1 ; RUN: opt %loadPolly -basicaa -polly-allow-differing-element-types -polly-scops -analyze < %s | FileCheck %s
2 ; RUN: opt %loadPolly -S -basicaa -polly-allow-differing-element-types -polly-codegen < %s | FileCheck --check-prefix=IR %s
5 ; CHECK-NEXT: i8 MemRef_A[*]; // Element size 1
6 ; CHECK-NEXT: i8 MemRef_B[*]; // Element size 1
9 ; CHECK-NEXT: Stmt_for_body3
10 ; CHECK-NEXT: Domain :=
11 ; CHECK-NEXT: { Stmt_for_body3[i0, i1] : 0 <= i0 <= 1023 and 0 <= i1 <= 1023 };
12 ; CHECK-NEXT: Schedule :=
13 ; CHECK-NEXT: { Stmt_for_body3[i0, i1] -> [i0, i1] };
14 ; CHECK-NEXT: MustWriteAccess := [Reduction Type: NONE] [Scalar: 0]
15 ; CHECK-NEXT: { Stmt_for_body3[i0, i1] -> MemRef_A[o0] : -16 <= o0 <= 15 };
16 ; CHECK-NEXT: ReadAccess := [Reduction Type: NONE] [Scalar: 0]
17 ; CHECK-NEXT: { Stmt_for_body3[i0, i1] -> MemRef_B[o0] : 64 <= o0 <= 95 };
19 ; IR: polly.loop_preheader:
20 ; IR: %[[r1:[a-zA-Z0-9]*]] = getelementptr i32, i32* %A, i64 -4
21 ; IR: %[[r2:[a-zA-Z0-9]*]] = bitcast i32* %scevgep to i8*
22 ; IR: %[[r3:[a-zA-Z0-9]*]] = getelementptr i64, i64* %B, i64 8
23 ; IR: %[[r4:[a-zA-Z0-9]*]] = bitcast i64* %scevgep8 to i8*
25 ; IR: polly.stmt.for.body3:
26 ; IR: call void @llvm.memmove.p0i8.p0i8.i64(i8* %[[r2]], i8* %[[r4]], i64 32, i32 4, i1 false)
30 ; void jd(int *restrict A, long *restrict B) {
31 ; for (int i = 0; i < 1024; i++)
32 ; for (int j = 0; j < 1024; j++)
33 ; memmove(A - 4, B + 8, 32);
36 target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
38 define void @jd(i32* noalias %A, i64* noalias %B) {
42 for.cond: ; preds = %for.inc5, %entry
43 %i.0 = phi i32 [ 0, %entry ], [ %inc6, %for.inc5 ]
44 %exitcond1 = icmp ne i32 %i.0, 1024
45 br i1 %exitcond1, label %for.body, label %for.end7
47 for.body: ; preds = %for.cond
50 for.cond1: ; preds = %for.inc, %for.body
51 %j.0 = phi i32 [ 0, %for.body ], [ %inc, %for.inc ]
52 %exitcond = icmp ne i32 %j.0, 1024
53 br i1 %exitcond, label %for.body3, label %for.end
55 for.body3: ; preds = %for.cond1
56 %add.ptr = getelementptr inbounds i32, i32* %A, i64 -4
57 %tmp = bitcast i32* %add.ptr to i8*
58 %add.ptr4 = getelementptr inbounds i64, i64* %B, i64 8
59 %tmp2 = bitcast i64* %add.ptr4 to i8*
60 call void @llvm.memmove.p0i8.p0i8.i64(i8* %tmp, i8* %tmp2, i64 32, i32 4, i1 false)
63 for.inc: ; preds = %for.body3
64 %inc = add nsw i32 %j.0, 1
67 for.end: ; preds = %for.cond1
70 for.inc5: ; preds = %for.end
71 %inc6 = add nsw i32 %i.0, 1
74 for.end7: ; preds = %for.cond
78 declare void @llvm.memmove.p0i8.p0i8.i64(i8* nocapture, i8* nocapture readonly, i64, i32, i1) #1