diff --git a/llvm/lib/Transforms/Utils/LoopUnrollRuntime.cpp b/llvm/lib/Transforms/Utils/LoopUnrollRuntime.cpp
--- a/llvm/lib/Transforms/Utils/LoopUnrollRuntime.cpp
+++ b/llvm/lib/Transforms/Utils/LoopUnrollRuntime.cpp
@@ -50,6 +50,10 @@
     "unroll-runtime-multi-exit", cl::init(false), cl::Hidden,
     cl::desc("Allow runtime unrolling for loops with multiple exits, when "
              "epilog is generated"));
+static cl::opt<unsigned> UnrollRuntimeCutoffPoint(
+    "unroll-runtime-cutoff-point", cl::init(0), cl::Hidden,
+    cl::desc("Number of exit/exiting blocks allowed to be runtime unrolled, "
+             "when epilog is generated"));
 
 /// Connect the unrolling prolog code to the original loop.
 /// The unrolling prolog code contains code to execute the
@@ -479,6 +483,15 @@
   // The second point is the increase in code size, but this is true
   // irrespective of multiple exits.
 
+  SmallVector<BasicBlock*, 4> ExitingBlocks;
+  L->getExitingBlocks(ExitingBlocks);
+
+  // Allow runtime unrolling for loops with at most UnrollRuntimeCutoffPoint of
+  // exit/exiting blocks.
+  if (ExitingBlocks.size() <= UnrollRuntimeCutoffPoint &&
+      OtherExits.size() < UnrollRuntimeCutoffPoint)
+    return true;
+
   // Note: Both the heuristics below are coarse grained. We are essentially
   // enabling unrolling of loops that have a single side exit other than the
   // normal LatchExit (i.e. exiting into a deoptimize block).
@@ -488,8 +501,6 @@
   // We avoid unrolling loops that have more than two exiting blocks. This
   // limits the total number of branches in the unrolled loop to be atmost
   // the unroll factor (since one of the exiting blocks is the latch block).
-  SmallVector<BasicBlock*, 4> ExitingBlocks;
-  L->getExitingBlocks(ExitingBlocks);
   if (ExitingBlocks.size() > 2)
     return false;
 
diff --git a/llvm/test/Transforms/LoopUnroll/runtime-loop-at-most-two-exits.ll b/llvm/test/Transforms/LoopUnroll/runtime-loop-at-most-two-exits.ll
new file mode 100644
--- /dev/null
+++ b/llvm/test/Transforms/LoopUnroll/runtime-loop-at-most-two-exits.ll
@@ -0,0 +1,175 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt < %s -loop-unroll -unroll-runtime=true -unroll-runtime-epilog=true -unroll-runtime-cutoff-point=2 -verify-loop-lcssa -verify-dom-info -verify-loop-info -S | FileCheck %s --check-prefix=ENABLED
+; RUN: opt < %s -loop-unroll -unroll-runtime=true -unroll-runtime-epilog=true -unroll-runtime-cutoff-point=0 -verify-loop-lcssa -verify-dom-info -verify-loop-info -S | FileCheck %s --check-prefix=DISABLED
+
+define i32 @test(i32* nocapture %a, i64 %n) {
+; ENABLED-LABEL: @test(
+; ENABLED-NEXT:  entry:
+; ENABLED-NEXT:    [[TMP0:%.*]] = add i64 [[N:%.*]], -1
