Adam Nemet | 67c8929 | 2016-07-20 04:03:43 +0000 | [diff] [blame] | 1 | ; RUN: opt -S -loop-vectorize -pass-remarks-missed=loop-vectorize \ |
| 2 | ; RUN: -pass-remarks-with-hotness < %s 2>&1 | \ |
| 3 | ; RUN: FileCheck -check-prefix=HOTNESS -check-prefix=BOTH %s |
| 4 | |
| 5 | ; RUN: opt -S -loop-vectorize -pass-remarks-missed=loop-vectorize < %s 2>&1 | \ |
| 6 | ; RUN: FileCheck -check-prefix=NO_HOTNESS -check-prefix=BOTH %s |
| 7 | |
| 8 | |
| 9 | ; RUN: opt -S -passes=loop-vectorize -pass-remarks-missed=loop-vectorize \ |
| 10 | ; RUN: -pass-remarks-with-hotness < %s 2>&1 | \ |
| 11 | ; RUN: FileCheck -check-prefix=HOTNESS -check-prefix=BOTH %s |
| 12 | |
Florian Hahn | 2e03213 | 2016-12-19 17:13:37 +0000 | [diff] [blame] | 13 | ; RUN: opt -S -passes=loop-vectorize \ |
| 14 | ; RUN: -pass-remarks-missed=loop-vectorize < %s 2>&1 | \ |
Adam Nemet | 67c8929 | 2016-07-20 04:03:43 +0000 | [diff] [blame] | 15 | ; RUN: FileCheck -check-prefix=NO_HOTNESS -check-prefix=BOTH %s |
| 16 | |
| 17 | |
| 18 | ; 1 void cold(char *A, char *B, char *C, char *D, char *E, int N) { |
| 19 | ; 2 for(int i = 0; i < N; i++) { |
| 20 | ; 3 A[i + 1] = A[i] + B[i]; |
| 21 | ; 4 C[i] = D[i] * E[i]; |
| 22 | ; 5 } |
| 23 | ; 6 } |
| 24 | ; 7 |
| 25 | ; 8 void hot(char *A, char *B, char *C, char *D, char *E, int N) { |
| 26 | ; 9 for(int i = 0; i < N; i++) { |
| 27 | ; 10 A[i + 1] = A[i] + B[i]; |
| 28 | ; 11 C[i] = D[i] * E[i]; |
| 29 | ; 12 } |
| 30 | ; 13 } |
| 31 | ; 14 |
| 32 | ; 15 void unknown(char *A, char *B, char *C, char *D, char *E, int N) { |
| 33 | ; 16 for(int i = 0; i < N; i++) { |
| 34 | ; 17 A[i + 1] = A[i] + B[i]; |
| 35 | ; 18 C[i] = D[i] * E[i]; |
| 36 | ; 19 } |
| 37 | ; 20 } |
| 38 | |
Adam Nemet | 9bfbf8b | 2016-11-11 22:51:46 +0000 | [diff] [blame] | 39 | ; HOTNESS: remark: /tmp/s.c:2:3: loop not vectorized (hotness: 300) |
| 40 | ; NO_HOTNESS: remark: /tmp/s.c:2:3: loop not vectorized{{$}} |
| 41 | ; HOTNESS: remark: /tmp/s.c:9:3: loop not vectorized (hotness: 5000) |
| 42 | ; NO_HOTNESS: remark: /tmp/s.c:9:3: loop not vectorized{{$}} |
| 43 | ; BOTH: remark: /tmp/s.c:16:3: loop not vectorized{{$}} |
Adam Nemet | 67c8929 | 2016-07-20 04:03:43 +0000 | [diff] [blame] | 44 | |
| 45 | ; ModuleID = '/tmp/s.c' |
| 46 | source_filename = "/tmp/s.c" |
| 47 | target datalayout = "e-m:o-i64:64-f80:128-n8:16:32:64-S128" |
Adam Nemet | 67c8929 | 2016-07-20 04:03:43 +0000 | [diff] [blame] | 48 | |
| 49 | ; Function Attrs: norecurse nounwind ssp uwtable |
| 50 | define void @cold(i8* nocapture %A, i8* nocapture readonly %B, i8* nocapture %C, i8* nocapture readonly %D, i8* nocapture readonly %E, i32 %N) local_unnamed_addr #0 !dbg !7 !prof !56 { |
| 51 | entry: |
| 52 | %cmp28 = icmp sgt i32 %N, 0, !dbg !9 |
| 53 | br i1 %cmp28, label %ph, label %for.cond.cleanup, !dbg !10, !prof !58 |
| 54 | |
| 55 | ph: |
| 56 | br label %for.body |
| 57 | |
| 58 | for.body: |
| 59 | %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %ph ] |
| 60 | %arrayidx = getelementptr inbounds i8, i8* %A, i64 %indvars.iv, !dbg !12 |
| 61 | %0 = load i8, i8* %arrayidx, align 1, !dbg !12, !tbaa !13 |
| 62 | %arrayidx2 = getelementptr inbounds i8, i8* %B, i64 %indvars.iv, !dbg !16 |
