1; RUN: opt < %s -S -loop-unroll -mtriple aarch64 -mcpu=cortex-a57 | FileCheck %s -check-prefix=EPILOG
2; RUN: opt < %s -S -loop-unroll -mtriple aarch64 -mcpu=cortex-a57 -unroll-runtime-epilog=false | FileCheck %s -check-prefix=PROLOG
3
4; Tests for unrolling loops with run-time trip counts
5
6; EPILOG:  %xtraiter = and i32 %n
7; EPILOG:  %lcmp.mod = icmp ne i32 %xtraiter, %n
8; EPILOG:  br i1 %lcmp.mod, label %for.body.preheader.new, label %for.end.loopexit.unr-lcssa
9
10; PROLOG:  %xtraiter = and i32 %n
11; PROLOG:  %lcmp.mod = icmp ne i32 %xtraiter, 0
12; PROLOG:  br i1 %lcmp.mod, label %for.body.prol.preheader, label %for.body.prol.loopexit
13
14; EPILOG:  for.body:
15; EPILOG:  for.body.epil:
16
17; PROLOG:  for.body.prol:
18; PROLOG:  for.body:
19
20define i32 @test(i32* nocapture %a, i32 %n) nounwind uwtable readonly {
21entry:
22  %cmp1 = icmp eq i32 %n, 0
23  br i1 %cmp1, label %for.end, label %for.body
24
25for.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, i32* %a, i64 %indvars.iv
29  %0 = load i32, i32* %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
35
36for.end:                                          ; preds = %for.body, %entry
37  %sum.0.lcssa = phi i32 [ 0, %entry ], [ %add, %for.body ]
38  ret i32 %sum.0.lcssa
39}
40
41
42