1//-------------------------------------------------------------------------------------------------- 2// WHEN CREATING A NEW TEST, PLEASE JUST COPY & PASTE WITHOUT EDITS. 3// 4// Set-up that's shared across all tests in this directory. In principle, this 5// config could be moved to lit.local.cfg. However, there are downstream users that 6// do not use these LIT config files. Hence why this is kept inline. 7// 8// DEFINE: %{sparsifier_opts} = enable-runtime-library=true 9// DEFINE: %{sparsifier_opts_sve} = enable-arm-sve=true %{sparsifier_opts} 10// DEFINE: %{compile} = mlir-opt %s --sparsifier="%{sparsifier_opts}" 11// DEFINE: %{compile_sve} = mlir-opt %s --sparsifier="%{sparsifier_opts_sve}" 12// DEFINE: %{run_libs} = -shared-libs=%mlir_c_runner_utils,%mlir_runner_utils 13// DEFINE: %{run_libs_sve} = -shared-libs=%native_mlir_runner_utils,%native_mlir_c_runner_utils 14// DEFINE: %{run_opts} = -e main -entry-point-result=void 15// DEFINE: %{run} = mlir-runner %{run_opts} %{run_libs} 16// DEFINE: %{run_sve} = %mcr_aarch64_cmd --march=aarch64 --mattr="+sve" %{run_opts} %{run_libs_sve} 17// 18// DEFINE: %{env} = 19//-------------------------------------------------------------------------------------------------- 20 21// RUN: %{compile} | %{run} | FileCheck %s 22// 23// Do the same run, but now with direct IR generation. 24// REDEFINE: %{sparsifier_opts} = enable-runtime-library=false 25// RUN: %{compile} | %{run} | FileCheck %s 26// 27// Do the same run, but now with direct IR generation and vectorization. 28// REDEFINE: %{sparsifier_opts} = enable-runtime-library=false vl=2 reassociate-fp-reductions=true enable-index-optimizations=true 29// RUN: %{compile} | %{run} | FileCheck %s 30// 31// Do the same run, but now with direct IR generation and VLA vectorization. 32// RUN: %if mlir_arm_sve_tests %{ %{compile_sve} | %{run_sve} | FileCheck %s %} 33 34#SparseVector = #sparse_tensor.encoding<{ map = (d0) -> (d0 : compressed) }> 35 36module { 37 38 // 39 // Sparse kernel. 40 // 41 func.func @sparse_dot(%a: tensor<1024xf32, #SparseVector>, 42 %b: tensor<1024xf32, #SparseVector>, 43 %x: tensor<f32>) -> tensor<f32> { 44 %dot = linalg.dot ins(%a, %b: tensor<1024xf32, #SparseVector>, 45 tensor<1024xf32, #SparseVector>) 46 outs(%x: tensor<f32>) -> tensor<f32> 47 return %dot : tensor<f32> 48 } 49 50 // 51 // Main driver. 52 // 53 func.func @main() { 54 // Setup two sparse vectors. 55 %d1 = arith.constant sparse< 56 [ [0], [1], [22], [23], [1022] ], [1.0, 2.0, 3.0, 4.0, 5.0] 57 > : tensor<1024xf32> 58 %d2 = arith.constant sparse< 59 [ [22], [1022], [1023] ], [6.0, 7.0, 8.0] 60 > : tensor<1024xf32> 61 %s1 = sparse_tensor.convert %d1 : tensor<1024xf32> to tensor<1024xf32, #SparseVector> 62 %s2 = sparse_tensor.convert %d2 : tensor<1024xf32> to tensor<1024xf32, #SparseVector> 63 64 // 65 // Verify the inputs. 66 // 67 // CHECK: ---- Sparse Tensor ---- 68 // CHECK-NEXT: nse = 5 69 // CHECK-NEXT: dim = ( 1024 ) 70 // CHECK-NEXT: lvl = ( 1024 ) 71 // CHECK-NEXT: pos[0] : ( 0, 5 ) 72 // CHECK-NEXT: crd[0] : ( 0, 1, 22, 23, 1022 ) 73 // CHECK-NEXT: values : ( 1, 2, 3, 4, 5 ) 74 // CHECK-NEXT: ---- 75 // 76 // CHECK: ---- Sparse Tensor ---- 77 // CHECK-NEXT: nse = 3 78 // CHECK-NEXT: dim = ( 1024 ) 79 // CHECK-NEXT: lvl = ( 1024 ) 80 // CHECK-NEXT: pos[0] : ( 0, 3 ) 81 // CHECK-NEXT: crd[0] : ( 22, 1022, 1023 ) 82 // CHECK-NEXT: values : ( 6, 7, 8 ) 83 // CHECK-NEXT: ---- 84 // 85 sparse_tensor.print %s1 : tensor<1024xf32, #SparseVector> 86 sparse_tensor.print %s2 : tensor<1024xf32, #SparseVector> 87 88 // Call the kernel and verify the output. 89 // 90 // CHECK: 53 91 // 92 %t = tensor.empty() : tensor<f32> 93 %z = arith.constant 0.0 : f32 94 %x = tensor.insert %z into %t[] : tensor<f32> 95 %0 = call @sparse_dot(%s1, %s2, %x) : (tensor<1024xf32, #SparseVector>, 96 tensor<1024xf32, #SparseVector>, 97 tensor<f32>) -> tensor<f32> 98 %1 = tensor.extract %0[] : tensor<f32> 99 vector.print %1 : f32 100 101 // Release the resources. 102 bufferization.dealloc_tensor %0 : tensor<f32> 103 bufferization.dealloc_tensor %s1 : tensor<1024xf32, #SparseVector> 104 bufferization.dealloc_tensor %s2 : tensor<1024xf32, #SparseVector> 105 106 return 107 } 108} 109