[llvm] r321210 - [ICP] Expose unconditional call promotion interface

Matthew Simpson via llvm-commits llvm-commits at lists.llvm.org
Wed Dec 20 11:26:37 PST 2017


Author: mssimpso
Date: Wed Dec 20 11:26:37 2017
New Revision: 321210

URL: http://llvm.org/viewvc/llvm-project?rev=321210&view=rev
Log:
[ICP] Expose unconditional call promotion interface

This patch modifies the indirect call promotion utilities by exposing and using
an unconditional call promotion interface. The unconditional promotion
interface (i.e., call promotion without creating an if-then-else) can be used
if it's known that an indirect call has only one possible callee. The existing
conditional promotion interface uses this unconditional interface to promote an
indirect call after it has been versioned and placed within the "then" block.

A consequence of unconditional promotion is that the fix-up operations for phi
nodes in the normal destination of invoke instructions are changed. This is
necessary because the existing implementation assumed that an invoke had been
versioned, creating a "merge" block where a return value bitcast could be
placed. In the new implementation, the edge between a promoted invoke's parent
block and its normal destination is split if needed to add a bitcast for the
return value. If the invoke is also versioned, the phi node merging the return
value of the promoted and original invoke instructions is placed in the "merge"
block.

Differential Revision: https://reviews.llvm.org/D40751

Modified:
    llvm/trunk/include/llvm/Transforms/Utils/CallPromotionUtils.h
    llvm/trunk/lib/Transforms/Utils/CallPromotionUtils.cpp
    llvm/trunk/test/Transforms/PGOProfile/icp_covariant_call_return.ll
    llvm/trunk/test/Transforms/PGOProfile/icp_covariant_invoke_return.ll
    llvm/trunk/test/Transforms/PGOProfile/icp_invoke.ll
    llvm/trunk/test/Transforms/PGOProfile/icp_invoke_nouse.ll
    llvm/trunk/test/Transforms/PGOProfile/icp_vararg.ll
    llvm/trunk/test/Transforms/PGOProfile/indirect_call_promotion.ll

Modified: llvm/trunk/include/llvm/Transforms/Utils/CallPromotionUtils.h
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/include/llvm/Transforms/Utils/CallPromotionUtils.h?rev=321210&r1=321209&r2=321210&view=diff
==============================================================================
--- llvm/trunk/include/llvm/Transforms/Utils/CallPromotionUtils.h (original)
+++ llvm/trunk/include/llvm/Transforms/Utils/CallPromotionUtils.h Wed Dec 20 11:26:37 2017
@@ -29,13 +29,23 @@ namespace llvm {
 bool isLegalToPromote(CallSite CS, Function *Callee,
                       const char **FailureReason = nullptr);
 
+/// Promote the given indirect call site to unconditionally call \p Callee.
+///
+/// This function promotes the given call site, returning the direct call or
+/// invoke instruction. If the function type of the call site doesn't match that
+/// of the callee, bitcast instructions are inserted where appropriate. If \p
+/// RetBitCast is non-null, it will be used to store the return value bitcast,
+/// if created.
+Instruction *promoteCall(CallSite CS, Function *Callee,
+                         CastInst **RetBitCast = nullptr);
+
 /// Promote the given indirect call site to conditionally call \p Callee.
 ///
 /// This function creates an if-then-else structure at the location of the call
-/// site. The original call site is promoted and moved into the "then" block. A
-/// clone of the indirect call site is placed in the "else" block and returned.
-/// If \p BranchWeights is non-null, it will be used to set !prof metadata on
-/// the new conditional branch.
+/// site. The original call site is moved into the "else" block. A clone of the
+/// indirect call site is promoted, placed in the "then" block, and returned. If
+/// \p BranchWeights is non-null, it will be used to set !prof metadata on the
+/// new conditional branch.
 Instruction *promoteCallWithIfThenElse(CallSite CS, Function *Callee,
                                        MDNode *BranchWeights = nullptr);
 

