File: uniform_divergent.mlir

package info (click to toggle)
swiftlang 6.0.3-2
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid, trixie
  • size: 2,519,992 kB
  • sloc: cpp: 9,107,863; ansic: 2,040,022; asm: 1,135,751; python: 296,500; objc: 82,456; f90: 60,502; lisp: 34,951; pascal: 19,946; sh: 18,133; perl: 7,482; ml: 4,937; javascript: 4,117; makefile: 3,840; awk: 3,535; xml: 914; fortran: 619; cs: 573; ruby: 573
file content (60 lines) | stat: -rw-r--r-- 2,055 bytes parent folder | download | duplicates (15)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
// RUN: mlir-opt %s -affine-super-vectorize="virtual-vector-size=128" -split-input-file | FileCheck %s

// Specific tests to check vectorization of uniform/divergent values.

// CHECK-LABEL: @uniform_arg
// CHECK-SAME:  %[[in:.*]]: memref<512xf32>,
// CHECK-SAME:  %[[uniform:.*]]: f32
func.func @uniform_arg(%in : memref<512xf32>, %uniform : f32) {
  affine.for %i = 0 to 512 {
    %ld = affine.load %in[%i] : memref<512xf32>
    %add = arith.addf %ld, %uniform : f32
  }
  return
}

// CHECK-NEXT: %[[bcast:.*]] = vector.broadcast %[[uniform]] : f32 to vector<128xf32>
// CHECK-NEXT: affine.for
// CHECK:        arith.addf %{{.*}}, %[[bcast]] : vector<128xf32>

// -----

// CHECK-LABEL: @multi_use_uniform_arg
// CHECK-SAME:  %[[in:.*]]: memref<512xf32>
// CHECK-SAME:  %[[uniform:.*]]: f32
func.func @multi_use_uniform_arg(%in : memref<512xf32>, %uniform : f32) {
  affine.for %i = 0 to 512 {
    %ld = affine.load %in[%i] : memref<512xf32>
    %user0 = arith.addf %ld, %uniform : f32
    %user1 = arith.addf %ld, %uniform : f32
  }
  return
}

// CHECK-NEXT: %[[bcast:.*]] = vector.broadcast %[[uniform]] : f32 to vector<128xf32>
// CHECK-NOT:  vector.broadcast
// CHECK-NEXT: affine.for
// CHECK:        arith.addf %{{.*}}, %[[bcast]] : vector<128xf32>
// CHECK:        arith.addf %{{.*}}, %[[bcast]] : vector<128xf32>

// -----

// CHECK-LABEL: @uniform_load
func.func @uniform_load(%A : memref<?x?xf32>, %C : memref<?x?xf32>) {
  %c0 = arith.constant 0 : index
  %N = memref.dim %A, %c0 : memref<?x?xf32>
  affine.for %i = 0 to %N {
    %uniform_ld = affine.load %A[%i, %i] : memref<?x?xf32>
    affine.for %j = 0 to %N {
      %b = affine.load %A[%i, %j] : memref<?x?xf32>
      %c = arith.addf %uniform_ld, %b : f32
    }
  }
  return
}

// CHECK:      affine.for
// CHECK-NEXT:   %[[uniform_ld:.*]] = affine.load %{{.*}}[%{{.*}}, %{{.*}}] : memref<?x?xf32>
// CHECK-NEXT:   %[[bcast:.*]] = vector.broadcast %[[uniform_ld]] : f32 to vector<128xf32>
// CHECK-NEXT:   affine.for
// CHECK:          arith.addf %[[bcast]], %{{.*}} : vector<128xf32>