1 //===- SjLjEHPrepare.cpp - Eliminate Invoke & Unwind instructions ---------===//
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 transformation is designed for use by code generators which use SjLj
11 // based exception handling.
12 //
13 //===----------------------------------------------------------------------===//
14 
15 #include "llvm/CodeGen/Passes.h"
16 #include "llvm/ADT/DenseMap.h"
17 #include "llvm/ADT/SetVector.h"
18 #include "llvm/ADT/SmallPtrSet.h"
19 #include "llvm/ADT/SmallVector.h"
20 #include "llvm/ADT/Statistic.h"
21 #include "llvm/IR/Constants.h"
22 #include "llvm/IR/DataLayout.h"
23 #include "llvm/IR/DerivedTypes.h"
24 #include "llvm/IR/IRBuilder.h"
25 #include "llvm/IR/Instructions.h"
26 #include "llvm/IR/Intrinsics.h"
27 #include "llvm/IR/LLVMContext.h"
28 #include "llvm/IR/Module.h"
29 #include "llvm/Pass.h"
30 #include "llvm/Support/CommandLine.h"
31 #include "llvm/Support/Debug.h"
32 #include "llvm/Support/raw_ostream.h"
33 #include "llvm/Target/TargetLowering.h"
34 #include "llvm/Transforms/Scalar.h"
35 #include "llvm/Transforms/Utils/BasicBlockUtils.h"
36 #include "llvm/Transforms/Utils/Local.h"
37 #include <set>
38 using namespace llvm;
39 
40 #define DEBUG_TYPE "sjljehprepare"
41 
42 STATISTIC(NumInvokes, "Number of invokes replaced");
43 STATISTIC(NumSpilled, "Number of registers live across unwind edges");
44 
45 namespace {
46 class SjLjEHPrepare : public FunctionPass {
47   const TargetMachine *TM;
48   Type *FunctionContextTy;
49   Constant *RegisterFn;
50   Constant *UnregisterFn;
51   Constant *BuiltinSetjmpFn;
52   Constant *FrameAddrFn;
53   Constant *StackAddrFn;
54   Constant *StackRestoreFn;
55   Constant *LSDAAddrFn;
56   Value *PersonalityFn;
57   Constant *CallSiteFn;
58   Constant *FuncCtxFn;
59   AllocaInst *FuncCtx;
60 
61 public:
62   static char ID; // Pass identification, replacement for typeid
63   explicit SjLjEHPrepare(const TargetMachine *TM) : FunctionPass(ID), TM(TM) {}
64   bool doInitialization(Module &M) override;
65   bool runOnFunction(Function &F) override;
66 
67   void getAnalysisUsage(AnalysisUsage &AU) const override {}
68   const char *getPassName() const override {
69     return "SJLJ Exception Handling preparation";
70   }
71 
72 private:
73   bool setupEntryBlockAndCallSites(Function &F);
74   void substituteLPadValues(LandingPadInst *LPI, Value *ExnVal, Value *SelVal);
75   Value *setupFunctionContext(Function &F, ArrayRef<LandingPadInst *> LPads);
76   void lowerIncomingArguments(Function &F);
77   void lowerAcrossUnwindEdges(Function &F, ArrayRef<InvokeInst *> Invokes);
78   void insertCallSiteStore(Instruction *I, int Number);
79 };
80 } // end anonymous namespace
81 
82 char SjLjEHPrepare::ID = 0;
83 
84 // Public Interface To the SjLjEHPrepare pass.
85 FunctionPass *llvm::createSjLjEHPreparePass(const TargetMachine *TM) {
86   return new SjLjEHPrepare(TM);
87 }
88 // doInitialization - Set up decalarations and types needed to process
89 // exceptions.
90 bool SjLjEHPrepare::doInitialization(Module &M) {
91   // Build the function context structure.
