| /llvm-project-15.0.7/mlir/test/Dialect/Tosa/ |
| H A D | broadcast.mlir | 6 // CHECK-NOT: reshape 14 // CHECK-DAG: %[[VAR0:.*]] = "tosa.reshape"(%arg0) {new_shape = [1, 1]} 23 // CHECK-DAG: %[[VAR0:.*]] = "tosa.reshape"(%arg1) {new_shape = [1, 1]} 32 // CHECK-DAG: %[[VAR0:.*]] = "tosa.reshape"(%arg1) {new_shape = [1, 1, 1, 1]} 41 // CHECK-DAG: %[[VAR0:.*]] = "tosa.reshape"(%arg1) {new_shape = [1, 1, 1, 1]} 50 // CHECK-DAG: %[[VAR0:.*]] = "tosa.reshape"(%arg1) {new_shape = [1, 1, 1, 1]} 59 // CHECK-DAG: %[[VAR0:.*]] = "tosa.reshape"(%arg1) {new_shape = [1, 1, 1, 1]} 68 // CHECK-DAG: %[[VAR0:.*]] = "tosa.reshape"(%arg1) {new_shape = [1, 1, 1, 1]} 77 // CHECK-DAG: %[[VAR0:.*]] = "tosa.reshape"(%arg1) {new_shape = [1, 1, 1, 1]} 104 // CHECK-DAG: %[[VAR0:.*]] = "tosa.reshape"(%arg0) {new_shape = [1, 1, 1, 1]} [all …]
|
| H A D | tosa-decompose-conv2d.mlir | 8 // CHECK: %[[VAR0:.*]] = "tosa.reshape"(%arg0) {new_shape = [400, 2]} 10 // CHECK: %[[VAR1:.*]] = "tosa.reshape"(%arg1) {new_shape = [3, 2]} 14 // CHECK: %[[VAR3:.*]] = "tosa.reshape"(%[[VAR2]]) {new_shape = [4, 10, 10, 3]} 26 // CHECK: %[[VAR0:.*]] = "tosa.reshape"(%arg0) {new_shape = [400, 2]} 28 // CHECK: %[[VAR1:.*]] = "tosa.reshape"(%arg1) {new_shape = [3, 2]} 33 // CHECK: %[[VAR3:.*]] = "tosa.reshape"(%[[VAR2]]) {new_shape = [4, 10, 10, 3]} 47 // CHECK: %[[VAL_3:.*]] = "tosa.reshape"(%[[VAL_0]]) {new_shape = [-1, 64]} : (tensor<?x1… 48 // CHECK: %[[VAL_4:.*]] = "tosa.reshape"(%[[VAL_1]]) {new_shape = [384, 64]} : (tensor<38… 50 // CHECK: %[[VAL_6:.*]] = "tosa.reshape"(%[[VAL_5]]) {new_shape = [-1, 14, 14, 384]} : (t…
|
| H A D | tosa-decompose-transpose-conv.mlir | 43 // CHECK-DAG: %[[RESW1:.+]] = "tosa.reshape"(%[[PADW]]) {new_shape = [5, 2, 2, 2, 3, 3]} 45 // CHECK-DAG: %[[RESW2:.+]] = "tosa.reshape"(%[[TRANS]]) {new_shape = [30, 2, 2, 3]} 57 // CHECK-DAG: %[[RESHAPE_OUT_1:.+]] = "tosa.reshape"(%[[CONV]]) {new_shape = [2, 18, 16, 2, 3, 5]} 59 // CHECK-DAG: %[[RESHAPE_OUT_2:.+]] = "tosa.reshape"(%[[TRANS_OUT]]) {new_shape = [2, 36, 48, 5]} 75 // CHECK-DAG: %[[RESW1:.+]] = "tosa.reshape"(%[[PADW]]) {new_shape = [5, 2, 2, 2, 3, 3]} 77 // CHECK-DAG: %[[RESW2:.+]] = "tosa.reshape"(%[[TRANS]]) {new_shape = [30, 2, 2, 3]} 89 // CHECK-DAG: %[[RESHAPE_OUT_1:.+]] = "tosa.reshape"(%[[CONV]]) {new_shape = [2, 18, 16, 2, 3, 5]} 91 // CHECK-DAG: %[[RESHAPE_OUT_2:.+]] = "tosa.reshape"(%[[TRANS_OUT]]) {new_shape = [2, 36, 48, 5]}
|
| H A D | tosa-decompose-depthwise.mlir | 8 // CHECK: %[[VAR0:.*]] = "tosa.reshape"(%arg0) {new_shape = [4, 10, 10, 2, 1]} 10 // CHECK: %[[VAR1:.*]] = "tosa.reshape"(%arg1) {new_shape = [1, 1, 1, 2, 3]} 14 // CHECK: %[[VAR3:.*]] = "tosa.reshape"(%[[VAR2]]) {new_shape = [4, 10, 10, 6]}
|
