Fix a FIXME: if we are inserting code for a PHI argument, split the critical

edge so that the code is not always executed for both operands.  This
prevents LSR from inserting code into loops whose exit blocks contain
PHI uses of IV expressions (which are outside of loops).  On gzip, for
example, we turn this ugly code:

.LBB_test_1:    ; loopentry
        add r27, r3, r28
        lhz r27, 3(r27)
        add r26, r4, r28
        lhz r26, 3(r26)
        add r25, r30, r28    ;; Only live if exiting the loop
        add r24, r29, r28    ;; Only live if exiting the loop
        cmpw cr0, r27, r26
        bne .LBB_test_5 ; loopexit

into this:

.LBB_test_1:    ; loopentry
        or r27, r28, r28
        add r28, r3, r27
        lhz r28, 3(r28)
        add r26, r4, r27
        lhz r26, 3(r26)
        cmpw cr0, r28, r26
        beq .LBB_test_3 ; shortcirc_next.0
.LBB_test_2:    ; loopentry.loopexit_crit_edge
        add r2, r30, r27
        add r8, r29, r27
        b .LBB_test_9   ; loopexit
.LBB_test_2:    ; shortcirc_next.0
        ...
        blt .LBB_test_1


into this:

.LBB_test_1:    ; loopentry
        or r27, r28, r28
        add r28, r3, r27
        lhz r28, 3(r28)
        add r26, r4, r27
        lhz r26, 3(r26)
        cmpw cr0, r28, r26
        beq .LBB_test_3 ; shortcirc_next.0
.LBB_test_2:    ; loopentry.loopexit_crit_edge
        add r2, r30, r27
        add r8, r29, r27
        b .LBB_t_3:    ; shortcirc_next.0
.LBB_test_3:    ; shortcirc_next.0
        ...
        blt .LBB_test_1


Next step: get the block out of the loop so that the loop is all
fall-throughs again.

llvm-svn: 22766

1 files changed, 19 insertions, 6 deletions

diff --git a/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp b/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp
index dc1dc18b899..75a681d98ec 100644
--- a/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp
+++ b/llvm/lib/Transforms/Scalar/LoopStrengthReduce.cpp
@@ -26,6 +26,7 @@
 #include "llvm/Analysis/ScalarEvolutionExpander.h"
 #include "llvm/Support/CFG.h"
 #include "llvm/Support/GetElementPtrTypeIterator.h"
+#include "llvm/Transforms/Utils/BasicBlockUtils.h"
 #include "llvm/Transforms/Utils/Local.h"
 #include "llvm/Target/TargetData.h"
 #include "llvm/ADT/Statistic.h"
@@ -391,7 +392,7 @@ namespace {
     // operands of Inst to use the new expression 'NewBase', with 'Imm' added
     // to it.
     void RewriteInstructionToUseNewBase(const SCEVHandle &NewBase,
-                                        SCEVExpander &Rewriter);
+                                        SCEVExpander &Rewriter, Pass *P);
 
     // Sort by the Base field.
     bool operator<(const BasedUser &BU) const { return Base < BU.Base; }
@@ -413,12 +414,12 @@ void BasedUser::dump() const {
 // operands of Inst to use the new expression 'NewBase', with 'Imm' added
 // to it.
 void BasedUser::RewriteInstructionToUseNewBase(const SCEVHandle &NewBase,
-                                               SCEVExpander &Rewriter) {
+                                               SCEVExpander &Rewriter,
+                                               Pass *P) {
   if (!isa<PHINode>(Inst)) {
     SCEVHandle NewValSCEV = SCEVAddExpr::get(NewBase, Imm);
     Value *NewVal = Rewriter.expandCodeFor(NewValSCEV, Inst,
                                            OperandValToReplace->getType());
-    
     // Replace the use of the operand Value with the new Phi we just created.
     Inst->replaceUsesOfWith(OperandValToReplace, NewVal);
     DEBUG(std::cerr << "    CHANGED: IMM =" << *Imm << "  Inst = " << *Inst);
@@ -434,7 +435,19 @@ void BasedUser::RewriteInstructionToUseNewBase(const SCEVHandle &NewBase,
   PHINode *PN = cast<PHINode>(Inst);
   for (unsigned i = 0, e = PN->getNumIncomingValues(); i != e; ++i) {
     if (PN->getIncomingValue(i) == OperandValToReplace) {
-      // FIXME: this should split any critical edges.
+      // If this is a critical edge, split the edge so that we do not insert the
+      // code on all predecessor/successor paths.
+      if (e != 1 &&
+          PN->getIncomingBlock(i)->getTerminator()->getNumSuccessors() > 1) {
+        TerminatorInst *PredTI = PN->getIncomingBlock(i)->getTerminator();
+        for (unsigned Succ = 0; ; ++Succ) {
+          assert(Succ != PredTI->getNumSuccessors() &&"Didn't find successor?");
+          if (PredTI->getSuccessor(Succ) == PN->getParent()) {
+            SplitCriticalEdge(PredTI, Succ, P);
+            break;
+          }
+        }
+      }
 
       Value *&Code = InsertedCode[PN->getIncomingBlock(i)];
       if (!Code) {
@@ -623,7 +636,7 @@ RemoveCommonExpressionsFromUseBases(std::vector<BasedUser> &Uses) {
       for (unsigned j = 0, e = AE->getNumOperands(); j != e; ++j)
         if (!SubExpressionUseCounts.count(AE->getOperand(j)))
           NewOps.push_back(AE->getOperand(j));
-      if (NewOps.size() == 0)
+      if (NewOps.empty())
         Uses[i].Base = Zero;
       else
         Uses[i].Base = SCEVAddExpr::get(NewOps);
@@ -783,7 +796,7 @@ void LoopStrengthReduce::StrengthReduceStridedIVUsers(const SCEVHandle &Stride,
         // Add BaseV to the PHI value if needed.
         RewriteExpr = SCEVAddExpr::get(RewriteExpr, SCEVUnknown::get(BaseV));
       
-      User.RewriteInstructionToUseNewBase(RewriteExpr, Rewriter);
+      User.RewriteInstructionToUseNewBase(RewriteExpr, Rewriter, this);
 
       // Mark old value we replaced as possibly dead, so that it is elminated
       // if we just replaced the last use of that value.