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 enable-buffer-initialization=true 25// RUN: %{compile} | %{run} | FileCheck %s 26// 27// Do the same run, but now with vectorization. 28// REDEFINE: %{sparsifier_opts} = enable-runtime-library=false vl=4 enable-buffer-initialization=true 29// RUN: %{compile} | %{run} | FileCheck %s 30// 31// Do the same run, but now with 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 36#trait_op = { 37 indexing_maps = [ 38 affine_map<(i) -> (i)>, // a 39 affine_map<(i) -> (i)> // x (out) 40 ], 41 iterator_types = ["parallel"], 42 doc = "x(i) = OP a(i)" 43} 44 45module { 46 // Performs sign operation (using semi-ring unary op) 47 // with semantics that 48 // > 0 : +1.0 49 // < 0 : -1.0 50 // +Inf: +1.0 51 // -Inf: -1.0 52 // +NaN: +NaN 53 // -NaN: -NaN 54 // +0.0: +0.0 55 // -0.0: -0.0 56 func.func @sparse_sign(%arg0: tensor<?xf64, #SparseVector>) 57 -> tensor<?xf64, #SparseVector> { 58 %c0 = arith.constant 0 : index 59 %d = tensor.dim %arg0, %c0 : tensor<?xf64, #SparseVector> 60 %xin = tensor.empty(%d) : tensor<?xf64, #SparseVector> 61 %0 = linalg.generic #trait_op 62 ins(%arg0: tensor<?xf64, #SparseVector>) 63 outs(%xin: tensor<?xf64, #SparseVector>) { 64 ^bb0(%a: f64, %x: f64) : 65 %result = sparse_tensor.unary %a : f64 to f64 66 present={ 67 ^bb1(%s: f64): 68 %z = arith.constant 0.0 : f64 69 %1 = arith.cmpf one, %s, %z : f64 70 %2 = arith.uitofp %1 : i1 to f64 71 %3 = math.copysign %2, %s : f64 72 %4 = arith.cmpf uno, %s, %s : f64 73 %5 = arith.select %4, %s, %3 : f64 74 sparse_tensor.yield %5 : f64 75 } 76 absent={} 77 linalg.yield %result : f64 78 } -> tensor<?xf64, #SparseVector> 79 return %0 : tensor<?xf64, #SparseVector> 80 } 81 82 // Driver method to call and verify sign kernel. 83 func.func @main() { 84 %c0 = arith.constant 0 : index 85 %du = arith.constant 0.0 : f64 86 87 %pnan = arith.constant 0x7FF0000001000000 : f64 88 %nnan = arith.constant 0xFFF0000001000000 : f64 89 %pinf = arith.constant 0x7FF0000000000000 : f64 90 %ninf = arith.constant 0xFFF0000000000000 : f64 91 92 // Setup sparse vector. 93 %v1 = arith.constant sparse< 94 [ [0], [3], [5], [11], [13], [17], [18], [20], [21], [28], [29], [31] ], 95 [ -1.5, 1.5, -10.2, 11.3, 1.0, -1.0, 96 0x7FF0000001000000, // +NaN 97 0xFFF0000001000000, // -NaN 98 0x7FF0000000000000, // +Inf 99 0xFFF0000000000000, // -Inf 100 -0.0, // -Zero 101 0.0 // +Zero 102 ] 103 > : tensor<32xf64> 104 %sv1 = sparse_tensor.convert %v1 105 : tensor<32xf64> to tensor<?xf64, #SparseVector> 106 107 // Call sign kernel. 108 %0 = call @sparse_sign(%sv1) : (tensor<?xf64, #SparseVector>) 109 -> tensor<?xf64, #SparseVector> 110 111 // 112 // Verify the results. 113 // 114 // CHECK: ---- Sparse Tensor ---- 115 // CHECK-NEXT: nse = 12 116 // CHECK-NEXT: dim = ( 32 ) 117 // CHECK-NEXT: lvl = ( 32 ) 118 // CHECK-NEXT: pos[0] : ( 0, 12 ) 119 // CHECK-NEXT: crd[0] : ( 0, 3, 5, 11, 13, 17, 18, 20, 21, 28, 29, 31 ) 120 // CHECK-NEXT: values : ( -1, 1, -1, 1, 1, -1, nan, -nan, 1, -1, -0, 0 ) 121 // CHECK-NEXT: ---- 122 // 123 sparse_tensor.print %0 : tensor<?xf64, #SparseVector> 124 125 // Release the resources. 126 bufferization.dealloc_tensor %sv1 : tensor<?xf64, #SparseVector> 127 bufferization.dealloc_tensor %0 : tensor<?xf64, #SparseVector> 128 return 129 } 130} 131 132 133