| H A D | canonicalize.mlir | 326 %0 = "tosa.reshape"(%arg0) {new_shape = [-1, 10]}: (tensor<?x10xf32>) -> tensor<?x10xf32> 332 // CHECK: %[[VAR0:.+]] = "tosa.reshape"(%arg0) {new_shape = [-1, 5]} 334 %0 = "tosa.reshape"(%arg0) {new_shape = [5, -1]}: (tensor<?x10xf32>) -> tensor<5x?xf32> 335 %1 = "tosa.reshape"(%0) {new_shape = [-1, 5]}: (tensor<5x?xf32>) -> tensor<?x5xf32> 344 %1 = "tosa.reshape"(%0) {new_shape = [1, 10]} : (tensor<10xi32>) -> tensor<1x10xi32> 354 %1 = "tosa.reshape"(%0) {new_shape = [1, 10]} : (tensor<10xi32>) -> tensor<1x10xi32> 360 //CHECK: "tosa.reshape" 362 %1 = "tosa.reshape"(%0) {new_shape = [1, 3]} : (tensor<3xi32>) -> tensor<1x3xi32> 408 %1 = "tosa.reshape"(%0) {new_shape = [1]} : (tensor<1x1xi1>) -> tensor<1xi1>
|
| H A D | ops.mlir | 340 %1 = "tosa.reshape"(%0) {new_shape = [21, 3]} : (tensor<1x21x3xi1>) -> tensor<21x3xi1> 348 %1 = "tosa.reshape"(%0) {new_shape = [21, 3]} : (tensor<1x21x3xi1>) -> tensor<21x3xi1> 356 %1 = "tosa.reshape"(%0) {new_shape = [21, 3]} : (tensor<1x21x3xf32>) -> tensor<21x3xf32> 364 %1 = "tosa.reshape"(%0) {new_shape = [21, 3]} : (tensor<1x21x3xf32>) -> tensor<21x3xf32> 372 %1 = "tosa.reshape"(%0) {new_shape = [21, 3]} : (tensor<1x21x3xf32>) -> tensor<21x3xf32> 380 %1 = "tosa.reshape"(%0) {new_shape = [21, 3]} : (tensor<1x21x3xf32>) -> tensor<21x3xf32> 407 // CHECK-LABEL: reshape 409 %0 = "tosa.reshape"(%arg0) {new_shape = [1, 819]} : (tensor<13x21x3xf32>) -> tensor<1x819xf32> 533 %4 = "tosa.reshape"(%2) {new_shape = [1]} : (tensor<i32>) -> tensor<1xi32>
|
| /llvm-project-15.0.7/clang/test/Analysis/ |
| H A D | malloc-interprocedural.c | 77 static char *reshape(char *in) { in reshape() function 83 v = reshape(v); in testThatRemoveDeadBindingsRunBeforeEachCall() 84 v = reshape(v);// expected-warning {{Potential leak of memory pointed to by 'v'}} in testThatRemoveDeadBindingsRunBeforeEachCall()
|
| /llvm-project-15.0.7/mlir/test/mlir-cpu-runner/ |
| H A D | memref-reshape.mlir | 53 %output = memref.reshape %input(%shape) 68 %output = memref.reshape %input(%shape) 83 %output = memref.reshape %input(%dyn_size_shape) 98 %output = memref.reshape %input(%dyn_size_shape)
|
| /llvm-project-15.0.7/mlir/docs/Tutorials/Toy/ |
| H A D | Ch-3.md | 179 A redundant reshape optimization similar to SimplifyRedundantTranspose can be 208 eliminating the reshape operation. 211 def ReshapeConstant : NativeCodeCall<"$0.reshape(($1.getType()).cast<ShapedType>())">; 217 We demonstrate these reshape optimizations using the following 233 %1 = toy.reshape(%0 : tensor<2xf64>) to tensor<2x1xf64> 234 %2 = toy.reshape(%1 : tensor<2x1xf64>) to tensor<2x1xf64> 235 %3 = toy.reshape(%2 : tensor<2x1xf64>) to tensor<2x1xf64> 255 As expected, no reshape operations remain after canonicalization.
