[RuntimeUnrolling] Populate the VMap entry correctly when default generated through...

During runtime unrolling on loops with multiple exits, we update the
exit blocks with the correct phi values from both original and remainder
loop.
In this process, we lookup the VMap for the mapped incoming phi values,
but did not update the VMap if a default entry was generated in the VMap
during the lookup. This default value is generated when constants or
values outside the current loop are looked up.
This patch fixes the assertion failure when null entries are present in
the VMap because of this lookup. Added a testcase that showcases the
problem.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@313358 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Transforms/Utils/LoopUnrollRuntime.cpp b/lib/Transforms/Utils/LoopUnrollRuntime.cpp
index 01c9362e4a4413e765a6e1b2b9a948df75e54fb8..c03fa153678a3f33d2bc691b30af96afe1e29942 100644 (file)
--- a/lib/Transforms/Utils/LoopUnrollRuntime.cpp
+++ b/lib/Transforms/Utils/LoopUnrollRuntime.cpp
@@ -782,11 +782,15 @@ bool llvm::UnrollRuntimeLoopRemainder(Loop *L, unsigned Count,
      // Add the incoming values from the remainder code to the end of the phi
      // node.
      for (unsigned i =0; i < oldNumOperands; i++){
-       Value *newVal = VMap[Phi->getIncomingValue(i)];
+       Value *newVal = VMap.lookup(Phi->getIncomingValue(i));
        // newVal can be a constant or derived from values outside the loop, and
-       // hence need not have a VMap value.
-       if (!newVal)
+       // hence need not have a VMap value. Also, since lookup already generated
+       // a default "null" VMap entry for this value, we need to populate that
+       // VMap entry correctly, with the mapped entry being itself.
+       if (!newVal) {
          newVal = Phi->getIncomingValue(i);
+         VMap[Phi->getIncomingValue(i)] = Phi->getIncomingValue(i);
+       }
        Phi->addIncoming(newVal,
                            cast<BasicBlock>(VMap[Phi->getIncomingBlock(i)]));
      }

diff --git a/test/Transforms/LoopUnroll/runtime-loop-multiple-exits.ll b/test/Transforms/LoopUnroll/runtime-loop-multiple-exits.ll
index 73672e14f78ac16d641527694824ab786690ebe5..ef983dffcd2b11e2386f757122909794eb758664 100644 (file)
--- a/test/Transforms/LoopUnroll/runtime-loop-multiple-exits.ll
+++ b/test/Transforms/LoopUnroll/runtime-loop-multiple-exits.ll
@@ -4,6 +4,8 @@
 ; 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
 
+; REQUIRES: asserts
+
 ; the third and fifth RUNs generate an epilog/prolog remainder block for all the test
 ; cases below (it does not generate a loop).
 
@@ -478,3 +480,45 @@ latch:                                              ; preds = %innerH
 exit:                                              ; preds = %latch
   ret void
 }
+
+declare i8 addrspace(1)* @foo(i32)
+; inner loop prolog unrolled
+; a value from outer loop is used in exit block of inner loop.
+; Don't create VMap entries for such values (%trip).
+define i8 addrspace(1)* @test9(i8* nocapture readonly %arg, i32 %n) {
+; PROLOG: test9(
+; PROLOG: header.prol:
+; PROLOG-NEXT: %phi.prol = phi i64 [ 0, %header.prol.preheader ], [ %iv.next.prol, %latch.prol ]
+; PROLOG: latch.prol:
+; PROLOG-NOT: trip
+; PROLOG:     br i1 %prol.iter.cmp, label %header.prol.loopexit.unr-lcssa, label %header.prol
+bb:
+  br label %outerloopHdr
+
+outerloopHdr:                                              ; preds = %outerLatch, %bb
+  %trip = add i32 %n, -1
+  %outercnd = icmp slt i32 0, %trip
+  br i1 %outercnd, label %preheader, label %outerLatch
+
+preheader:                                              ; preds = %outerloopHdr
+  %tmp4 = zext i32 0 to i64
+  br label %header
+
+header:                                              ; preds = %latch, %preheader
+  %phi = phi i64 [ %tmp4, %preheader ], [ %iv.next, %latch ]
+  %tmp7 = trunc i64 %phi to i32
+  br i1 true, label %latch, label %innerexit
+
+innerexit:                                              ; preds = %header
+  %tmp9 = call i8 addrspace(1)* @foo(i32 %trip)
+  ret i8 addrspace(1)* %tmp9
+
+latch:                                             ; preds = %header
+  %tmp11 = add nsw i32 %tmp7, 1
+  %innercnd = icmp slt i32 %tmp11, %trip
+  %iv.next = add nuw nsw i64 %phi, 1
+  br i1 %innercnd, label %header, label %outerLatch
+
+outerLatch:                                             ; preds = %latch, %outerloopHdr
+  br label %outerloopHdr
+}