+; ENABLED-NEXT:    [[XTRAITER:%.*]] = and i64 [[N]], 7
+; ENABLED-NEXT:    [[TMP1:%.*]] = icmp ult i64 [[TMP0]], 7
+; ENABLED-NEXT:    br i1 [[TMP1]], label [[FOR_END_UNR_LCSSA:%.*]], label [[ENTRY_NEW:%.*]]
+; ENABLED:       entry.new:
+; ENABLED-NEXT:    [[UNROLL_ITER:%.*]] = sub i64 [[N]], [[XTRAITER]]
+; ENABLED-NEXT:    br label [[HEADER:%.*]]
+; ENABLED:       header:
+; ENABLED-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ 0, [[ENTRY_NEW]] ], [ [[INDVARS_IV_NEXT_7:%.*]], [[FOR_BODY_7:%.*]] ]
+; ENABLED-NEXT:    [[SUM_02:%.*]] = phi i32 [ 0, [[ENTRY_NEW]] ], [ [[ADD_7:%.*]], [[FOR_BODY_7]] ]
+; ENABLED-NEXT:    [[NITER:%.*]] = phi i64 [ [[UNROLL_ITER]], [[ENTRY_NEW]] ], [ [[NITER_NSUB_7:%.*]], [[FOR_BODY_7]] ]
+; ENABLED-NEXT:    [[CMP:%.*]] = icmp eq i64 [[N]], 42
+; ENABLED-NEXT:    br i1 [[CMP]], label [[FOR_EXIT2_LOOPEXIT:%.*]], label [[FOR_BODY:%.*]]
+; ENABLED:       for.body:
+; ENABLED-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[A:%.*]], i64 [[INDVARS_IV]]
+; ENABLED-NEXT:    [[TMP2:%.*]] = load i32, i32* [[ARRAYIDX]], align 4
+; ENABLED-NEXT:    [[ADD:%.*]] = add nsw i32 [[TMP2]], [[SUM_02]]
+; ENABLED-NEXT:    [[INDVARS_IV_NEXT:%.*]] = add nuw nsw i64 [[INDVARS_IV]], 1
+; ENABLED-NEXT:    [[NITER_NSUB:%.*]] = sub i64 [[NITER]], 1
+; ENABLED-NEXT:    [[CMP_1:%.*]] = icmp eq i64 [[N]], 42
+; ENABLED-NEXT:    br i1 [[CMP_1]], label [[FOR_EXIT2_LOOPEXIT]], label [[FOR_BODY_1:%.*]]
+; ENABLED:       for.end.unr-lcssa.loopexit:
+; ENABLED-NEXT:    [[SUM_0_LCSSA_PH_PH:%.*]] = phi i32 [ [[ADD_7]], [[FOR_BODY_7]] ]
+; ENABLED-NEXT:    [[INDVARS_IV_UNR_PH:%.*]] = phi i64 [ [[INDVARS_IV_NEXT_7]], [[FOR_BODY_7]] ]
+; ENABLED-NEXT:    [[SUM_02_UNR_PH:%.*]] = phi i32 [ [[ADD_7]], [[FOR_BODY_7]] ]
+; ENABLED-NEXT:    br label [[FOR_END_UNR_LCSSA]]
+; ENABLED:       for.end.unr-lcssa:
+; ENABLED-NEXT:    [[SUM_0_LCSSA_PH:%.*]] = phi i32 [ undef, [[ENTRY:%.*]] ], [ [[SUM_0_LCSSA_PH_PH]], [[FOR_END_UNR_LCSSA_LOOPEXIT:%.*]] ]
+; ENABLED-NEXT:    [[INDVARS_IV_UNR:%.*]] = phi i64 [ 0, [[ENTRY]] ], [ [[INDVARS_IV_UNR_PH]], [[FOR_END_UNR_LCSSA_LOOPEXIT]] ]
+; ENABLED-NEXT:    [[SUM_02_UNR:%.*]] = phi i32 [ 0, [[ENTRY]] ], [ [[SUM_02_UNR_PH]], [[FOR_END_UNR_LCSSA_LOOPEXIT]] ]
+; ENABLED-NEXT:    [[LCMP_MOD:%.*]] = icmp ne i64 [[XTRAITER]], 0
+; ENABLED-NEXT:    br i1 [[LCMP_MOD]], label [[HEADER_EPIL_PREHEADER:%.*]], label [[FOR_END:%.*]]
+; ENABLED:       header.epil.preheader:
+; ENABLED-NEXT:    br label [[HEADER_EPIL:%.*]]
+; ENABLED:       header.epil:
+; ENABLED-NEXT:    [[INDVARS_IV_EPIL:%.*]] = phi i64 [ [[INDVARS_IV_NEXT_EPIL:%.*]], [[FOR_BODY_EPIL:%.*]] ], [ [[INDVARS_IV_UNR]], [[HEADER_EPIL_PREHEADER]] ]
+; ENABLED-NEXT:    [[SUM_02_EPIL:%.*]] = phi i32 [ [[ADD_EPIL:%.*]], [[FOR_BODY_EPIL]] ], [ [[SUM_02_UNR]], [[HEADER_EPIL_PREHEADER]] ]
