xref: /llvm-project/llvm/test/Transforms/SLPVectorizer/X86/jumbled-load-multiuse.ll (revision 40105a993399699fe351789c7eb2a0e6d36f440a)
1; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
2; RUN: opt < %s -passes=slp-vectorizer -S -mtriple=x86_64-unknown-linux -mattr=+sse4.2 | FileCheck %s
3
4@a = common local_unnamed_addr global [4 x i32] zeroinitializer, align 4
5@b = common local_unnamed_addr global [4 x i32] zeroinitializer, align 4
6
7define i32 @fn1() {
8; CHECK-LABEL: @fn1(
9; CHECK-NEXT:  entry:
10; CHECK-NEXT:    [[TMP0:%.*]] = load <4 x i32>, ptr @b, align 4
11; CHECK-NEXT:    [[TMP1:%.*]] = icmp sgt <4 x i32> [[TMP0]], zeroinitializer
12; CHECK-NEXT:    [[TMP2:%.*]] = shufflevector <4 x i32> [[TMP0]], <4 x i32> <i32 8, i32 poison, i32 ptrtoint (ptr @fn1 to i32), i32 poison>, <4 x i32> <i32 4, i32 1, i32 6, i32 6>
13; CHECK-NEXT:    [[TMP3:%.*]] = select <4 x i1> [[TMP1]], <4 x i32> [[TMP2]], <4 x i32> <i32 0, i32 6, i32 0, i32 0>
14; CHECK-NEXT:    [[TMP4:%.*]] = shufflevector <4 x i32> [[TMP3]], <4 x i32> poison, <4 x i32> <i32 1, i32 2, i32 3, i32 0>
15; CHECK-NEXT:    store <4 x i32> [[TMP4]], ptr @a, align 4
16; CHECK-NEXT:    ret i32 0
17;
18  entry:
19  %0 = load i32, ptr @b, align 4
20  %cmp = icmp sgt i32 %0, 0
21  %cond = select i1 %cmp, i32 8, i32 0
22  store i32 %cond, ptr getelementptr inbounds ([4 x i32], ptr @a, i64 0, i32 3), align 4
23  %1 = load i32, ptr getelementptr ([4 x i32], ptr @b, i64 0, i32 1), align 4
24  %cmp1 = icmp sgt i32 %1, 0
25  %. = select i1 %cmp1, i32 %1, i32 6
26  store i32 %., ptr @a, align 4
27  %2 = load i32, ptr getelementptr ([4 x i32], ptr @b, i64 0, i32 2), align 4
28  %cmp4 = icmp sgt i32 %2, 0
29  %3 = select i1 %cmp4, i32 ptrtoint (ptr @fn1 to i32), i32 0
30  store i32 %3, ptr getelementptr inbounds ([4 x i32], ptr @a, i64 0, i32 1), align 4
31  %4 = load i32, ptr getelementptr ([4 x i32], ptr @b, i64 0, i32 3), align 4
32  %cmp6 = icmp sgt i32 %4, 0
33  %5 = select i1 %cmp6, i32 ptrtoint (ptr @fn1 to i32), i32 0
34  store i32 %5, ptr getelementptr inbounds ([4 x i32], ptr @a, i64 0, i32 2), align 4
35  ret i32 0
36}
37