Modified: llvm/trunk/lib/Transforms/Utils/CallPromotionUtils.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Transforms/Utils/CallPromotionUtils.cpp?rev=321210&r1=321209&r2=321210&view=diff
==============================================================================
--- llvm/trunk/lib/Transforms/Utils/CallPromotionUtils.cpp (original)
+++ llvm/trunk/lib/Transforms/Utils/CallPromotionUtils.cpp Wed Dec 20 11:26:37 2017
@@ -23,10 +23,30 @@ using namespace llvm;
 /// Fix-up phi nodes in an invoke instruction's normal destination.
 ///
 /// After versioning an invoke instruction, values coming from the original
-/// block will now either be coming from the original block or the "else" block.
+/// block will now be coming from the "merge" block. For example, in the code
+/// below:
+///
+///   then_bb:
+///     %t0 = invoke i32 %ptr() to label %merge_bb unwind label %unwind_dst
+///
+///   else_bb:
+///     %t1 = invoke i32 %ptr() to label %merge_bb unwind label %unwind_dst
+///
+///   merge_bb:
+///     %t2 = phi i32 [ %t0, %then_bb ], [ %t1, %else_bb ]
+///     br %normal_dst
+///
+///   normal_dst:
+///     %t3 = phi i32 [ %x, %orig_bb ], ...
+///
+/// "orig_bb" is no longer a predecessor of "normal_dst", so the phi nodes in
+/// "normal_dst" must be fixed to refer to "merge_bb":
+///
+///    normal_dst:
+///      %t3 = phi i32 [ %x, %merge_bb ], ...
+///
 static void fixupPHINodeForNormalDest(InvokeInst *Invoke, BasicBlock *OrigBlock,
-                                      BasicBlock *ElseBlock,
-                                      Instruction *NewInst) {
+                                      BasicBlock *MergeBlock) {
   for (auto &I : *Invoke->getNormalDest()) {
     auto *Phi = dyn_cast<PHINode>(&I);
     if (!Phi)
@@ -34,13 +54,7 @@ static void fixupPHINodeForNormalDest(In
     int Idx = Phi->getBasicBlockIndex(OrigBlock);
     if (Idx == -1)
       continue;
-    Value *V = Phi->getIncomingValue(Idx);
-    if (dyn_cast<Instruction>(V) == Invoke) {
-      Phi->setIncomingBlock(Idx, ElseBlock);
-      Phi->addIncoming(NewInst, OrigBlock);
-      continue;
-    }
-    Phi->addIncoming(V, ElseBlock);
+    Phi->setIncomingBlock(Idx, MergeBlock);
   }
 }
 
@@ -48,6 +62,23 @@ static void fixupPHINodeForNormalDest(In
 ///
 /// After versioning an invoke instruction, values coming from the original
 /// block will now be coming from either the "then" block or the "else" block.
+/// For example, in the code below:
+///
+///   then_bb:
+///     %t0 = invoke i32 %ptr() to label %merge_bb unwind label %unwind_dst
+///
+///   else_bb:
+///     %t1 = invoke i32 %ptr() to label %merge_bb unwind label %unwind_dst
+///
+///   unwind_dst:
+///     %t3 = phi i32 [ %x, %orig_bb ], ...
+///
+/// "orig_bb" is no longer a predecessor of "unwind_dst", so the phi nodes in
+/// "unwind_dst" must be fixed to refer to "then_bb" and "else_bb":
+///
+///   unwind_dst:
+///     %t3 = phi i32 [ %x, %then_bb ], [ %x, %else_bb ], ...
+///
 static void fixupPHINodeForUnwindDest(InvokeInst *Invoke, BasicBlock *OrigBlock,
                                       BasicBlock *ThenBlock,
                                       BasicBlock *ElseBlock) {
@@ -64,44 +95,26 @@ static void fixupPHINodeForUnwindDest(In
   }
 }
 
-/// Get the phi node having the returned value of a call or invoke instruction
-/// as it's operand.
-static bool getRetPhiNode(Instruction *Inst, BasicBlock *Block) {
-  BasicBlock *FromBlock = Inst->getParent();
-  for (auto &I : *Block) {
-    PHINode *PHI = dyn_cast<PHINode>(&I);
-    if (!PHI)
-      break;
-    int Idx = PHI->getBasicBlockIndex(FromBlock);
-    if (Idx == -1)
-      continue;
-    auto *V = PHI->getIncomingValue(Idx);
-    if (V == Inst)
-      return true;
-  }
-  return false;
-}
-
 /// Create a phi node for the returned value of a call or invoke instruction.
 ///
 /// After versioning a call or invoke instruction that returns a value, we have
 /// to merge the value of the original and new instructions. We do this by
 /// creating a phi node and replacing uses of the original instruction with this
 /// phi node.
-static void createRetPHINode(Instruction *OrigInst, Instruction *NewInst) {
+///
+/// For example, if \p OrigInst is defined in "else_bb" and \p NewInst is
+/// defined in "then_bb", we create the following phi node:
+///
+///   ; Uses of the original instruction are replaced by uses of the phi node.
+///   %t0 = phi i32 [ %orig_inst, %else_bb ], [ %new_inst, %then_bb ],
+///
+static void createRetPHINode(Instruction *OrigInst, Instruction *NewInst,
+                             BasicBlock *MergeBlock, IRBuilder<> &Builder) {
 
   if (OrigInst->getType()->isVoidTy() || OrigInst->use_empty())
     return;
 