| 63 | %1 = load i8, i8* %arrayidx2, align 1, !dbg !16, !tbaa !13 |
| 64 | %add = add i8 %1, %0, !dbg !17 |
| 65 | %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1, !dbg !10 |
| 66 | %arrayidx7 = getelementptr inbounds i8, i8* %A, i64 %indvars.iv.next, !dbg !18 |
| 67 | store i8 %add, i8* %arrayidx7, align 1, !dbg !19, !tbaa !13 |
| 68 | %arrayidx9 = getelementptr inbounds i8, i8* %D, i64 %indvars.iv, !dbg !20 |
| 69 | %2 = load i8, i8* %arrayidx9, align 1, !dbg !20, !tbaa !13 |
| 70 | %arrayidx12 = getelementptr inbounds i8, i8* %E, i64 %indvars.iv, !dbg !21 |
| 71 | %3 = load i8, i8* %arrayidx12, align 1, !dbg !21, !tbaa !13 |
| 72 | %mul = mul i8 %3, %2, !dbg !22 |
| 73 | %arrayidx16 = getelementptr inbounds i8, i8* %C, i64 %indvars.iv, !dbg !23 |
| 74 | store i8 %mul, i8* %arrayidx16, align 1, !dbg !24, !tbaa !13 |
| 75 | %lftr.wideiv = trunc i64 %indvars.iv.next to i32, !dbg !10 |
| 76 | %exitcond = icmp eq i32 %lftr.wideiv, %N, !dbg !10 |
| 77 | br i1 %exitcond, label %for.cond.cleanup, label %for.body, !dbg !10, !llvm.loop !25, !prof !59 |
| 78 | |
| 79 | for.cond.cleanup: |
| 80 | ret void, !dbg !11 |
| 81 | } |
| 82 | |
| 83 | ; Function Attrs: norecurse nounwind ssp uwtable |
| 84 | define void @hot(i8* nocapture %A, i8* nocapture readonly %B, i8* nocapture %C, i8* nocapture readonly %D, i8* nocapture readonly %E, i32 %N) local_unnamed_addr #0 !dbg !26 !prof !57 { |
| 85 | entry: |
| 86 | %cmp28 = icmp sgt i32 %N, 0, !dbg !27 |
| 87 | br i1 %cmp28, label %ph, label %for.cond.cleanup, !dbg !28, !prof !58 |
| 88 | |
| 89 | ph: |
| 90 | br label %for.body |
| 91 | |
| 92 | for.body: |
| 93 | %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %ph ] |
| 94 | %arrayidx = getelementptr inbounds i8, i8* %A, i64 %indvars.iv, !dbg !30 |
| 95 | %0 = load i8, i8* %arrayidx, align 1, !dbg !30, !tbaa !13 |
| 96 | %arrayidx2 = getelementptr inbounds i8, i8* %B, i64 %indvars.iv, !dbg !31 |
| 97 | %1 = load i8, i8* %arrayidx2, align 1, !dbg !31, !tbaa !13 |
| 98 | %add = add i8 %1, %0, !dbg !32 |
| 99 | %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1, !dbg !28 |
| 100 | %arrayidx7 = getelementptr inbounds i8, i8* %A, i64 %indvars.iv.next, !dbg !33 |
| 101 | store i8 %add, i8* %arrayidx7, align 1, !dbg !34, !tbaa !13 |
| 102 | %arrayidx9 = getelementptr inbounds i8, i8* %D, i64 %indvars.iv, !dbg !35 |
| 103 | %2 = load i8, i8* %arrayidx9, align 1, !dbg !35, !tbaa !13 |
| 104 | %arrayidx12 = getelementptr inbounds i8, i8* %E, i64 %indvars.iv, !dbg !36 |
| 105 | %3 = load i8, i8* %arrayidx12, align 1, !dbg !36, !tbaa !13 |
| 106 | %mul = mul i8 %3, %2, !dbg !37 |
| 107 | %arrayidx16 = getelementptr inbounds i8, i8* %C, i64 %indvars.iv, !dbg !38 |
| 108 | store i8 %mul, i8* %arrayidx16, align 1, !dbg !39, !tbaa !13 |
| 109 | %lftr.wideiv = trunc i64 %indvars.iv.next to i32, !dbg !28 |
| 110 | %exitcond = icmp eq i32 %lftr.wideiv, %N, !dbg !28 |
| 111 | br i1 %exitcond, label %for.cond.cleanup, label %for.body, !dbg !28, !llvm.loop !40, !prof !59 |
| 112 | |
| 113 | for.cond.cleanup: |
| 114 | ret void, !dbg !29 |
| 115 | } |
| 116 | |
| 117 | ; Function Attrs: norecurse nounwind ssp uwtable |
