_dense_9x4(%1) : (tensor<9x4xf64>) -> ()
// CHECK-NEXT: ( ( 1, 0, 3, 0 ), ( 0, 2, 0, 0 ), ( 1, 0, 1, 1 ), ( 0, 0.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 0, 0, 1.5, 1 ), ( 0, 3.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 3, 0],
// CHECK-NEXT: [0, 2, 0, 0],
// CHECK-NEXT: [1, 0, 1, 1],
// CHECK-NEXT: [0, 0.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 1.5, 1],
// CHECK-NEXT: [0, 3.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [1, 0.5, 0, 0]]
// CHECK-NEXT: 18 // CHECK-NEXT: 18
// CHECK-NEXT: ( 1, 3, 2, 1, 1, 1, 0.5, 1, 5, 2, 1.5, 1, 3.5, 1, 5, 2, 1, 0.5 ) // CHECK: [1, 3, 2, 1, 1, 1, 0.5, 1, 5, 2, 1.5, 1, 3.5, 1, 5, 2, 1, 0.5
%2 = call @concat_mix_sparse(%m24, %sm34cd, %sm44dc) %2 = call @concat_mix_sparse(%m24, %sm34cd, %sm44dc)
: (tensor<2x4xf64>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_C_C> : (tensor<2x4xf64>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_C_C>
call @dump_mat_9x4(%2) : (tensor<9x4xf64, #MAT_C_C>) -> () call @dump_mat_9x4(%2) : (tensor<9x4xf64, #MAT_C_C>) -> ()
// CHECK-NEXT: ( ( 1, 0, 3, 0 ), ( 0, 2, 0, 0 ), ( 1, 0, 1, 1 ), ( 0, 0.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 0, 0, 1.5, 1 ), ( 0, 3.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 3, 0],
// CHECK-NEXT: [0, 2, 0, 0],
// CHECK-NEXT: [1, 0, 1, 1],
// CHECK-NEXT: [0, 0.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 1.5, 1],
// CHECK-NEXT: [0, 3.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [1, 0.5, 0, 0]]
%3 = call @concat_mix_dense(%m24, %sm34cd, %sm44dc) %3 = call @concat_mix_dense(%m24, %sm34cd, %sm44dc)
: (tensor<2x4xf64>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64> : (tensor<2x4xf64>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64>
call @dump_mat_dense_9x4(%3) : (tensor<9x4xf64>) -> () call @dump_mat_dense_9x4(%3) : (tensor<9x4xf64>) -> ()
// CHECK-NEXT: ( ( 1, 0, 3, 0 ), ( 0, 2, 0, 0 ), ( 1, 0, 1, 1 ), ( 0, 0.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 0, 0, 1.5, 1 ), ( 0, 3.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 3, 0],
// CHECK-NEXT: [0, 2, 0, 0],
// CHECK-NEXT: [1, 0, 1, 1],
// CHECK-NEXT: [0, 0.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 1.5, 1],
// CHECK-NEXT: [0, 3.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [1, 0.5, 0, 0]]
// CHECK-NEXT: 18 // CHECK-NEXT: 18
// CHECK-NEXT: ( 1, 1, 1, 1, 1, 2, 0.5, 5, 3.5, 5, 0.5, 3, 1, 2, 1.5, 2, 1, 1 ) // CHECK: [1, 1, 1, 1, 1, 2, 0.5, 5, 3.5, 5, 0.5, 3, 1, 2, 1.5, 2, 1, 1
%4 = call @concat_sparse_sparse_perm(%sm24ccp, %sm34cd, %sm44dc) %4 = call @concat_sparse_sparse_perm(%sm24ccp, %sm34cd, %sm44dc)
: (tensor<2x4xf64, #MAT_C_C_P>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_C_C_P> : (tensor<2x4xf64, #MAT_C_C_P>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_C_C_P>
call @dump_mat_perm_9x4(%4) : (tensor<9x4xf64, #MAT_C_C_P>) -> () call @dump_mat_perm_9x4(%4) : (tensor<9x4xf64, #MAT_C_C_P>) -> ()
// CHECK-NEXT: ( ( 1, 0, 3, 0 ), ( 0, 2, 0, 0 ), ( 1, 0, 1, 1 ), ( 0, 0.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 0, 0, 1.5, 1 ), ( 0, 3.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 3, 0],