-  BasicBlock *RetValBB = NewInst->getParent();
-  if (auto *Invoke = dyn_cast<InvokeInst>(NewInst))
-    RetValBB = Invoke->getNormalDest();
-  BasicBlock *PhiBB = RetValBB->getSingleSuccessor();
-
-  if (getRetPhiNode(OrigInst, PhiBB))
-    return;
-
-  IRBuilder<> Builder(&PhiBB->front());
+  Builder.SetInsertPoint(&MergeBlock->front());
   PHINode *Phi = Builder.CreatePHI(OrigInst->getType(), 0);
   SmallVector<User *, 16> UsersToUpdate;
   for (User *U : OrigInst->users())
@@ -109,7 +122,7 @@ static void createRetPHINode(Instruction
   for (User *U : UsersToUpdate)
     U->replaceUsesOfWith(OrigInst, Phi);
   Phi->addIncoming(OrigInst, OrigInst->getParent());
-  Phi->addIncoming(NewInst, RetValBB);
+  Phi->addIncoming(NewInst, NewInst->getParent());
 }
 
 /// Cast a call or invoke instruction to the given type.
@@ -118,7 +131,41 @@ static void createRetPHINode(Instruction
 /// that of the callee. If this is the case, we have to cast the returned value
 /// to the correct type. The location of the cast depends on if we have a call
 /// or invoke instruction.
-Instruction *createRetBitCast(CallSite CS, Type *RetTy) {
+///
+/// For example, if the call instruction below requires a bitcast after
+/// promotion:
+///
+///   orig_bb:
+///     %t0 = call i32 @func()
+///     ...
+///
+/// The bitcast is placed after the call instruction:
+///
+///   orig_bb:
+///     ; Uses of the original return value are replaced by uses of the bitcast.
+///     %t0 = call i32 @func()
+///     %t1 = bitcast i32 %t0 to ...
+///     ...
+///
+/// A similar transformation is performed for invoke instructions. However,
+/// since invokes are terminating, a new block is created for the bitcast. For
+/// example, if the invoke instruction below requires a bitcast after promotion:
+///
+///   orig_bb:
+///     %t0 = invoke i32 @func() to label %normal_dst unwind label %unwind_dst
+///
+/// The edge between the original block and the invoke's normal destination is
+/// split, and the bitcast is placed there:
+///
+///   orig_bb:
+///     %t0 = invoke i32 @func() to label %split_bb unwind label %unwind_dst
+///
+///   split_bb:
+///     ; Uses of the original return value are replaced by uses of the bitcast.
+///     %t1 = bitcast i32 %t0 to ...
+///     br label %normal_dst
+///
+static void createRetBitCast(CallSite CS, Type *RetTy, CastInst **RetBitCast) {
 
   // Save the users of the calling instruction. These uses will be changed to
   // use the bitcast after we create it.
@@ -130,19 +177,20 @@ Instruction *createRetBitCast(CallSite C
   // value. The location depends on if we have a call or invoke instruction.
   Instruction *InsertBefore = nullptr;
   if (auto *Invoke = dyn_cast<InvokeInst>(CS.getInstruction()))
-    InsertBefore = &*Invoke->getNormalDest()->getFirstInsertionPt();
+    InsertBefore =
+        &SplitEdge(Invoke->getParent(), Invoke->getNormalDest())->front();
   else
     InsertBefore = &*std::next(CS.getInstruction()->getIterator());
 
   // Bitcast the return value to the correct type.
   auto *Cast = CastInst::Create(Instruction::BitCast, CS.getInstruction(),
                                 RetTy, "", InsertBefore);
+  if (RetBitCast)
+    *RetBitCast = Cast;
 
   // Replace all the original uses of the calling instruction with the bitcast.
   for (User *U : UsersToUpdate)
     U->replaceUsesOfWith(CS.getInstruction(), Cast);
-
-  return Cast;
 }
 