| 118 | define void @unknown(i8* nocapture %A, i8* nocapture readonly %B, i8* nocapture %C, i8* nocapture readonly %D, i8* nocapture readonly %E, i32 %N) local_unnamed_addr #0 !dbg !41 { |
| 119 | entry: |
| 120 | %cmp28 = icmp sgt i32 %N, 0, !dbg !42 |
| 121 | br i1 %cmp28, label %for.body, label %for.cond.cleanup, !dbg !43 |
| 122 | |
| 123 | for.cond.cleanup: ; preds = %for.body, %entry |
| 124 | ret void, !dbg !44 |
| 125 | |
| 126 | for.body: ; preds = %entry, %for.body |
| 127 | %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %entry ] |
| 128 | %arrayidx = getelementptr inbounds i8, i8* %A, i64 %indvars.iv, !dbg !45 |
| 129 | %0 = load i8, i8* %arrayidx, align 1, !dbg !45, !tbaa !13 |
| 130 | %arrayidx2 = getelementptr inbounds i8, i8* %B, i64 %indvars.iv, !dbg !46 |
| 131 | %1 = load i8, i8* %arrayidx2, align 1, !dbg !46, !tbaa !13 |
| 132 | %add = add i8 %1, %0, !dbg !47 |
| 133 | %indvars.iv.next = add nuw nsw i64 %indvars.iv, 1, !dbg !43 |
| 134 | %arrayidx7 = getelementptr inbounds i8, i8* %A, i64 %indvars.iv.next, !dbg !48 |
| 135 | store i8 %add, i8* %arrayidx7, align 1, !dbg !49, !tbaa !13 |
| 136 | %arrayidx9 = getelementptr inbounds i8, i8* %D, i64 %indvars.iv, !dbg !50 |
| 137 | %2 = load i8, i8* %arrayidx9, align 1, !dbg !50, !tbaa !13 |
| 138 | %arrayidx12 = getelementptr inbounds i8, i8* %E, i64 %indvars.iv, !dbg !51 |
| 139 | %3 = load i8, i8* %arrayidx12, align 1, !dbg !51, !tbaa !13 |
| 140 | %mul = mul i8 %3, %2, !dbg !52 |
| 141 | %arrayidx16 = getelementptr inbounds i8, i8* %C, i64 %indvars.iv, !dbg !53 |
| 142 | store i8 %mul, i8* %arrayidx16, align 1, !dbg !54, !tbaa !13 |
| 143 | %lftr.wideiv = trunc i64 %indvars.iv.next to i32, !dbg !43 |
| 144 | %exitcond = icmp eq i32 %lftr.wideiv, %N, !dbg !43 |
| 145 | br i1 %exitcond, label %for.cond.cleanup, label %for.body, !dbg !43, !llvm.loop !55 |
| 146 | } |
| 147 | |
| 148 | attributes #0 = { norecurse nounwind ssp uwtable "disable-tail-calls"="false" "less-precise-fpmad"="false" "no-frame-pointer-elim"="true" "no-frame-pointer-elim-non-leaf" "no-infs-fp-math"="false" "no-jump-tables"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "target-cpu"="core2" "target-features"="+cx16,+fxsr,+mmx,+sse,+sse2,+sse3,+ssse3,+x87" "unsafe-fp-math"="false" "use-soft-float"="false" } |
| 149 | |
| 150 | !llvm.dbg.cu = !{!0} |
| 151 | !llvm.module.flags = !{!3, !4, !5} |
| 152 | !llvm.ident = !{!6} |
| 153 | |
| 154 | !0 = distinct !DICompileUnit(language: DW_LANG_C99, file: !1, producer: "clang version 3.9.0 (trunk 273572) (llvm/trunk 273585)", isOptimized: true, runtimeVersion: 0, emissionKind: LineTablesOnly, enums: !2) |
| 155 | !1 = !DIFile(filename: "/tmp/s.c", directory: "/tmp") |
| 156 | !2 = !{} |
| 157 | !3 = !{i32 2, !"Dwarf Version", i32 2} |
| 158 | !4 = !{i32 2, !"Debug Info Version", i32 3} |
| 159 | !5 = !{i32 1, !"PIC Level", i32 2} |
| 160 | !6 = !{!"clang version 3.9.0 (trunk 273572) (llvm/trunk 273585)"} |
| 161 | !7 = distinct !DISubprogram(name: "cold", scope: !1, file: !1, line: 1, type: !8, isLocal: false, isDefinition: true, scopeLine: 1, flags: DIFlagPrototyped, isOptimized: true, unit: !0, variables: !2) |
| 162 | !8 = !DISubroutineType(types: !2) |
| 163 | !9 = !DILocation(line: 2, column: 20, scope: !7) |