|
| /llvm-project-15.0.7/mlir/test/Examples/Toy/Ch4/ |
| H A D | shape_inference.mlir | 13 %1 = toy.reshape(%0 : tensor<2x3xf64>) to tensor<2x3xf64> 15 %3 = toy.reshape(%2 : tensor<6xf64>) to tensor<2x3xf64>
|
| H A D | codegen.toy | 25 # CHECK-NEXT: [[VAL_6:%.*]] = toy.reshape([[VAL_5]] : tensor<2x3xf64>) to tensor<2x3xf64> 27 # CHECK-NEXT: [[VAL_8:%.*]] = toy.reshape([[VAL_7]] : tensor<6xf64>) to tensor<2x3xf64>
|
| /llvm-project-15.0.7/mlir/test/Examples/Toy/Ch6/ |
| H A D | shape_inference.mlir | 13 %1 = toy.reshape(%0 : tensor<2x3xf64>) to tensor<2x3xf64> 15 %3 = toy.reshape(%2 : tensor<6xf64>) to tensor<2x3xf64>
|
| H A D | codegen.toy | 25 # CHECK-NEXT: [[VAL_6:%.*]] = toy.reshape([[VAL_5]] : tensor<2x3xf64>) to tensor<2x3xf64> 27 # CHECK-NEXT: [[VAL_8:%.*]] = toy.reshape([[VAL_7]] : tensor<6xf64>) to tensor<2x3xf64>
|
| /llvm-project-15.0.7/mlir/test/Examples/Toy/Ch5/ |
| H A D | shape_inference.mlir | 13 %1 = toy.reshape(%0 : tensor<2x3xf64>) to tensor<2x3xf64> 15 %3 = toy.reshape(%2 : tensor<6xf64>) to tensor<2x3xf64>
|
| H A D | codegen.toy | 25 # CHECK-NEXT: [[VAL_6:%.*]] = toy.reshape([[VAL_5]] : tensor<2x3xf64>) to tensor<2x3xf64> 27 # CHECK-NEXT: [[VAL_8:%.*]] = toy.reshape([[VAL_7]] : tensor<6xf64>) to tensor<2x3xf64>
|
| /llvm-project-15.0.7/mlir/test/Examples/Toy/Ch7/ |
| H A D | shape_inference.mlir | 13 %1 = toy.reshape(%0 : tensor<2x3xf64>) to tensor<2x3xf64> 15 %3 = toy.reshape(%2 : tensor<6xf64>) to tensor<2x3xf64>
|
| H A D | codegen.toy | 25 # CHECK-NEXT: [[VAL_6:%.*]] = toy.reshape([[VAL_5]] : tensor<2x3xf64>) to tensor<2x3xf64> 27 # CHECK-NEXT: [[VAL_8:%.*]] = toy.reshape([[VAL_7]] : tensor<6xf64>) to tensor<2x3xf64>
|
| /llvm-project-15.0.7/mlir/test/Examples/Toy/Ch3/ |
| H A D | codegen.toy | 25 # CHECK-NEXT: [[VAL_6:%.*]] = toy.reshape([[VAL_5]] : tensor<2x3xf64>) to tensor<2x3xf64> 27 # CHECK-NEXT: [[VAL_8:%.*]] = toy.reshape([[VAL_7]] : tensor<6xf64>) to tensor<2x3xf64>
|
| /llvm-project-15.0.7/mlir/test/Examples/Toy/Ch2/ |
| H A D | codegen.toy | 25 # CHECK-NEXT: [[VAL_6:%.*]] = toy.reshape([[VAL_5]] : tensor<2x3xf64>) to tensor<2x3xf64> 27 # CHECK-NEXT: [[VAL_8:%.*]] = toy.reshape([[VAL_7]] : tensor<6xf64>) to tensor<2x3xf64>
|