 /// Predicate and clone the given call site.
@@ -152,21 +200,78 @@ Instruction *createRetBitCast(CallSite C
 /// callee. The original call site is moved into the "else" block, and a clone
 /// of the call site is placed in the "then" block. The cloned instruction is
 /// returned.
+///
+/// For example, the call instruction below:
+///
+///   orig_bb:
+///     %t0 = call i32 %ptr()
+///     ...
+///
+/// Is replace by the following:
+///
+///   orig_bb:
+///     %cond = icmp eq i32 ()* %ptr, @func
+///     br i1 %cond, %then_bb, %else_bb
+///
+///   then_bb:
+///     ; The clone of the original call instruction is placed in the "then"
+///     ; block. It is not yet promoted.
+///     %t1 = call i32 %ptr()
+///     br merge_bb
+///
+///   else_bb:
+///     ; The original call instruction is moved to the "else" block.
+///     %t0 = call i32 %ptr()
+///     br merge_bb
+///
+///   merge_bb:
+///     ; Uses of the original call instruction are replaced by uses of the phi
+///     ; node.
+///     %t2 = phi i32 [ %t0, %else_bb ], [ %t1, %then_bb ]
+///     ...
+///
+/// A similar transformation is performed for invoke instructions. However,
+/// since invokes are terminating, more work is required. For example, the
+/// invoke instruction below:
+///
+///   orig_bb:
+///     %t0 = invoke %ptr() to label %normal_dst unwind label %unwind_dst
+///
+/// Is replace by the following:
+///
+///   orig_bb:
+///     %cond = icmp eq i32 ()* %ptr, @func
+///     br i1 %cond, %then_bb, %else_bb
+///
+///   then_bb:
+///     ; The clone of the original invoke instruction is placed in the "then"
+///     ; block, and its normal destination is set to the "merge" block. It is
+///     ; not yet promoted.
+///     %t1 = invoke i32 %ptr() to label %merge_bb unwind label %unwind_dst
+///
+///   else_bb:
+///     ; The original invoke instruction is moved into the "else" block, and
+///     ; its normal destination is set to the "merge" block.
+///     %t0 = invoke i32 %ptr() to label %merge_bb unwind label %unwind_dst
+///
+///   merge_bb:
+///     ; Uses of the original invoke instruction are replaced by uses of the
+///     ; phi node, and the merge block branches to the normal destination.
+///     %t2 = phi i32 [ %t0, %else_bb ], [ %t1, %then_bb ]
+///     br %normal_dst
+///
 static Instruction *versionCallSite(CallSite CS, Value *Callee,
-                                    MDNode *BranchWeights,
-                                    BasicBlock *&ThenBlock,
-                                    BasicBlock *&ElseBlock,
-                                    BasicBlock *&MergeBlock) {
+                                    MDNode *BranchWeights) {
 
   IRBuilder<> Builder(CS.getInstruction());
   Instruction *OrigInst = CS.getInstruction();
+  BasicBlock *OrigBlock = OrigInst->getParent();
 
   // Create the compare. The called value and callee must have the same type to
   // be compared.
-  auto *LHS =
-      Builder.CreateBitCast(CS.getCalledValue(), Builder.getInt8PtrTy());
-  auto *RHS = Builder.CreateBitCast(Callee, Builder.getInt8PtrTy());
-  auto *Cond = Builder.CreateICmpEQ(LHS, RHS);
+  if (CS.getCalledValue()->getType() != Callee->getType())
+    Callee = Builder.CreateBitCast(Callee, CS.getCalledValue()->getType());
+  auto *Cond = Builder.CreateICmpEQ(CS.getCalledValue(), Callee);
 
   // Create an if-then-else structure. The original instruction is moved into
   // the "else" block, and a clone of the original instruction is placed in the
@@ -175,9 +280,9 @@ static Instruction *versionCallSite(Call
   TerminatorInst *ElseTerm = nullptr;
   SplitBlockAndInsertIfThenElse(Cond, CS.getInstruction(), &ThenTerm, &ElseTerm,
                                 BranchWeights);
-  ThenBlock = ThenTerm->getParent();
-  ElseBlock = ElseTerm->getParent();
-  MergeBlock = OrigInst->getParent();
+  BasicBlock *ThenBlock = ThenTerm->getParent();
+  BasicBlock *ElseBlock = ElseTerm->getParent();
+  BasicBlock *MergeBlock = OrigInst->getParent();
 
   ThenBlock->setName("if.true.direct_targ");
   ElseBlock->setName("if.false.orig_indirect");
@@ -188,7 +293,8 @@ static Instruction *versionCallSite(Call
   NewInst->insertBefore(ThenTerm);
 
