Merge information about the number of zero, one, and sign bits of live-out registers

at phis. This enables us to eliminate a lot of pointless zexts during the DAGCombine
phase. This fixes <rdar://problem/8760114>.

llvm-svn: 126170

include/llvm/CodeGen/FunctionLoweringInfo.h

@@ -101,14 +101,30 @@ public:
 #endif
 
   struct LiveOutInfo {
-    unsigned NumSignBits;
+    unsigned NumSignBits : 31;
+    bool IsValid : 1;
     APInt KnownOne, KnownZero;
-    LiveOutInfo() : NumSignBits(0), KnownOne(1, 0), KnownZero(1, 0) {}
+    LiveOutInfo() : NumSignBits(0), IsValid(false), KnownOne(1, 0),
+                    KnownZero(1, 0) {}
   };
   
   /// LiveOutRegInfo - Information about live out vregs.
   IndexedMap<LiveOutInfo, VirtReg2IndexFunctor> LiveOutRegInfo;
 
+  /// VisitedBBs - Basic blocks that have been visited by reverse postorder.
+  DenseSet<const BasicBlock*> VisitedBBs;
+
+  /// AllPredsVisited - Tracks whether all predecessors of the current basic
+  /// block have already been visited.
+  bool AllPredsVisited;
+
+  /// PHIDestRegs - Virtual registers that are the destinations of PHIs.
+  DenseSet<unsigned> PHIDestRegs;
+
+  /// PHISrcToDestMap - Maps the virtual register defining a PHI's source to the
+  /// virtual register defining its destination.
+  DenseMap<unsigned, unsigned> PHISrcToDestMap;
+
   /// PHINodesToUpdate - A list of phi instructions whose operand list will
   /// be updated after processing the current basic block.
   /// TODO: This isn't per-function state, it's per-basic-block state. But

lib/CodeGen/SelectionDAG/FunctionLoweringInfo.cpp

@@ -127,10 +127,13 @@ void FunctionLoweringInfo::set(const Function &fn, MachineFunction &mf) {
     for (BasicBlock::const_iterator I = BB->begin(), E = BB->end(); I != E; ++I) {
       // Mark values used outside their block as exported, by allocating
       // a virtual register for them.
-      if (isUsedOutsideOfDefiningBlock(I))
+      if (!EnableFastISel && isa<PHINode>(I)) {
+        PHIDestRegs.insert(InitializeRegForValue(I));
+      } else if (isUsedOutsideOfDefiningBlock(I)) {
         if (!isa<AllocaInst>(I) ||
             !StaticAllocaMap.count(cast<AllocaInst>(I)))
           InitializeRegForValue(I);
+      }
 
       // Collect llvm.dbg.declare information. This is done now instead of
       // during the initial isel pass through the IR so that it is done
@@ -219,6 +222,9 @@ void FunctionLoweringInfo::clear() {
   CatchInfoFound.clear();
 #endif
   LiveOutRegInfo.clear();
+  VisitedBBs.clear();
+  PHIDestRegs.clear();
+  PHISrcToDestMap.clear();
   ArgDbgValues.clear();
   ByValArgFrameIndexMap.clear();
   RegFixups.clear();

lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp

@@ -644,7 +644,10 @@ SDValue RegsForValue::getCopyFromRegs(SelectionDAG &DAG,
           !RegisterVT.isInteger() || RegisterVT.isVector() ||
           !FuncInfo.LiveOutRegInfo.inBounds(Regs[Part+i]))
         continue;
-      
+
+      if (FuncInfo.PHIDestRegs.count(Regs[Part+i]) && !FuncInfo.AllPredsVisited)
+        continue;
+
       const FunctionLoweringInfo::LiveOutInfo &LOI =
         FuncInfo.LiveOutRegInfo[Regs[Part+i]];
 
@@ -6466,6 +6469,9 @@ SelectionDAGBuilder::HandlePHINodesInSuccessorBlocks(const BasicBlock *LLVMBB) {
         }
       }
 
+      if (!EnableFastISel)
+        FuncInfo.PHISrcToDestMap[Reg] = FuncInfo.ValueMap[PN];
+
       // Remember that this register needs to added to the machine PHI node as
       // the input for this MBB.
       SmallVector<EVT, 4> ValueVTs;

lib/CodeGen/SelectionDAG/SelectionDAGISel.cpp

@@ -471,6 +471,13 @@ void SelectionDAGISel::ComputeLiveOutVRegInfo() {
     if (!TargetRegisterInfo::isVirtualRegister(DestReg))
       continue;
 