+; ENABLED-NEXT:    [[EPIL_ITER:%.*]] = phi i64 [ [[XTRAITER]], [[HEADER_EPIL_PREHEADER]] ], [ [[EPIL_ITER_SUB:%.*]], [[FOR_BODY_EPIL]] ]
+; ENABLED-NEXT:    [[CMP_EPIL:%.*]] = icmp eq i64 [[N]], 42
+; ENABLED-NEXT:    br i1 [[CMP_EPIL]], label [[FOR_EXIT2_LOOPEXIT2:%.*]], label [[FOR_BODY_EPIL]]
+; ENABLED:       for.body.epil:
+; ENABLED-NEXT:    [[ARRAYIDX_EPIL:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV_EPIL]]
+; ENABLED-NEXT:    [[TMP3:%.*]] = load i32, i32* [[ARRAYIDX_EPIL]], align 4
+; ENABLED-NEXT:    [[ADD_EPIL]] = add nsw i32 [[TMP3]], [[SUM_02_EPIL]]
+; ENABLED-NEXT:    [[INDVARS_IV_NEXT_EPIL]] = add i64 [[INDVARS_IV_EPIL]], 1
+; ENABLED-NEXT:    [[EXITCOND_EPIL:%.*]] = icmp eq i64 [[INDVARS_IV_NEXT_EPIL]], [[N]]
+; ENABLED-NEXT:    [[EPIL_ITER_SUB]] = sub i64 [[EPIL_ITER]], 1
+; ENABLED-NEXT:    [[EPIL_ITER_CMP:%.*]] = icmp ne i64 [[EPIL_ITER_SUB]], 0
+; ENABLED-NEXT:    br i1 [[EPIL_ITER_CMP]], label [[HEADER_EPIL]], label [[FOR_END_EPILOG_LCSSA:%.*]], [[LOOP0:!llvm.loop !.*]]
+; ENABLED:       for.end.epilog-lcssa:
+; ENABLED-NEXT:    [[SUM_0_LCSSA_PH1:%.*]] = phi i32 [ [[ADD_EPIL]], [[FOR_BODY_EPIL]] ]
+; ENABLED-NEXT:    br label [[FOR_END]]
+; ENABLED:       for.end:
+; ENABLED-NEXT:    [[SUM_0_LCSSA:%.*]] = phi i32 [ [[SUM_0_LCSSA_PH]], [[FOR_END_UNR_LCSSA]] ], [ [[SUM_0_LCSSA_PH1]], [[FOR_END_EPILOG_LCSSA]] ]
+; ENABLED-NEXT:    ret i32 [[SUM_0_LCSSA]]
+; ENABLED:       for.exit2.loopexit:
+; ENABLED-NEXT:    [[RETVAL_PH:%.*]] = phi i32 [ [[SUM_02]], [[HEADER]] ], [ [[ADD]], [[FOR_BODY]] ], [ [[ADD_1:%.*]], [[FOR_BODY_1]] ], [ [[ADD_2:%.*]], [[FOR_BODY_2:%.*]] ], [ [[ADD_3:%.*]], [[FOR_BODY_3:%.*]] ], [ [[ADD_4:%.*]], [[FOR_BODY_4:%.*]] ], [ [[ADD_5:%.*]], [[FOR_BODY_5:%.*]] ], [ [[ADD_6:%.*]], [[FOR_BODY_6:%.*]] ]
+; ENABLED-NEXT:    br label [[FOR_EXIT2:%.*]]
+; ENABLED:       for.exit2.loopexit2:
+; ENABLED-NEXT:    [[RETVAL_PH3:%.*]] = phi i32 [ [[SUM_02_EPIL]], [[HEADER_EPIL]] ]
+; ENABLED-NEXT:    br label [[FOR_EXIT2]]
+; ENABLED:       for.exit2:
+; ENABLED-NEXT:    [[RETVAL:%.*]] = phi i32 [ [[RETVAL_PH]], [[FOR_EXIT2_LOOPEXIT]] ], [ [[RETVAL_PH3]], [[FOR_EXIT2_LOOPEXIT2]] ]
+; ENABLED-NEXT:    ret i32 [[RETVAL]]
+; ENABLED:       for.body.1:
+; ENABLED-NEXT:    [[ARRAYIDX_1:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV_NEXT]]
+; ENABLED-NEXT:    [[TMP4:%.*]] = load i32, i32* [[ARRAYIDX_1]], align 4
+; ENABLED-NEXT:    [[ADD_1]] = add nsw i32 [[TMP4]], [[ADD]]
+; ENABLED-NEXT:    [[INDVARS_IV_NEXT_1:%.*]] = add nuw nsw i64 [[INDVARS_IV_NEXT]], 1
+; ENABLED-NEXT:    [[NITER_NSUB_1:%.*]] = sub i64 [[NITER_NSUB]], 1
+; ENABLED-NEXT:    [[CMP_2:%.*]] = icmp eq i64 [[N]], 42