   // If the original call site is an invoke instruction, we have extra work to
-  // do since invoke instructions are terminating.
+  // do since invoke instructions are terminating. We have to fix-up phi nodes
+  // in the invoke's normal and unwind destinations.
   if (auto *OrigInvoke = dyn_cast<InvokeInst>(OrigInst)) {
     auto *NewInvoke = cast<InvokeInst>(NewInst);
 
@@ -201,11 +307,19 @@ static Instruction *versionCallSite(Call
     Builder.SetInsertPoint(MergeBlock);
     Builder.CreateBr(OrigInvoke->getNormalDest());
 
-    // Now set the normal destination of new the invoke instruction to be the
+    // Fix-up phi nodes in the original invoke's normal and unwind destinations.
+    fixupPHINodeForNormalDest(OrigInvoke, OrigBlock, MergeBlock);
+    fixupPHINodeForUnwindDest(OrigInvoke, MergeBlock, ThenBlock, ElseBlock);
+
+    // Now set the normal destinations of the invoke instructions to be the
     // "merge" block.
+    OrigInvoke->setNormalDest(MergeBlock);
     NewInvoke->setNormalDest(MergeBlock);
   }
 
+  // Create a phi node for the returned value of the call site.
+  createRetPHINode(OrigInst, NewInst, MergeBlock, Builder);
+
   return NewInst;
 }
 
@@ -253,7 +367,8 @@ bool llvm::isLegalToPromote(CallSite CS,
   return true;
 }
 
-static void promoteCall(CallSite CS, Function *Callee, Instruction *&Cast) {
+Instruction *llvm::promoteCall(CallSite CS, Function *Callee,
+                               CastInst **RetBitCast) {
   assert(!CS.getCalledFunction() && "Only indirect call sites can be promoted");
 
   // Set the called function of the call site to be the given callee.
@@ -268,7 +383,7 @@ static void promoteCall(CallSite CS, Fun
   // If the function type of the call site matches that of the callee, no
   // additional work is required.
   if (CS.getFunctionType() == Callee->getFunctionType())
-    return;
+    return CS.getInstruction();
 
   // Save the return types of the call site and callee.
   Type *CallSiteRetTy = CS.getInstruction()->getType();
@@ -294,7 +409,9 @@ static void promoteCall(CallSite CS, Fun
   // If the return type of the call site doesn't match that of the callee, cast
   // the returned value to the appropriate type.
   if (!CallSiteRetTy->isVoidTy() && CallSiteRetTy != CalleeRetTy)
-    Cast = createRetBitCast(CS, CallSiteRetTy);
+    createRetBitCast(CS, CallSiteRetTy, RetBitCast);
+
+  return CS.getInstruction();
 }
 
 Instruction *llvm::promoteCallWithIfThenElse(CallSite CS, Function *Callee,
@@ -303,26 +420,10 @@ Instruction *llvm::promoteCallWithIfThen
   // Version the indirect call site. If the called value is equal to the given
   // callee, 'NewInst' will be executed, otherwise the original call site will
   // be executed.
-  BasicBlock *ThenBlock, *ElseBlock, *MergeBlock;
-  Instruction *NewInst = versionCallSite(CS, Callee, BranchWeights, ThenBlock,
-                                         ElseBlock, MergeBlock);
+  Instruction *NewInst = versionCallSite(CS, Callee, BranchWeights);
 
   // Promote 'NewInst' so that it directly calls the desired function.
-  Instruction *Cast = NewInst;
-  promoteCall(CallSite(NewInst), Callee, Cast);
-
-  // If the original call site is an invoke instruction, we have to fix-up phi
-  // nodes in the invoke's normal and unwind destinations.
-  if (auto *OrigInvoke = dyn_cast<InvokeInst>(CS.getInstruction())) {
-    fixupPHINodeForNormalDest(OrigInvoke, MergeBlock, ElseBlock, Cast);
-    fixupPHINodeForUnwindDest(OrigInvoke, MergeBlock, ThenBlock, ElseBlock);
-  }
-
-  // Create a phi node for the returned value of the call site.
-  createRetPHINode(CS.getInstruction(), Cast ? Cast : NewInst);
-
-  // Return the new direct call.
-  return NewInst;
+  return promoteCall(CallSite(NewInst), Callee);
 }
 