| /llvm-project-15.0.7/mlir/examples/toy/Ch3/include/toy/ |
| H A D | Ops.td | 128 %1 = toy.reshape(%0 : tensor<f64>) to tensor<2x2xf64> 243 def ReshapeOp : Toy_Op<"reshape", [NoSideEffect]> { 244 let summary = "tensor reshape operation"; 250 %0 = toy.reshape (%arg1 : tensor<10xf64>) to tensor<5x2xf64> 256 // We expect that the reshape operation returns a statically shaped tensor.
|
| /llvm-project-15.0.7/mlir/examples/toy/Ch2/include/toy/ |
| H A D | Ops.td | 129 %1 = toy.reshape(%0 : tensor<f64>) to tensor<2x2xf64> 244 def ReshapeOp : Toy_Op<"reshape"> { 245 let summary = "tensor reshape operation"; 251 %0 = toy.reshape (%arg1 : tensor<10xf64>) to tensor<5x2xf64> 257 // We expect that the reshape operation returns a statically shaped tensor.
|
| /llvm-project-15.0.7/mlir/test/Dialect/Linalg/ |
| H A D | fusion-push-reshape.mlir | 1 // RUN: mlir-opt %s -test-linalg-elementwise-fusion-patterns=fuse-with-reshape-by-collapsing -split… 6 // CHECK-LABEL: func @reshape 14 func.func @reshape(%A: tensor<?x16xf32>, %B: tensor<16xf32>, %init: tensor<?x112x16xf32>) -> tensor…
|
| /llvm-project-15.0.7/mlir/examples/toy/Ch4/include/toy/ |
| H A D | Ops.td | 158 %1 = toy.reshape(%0 : tensor<f64>) to tensor<2x2xf64> 275 def ReshapeOp : Toy_Op<"reshape", [NoSideEffect]> { 276 let summary = "tensor reshape operation"; 282 %0 = toy.reshape (%arg1 : tensor<10xf64>) to tensor<5x2xf64> 288 // We expect that the reshape operation returns a statically shaped tensor.
|
| /llvm-project-15.0.7/mlir/examples/toy/Ch6/include/toy/ |
| H A D | Ops.td | 158 %1 = toy.reshape(%0 : tensor<f64>) to tensor<2x2xf64> 276 def ReshapeOp : Toy_Op<"reshape", [NoSideEffect]> { 277 let summary = "tensor reshape operation"; 283 %0 = toy.reshape (%arg1 : tensor<10xf64>) to tensor<5x2xf64> 296 // We expect that the reshape operation returns a statically shaped tensor.
|
| /llvm-project-15.0.7/mlir/examples/toy/Ch5/include/toy/ |
| H A D | Ops.td | 158 %1 = toy.reshape(%0 : tensor<f64>) to tensor<2x2xf64> 276 def ReshapeOp : Toy_Op<"reshape", [NoSideEffect]> { 277 let summary = "tensor reshape operation"; 283 %0 = toy.reshape (%arg1 : tensor<10xf64>) to tensor<5x2xf64> 289 // We expect that the reshape operation returns a statically shaped tensor.
|