brintos

brintos / llvm-project-archived public Read only

0
0
Text · 4.0 KiB · 2a24028 Raw
106 lines · plain
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, this5// config could be moved to lit.local.cfg. However, there are downstream users that6//  do not use these LIT config files. Hence why this is kept inline.7//8// DEFINE: %{sparsifier_opts} = enable-runtime-library=true9// 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_utils13// DEFINE: %{run_libs_sve} = -shared-libs=%native_mlir_runner_utils,%native_mlir_c_runner_utils14// DEFINE: %{run_opts} = -e main -entry-point-result=void15// 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// REDEFINE: %{env} = TENSOR0="%mlir_src_dir/test/Integration/data/test.mtx"22// RUN: %{compile} | env %{env} %{run} | FileCheck %s23//24// Do the same run, but now with direct IR generation and vectorization.25// REDEFINE: %{sparsifier_opts} = enable-runtime-library=false vl=2 reassociate-fp-reductions=true enable-index-optimizations=true26// RUN: %{compile} | env %{env} %{run} | FileCheck %s27//28// Do the same run, but now with direct IR generation and, if available, VLA29// vectorization.30// RUN: %if mlir_arm_sve_tests %{ %{compile_sve} | env %{env} %{run_sve} | FileCheck %s %}31 32!Filename = !llvm.ptr33 34#DCSR = #sparse_tensor.encoding<{35  map = (d0, d1) -> (d0 : compressed, d1 : compressed)36}>37 38#eltwise_mult = {39  indexing_maps = [40    affine_map<(i,j) -> (i,j)>  // X (out)41  ],42  iterator_types = ["parallel", "parallel"],43  doc = "X(i,j) *= X(i,j)"44}45 46//47// Integration test that lowers a kernel annotated as sparse to48// actual sparse code, initializes a matching sparse storage scheme49// from file, and runs the resulting code with the JIT compiler.50//51module {52  //53  // A kernel that multiplies a sparse matrix A with itself54  // in an element-wise fashion. In this operation, we have55  // a sparse tensor as output, but although the values of the56  // sparse tensor change, its nonzero structure remains the same.57  //58  func.func @kernel_eltwise_mult(%argx: tensor<?x?xf64, #DCSR>)59    -> tensor<?x?xf64, #DCSR> {60    %0 = linalg.generic #eltwise_mult61      outs(%argx: tensor<?x?xf64, #DCSR>) {62      ^bb(%x: f64):63        %0 = arith.mulf %x, %x : f6464        linalg.yield %0 : f6465    } -> tensor<?x?xf64, #DCSR>66    return %0 : tensor<?x?xf64, #DCSR>67  }68 69  func.func private @getTensorFilename(index) -> (!Filename)70 71  //72  // Main driver that reads matrix from file and calls the sparse kernel.73  //74  func.func @main() {75    %d0 = arith.constant 0.0 : f6476    %c0 = arith.constant 0 : index77 78    // Read the sparse matrix from file, construct sparse storage.79    %fileName = call @getTensorFilename(%c0) : (index) -> (!Filename)80    %x = sparse_tensor.new %fileName : !Filename to tensor<?x?xf64, #DCSR>81 82    // Call kernel.83    %0 = call @kernel_eltwise_mult(%x) : (tensor<?x?xf64, #DCSR>) -> tensor<?x?xf64, #DCSR>84 85    // Print the result for verification.86    //87    // CHECK:      ---- Sparse Tensor ----88    // CHECK-NEXT: nse = 989    // CHECK-NEXT: dim = ( 5, 5 )90    // CHECK-NEXT: lvl = ( 5, 5 )91    // CHECK-NEXT: pos[0] : ( 0, 5 )92    // CHECK-NEXT: crd[0] : ( 0, 1, 2, 3, 4 )93    // CHECK-NEXT: pos[1] : ( 0, 2, 4, 5, 7, 9 )94    // CHECK-NEXT: crd[1] : ( 0, 3, 1, 4, 2, 0, 3, 1, 4 )95    // CHECK-NEXT: values : ( 1, 1.96, 4, 6.25, 9, 16.81, 16, 27.04, 25 )96    // CHECK-NEXT: ----97    //98    sparse_tensor.print %0 : tensor<?x?xf64, #DCSR>99 100    // Release the resources.101    bufferization.dealloc_tensor %x : tensor<?x?xf64, #DCSR>102 103    return104  }105}106