// CHECK-NEXT: [0, 2, 0, 0],
// CHECK-NEXT: [1, 0, 1, 1],
// CHECK-NEXT: [0, 0.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 1.5, 1],
// CHECK-NEXT: [0, 3.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [1, 0.5, 0, 0]]
%5 = call @concat_sparse_dense_perm(%sm24ccp, %sm34cdp, %sm44dc) %5 = call @concat_sparse_dense_perm(%sm24ccp, %sm34cdp, %sm44dc)
: (tensor<2x4xf64, #MAT_C_C_P>, tensor<3x4xf64, #MAT_C_D_P>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64> : (tensor<2x4xf64, #MAT_C_C_P>, tensor<3x4xf64, #MAT_C_D_P>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64>
call @dump_mat_dense_9x4(%5) : (tensor<9x4xf64>) -> () call @dump_mat_dense_9x4(%5) : (tensor<9x4xf64>) -> ()
// CHECK-NEXT: ( ( 1, 0, 3, 0 ), ( 0, 2, 0, 0 ), ( 1, 0, 1, 1 ), ( 0, 0.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 0, 0, 1.5, 1 ), ( 0, 3.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 3, 0],
// CHECK-NEXT: [0, 2, 0, 0],
// CHECK-NEXT: [1, 0, 1, 1],
// CHECK-NEXT: [0, 0.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 1.5, 1],
// CHECK-NEXT: [0, 3.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [1, 0.5, 0, 0]]
// CHECK-NEXT: 18 // CHECK-NEXT: 18
// CHECK-NEXT: ( 1, 3, 2, 1, 1, 1, 0.5, 1, 5, 2, 1.5, 1, 3.5, 1, 5, 2, 1, 0.5 ) // CHECK: [1, 3, 2, 1, 1, 1, 0.5, 1, 5, 2, 1.5, 1, 3.5, 1, 5, 2, 1, 0.5
%6 = call @concat_mix_sparse_perm(%m24, %sm34cdp, %sm44dc) %6 = call @concat_mix_sparse_perm(%m24, %sm34cdp, %sm44dc)
: (tensor<2x4xf64>, tensor<3x4xf64, #MAT_C_D_P>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_C_C> : (tensor<2x4xf64>, tensor<3x4xf64, #MAT_C_D_P>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_C_C>
call @dump_mat_9x4(%6) : (tensor<9x4xf64, #MAT_C_C>) -> () call @dump_mat_9x4(%6) : (tensor<9x4xf64, #MAT_C_C>) -> ()
// CHECK-NEXT: ( ( 1, 0, 3, 0 ), ( 0, 2, 0, 0 ), ( 1, 0, 1, 1 ), ( 0, 0.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 0, 0, 1.5, 1 ), ( 0, 3.5, 0, 0 ), ( 1, 5, 2, 0 ), ( 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 3, 0],
// CHECK-NEXT: [0, 2, 0, 0],
// CHECK-NEXT: [1, 0, 1, 1],
// CHECK-NEXT: [0, 0.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 1.5, 1],
// CHECK-NEXT: [0, 3.5, 0, 0],
// CHECK-NEXT: [1, 5, 2, 0],
// CHECK-NEXT: [1, 0.5, 0, 0]]
%7 = call @concat_mix_dense_perm(%m24, %sm34cd, %sm44dcp) %7 = call @concat_mix_dense_perm(%m24, %sm34cd, %sm44dcp)
: (tensor<2x4xf64>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C_P>) -> tensor<9x4xf64> : (tensor<2x4xf64>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C_P>) -> tensor<9x4xf64>
call @dump_mat_dense_9x4(%7) : (tensor<9x4xf64>) -> () call @dump_mat_dense_9x4(%7) : (tensor<9x4xf64>) -> ()
// CHECK-NEXT: ( ( 1, 0, 1, 0, 1, 0, 0, 1.5, 1 ), ( 3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0 ), ( 0, 2, 0, 0, 1, 1, 5, 2, 0 ), ( 0, 0, 5, 2, 0, 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 1, 0, 1, 0, 0, 1.5, 1],
// CHECK-NEXT: [3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0],
// CHECK-NEXT: [0, 2, 0, 0, 1, 1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 5, 2, 0, 1, 0.5, 0, 0]]