+    bool IsPHI = false;
+    DenseMap<unsigned, unsigned>::const_iterator It = FuncInfo->PHISrcToDestMap.find(DestReg);
+    if (It != FuncInfo->PHISrcToDestMap.end()) {
+      IsPHI = true;
+      DestReg = It->second;
+    }
+
     // Ignore non-scalar or non-integer values.
     SDValue Src = N->getOperand(2);
     EVT SrcVT = Src.getValueType();
@@ -482,14 +489,27 @@ void SelectionDAGISel::ComputeLiveOutVRegInfo() {
     CurDAG->ComputeMaskedBits(Src, Mask, KnownZero, KnownOne);
 
     // Only install this information if it tells us something.
-    if (NumSignBits != 1 || KnownZero != 0 || KnownOne != 0) {
-      FuncInfo->LiveOutRegInfo.grow(DestReg);
+    if (!IsPHI && NumSignBits == 1 && KnownZero == 0 && KnownOne == 0)
+      continue;
+
+    FuncInfo->LiveOutRegInfo.grow(DestReg);
+    FunctionLoweringInfo::LiveOutInfo &LOI = FuncInfo->LiveOutRegInfo[DestReg];
+
+    // If this is a PHI and there is existing information, merge it with the
+    // information from this block.
+    if (IsPHI && LOI.IsValid) {
       FunctionLoweringInfo::LiveOutInfo &LOI =
         FuncInfo->LiveOutRegInfo[DestReg];
-      LOI.NumSignBits = NumSignBits;
-      LOI.KnownOne = KnownOne;
-      LOI.KnownZero = KnownZero;
+      LOI.NumSignBits = std::min(LOI.NumSignBits, NumSignBits);
+      LOI.KnownOne &= KnownOne;
+      LOI.KnownZero &= KnownZero;
+      continue;
     }
+
+    LOI.NumSignBits = NumSignBits;
+    LOI.KnownOne = KnownOne;
+    LOI.KnownZero = KnownZero;
+    LOI.IsValid = true;
   } while (!Worklist.empty());
 }
 
@@ -840,6 +860,21 @@ void SelectionDAGISel::SelectAllBasicBlocks(const Function &Fn) {
 #ifndef NDEBUG
     CheckLineNumbers(LLVMBB);
 #endif
+
+    if (EnableFastISel) {
+      FuncInfo->AllPredsVisited = false;
+    } else {
+      FuncInfo->AllPredsVisited = true;
+      for (const_pred_iterator PI = pred_begin(LLVMBB), PE = pred_end(LLVMBB);
+           PI != PE; ++PI) {
+        if (!FuncInfo->VisitedBBs.count(*PI)) {
+          FuncInfo->AllPredsVisited = false;
+          break;
+        }
+      }
+      FuncInfo->VisitedBBs.insert(LLVMBB);
+    }
+
     FuncInfo->MBB = FuncInfo->MBBMap[LLVMBB];
     FuncInfo->InsertPt = FuncInfo->MBB->getFirstNonPHI();
 

test/CodeGen/X86/phi-constants.ll

@@ -0,0 +1,35 @@
+; RUN: llc < %s -march=x86-64 | FileCheck %s
+
+%"class.std::bitset" = type { [8 x i8] }
+
+define zeroext i1 @_Z3fooPjmS_mRSt6bitsetILm32EE(i32* nocapture %a, i64 %asize, i32* nocapture %b, i64 %bsize, %"class.std::bitset"* %bits) nounwind readonly ssp noredzone {
+entry:
+  %tmp.i.i.i.i = bitcast %"class.std::bitset"* %bits to i64*
+  br label %for.cond
+
+for.cond:                                         ; preds = %for.inc, %entry
+  %0 = phi i32 [ 0, %entry ], [ %inc, %for.inc ]
+  %conv = zext i32 %0 to i64
+  %cmp = icmp eq i64 %conv, %bsize
+  br i1 %cmp, label %return, label %for.body
+
+for.body:                                         ; preds = %for.cond
+  %arrayidx = getelementptr inbounds i32* %b, i64 %conv
+  %tmp5 = load i32* %arrayidx, align 4
+  %conv6 = zext i32 %tmp5 to i64
+  %rem.i.i.i.i = and i64 %conv6, 63
+  %tmp3.i = load i64* %tmp.i.i.i.i, align 8
+  %shl.i.i = shl i64 1, %rem.i.i.i.i
+  %and.i = and i64 %shl.i.i, %tmp3.i
+  %cmp.i = icmp eq i64 %and.i, 0
+  br i1 %cmp.i, label %for.inc, label %return
+
+for.inc:                                          ; preds = %for.body
+  %inc = add i32 %0, 1
+  br label %for.cond
+
+return:                                           ; preds = %for.body, %for.cond
+; CHECK-NOT: and
+  %retval.0 = phi i1 [ true, %for.body ], [ false, %for.cond ]
+  ret i1 %retval.0
+}