| 164 | !10 = !DILocation(line: 2, column: 3, scope: !7) |
| 165 | !11 = !DILocation(line: 6, column: 1, scope: !7) |
| 166 | !12 = !DILocation(line: 3, column: 16, scope: !7) |
| 167 | !13 = !{!14, !14, i64 0} |
| 168 | !14 = !{!"omnipotent char", !15, i64 0} |
| 169 | !15 = !{!"Simple C/C++ TBAA"} |
| 170 | !16 = !DILocation(line: 3, column: 23, scope: !7) |
| 171 | !17 = !DILocation(line: 3, column: 21, scope: !7) |
| 172 | !18 = !DILocation(line: 3, column: 5, scope: !7) |
| 173 | !19 = !DILocation(line: 3, column: 14, scope: !7) |
| 174 | !20 = !DILocation(line: 4, column: 12, scope: !7) |
| 175 | !21 = !DILocation(line: 4, column: 19, scope: !7) |
| 176 | !22 = !DILocation(line: 4, column: 17, scope: !7) |
| 177 | !23 = !DILocation(line: 4, column: 5, scope: !7) |
| 178 | !24 = !DILocation(line: 4, column: 10, scope: !7) |
| 179 | !25 = distinct !{!25, !10} |
| 180 | !26 = distinct !DISubprogram(name: "hot", scope: !1, file: !1, line: 8, type: !8, isLocal: false, isDefinition: true, scopeLine: 8, flags: DIFlagPrototyped, isOptimized: true, unit: !0, variables: !2) |
| 181 | !27 = !DILocation(line: 9, column: 20, scope: !26) |
| 182 | !28 = !DILocation(line: 9, column: 3, scope: !26) |
| 183 | !29 = !DILocation(line: 13, column: 1, scope: !26) |
| 184 | !30 = !DILocation(line: 10, column: 16, scope: !26) |
| 185 | !31 = !DILocation(line: 10, column: 23, scope: !26) |
| 186 | !32 = !DILocation(line: 10, column: 21, scope: !26) |
| 187 | !33 = !DILocation(line: 10, column: 5, scope: !26) |
| 188 | !34 = !DILocation(line: 10, column: 14, scope: !26) |
| 189 | !35 = !DILocation(line: 11, column: 12, scope: !26) |
| 190 | !36 = !DILocation(line: 11, column: 19, scope: !26) |
| 191 | !37 = !DILocation(line: 11, column: 17, scope: !26) |
| 192 | !38 = !DILocation(line: 11, column: 5, scope: !26) |
| 193 | !39 = !DILocation(line: 11, column: 10, scope: !26) |
| 194 | !40 = distinct !{!40, !28} |
| 195 | !41 = distinct !DISubprogram(name: "unknown", scope: !1, file: !1, line: 15, type: !8, isLocal: false, isDefinition: true, scopeLine: 15, flags: DIFlagPrototyped, isOptimized: true, unit: !0, variables: !2) |
| 196 | !42 = !DILocation(line: 16, column: 20, scope: !41) |
| 197 | !43 = !DILocation(line: 16, column: 3, scope: !41) |
| 198 | !44 = !DILocation(line: 20, column: 1, scope: !41) |
| 199 | !45 = !DILocation(line: 17, column: 16, scope: !41) |
| 200 | !46 = !DILocation(line: 17, column: 23, scope: !41) |
| 201 | !47 = !DILocation(line: 17, column: 21, scope: !41) |
| 202 | !48 = !DILocation(line: 17, column: 5, scope: !41) |
| 203 | !49 = !DILocation(line: 17, column: 14, scope: !41) |
| 204 | !50 = !DILocation(line: 18, column: 12, scope: !41) |
| 205 | !51 = !DILocation(line: 18, column: 19, scope: !41) |
| 206 | !52 = !DILocation(line: 18, column: 17, scope: !41) |
| 207 | !53 = !DILocation(line: 18, column: 5, scope: !41) |
| 208 | !54 = !DILocation(line: 18, column: 10, scope: !41) |
| 209 | !55 = distinct !{!55, !43} |
| 210 | !56 = !{!"function_entry_count", i64 3} |
| 211 | !57 = !{!"function_entry_count", i64 50} |
| 212 | !58 = !{!"branch_weights", i32 99, i32 1} |
| 213 | !59 = !{!"branch_weights", i32 1, i32 99} |