// CHECK-NEXT: 18 // CHECK-NEXT: 18
// CHECK-NEXT: ( 1, 1, 1, 1.5, 1, 3.1, 1, 0.5, 3.5, 2, 1, 1, 5, 2, 5, 2, 1, 0.5 ) // CHECK: [1, 1, 1, 1.5, 1, 3.1, 1, 0.5, 3.5, 2, 1, 1, 5, 2, 5, 2, 1, 0.5
%8 = call @concat_sparse_sparse_dim1(%sm42cc, %sm43cd, %sm44dc) %8 = call @concat_sparse_sparse_dim1(%sm42cc, %sm43cd, %sm44dc)
: (tensor<4x2xf64, #MAT_C_C>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64, #MAT_C_C> : (tensor<4x2xf64, #MAT_C_C>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64, #MAT_C_C>
call @dump_mat_4x9(%8) : (tensor<4x9xf64, #MAT_C_C>) -> () call @dump_mat_4x9(%8) : (tensor<4x9xf64, #MAT_C_C>) -> ()
// CHECK-NEXT: ( ( 1, 0, 1, 0, 1, 0, 0, 1.5, 1 ), ( 3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0 ), ( 0, 2, 0, 0, 1, 1, 5, 2, 0 ), ( 0, 0, 5, 2, 0, 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 1, 0, 1, 0, 0, 1.5, 1],
// CHECK-NEXT: [3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0],
// CHECK-NEXT: [0, 2, 0, 0, 1, 1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 5, 2, 0, 1, 0.5, 0, 0]]
%9 = call @concat_sparse_dense_dim1(%sm42cc, %sm43cd, %sm44dc) %9 = call @concat_sparse_dense_dim1(%sm42cc, %sm43cd, %sm44dc)
: (tensor<4x2xf64, #MAT_C_C>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64> : (tensor<4x2xf64, #MAT_C_C>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64>
call @dump_mat_dense_4x9(%9) : (tensor<4x9xf64>) -> () call @dump_mat_dense_4x9(%9) : (tensor<4x9xf64>) -> ()
// CHECK-NEXT: ( ( 1, 0, 1, 0, 1, 0, 0, 1.5, 1 ), ( 3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0 ), ( 0, 2, 0, 0, 1, 1, 5, 2, 0 ), ( 0, 0, 5, 2, 0, 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 1, 0, 1, 0, 0, 1.5, 1],
// CHECK-NEXT: [3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0],
// CHECK-NEXT: [0, 2, 0, 0, 1, 1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 5, 2, 0, 1, 0.5, 0, 0]]
// CHECK-NEXT: 18 // CHECK-NEXT: 18
// CHECK-NEXT: ( 1, 1, 1, 1.5, 1, 3.1, 1, 0.5, 3.5, 2, 1, 1, 5, 2, 5, 2, 1, 0.5 ) // CHECK: [1, 1, 1, 1.5, 1, 3.1, 1, 0.5, 3.5, 2, 1, 1, 5, 2, 5, 2, 1, 0.5
%10 = call @concat_mix_sparse_dim1(%m42, %sm43cd, %sm44dc) %10 = call @concat_mix_sparse_dim1(%m42, %sm43cd, %sm44dc)
: (tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64, #MAT_C_C> : (tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64, #MAT_C_C>
call @dump_mat_4x9(%10) : (tensor<4x9xf64, #MAT_C_C>) -> () call @dump_mat_4x9(%10) : (tensor<4x9xf64, #MAT_C_C>) -> ()
// CHECK-NEXT: ( ( 1, 0, 1, 0, 1, 0, 0, 1.5, 1 ), ( 3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0 ), ( 0, 2, 0, 0, 1, 1, 5, 2, 0 ), ( 0, 0, 5, 2, 0, 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 1, 0, 1, 0, 0, 1.5, 1],
// CHECK-NEXT: [3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0],
// CHECK-NEXT: [0, 2, 0, 0, 1, 1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 5, 2, 0, 1, 0.5, 0, 0]]
%11 = call @concat_mix_dense_dim1(%m42, %sm43cd, %sm44dc) %11 = call @concat_mix_dense_dim1(%m42, %sm43cd, %sm44dc)
: (tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64> : (tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64>
