1 ; RUN: opt < %s -S -passes=loop-unroll -mtriple aarch64 -mcpu=cortex-a57 -unroll-runtime-epilog=true | FileCheck %s -check-prefix=EPILOG
2 ; RUN: opt < %s -S -passes=loop-unroll -mtriple aarch64 -mcpu=cortex-a57 -unroll-runtime-epilog=false | FileCheck %s -check-prefix=PROLOG
3 ; RUN: opt < %s -S -passes=loop-unroll -mtriple aarch64 -mcpu=cortex-r82 -unroll-runtime-epilog=true | FileCheck %s -check-prefix=EPILOG
4 ; RUN: opt < %s -S -passes=loop-unroll -mtriple aarch64 -mcpu=cortex-r82 -unroll-runtime-epilog=false | FileCheck %s -check-prefix=PROLOG
6 ; Tests for unrolling loops with run-time trip counts
8 ; EPILOG: %xtraiter = and i32 %n
10 ; EPILOG: %lcmp.mod = icmp ne i32 %xtraiter, 0
11 ; EPILOG: br i1 %lcmp.mod, label %for.body.epil.preheader, label %for.end.loopexit
12 ; EPILOG: for.body.epil:
14 ; PROLOG: %xtraiter = and i32 %n
15 ; PROLOG: %lcmp.mod = icmp ne i32 %xtraiter, 0
16 ; PROLOG: br i1 %lcmp.mod, label %for.body.prol.preheader, label %for.body.prol.loopexit
17 ; PROLOG: for.body.prol:
20 define i32 @test(ptr nocapture %a, i32 %n) nounwind uwtable readonly {
22 %cmp1 = icmp eq i32 %n, 0
23 br i1 %cmp1, label %for.end, label %for.body
25 for.body: ; preds = %for.body, %entry
26 %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %entry ]
27 %sum.02 = phi i32 [ %add, %for.body ], [ 0, %entry ]
28 %arrayidx = getelementptr inbounds i32, ptr %a, i64 %indvars.iv
29 %0 = load i32, ptr %arrayidx, align 4
30 %add = add nsw i32 %0, %sum.02
31 %indvars.iv.next = add i64 %indvars.iv, 1
32 %lftr.wideiv = trunc i64 %indvars.iv.next to i32
33 %exitcond = icmp eq i32 %lftr.wideiv, %n
34 br i1 %exitcond, label %for.end, label %for.body
36 for.end: ; preds = %for.body, %entry
37 %sum.0.lcssa = phi i32 [ 0, %entry ], [ %add, %for.body ]