 #undef DEBUG_TYPE

Modified: llvm/trunk/test/Transforms/PGOProfile/icp_covariant_call_return.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/PGOProfile/icp_covariant_call_return.ll?rev=321210&r1=321209&r2=321210&view=diff
==============================================================================
--- llvm/trunk/test/Transforms/PGOProfile/icp_covariant_call_return.ll (original)
+++ llvm/trunk/test/Transforms/PGOProfile/icp_covariant_call_return.ll Wed Dec 20 11:26:37 2017
@@ -22,8 +22,7 @@ entry:
   %vtable = load %struct.Base* (%struct.B*)**, %struct.Base* (%struct.B*)*** %tmp2, align 8
   %vfn = getelementptr inbounds %struct.Base* (%struct.B*)*, %struct.Base* (%struct.B*)** %vtable, i64 0
   %tmp3 = load %struct.Base* (%struct.B*)*, %struct.Base* (%struct.B*)** %vfn, align 8
-; ICALL-PROM:  [[BITCAST:%[0-9]+]] = bitcast %struct.Base* (%struct.B*)* %tmp3 to i8*
-; ICALL-PROM:  [[CMP:%[0-9]+]] = icmp eq i8* [[BITCAST]], bitcast (%struct.Derived* (%struct.D*)* @_ZN1D4funcEv to i8*)
+; ICALL-PROM:  [[CMP:%[0-9]+]] = icmp eq %struct.Base* (%struct.B*)* %tmp3, bitcast (%struct.Derived* (%struct.D*)* @_ZN1D4funcEv to %struct.Base* (%struct.B*)*)
 ; ICALL-PROM:  br i1 [[CMP]], label %if.true.direct_targ, label %if.false.orig_indirect, !prof [[BRANCH_WEIGHT:![0-9]+]]
 ; ICALL-PROM:if.true.direct_targ:
 ; ICALL-PROM:  [[ARG_BITCAST:%[0-9]+]] = bitcast %struct.B* %tmp1 to %struct.D*

Modified: llvm/trunk/test/Transforms/PGOProfile/icp_covariant_invoke_return.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/PGOProfile/icp_covariant_invoke_return.ll?rev=321210&r1=321209&r2=321210&view=diff
==============================================================================
--- llvm/trunk/test/Transforms/PGOProfile/icp_covariant_invoke_return.ll (original)
+++ llvm/trunk/test/Transforms/PGOProfile/icp_covariant_invoke_return.ll Wed Dec 20 11:26:37 2017
@@ -32,18 +32,19 @@ invoke.cont:
   %vtable = load %struct.Base* (%struct.B*)**, %struct.Base* (%struct.B*)*** %tmp2, align 8
   %vfn = getelementptr inbounds %struct.Base* (%struct.B*)*, %struct.Base* (%struct.B*)** %vtable, i64 0
   %tmp3 = load %struct.Base* (%struct.B*)*, %struct.Base* (%struct.B*)** %vfn, align 8
-; ICALL-PROM:  [[BITCAST:%[0-9]+]] = bitcast %struct.Base* (%struct.B*)* %tmp3 to i8*
-; ICALL-PROM:  [[CMP:%[0-9]+]] = icmp eq i8* [[BITCAST]], bitcast (%struct.Derived* (%struct.D*)* @_ZN1D4funcEv to i8*)
+; ICALL-PROM:  [[CMP:%[0-9]+]] = icmp eq %struct.Base* (%struct.B*)* %tmp3, bitcast (%struct.Derived* (%struct.D*)* @_ZN1D4funcEv to %struct.Base* (%struct.B*)*)
 ; ICALL-PROM:  br i1 [[CMP]], label %if.true.direct_targ, label %if.false.orig_indirect, !prof [[BRANCH_WEIGHT:![0-9]+]]
 ; ICALL-PROM:if.true.direct_targ:
 ; ICALL-PROM:  [[ARG_BITCAST:%[0-9]+]] = bitcast %struct.B* %tmp1 to %struct.D*
 ; ICALL-PROM:  [[DIRCALL_RET:%[0-9]+]] = invoke %struct.Derived* @_ZN1D4funcEv(%struct.D* [[ARG_BITCAST]])
-; ICALL-PROM:          to label %if.end.icp unwind label %lpad
+; ICALL-PROM:          to label %if.true.direct_targ.if.end.icp_crit_edge unwind label %lpad
+; ICALL-PROM:if.true.direct_targ.if.end.icp_crit_edge:
+; ICALL-PROM:  [[DIRCALL_RET_CAST:%[0-9]+]] = bitcast %struct.Derived* [[DIRCALL_RET]] to %struct.Base*
+; ICALL-PROM:  br label %if.end.icp
 ; ICALL-PROM:if.false.orig_indirect:
 ; ICAll-PROM:  %call2 = invoke %struct.Base* %tmp3(%struct.B* %tmp1)
 ; ICAll-PROM:          to label %invoke.cont1 unwind label %lpad
 ; ICALL-PROM:if.end.icp:
-; ICALL-PROM:  [[DIRCALL_RET_CAST:%[0-9]+]] = bitcast %struct.Derived* [[DIRCALL_RET]] to %struct.Base*
 ; ICALL-PROM:  br label %invoke.cont1
   %call2 = invoke %struct.Base* %tmp3(%struct.B* %tmp1)
           to label %invoke.cont1 unwind label %lpad, !prof !1