+; ENABLED-NEXT:    br i1 [[CMP_2]], label [[FOR_EXIT2_LOOPEXIT]], label [[FOR_BODY_2]]
+; ENABLED:       for.body.2:
+; ENABLED-NEXT:    [[ARRAYIDX_2:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV_NEXT_1]]
+; ENABLED-NEXT:    [[TMP5:%.*]] = load i32, i32* [[ARRAYIDX_2]], align 4
+; ENABLED-NEXT:    [[ADD_2]] = add nsw i32 [[TMP5]], [[ADD_1]]
+; ENABLED-NEXT:    [[INDVARS_IV_NEXT_2:%.*]] = add nuw nsw i64 [[INDVARS_IV_NEXT_1]], 1
+; ENABLED-NEXT:    [[NITER_NSUB_2:%.*]] = sub i64 [[NITER_NSUB_1]], 1
+; ENABLED-NEXT:    [[CMP_3:%.*]] = icmp eq i64 [[N]], 42
+; ENABLED-NEXT:    br i1 [[CMP_3]], label [[FOR_EXIT2_LOOPEXIT]], label [[FOR_BODY_3]]
+; ENABLED:       for.body.3:
+; ENABLED-NEXT:    [[ARRAYIDX_3:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV_NEXT_2]]
+; ENABLED-NEXT:    [[TMP6:%.*]] = load i32, i32* [[ARRAYIDX_3]], align 4
+; ENABLED-NEXT:    [[ADD_3]] = add nsw i32 [[TMP6]], [[ADD_2]]
+; ENABLED-NEXT:    [[INDVARS_IV_NEXT_3:%.*]] = add nuw nsw i64 [[INDVARS_IV_NEXT_2]], 1
+; ENABLED-NEXT:    [[NITER_NSUB_3:%.*]] = sub i64 [[NITER_NSUB_2]], 1
+; ENABLED-NEXT:    [[CMP_4:%.*]] = icmp eq i64 [[N]], 42
+; ENABLED-NEXT:    br i1 [[CMP_4]], label [[FOR_EXIT2_LOOPEXIT]], label [[FOR_BODY_4]]
+; ENABLED:       for.body.4:
+; ENABLED-NEXT:    [[ARRAYIDX_4:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV_NEXT_3]]
+; ENABLED-NEXT:    [[TMP7:%.*]] = load i32, i32* [[ARRAYIDX_4]], align 4
+; ENABLED-NEXT:    [[ADD_4]] = add nsw i32 [[TMP7]], [[ADD_3]]
+; ENABLED-NEXT:    [[INDVARS_IV_NEXT_4:%.*]] = add nuw nsw i64 [[INDVARS_IV_NEXT_3]], 1
+; ENABLED-NEXT:    [[NITER_NSUB_4:%.*]] = sub i64 [[NITER_NSUB_3]], 1
+; ENABLED-NEXT:    [[CMP_5:%.*]] = icmp eq i64 [[N]], 42
+; ENABLED-NEXT:    br i1 [[CMP_5]], label [[FOR_EXIT2_LOOPEXIT]], label [[FOR_BODY_5]]
+; ENABLED:       for.body.5:
+; ENABLED-NEXT:    [[ARRAYIDX_5:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV_NEXT_4]]
+; ENABLED-NEXT:    [[TMP8:%.*]] = load i32, i32* [[ARRAYIDX_5]], align 4
+; ENABLED-NEXT:    [[ADD_5]] = add nsw i32 [[TMP8]], [[ADD_4]]
+; ENABLED-NEXT:    [[INDVARS_IV_NEXT_5:%.*]] = add nuw nsw i64 [[INDVARS_IV_NEXT_4]], 1
+; ENABLED-NEXT:    [[NITER_NSUB_5:%.*]] = sub i64 [[NITER_NSUB_4]], 1
+; ENABLED-NEXT:    [[CMP_6:%.*]] = icmp eq i64 [[N]], 42
+; ENABLED-NEXT:    br i1 [[CMP_6]], label [[FOR_EXIT2_LOOPEXIT]], label [[FOR_BODY_6]]
+; ENABLED:       for.body.6:
+; ENABLED-NEXT:    [[ARRAYIDX_6:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV_NEXT_5]]
+; ENABLED-NEXT:    [[TMP9:%.*]] = load i32, i32* [[ARRAYIDX_6]], align 4
+; ENABLED-NEXT:    [[ADD_6]] = add nsw i32 [[TMP9]], [[ADD_5]]
+; ENABLED-NEXT:    [[INDVARS_IV_NEXT_6:%.*]] = add nuw nsw i64 [[INDVARS_IV_NEXT_5]], 1
+; ENABLED-NEXT:    [[NITER_NSUB_6:%.*]] = sub i64 [[NITER_NSUB_5]], 1