call @dump_mat_dense_4x9(%11) : (tensor<4x9xf64>) -> () call @dump_mat_dense_4x9(%11) : (tensor<4x9xf64>) -> ()
// CHECK-NEXT: ( ( 1, 0, 1, 0, 1, 0, 0, 1.5, 1 ), ( 3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0 ), ( 0, 2, 0, 0, 1, 1, 5, 2, 0 ), ( 0, 0, 5, 2, 0, 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 1, 0, 1, 0, 0, 1.5, 1],
// CHECK-NEXT: [3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0],
// CHECK-NEXT: [0, 2, 0, 0, 1, 1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 5, 2, 0, 1, 0.5, 0, 0]]
// CHECK-NEXT: 18 // CHECK-NEXT: 18
// CHECK-NEXT: ( 1, 3.1, 2, 1, 1, 5, 2, 1, 0.5, 1, 1, 1, 3.5, 5, 0.5, 1.5, 2, 1 ) // CHECK: [1, 3.1, 2, 1, 1, 5, 2, 1, 0.5, 1, 1, 1, 3.5, 5, 0.5, 1.5, 2, 1
%12 = call @concat_sparse_sparse_perm_dim1(%sm42ccp, %sm43cd, %sm44dc) %12 = call @concat_sparse_sparse_perm_dim1(%sm42ccp, %sm43cd, %sm44dc)
: (tensor<4x2xf64, #MAT_C_C_P>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64, #MAT_C_C_P> : (tensor<4x2xf64, #MAT_C_C_P>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64, #MAT_C_C_P>
call @dump_mat_perm_4x9(%12) : (tensor<4x9xf64, #MAT_C_C_P>) -> () call @dump_mat_perm_4x9(%12) : (tensor<4x9xf64, #MAT_C_C_P>) -> ()
// CHECK-NEXT: ( ( 1, 0, 1, 0, 1, 0, 0, 1.5, 1 ), ( 3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0 ), ( 0, 2, 0, 0, 1, 1, 5, 2, 0 ), ( 0, 0, 5, 2, 0, 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 1, 0, 1, 0, 0, 1.5, 1],
// CHECK-NEXT: [3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0],
// CHECK-NEXT: [0, 2, 0, 0, 1, 1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 5, 2, 0, 1, 0.5, 0, 0]]
%13 = call @concat_sparse_dense_perm_dim1(%sm42ccp, %sm43cdp, %sm44dc) %13 = call @concat_sparse_dense_perm_dim1(%sm42ccp, %sm43cdp, %sm44dc)
: (tensor<4x2xf64, #MAT_C_C_P>, tensor<4x3xf64, #MAT_C_D_P>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64> : (tensor<4x2xf64, #MAT_C_C_P>, tensor<4x3xf64, #MAT_C_D_P>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64>
call @dump_mat_dense_4x9(%13) : (tensor<4x9xf64>) -> () call @dump_mat_dense_4x9(%13) : (tensor<4x9xf64>) -> ()
// CHECK-NEXT: ( ( 1, 0, 1, 0, 1, 0, 0, 1.5, 1 ), ( 3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0 ), ( 0, 2, 0, 0, 1, 1, 5, 2, 0 ), ( 0, 0, 5, 2, 0, 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 1, 0, 1, 0, 0, 1.5, 1],
// CHECK-NEXT: [3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0],
// CHECK-NEXT: [0, 2, 0, 0, 1, 1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 5, 2, 0, 1, 0.5, 0, 0]]
// CHECK-NEXT: 18 // CHECK-NEXT: 18
// CHECK-NEXT: ( 1, 1, 1, 1.5, 1, 3.1, 1, 0.5, 3.5, 2, 1, 1, 5, 2, 5, 2, 1, 0.5 ) // CHECK: [1, 1, 1, 1.5, 1, 3.1, 1, 0.5, 3.5, 2, 1, 1, 5, 2, 5, 2, 1, 0.5
%14 = call @concat_mix_sparse_perm_dim1(%m42, %sm43cdp, %sm44dc) %14 = call @concat_mix_sparse_perm_dim1(%m42, %sm43cdp, %sm44dc)
: (tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D_P>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64, #MAT_C_C> : (tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D_P>, tensor<4x4xf64, #MAT_D_C>) -> tensor<4x9xf64, #MAT_C_C>