Modified: llvm/trunk/test/Transforms/PGOProfile/icp_invoke.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/PGOProfile/icp_invoke.ll?rev=321210&r1=321209&r2=321210&view=diff
==============================================================================
--- llvm/trunk/test/Transforms/PGOProfile/icp_invoke.ll (original)
+++ llvm/trunk/test/Transforms/PGOProfile/icp_invoke.ll Wed Dec 20 11:26:37 2017
@@ -20,8 +20,7 @@ entry:
 define i32 @_Z3goov() personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) {
 entry:
   %tmp = load void ()*, void ()** @foo1, align 8
-; ICP:  [[BITCAST_IC1:%[0-9]+]] = bitcast void ()* %tmp to i8*
-; ICP:  [[CMP_IC1:%[0-9]+]] = icmp eq i8* [[BITCAST_IC1]], bitcast (void ()* @_ZL4bar1v to i8*)
+; ICP:  [[CMP_IC1:%[0-9]+]] = icmp eq void ()* %tmp, @_ZL4bar1v
 ; ICP:  br i1 [[CMP_IC1]], label %[[TRUE_LABEL_IC1:.*]], label %[[FALSE_LABEL_IC1:.*]], !prof [[BRANCH_WEIGHT:![0-9]+]]
 ; ICP:[[TRUE_LABEL_IC1]]:
 ; ICP:  invoke void @_ZL4bar1v()
@@ -49,17 +48,19 @@ catch:
 
 try.cont:
   %tmp6 = load i32 ()*, i32 ()** @foo2, align 8
-; ICP:  [[BITCAST_IC2:%[0-9]+]] = bitcast i32 ()* %tmp6 to i8*
-; ICP:  [[CMP_IC2:%[0-9]+]] = icmp eq i8* [[BITCAST_IC2]], bitcast (i32 ()* @_ZL4bar2v to i8*)
+; ICP:  [[CMP_IC2:%[0-9]+]] = icmp eq i32 ()* %tmp6, @_ZL4bar2v
 ; ICP:  br i1 [[CMP_IC2]], label %[[TRUE_LABEL_IC2:.*]], label %[[FALSE_LABEL_IC2:.*]], !prof [[BRANCH_WEIGHT:![0-9]+]]
 ; ICP:[[TRUE_LABEL_IC2]]:
-; ICP:  [[RESULT_IC2:%[0-9]+]] = invoke i32 @_ZL4bar2v()
-; ICP:          to label %[[DCALL_NORMAL_DEST_IC2:.*]] unwind label %lpad1
+; ICP:  [[RESULT_IC2_0:%[0-9]+]] = invoke i32 @_ZL4bar2v()
+; ICP:          to label %[[MERGE_BB:.*]] unwind label %lpad1
 ; ICP:[[FALSE_LABEL_IC2]]:
+; ICP:  [[RESULT_IC2_1:%.+]] = invoke i32 %tmp6()
+; ICP:          to label %[[MERGE_BB]] unwind label %lpad1
   %call = invoke i32 %tmp6()
           to label %try.cont8 unwind label %lpad1, !prof !3
 
-; ICP:[[DCALL_NORMAL_DEST_IC2]]:
+; ICP:[[MERGE_BB]]:
+; ICP:  [[MERGE_PHI:%.+]] = phi i32 [ [[RESULT_IC2_1]], %[[FALSE_LABEL_IC2]] ], [ [[RESULT_IC2_0]], %[[TRUE_LABEL_IC2]] ]
 ; ICP:  br label %try.cont8
 lpad1:
   %tmp7 = landingpad { i8*, i32 }
@@ -77,7 +78,7 @@ catch6:
 
 try.cont8:
   %i.0 = phi i32 [ undef, %catch6 ], [ %call, %try.cont ]
-; ICP:  %i.0 = phi i32 [ undef, %catch6 ], [ %call, %[[FALSE_LABEL_IC2]] ], [ [[RESULT_IC2]], %[[DCALL_NORMAL_DEST_IC2]] ]
+; ICP:  %i.0 = phi i32 [ undef, %catch6 ], [ [[MERGE_PHI]], %[[MERGE_BB]] ]
   ret i32 %i.0
 
 eh.resume:

Modified: llvm/trunk/test/Transforms/PGOProfile/icp_invoke_nouse.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/PGOProfile/icp_invoke_nouse.ll?rev=321210&r1=321209&r2=321210&view=diff
==============================================================================
--- llvm/trunk/test/Transforms/PGOProfile/icp_invoke_nouse.ll (original)
+++ llvm/trunk/test/Transforms/PGOProfile/icp_invoke_nouse.ll Wed Dec 20 11:26:37 2017
@@ -18,8 +18,7 @@ entry:
 
 if.end:                                           ; preds = %entry
   %fptr = load i32 ()*, i32 ()** @pfptr, align 8
-; ICP:  [[BITCAST_IC1:%[0-9]+]] = bitcast i32 ()* %fptr to i8*
-; ICP:  [[CMP_IC1:%[0-9]+]] = icmp eq i8* [[BITCAST_IC1]], bitcast (i32 ()* @_ZL4bar1v to i8*)
+; ICP:  [[CMP_IC1:%[0-9]+]] = icmp eq i32 ()* %fptr, @_ZL4bar1v
 ; ICP:  br i1 [[CMP_IC1]], label %[[TRUE_LABEL_IC1:.*]], label %[[FALSE_LABEL_IC1:.*]], !prof [[BRANCH_WEIGHT:![0-9]+]]
 ; ICP:[[TRUE_LABEL_IC1]]:
 ; ICP:  invoke i32 @_ZL4bar1v()

Modified: llvm/trunk/test/Transforms/PGOProfile/icp_vararg.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/PGOProfile/icp_vararg.ll?rev=321210&r1=321209&r2=321210&view=diff
==============================================================================
--- llvm/trunk/test/Transforms/PGOProfile/icp_vararg.ll (original)
+++ llvm/trunk/test/Transforms/PGOProfile/icp_vararg.ll Wed Dec 20 11:26:37 2017
@@ -13,8 +13,7 @@ entry:
 define i32 @bar() #1 {
 entry:
   %tmp = load i32 (i32, ...)*, i32 (i32, ...)** @foo, align 8
-; ICALL-PROM:  [[BITCAST:%[0-9]+]] = bitcast i32 (i32, ...)* %tmp to i8*
-; ICALL-PROM:  [[CMP:%[0-9]+]] = icmp eq i8* [[BITCAST]], bitcast (i32 (i32, ...)* @va_func to i8*)
+; ICALL-PROM:  [[CMP:%[0-9]+]] = icmp eq i32 (i32, ...)* %tmp, @va_func
 ; ICALL-PROM:  br i1 [[CMP]], label %if.true.direct_targ, label %if.false.orig_indirect, !prof [[BRANCH_WEIGHT:![0-9]+]]
 ; ICALL-PROM:if.true.direct_targ:
 ; ICALL-PROM:  [[DIRCALL_RET:%[0-9]+]] = call i32 (i32, ...) @va_func(i32 3, i32 12, i32 22, i32 4)

Modified: llvm/trunk/test/Transforms/PGOProfile/indirect_call_promotion.ll
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/test/Transforms/PGOProfile/indirect_call_promotion.ll?rev=321210&r1=321209&r2=321210&view=diff
==============================================================================
--- llvm/trunk/test/Transforms/PGOProfile/indirect_call_promotion.ll (original)
+++ llvm/trunk/test/Transforms/PGOProfile/indirect_call_promotion.ll Wed Dec 20 11:26:37 2017
@@ -43,8 +43,7 @@ entry:
 define i32 @bar() {
 entry:
   %tmp = load i32 ()*, i32 ()** @foo, align 8
-; ICALL-PROM:   [[BITCAST:%[0-9]+]] = bitcast i32 ()* %tmp to i8*
-; ICALL-PROM:   [[CMP:%[0-9]+]] = icmp eq i8* [[BITCAST]], bitcast (i32 ()* @func4 to i8*)
+; ICALL-PROM:   [[CMP:%[0-9]+]] = icmp eq i32 ()* %tmp, @func4
 ; ICALL-PROM:   br i1 [[CMP]], label %if.true.direct_targ, label %if.false.orig_indirect, !prof [[BRANCH_WEIGHT:![0-9]+]]
 ; ICALL-PROM: if.true.direct_targ:
 ; ICALL-PROM:   [[DIRCALL_RET:%[0-9]+]] = call i32 @func4()




More information about the llvm-commits mailing list