+; ENABLED-NEXT:    [[CMP_7:%.*]] = icmp eq i64 [[N]], 42
+; ENABLED-NEXT:    br i1 [[CMP_7]], label [[FOR_EXIT2_LOOPEXIT]], label [[FOR_BODY_7]]
+; ENABLED:       for.body.7:
+; ENABLED-NEXT:    [[ARRAYIDX_7:%.*]] = getelementptr inbounds i32, i32* [[A]], i64 [[INDVARS_IV_NEXT_6]]
+; ENABLED-NEXT:    [[TMP10:%.*]] = load i32, i32* [[ARRAYIDX_7]], align 4
+; ENABLED-NEXT:    [[ADD_7]] = add nsw i32 [[TMP10]], [[ADD_6]]
+; ENABLED-NEXT:    [[INDVARS_IV_NEXT_7]] = add i64 [[INDVARS_IV_NEXT_6]], 1
+; ENABLED-NEXT:    [[NITER_NSUB_7]] = sub i64 [[NITER_NSUB_6]], 1
+; ENABLED-NEXT:    [[NITER_NCMP_7:%.*]] = icmp eq i64 [[NITER_NSUB_7]], 0
+; ENABLED-NEXT:    br i1 [[NITER_NCMP_7]], label [[FOR_END_UNR_LCSSA_LOOPEXIT]], label [[HEADER]]
+;
+; DISABLED-LABEL: @test(
+; DISABLED-NEXT:  entry:
+; DISABLED-NEXT:    br label [[HEADER:%.*]]
+; DISABLED:       header:
+; DISABLED-NEXT:    [[INDVARS_IV:%.*]] = phi i64 [ [[INDVARS_IV_NEXT:%.*]], [[FOR_BODY:%.*]] ], [ 0, [[ENTRY:%.*]] ]
+; DISABLED-NEXT:    [[SUM_02:%.*]] = phi i32 [ [[ADD:%.*]], [[FOR_BODY]] ], [ 0, [[ENTRY]] ]
+; DISABLED-NEXT:    [[CMP:%.*]] = icmp eq i64 [[N:%.*]], 42
+; DISABLED-NEXT:    br i1 [[CMP]], label [[FOR_EXIT2:%.*]], label [[FOR_BODY]]
+; DISABLED:       for.body:
+; DISABLED-NEXT:    [[ARRAYIDX:%.*]] = getelementptr inbounds i32, i32* [[A:%.*]], i64 [[INDVARS_IV]]
+; DISABLED-NEXT:    [[TMP0:%.*]] = load i32, i32* [[ARRAYIDX]], align 4
+; DISABLED-NEXT:    [[ADD]] = add nsw i32 [[TMP0]], [[SUM_02]]
+; DISABLED-NEXT:    [[INDVARS_IV_NEXT]] = add i64 [[INDVARS_IV]], 1
+; DISABLED-NEXT:    [[EXITCOND:%.*]] = icmp eq i64 [[INDVARS_IV_NEXT]], [[N]]
+; DISABLED-NEXT:    br i1 [[EXITCOND]], label [[FOR_END:%.*]], label [[HEADER]]
+; DISABLED:       for.end:
+; DISABLED-NEXT:    [[SUM_0_LCSSA:%.*]] = phi i32 [ [[ADD]], [[FOR_BODY]] ]
+; DISABLED-NEXT:    ret i32 [[SUM_0_LCSSA]]
+; DISABLED:       for.exit2:
+; DISABLED-NEXT:    [[RETVAL:%.*]] = phi i32 [ [[SUM_02]], [[HEADER]] ]
+; DISABLED-NEXT:    ret i32 [[RETVAL]]
+;
+entry:
+  br label %header
+
+header:
+  %indvars.iv = phi i64 [ %indvars.iv.next, %for.body ], [ 0, %entry ]
+  %sum.02 = phi i32 [ %add, %for.body ], [ 0, %entry ]
+  %cmp = icmp eq i64 %n, 42
+  br i1 %cmp, label %for.exit2, label %for.body
+
+for.body:
+  %arrayidx = getelementptr inbounds i32, i32* %a, i64 %indvars.iv
+  %0 = load i32, i32* %arrayidx, align 4
+  %add = add nsw i32 %0, %sum.02
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %exitcond = icmp eq i64 %indvars.iv.next, %n
+  br i1 %exitcond, label %for.end, label %header
+
+for.end:
+  %sum.0.lcssa = phi i32 [ %add, %for.body ]
+  ret i32 %sum.0.lcssa
+
+for.exit2:
+  %retval = phi i32 [ %sum.02, %header ]
+  ret i32 %retval
+}