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
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
|
;=========================== begin_copyright_notice ============================
;
; Copyright (C) 2023-2024 Intel Corporation
;
; SPDX-License-Identifier: MIT
;
;============================ end_copyright_notice =============================
; REQUIRES: opaque-ptr-fix, llvm-14-plus
; RUN: igc_opt --opaque-pointers -debugify --igc-gep-loop-strength-reduction -check-debugify -S < %s 2>&1 | FileCheck %s
;
; Input:
;
; struct Foo {
; int bar[100];
; int baz[100];
; };
; kernel void test(global struct Foo* p, int n) {
; for (int i = 0; i < n; ++i) {
; p->baz[i] = p->bar[i];
; }
; }
;
; Two unrelated GEP with multiple indices. Reduce them to two new
; induction variables. Output:
;
; kernel void test(global struct Foo* p, int n) {
; global int* tmp1 = p->bar;
; global int* tmp2 = p->baz;
; for (int i = 0; i < n; ++i, ++tmp1, ++tmp2) {
; *tmp2 = *tmp1;
; }
; }
; Debug-info related check
; CHECK: CheckModuleDebugify: PASS
%struct.Foo = type { [100 x i32], [100 x i32] }
define spir_kernel void @test(%struct.Foo addrspace(1)* %p, i32 %n) {
entry:
%cmp1 = icmp slt i32 0, %n
br i1 %cmp1, label %for.body.lr.ph, label %for.end
; CHECK-LABEL: for.body.lr.ph:
; CHECK: [[GEP1_PHI1:%.*]] = getelementptr %struct.Foo, %struct.Foo addrspace(1)* %p, i64 0, i32 0, i64 0
; CHECK: [[GEP2_PHI1:%.*]] = getelementptr %struct.Foo, %struct.Foo addrspace(1)* %p, i64 0, i32 1, i64 0
; CHECK: br label %for.body
for.body.lr.ph: ; preds = %entry
br label %for.body
; CHECK-LABEL: for.body:
; CHECK: [[GEP2:%.*]] = phi i32 addrspace(1)* [ [[GEP2_PHI1]], %for.body.lr.ph ], [ [[GEP2_PHI2:%.*]], %for.body ]
; CHECK: [[GEP1:%.*]] = phi i32 addrspace(1)* [ [[GEP1_PHI1]], %for.body.lr.ph ], [ [[GEP1_PHI2:%.*]], %for.body ]
; CHECK: %i.02 = phi i32 [ 0, %for.body.lr.ph ], [ %inc, %for.body ]
; CHECK-NOT: zext
; CHECK-NOT: getelementptr inbounds %struct.Foo, %struct.Foo addrspace(1)* %p
; CHECK: [[LOAD:%.*]] = load i32, i32 addrspace(1)* [[GEP1]], align 4
; CHECK-NOT: getelementptr inbounds %struct.Foo, %struct.Foo addrspace(1)* %p
; CHECK: store i32 [[LOAD]], i32 addrspace(1)* [[GEP2]], align 4
; CHECK: %inc = add nuw nsw i32 %i.02, 1
; CHECK: %cmp = icmp slt i32 %inc, %n
; CHECK: [[GEP1_PHI2]] = getelementptr i32, i32 addrspace(1)* [[GEP1]], i64 1
; CHECK: [[GEP2_PHI2]] = getelementptr i32, i32 addrspace(1)* [[GEP2]], i64 1
; CHECK: br i1 %cmp, label %for.body, label %for.cond.for.end_crit_edge
for.body: ; preds = %for.body.lr.ph, %for.body
%i.02 = phi i32 [ 0, %for.body.lr.ph ], [ %inc, %for.body ]
%idxprom = zext i32 %i.02 to i64
%arrayidx = getelementptr inbounds %struct.Foo, %struct.Foo addrspace(1)* %p, i64 0, i32 0, i64 %idxprom
%0 = load i32, i32 addrspace(1)* %arrayidx, align 4
%arrayidx2 = getelementptr inbounds %struct.Foo, %struct.Foo addrspace(1)* %p, i64 0, i32 1, i64 %idxprom
store i32 %0, i32 addrspace(1)* %arrayidx2, align 4
%inc = add nuw nsw i32 %i.02, 1
%cmp = icmp slt i32 %inc, %n
br i1 %cmp, label %for.body, label %for.cond.for.end_crit_edge
for.cond.for.end_crit_edge: ; preds = %for.body
br label %for.end
for.end: ; preds = %for.cond.for.end_crit_edge, %entry
ret void
}
!igc.functions = !{!0}
!0 = !{void (%struct.Foo addrspace(1)*, i32)* @test, !1}
!1 = !{!2}
!2 = !{!"function_type", i32 0}
|
