40 lines · plain
1; RUN: opt < %s -S -passes=loop-unroll -unroll-threshold=25 -unroll-partial-threshold=25 -unroll-runtime -unroll-runtime-epilog=true -unroll-count=8 | FileCheck %s -check-prefix=EPILOG2; RUN: opt < %s -S -passes=loop-unroll -unroll-threshold=25 -unroll-partial-threshold=25 -unroll-runtime -unroll-runtime-epilog=false | FileCheck %s -check-prefix=PROLOG3 4; RUN: opt < %s -S -passes='require<opt-remark-emit>,loop-unroll' -unroll-threshold=25 -unroll-partial-threshold=25 -unroll-runtime -unroll-runtime-epilog=true -unroll-count=8 | FileCheck %s -check-prefix=EPILOG5; RUN: opt < %s -S -passes='require<opt-remark-emit>,loop-unroll' -unroll-threshold=25 -unroll-partial-threshold=25 -unroll-runtime -unroll-runtime-epilog=false | FileCheck %s -check-prefix=PROLOG6 7; Choose a smaller, power-of-two, unroll count if the loop is too large.8; This test makes sure we're not unrolling 'odd' counts9 10; EPILOG: for.body:11; EPILOG: br i1 %niter.ncmp.3, label %for.end.loopexit.unr-lcssa{{.*}}, label %for.body12; EPILOG-NOT: br i1 %niter.ncmp.4, label %for.end.loopexit.unr-lcssa{{.*}}, label %for.body13; EPILOG: for.body.epil:14 15; PROLOG: for.body.prol:16; PROLOG: for.body:17; PROLOG: br i1 %exitcond.3, label %for.end.loopexit{{.*}}, label %for.body18; PROLOG-NOT: br i1 %exitcond.4, label %for.end.loopexit{{.*}}, label %for.body19 20define i32 @test(ptr nocapture %a, i32 %n) nounwind uwtable readonly {21entry:22 %cmp1 = icmp eq i32 %n, 023 br i1 %cmp1, label %for.end, label %for.body24 25for.body: ; preds = %for.body, %entry26 %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.iv29 %0 = load i32, ptr %arrayidx, align 430 %add = add nsw i32 %0, %sum.0231 %indvars.iv.next = add i64 %indvars.iv, 132 %lftr.wideiv = trunc i64 %indvars.iv.next to i3233 %exitcond = icmp eq i32 %lftr.wideiv, %n34 br i1 %exitcond, label %for.end, label %for.body35 36for.end: ; preds = %for.body, %entry37 %sum.0.lcssa = phi i32 [ 0, %entry ], [ %add, %for.body ]38 ret i32 %sum.0.lcssa39}40