call @dump_mat_4x9(%14) : (tensor<4x9xf64, #MAT_C_C>) -> () call @dump_mat_4x9(%14) : (tensor<4x9xf64, #MAT_C_C>) -> ()
// CHECK-NEXT: ( ( 1, 0, 1, 0, 1, 0, 0, 1.5, 1 ), ( 3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0 ), ( 0, 2, 0, 0, 1, 1, 5, 2, 0 ), ( 0, 0, 5, 2, 0, 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 1, 0, 1, 0, 0, 1.5, 1],
// CHECK-NEXT: [3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0],
// CHECK-NEXT: [0, 2, 0, 0, 1, 1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 5, 2, 0, 1, 0.5, 0, 0]]
%15 = call @concat_mix_dense_perm_dim1(%m42, %sm43cd, %sm44dcp) %15 = call @concat_mix_dense_perm_dim1(%m42, %sm43cd, %sm44dcp)
: (tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C_P>) -> tensor<4x9xf64> : (tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C_P>) -> tensor<4x9xf64>
call @dump_mat_dense_4x9(%15) : (tensor<4x9xf64>) -> () call @dump_mat_dense_4x9(%15) : (tensor<4x9xf64>) -> ()
// CHECK-NEXT: ( ( 1, 0, 1, 0, 1, 0, 0, 1.5, 1 ), ( 3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0 ), ( 0, 2, 0, 0, 1, 1, 5, 2, 0 ), ( 0, 0, 5, 2, 0, 1, 0.5, 0, 0 ) ) // CHECK: {{\[}}[1, 0, 1, 0, 1, 0, 0, 1.5, 1],
// CHECK-NEXT: [3.1, 0, 1, 0, 0.5, 0, 3.5, 0, 0],
// CHECK-NEXT: [0, 2, 0, 0, 1, 1, 5, 2, 0],
// CHECK-NEXT: [0, 0, 5, 2, 0, 1, 0.5, 0, 0]]
// CHECK-NEXT: 18 // CHECK-NEXT: 18
// CHECK-NEXT: ( 1, 1, 1, 1.5, 1, 3.1, 1, 0.5, 3.5, 2, 1, 1, 5, 2, 5, 2, 1, 0.5 ) // CHECK: [1, 1, 1, 1.5, 1, 3.1, 1, 0.5, 3.5, 2, 1, 1, 5, 2, 5, 2, 1, 0.5
%16 = call @concat_mix_sparse_dyn(%m42, %sm43cd, %sm44dc) %16 = call @concat_mix_sparse_dyn(%m42, %sm43cd, %sm44dc)
: (tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<?x?xf64, #MAT_C_C> : (tensor<4x2xf64>, tensor<4x3xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<?x?xf64, #MAT_C_C>
call @dump_mat_dyn(%16) : (tensor<?x?xf64, #MAT_C_C>) -> () call @dump_mat_dyn(%16) : (tensor<?x?xf64, #MAT_C_C>) -> ()
// CHECK-NEXT: 36 // CHECK-NEXT: 36
// CHECK-NEXT: ( 1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 2, 0, 0.5, 5, 0, 3.5, 5, 0.5, 3, 0, 1, 0, 2, 1.5, 0, 2, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0 ) // CHECK: [1, 0, 1, 0, 1, 0, 0, 1, 1, 0, 2, 0, 0.5, 5, 0, 3.5, 5, 0.5, 3, 0, 1, 0, 2, 1.5, 0, 2, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0
%17 = call @concat_sparse_annotated_dense(%sm24cc, %sm34cd, %sm44dc) %17 = call @concat_sparse_annotated_dense(%sm24cc, %sm34cd, %sm44dc)
: (tensor<2x4xf64, #MAT_C_C>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_D_D> : (tensor<2x4xf64, #MAT_C_C>, tensor<3x4xf64, #MAT_C_D>, tensor<4x4xf64, #MAT_D_C>) -> tensor<9x4xf64, #MAT_D_D>
call @dump_mat_annotated_dense_9x4(%17) : (tensor<9x4xf64, #MAT_D_D>) -> () call @dump_mat_annotated_dense_9x4(%17) : (tensor<9x4xf64, #MAT_D_D>) -> ()
// Release resources. // Release resources.
bufferization.dealloc_tensor %sm24cc : tensor<2x4xf64, #MAT_C_C> bufferization.dealloc_tensor %sm24cc : tensor<2x4xf64, #MAT_C_C>
bufferization.dealloc_tensor %sm34cd : tensor<3x4xf64, #MAT_C_D> bufferization.dealloc_tensor %sm34cd : tensor<3x4xf64, #MAT_C_D>
Show All 29 Lines