From 70ffd65ca97bd7010108ad8c1369c105fb78714a Mon Sep 17 00:00:00 2001
From: Anna Thomas <anna@azul.com>
Date: Mon, 10 Jul 2017 15:29:38 +0000
Subject: [PATCH] [LoopUnrollRuntime] Remove strict assert about VMap
 requirement

When unrolling under multiple exits which is under off-by-default option,
the assert that checks for VMap entry in loop exit values is too strong.
(assert if VMap entry did not exist, the value should be a
constant). However, values derived from
constants or from values outside loop, does not have a VMap entry too.

Removed the assert and added a testcase showcasing the property for
non-constant values.

llvm-svn: 307542
---
 llvm/lib/Transforms/Utils/LoopUnrollRuntime.cpp    |  7 ++--
 .../LoopUnroll/runtime-loop-multiple-exits.ll      | 43 +++++++++++++++++++++-
 2 files changed, 45 insertions(+), 5 deletions(-)

diff --git a/llvm/lib/Transforms/Utils/LoopUnrollRuntime.cpp b/llvm/lib/Transforms/Utils/LoopUnrollRuntime.cpp
index 8733a13..59408ff 100644
--- a/llvm/lib/Transforms/Utils/LoopUnrollRuntime.cpp
+++ b/llvm/lib/Transforms/Utils/LoopUnrollRuntime.cpp
@@ -710,11 +710,10 @@ bool llvm::UnrollRuntimeLoopRemainder(Loop *L, unsigned Count,
      // node.
      for (unsigned i =0; i < oldNumOperands; i++){
        Value *newVal = VMap[Phi->getIncomingValue(i)];
-       if (!newVal) {
-         assert(isa<Constant>(Phi->getIncomingValue(i)) &&
-                "VMap should exist for all values except constants!");
+       // newVal can be a constant or derived from values outside the loop, and
+       // hence need not have a VMap value.
+       if (!newVal)
          newVal = Phi->getIncomingValue(i);
-       }
        Phi->addIncoming(newVal,
                            cast<BasicBlock>(VMap[Phi->getIncomingBlock(i)]));
      }
diff --git a/llvm/test/Transforms/LoopUnroll/runtime-loop-multiple-exits.ll b/llvm/test/Transforms/LoopUnroll/runtime-loop-multiple-exits.ll
index eaff2d6a..18548c8 100644
--- a/llvm/test/Transforms/LoopUnroll/runtime-loop-multiple-exits.ll
+++ b/llvm/test/Transforms/LoopUnroll/runtime-loop-multiple-exits.ll
@@ -1,9 +1,10 @@
+; RUN: opt < %s -loop-unroll -unroll-runtime=true -unroll-runtime-epilog=true -unroll-runtime-multi-exit=true -verify-dom-info -verify-loop-info -S | FileCheck %s -check-prefix=EPILOG-NO-IC
 ; RUN: opt < %s -loop-unroll -unroll-runtime=true -unroll-runtime-epilog=true -unroll-runtime-multi-exit=true -verify-dom-info -verify-loop-info -instcombine -S | FileCheck %s -check-prefix=EPILOG
 ; RUN: opt < %s -loop-unroll -unroll-runtime -unroll-count=2 -unroll-runtime-epilog=true -unroll-runtime-multi-exit=true -verify-dom-info -verify-loop-info -instcombine
 ; RUN: opt < %s -loop-unroll -unroll-runtime=true -unroll-runtime-epilog=false -unroll-runtime-multi-exit=true -verify-dom-info -verify-loop-info -instcombine -S | FileCheck %s -check-prefix=PROLOG
 ; RUN: opt < %s -loop-unroll -unroll-runtime -unroll-runtime-epilog=false -unroll-count=2 -unroll-runtime-multi-exit=true -verify-dom-info -verify-loop-info -instcombine
 
-; the second and fourth RUNs generate an epilog/prolog remainder block for all the test
+; the third and fifth RUNs generate an epilog/prolog remainder block for all the test
 ; cases below (it does not generate a loop).
 
 ; test with three exiting and three exit blocks.
@@ -393,3 +394,43 @@ exit_true:
 exit_false:
   ret i32 %addx
 }
+
+; test when value in exit block does not have VMap.
+define i32 @test7(i32 %arg, i32 %arg1, i32 %arg2) {
+; EPILOG-NO-IC: test7(
+; EPILOG-NO-IC: loopexit1.loopexit:
+; EPILOG-NO-IC-NEXT:  %sext3.ph = phi i32 [ %shft, %header ], [ %shft, %latch ], [ %shft, %latch.1 ], [ %shft, %latch.2 ], [ %shft, %latch.3 ], [ %shft, %latch.4 ], [ %shft, %latch.5 ], [ %shft, %latch.6 ]
+; EPILOG-NO-IC-NEXT:  br label %loopexit1
+; EPILOG-NO-IC: loopexit1.loopexit1:
+; EPILOG-NO-IC-NEXT:  %sext3.ph2 = phi i32 [ %shft, %header.epil ]
+; EPILOG-NO-IC-NEXT:  br label %loopexit1
+; EPILOG-NO-IC: loopexit1:
+; EPILOG-NO-IC-NEXT:   %sext3 = phi i32 [ %sext3.ph, %loopexit1.loopexit ], [ %sext3.ph2, %loopexit1.loopexit1 ]
+bb:
+  %tmp = icmp slt i32 undef, 2
+  %sext = sext i32 undef to i64
+  %shft = ashr exact i32 %arg, 16
+  br i1 %tmp, label %loopexit2, label %preheader
+
+preheader:                                              ; preds = %bb2
+  br label %header
+
+header:                                              ; preds = %latch, %preheader
+  %tmp6 = phi i64 [ 1, %preheader ], [ %add, %latch ]
+  br i1 false, label %loopexit1, label %latch
+
+latch:                                              ; preds = %header
+  %add = add nuw nsw i64 %tmp6, 1
+  %tmp9 = icmp slt i64 %add, %sext
+  br i1 %tmp9, label %header, label %latchexit
+
+latchexit:                                             ; preds = %latch
+  unreachable
+
+loopexit2:                                             ; preds = %bb2
+ ret i32 %shft
+
+loopexit1:                                             ; preds = %header
+  %sext3 = phi i32 [ %shft, %header ]
+  ret i32 %sext3
+}
-- 
2.7.4