122 lines · plain
1// RUN: mlir-opt %s --transform-interpreter --split-input-file -canonicalize | FileCheck %s2 3// This is a simple tile-and-fuse example with a single fusion group.4 5module {6 // CHECK: func @foo7 // CHECK: scf.forall {{.*}} {8 // CHECK: linalg.fill9 // CHECK: linalg.matmul10 // CHECK: linalg.generic11 // CHECK: }12 func.func @foo(%A: tensor<?x?xf32>, %B: tensor<?x?xf32>, %C: tensor<?xf32>,13 %D: tensor<?x?xf32>, %sz0: index, %sz1: index)14 -> tensor<?x?xf32>15 {16 %cst = arith.constant 0.000000e+00 : f3217 %5 = linalg.fill18 {__producer__}19 ins(%cst : f32)20 outs(%D : tensor<?x?xf32>) -> tensor<?x?xf32>21 %6 = linalg.matmul22 {__producer__}23 ins(%A, %B : tensor<?x?xf32>, tensor<?x?xf32>)24 outs(%5 : tensor<?x?xf32>) -> tensor<?x?xf32>25 %7 = linalg.generic26 {__root__,27 indexing_maps = [affine_map<(d0, d1) -> (d0)>,28 affine_map<(d0, d1) -> (d0, d1)>,29 affine_map<(d0, d1) -> (d0, d1)>],30 iterator_types = ["parallel", "parallel"]31 }32 ins(%C, %6 : tensor<?xf32>, tensor<?x?xf32>)33 outs(%D : tensor<?x?xf32>) {34 ^bb0(%arg2: f32, %arg3: f32, %arg4: f32):35 %16 = arith.maximumf %arg3, %cst : f3236 %17 = arith.cmpf ogt, %arg2, %cst : f3237 %18 = arith.select %17, %cst, %16 : f3238 linalg.yield %18 : f3239 } -> tensor<?x?xf32>40 return %7 : tensor<?x?xf32>41 }42 43 module attributes {transform.with_named_sequence} {44 transform.named_sequence @__transform_main(%arg1: !transform.any_op {transform.readonly}) {45 // Find the root and all producers.46 %root = transform.structured.match attributes{"__root__"} in %arg1 : (!transform.any_op) -> !transform.any_op47 %producers = transform.structured.match attributes{"__producer__"} in %arg1 : (!transform.any_op) -> !transform.any_op48 49 // Tile the root.50 %tiled_op, %forall_op = transform.structured.tile_using_forall %root num_threads [10, 20]51 : (!transform.any_op) -> (!transform.any_op, !transform.any_op)52 53 // Fuse all producers.54 transform.structured.fuse_into_containing_op %producers into %forall_op55 : (!transform.any_op, !transform.any_op) -> (!transform.any_op, !transform.any_op)56 transform.yield57 }58 }59}60 61// -----62 63// Inverse the order of the payload ops passed to the tile_using_forall64// op. Fusion should still work.65 66module {67 // CHECK: func @foo68 // CHECK: scf.forall {{.*}} {69 // CHECK: linalg.fill70 // CHECK: linalg.matmul71 // CHECK: linalg.generic72 // CHECK: }73 func.func @foo(%A: tensor<?x?xf32>, %B: tensor<?x?xf32>, %C: tensor<?xf32>,74 %D: tensor<?x?xf32>, %sz0: index, %sz1: index)75 -> tensor<?x?xf32>76 {77 %cst = arith.constant 0.000000e+00 : f3278 %5 = linalg.fill79 {__producer__}80 ins(%cst : f32)81 outs(%D : tensor<?x?xf32>) -> tensor<?x?xf32>82 %6 = linalg.matmul83 {__producer__}84 ins(%A, %B : tensor<?x?xf32>, tensor<?x?xf32>)85 outs(%5 : tensor<?x?xf32>) -> tensor<?x?xf32>86 %7 = linalg.generic87 {__root__,88 indexing_maps = [affine_map<(d0, d1) -> (d0)>,89 affine_map<(d0, d1) -> (d0, d1)>,90 affine_map<(d0, d1) -> (d0, d1)>],91 iterator_types = ["parallel", "parallel"]92 }93 ins(%C, %6 : tensor<?xf32>, tensor<?x?xf32>)94 outs(%D : tensor<?x?xf32>) {95 ^bb0(%arg2: f32, %arg3: f32, %arg4: f32):96 %16 = arith.maximumf %arg3, %cst : f3297 %17 = arith.cmpf ogt, %arg2, %cst : f3298 %18 = arith.select %17, %cst, %16 : f3299 linalg.yield %18 : f32100 } -> tensor<?x?xf32>101 return %7 : tensor<?x?xf32>102 }103 104 module attributes {transform.with_named_sequence} {105 transform.named_sequence @__transform_main(%arg1: !transform.any_op {transform.readonly}) {106 // Find the root and all producers.107 %root = transform.structured.match attributes{"__root__"} in %arg1 : (!transform.any_op) -> !transform.any_op108 %producers = transform.structured.match attributes{"__producer__"} in %arg1 : (!transform.any_op) -> !transform.any_op109 %reversed_producers = transform.test_reverse_payload_ops %producers : (!transform.any_op) -> !transform.any_op110 111 // Tile the root.112 %tiled_op, %forall_op = transform.structured.tile_using_forall %root num_threads [10, 20]113 : (!transform.any_op) -> (!transform.any_op, !transform.any_op)114 115 // Fuse all producers.116 transform.structured.fuse_into_containing_op %reversed_producers into %forall_op117 : (!transform.any_op, !transform.any_op) -> (!transform.any_op, !transform.any_op)118 transform.yield119 }120 }121}122