1 //===- GlobalSplit.cpp - global variable splitter -------------------------===//
2 //
3 //                     The LLVM Compiler Infrastructure
4 //
5 // This file is distributed under the University of Illinois Open Source
6 // License. See LICENSE.TXT for details.
7 //
8 //===----------------------------------------------------------------------===//
9 //
10 // This pass uses inrange annotations on GEP indices to split globals where
11 // beneficial. Clang currently attaches these annotations to references to
12 // virtual table globals under the Itanium ABI for the benefit of the
13 // whole-program virtual call optimization and control flow integrity passes.
14 //
15 //===----------------------------------------------------------------------===//
16 
17 #include "llvm/Transforms/IPO.h"
18 #include "llvm/Transforms/IPO/GlobalSplit.h"
19 #include "llvm/ADT/StringExtras.h"
20 #include "llvm/IR/Constants.h"
21 #include "llvm/IR/GlobalVariable.h"
22 #include "llvm/IR/Intrinsics.h"
23 #include "llvm/IR/Module.h"
24 #include "llvm/IR/Operator.h"
25 #include "llvm/Pass.h"
26 
27 #include <set>
28 
29 using namespace llvm;
30 
31 namespace {
32 
33 bool splitGlobal(GlobalVariable &GV) {
34   // If the address of the global is taken outside of the module, we cannot
35   // apply this transformation.
36   if (!GV.hasLocalLinkage())
37     return false;
38 
39   // We currently only know how to split ConstantStructs.
40   auto *Init = dyn_cast_or_null<ConstantStruct>(GV.getInitializer());
41   if (!Init)
42     return false;
43 
44   // Verify that each user of the global is an inrange getelementptr constant.
45   // From this it follows that any loads from or stores to that global must use
46   // a pointer derived from an inrange getelementptr constant, which is
47   // sufficient to allow us to apply the splitting transform.
48   for (User *U : GV.users()) {
49     if (!isa<Constant>(U))
50       return false;
51 
52     auto *GEP = dyn_cast<GEPOperator>(U);
53     if (!GEP || !GEP->getInRangeIndex() || *GEP->getInRangeIndex() != 1 ||
54         !isa<ConstantInt>(GEP->getOperand(1)) ||
55         !cast<ConstantInt>(GEP->getOperand(1))->isZero() ||
56         !isa<ConstantInt>(GEP->getOperand(2)))
57       return false;
58   }
59 
60   SmallVector<MDNode *, 2> Types;
61   GV.getMetadata(LLVMContext::MD_type, Types);
62 
63   const DataLayout &DL = GV.getParent()->getDataLayout();
64   const StructLayout *SL = DL.getStructLayout(Init->getType());
65 
66   IntegerType *Int32Ty = Type::getInt32Ty(GV.getContext());
67 
68   std::vector<GlobalVariable *> SplitGlobals(Init->getNumOperands());
69   for (unsigned I = 0; I != Init->getNumOperands(); ++I) {
70     // Build a global representing this split piece.
71     auto *SplitGV =
72         new GlobalVariable(*GV.getParent(), Init->getOperand(I)->getType(),
73                            GV.isConstant(), GlobalValue::PrivateLinkage,
74                            Init->getOperand(I), GV.getName() + "." + utostr(I));
75     SplitGlobals[I] = SplitGV;
76 
77     unsigned SplitBegin = SL->getElementOffset(I);
78     unsigned SplitEnd = (I == Init->getNumOperands() - 1)
79                             ? SL->getSizeInBytes()
80                             : SL->getElementOffset(I + 1);
81 
82     // Rebuild type metadata, adjusting by the split offset.
83     // FIXME: See if we can use DW_OP_piece to preserve debug metadata here.
84     for (MDNode *Type : Types) {
85       uint64_t ByteOffset = cast<ConstantInt>(
86               cast<ConstantAsMetadata>(Type->getOperand(0))->getValue())
87               ->getZExtValue();
88       if (ByteOffset < SplitBegin || ByteOffset >= SplitEnd)
89         continue;
90       SplitGV->addMetadata(
91           LLVMContext::MD_type,
92           *MDNode::get(GV.getContext(),
93                        {ConstantAsMetadata::get(
94                             ConstantInt::get(Int32Ty, ByteOffset - SplitBegin)),
95                         Type->getOperand(1)}));
96     }
97   }
98 
99   for (User *U : GV.users()) {
100     auto *GEP = cast<GEPOperator>(U);
101     unsigned I = cast<ConstantInt>(GEP->getOperand(2))->getZExtValue();
102     if (I >= SplitGlobals.size())
103       continue;
104 
105     SmallVector<Value *, 4> Ops;
106     Ops.push_back(ConstantInt::get(Int32Ty, 0));
107     for (unsigned I = 3; I != GEP->getNumOperands(); ++I)
108       Ops.push_back(GEP->getOperand(I));
109 
110     auto *NewGEP = ConstantExpr::getGetElementPtr(
111         SplitGlobals[I]->getInitializer()->getType(), SplitGlobals[I], Ops,
112         GEP->isInBounds());
113     GEP->replaceAllUsesWith(NewGEP);
114   }
115 
116   // Finally, remove the original global. Any remaining uses refer to invalid
117   // elements of the global, so replace with undef.
118   if (!GV.use_empty())
119     GV.replaceAllUsesWith(UndefValue::get(GV.getType()));
120   GV.eraseFromParent();
121   return true;
122 }
123 
124 bool splitGlobals(Module &M) {
125   // First, see if the module uses either of the llvm.type.test or
126   // llvm.type.checked.load intrinsics, which indicates that splitting globals
127   // may be beneficial.
128   Function *TypeTestFunc =
129       M.getFunction(Intrinsic::getName(Intrinsic::type_test));
130   Function *TypeCheckedLoadFunc =
131       M.getFunction(Intrinsic::getName(Intrinsic::type_checked_load));
132   if ((!TypeTestFunc || TypeTestFunc->use_empty()) &&
133       (!TypeCheckedLoadFunc || TypeCheckedLoadFunc->use_empty()))
134     return false;
135 
136   bool Changed = false;
137   for (auto I = M.global_begin(); I != M.global_end();) {
138     GlobalVariable &GV = *I;
139     ++I;
140     Changed |= splitGlobal(GV);
141   }
142   return Changed;
143 }
144 
145 struct GlobalSplit : public ModulePass {
146   static char ID;
147   GlobalSplit() : ModulePass(ID) {
148     initializeGlobalSplitPass(*PassRegistry::getPassRegistry());
149   }
150   bool runOnModule(Module &M) {
151     if (skipModule(M))
152       return false;
153 
154     return splitGlobals(M);
155   }
156 };
157 
158 }
159 
160 INITIALIZE_PASS(GlobalSplit, "globalsplit", "Global splitter", false, false)
161 char GlobalSplit::ID = 0;
162 
163 ModulePass *llvm::createGlobalSplitPass() {
164   return new GlobalSplit;
165 }
166 
167 PreservedAnalyses GlobalSplitPass::run(Module &M, ModuleAnalysisManager &AM) {
168   if (!splitGlobals(M))
169     return PreservedAnalyses::all();
170   return PreservedAnalyses::none();
171 }
172