1; RUN: opt < %s -loop-vectorize -force-vector-interleave=1 -force-vector-width=4 -dce -instcombine -S | FileCheck %s 2 3target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128-n8:16:32:64-S128" 4 5; Make sure we vectorize this loop: 6; int foo(float *a, float *b, int n) { 7; for (int i=0; i<n; ++i) 8; a[i] = b[i] * 3; 9; } 10 11;CHECK-LABEL: define i32 @foo 12;CHECK: for.body.preheader: 13;CHECK: br i1 %cmp.zero, label %scalar.ph, label %vector.memcheck, !dbg [[BODY_LOC:![0-9]+]] 14;CHECK: vector.memcheck: 15;CHECK: br i1 %memcheck.conflict, label %scalar.ph, label %vector.ph, !dbg [[BODY_LOC]] 16;CHECK: load <4 x float> 17define i32 @foo(float* nocapture %a, float* nocapture %b, i32 %n) nounwind uwtable ssp { 18entry: 19 %cmp6 = icmp sgt i32 %n, 0, !dbg !6 20 br i1 %cmp6, label %for.body, label %for.end, !dbg !6 21 22for.body: ; preds = %entry, %for.body 23 %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %entry ], !dbg !7 24 %arrayidx = getelementptr inbounds float, float* %b, i64 %indvars.iv, !dbg !7 25 %0 = load float, float* %arrayidx, align 4, !dbg !7 26 %mul = fmul float %0, 3.000000e+00, !dbg !7 27 %arrayidx2 = getelementptr inbounds float, float* %a, i64 %indvars.iv, !dbg !7 28 store float %mul, float* %arrayidx2, align 4, !dbg !7 29 %indvars.iv.next = add i64 %indvars.iv, 1, !dbg !7 30 %lftr.wideiv = trunc i64 %indvars.iv.next to i32, !dbg !7 31 %exitcond = icmp eq i32 %lftr.wideiv, %n, !dbg !7 32 br i1 %exitcond, label %for.end, label %for.body, !dbg !7 33 34for.end: ; preds = %for.body, %entry 35 ret i32 undef, !dbg !8 36} 37 38; Make sure that we try to vectorize loops with a runtime check if the 39; dependency check fails. 40 41; CHECK-LABEL: test_runtime_check 42; CHECK: <4 x float> 43define void @test_runtime_check(float* %a, float %b, i64 %offset, i64 %offset2, i64 %n) { 44entry: 45 br label %for.body 46 47for.body: 48 %iv = phi i64 [ 0, %entry ], [ %iv.next, %for.body ] 49 %ind.sum = add i64 %iv, %offset 50 %arr.idx = getelementptr inbounds float, float* %a, i64 %ind.sum 51 %l1 = load float, float* %arr.idx, align 4 52 %ind.sum2 = add i64 %iv, %offset2 53 %arr.idx2 = getelementptr inbounds float, float* %a, i64 %ind.sum2 54 %l2 = load float, float* %arr.idx2, align 4 55 %m = fmul fast float %b, %l2 56 %ad = fadd fast float %l1, %m 57 store float %ad, float* %arr.idx, align 4 58 %iv.next = add nuw nsw i64 %iv, 1 59 %exitcond = icmp eq i64 %iv.next, %n 60 br i1 %exitcond, label %loopexit, label %for.body 61 62loopexit: 63 ret void 64} 65 66; CHECK: [[BODY_LOC]] = !DILocation(line: 101, column: 1, scope: !{{.*}}) 67 68!llvm.module.flags = !{!0, !1} 69!llvm.dbg.cu = !{!9} 70!0 = !{i32 2, !"Dwarf Version", i32 4} 71!1 = !{i32 2, !"Debug Info Version", i32 3} 72 73!2 = !{} 74!3 = !DISubroutineType(types: !2) 75!4 = !DIFile(filename: "test.cpp", directory: "/tmp") 76!5 = distinct !DISubprogram(name: "foo", scope: !4, file: !4, line: 99, type: !3, isLocal: false, isDefinition: true, scopeLine: 100, flags: DIFlagPrototyped, isOptimized: false, unit: !9, variables: !2) 77!6 = !DILocation(line: 100, column: 1, scope: !5) 78!7 = !DILocation(line: 101, column: 1, scope: !5) 79!8 = !DILocation(line: 102, column: 1, scope: !5) 80!9 = distinct !DICompileUnit(language: DW_LANG_C99, producer: "clang", 81 file: !10, 82 isOptimized: true, flags: "-O2", 83 splitDebugFilename: "abc.debug", emissionKind: 2) 84!10 = !DIFile(filename: "path/to/file", directory: "/path/to/dir") 85!11 = !{i32 2, !"Debug Info Version", i32 3} 86