92   // builtin_setjmp uses a five word jbuf
93   Type *VoidPtrTy = Type::getInt8PtrTy(M.getContext());
94   Type *Int32Ty = Type::getInt32Ty(M.getContext());
95   FunctionContextTy = StructType::get(VoidPtrTy,                  // __prev
96                                       Int32Ty,                    // call_site
97                                       ArrayType::get(Int32Ty, 4), // __data
98                                       VoidPtrTy, // __personality
99                                       VoidPtrTy, // __lsda
100                                       ArrayType::get(VoidPtrTy, 5), // __jbuf
101                                       NULL);
102   RegisterFn = M.getOrInsertFunction(
103       "_Unwind_SjLj_Register", Type::getVoidTy(M.getContext()),
104       PointerType::getUnqual(FunctionContextTy), (Type *)nullptr);
105   UnregisterFn = M.getOrInsertFunction(
106       "_Unwind_SjLj_Unregister", Type::getVoidTy(M.getContext()),
107       PointerType::getUnqual(FunctionContextTy), (Type *)nullptr);
108   FrameAddrFn = Intrinsic::getDeclaration(&M, Intrinsic::frameaddress);
109   StackAddrFn = Intrinsic::getDeclaration(&M, Intrinsic::stacksave);
110   StackRestoreFn = Intrinsic::getDeclaration(&M, Intrinsic::stackrestore);
111   BuiltinSetjmpFn = Intrinsic::getDeclaration(&M, Intrinsic::eh_sjlj_setjmp);
112   LSDAAddrFn = Intrinsic::getDeclaration(&M, Intrinsic::eh_sjlj_lsda);
113   CallSiteFn = Intrinsic::getDeclaration(&M, Intrinsic::eh_sjlj_callsite);
114   FuncCtxFn = Intrinsic::getDeclaration(&M, Intrinsic::eh_sjlj_functioncontext);
115   PersonalityFn = nullptr;
116 
117   return true;
118 }
119 
120 /// insertCallSiteStore - Insert a store of the call-site value to the
121 /// function context
122 void SjLjEHPrepare::insertCallSiteStore(Instruction *I, int Number) {
123   IRBuilder<> Builder(I);
124 
125   // Get a reference to the call_site field.
126   Type *Int32Ty = Type::getInt32Ty(I->getContext());
127   Value *Zero = ConstantInt::get(Int32Ty, 0);
128   Value *One = ConstantInt::get(Int32Ty, 1);
129   Value *Idxs[2] = { Zero, One };
130   Value *CallSite = Builder.CreateGEP(FuncCtx, Idxs, "call_site");
131 
132   // Insert a store of the call-site number
133   ConstantInt *CallSiteNoC =
134       ConstantInt::get(Type::getInt32Ty(I->getContext()), Number);
135   Builder.CreateStore(CallSiteNoC, CallSite, true /*volatile*/);
136 }
137 
138 /// MarkBlocksLiveIn - Insert BB and all of its predescessors into LiveBBs until
139 /// we reach blocks we've already seen.
140 static void MarkBlocksLiveIn(BasicBlock *BB,
141                              SmallPtrSet<BasicBlock *, 64> &LiveBBs) {
142   if (!LiveBBs.insert(BB))
143     return; // already been here.
144 
145   for (pred_iterator PI = pred_begin(BB), E = pred_end(BB); PI != E; ++PI)
146     MarkBlocksLiveIn(*PI, LiveBBs);
147 }
148 
149 /// substituteLPadValues - Substitute the values returned by the landingpad
150 /// instruction with those returned by the personality function.
151 void SjLjEHPrepare::substituteLPadValues(LandingPadInst *LPI, Value *ExnVal,
152                                          Value *SelVal) {
153   SmallVector<Value *, 8> UseWorkList(LPI->user_begin(), LPI->user_end());
154   while (!UseWorkList.empty()) {
155     Value *Val = UseWorkList.pop_back_val();
156     ExtractValueInst *EVI = dyn_cast<ExtractValueInst>(Val);
157     if (!EVI)
158       continue;
159     if (EVI->getNumIndices() != 1)
160       continue;
161     if (*EVI->idx_begin() == 0)
162       EVI->replaceAllUsesWith(ExnVal);
163     else if (*EVI->idx_begin() == 1)
164       EVI->replaceAllUsesWith(SelVal);
165     if (EVI->getNumUses() == 0)
166       EVI->eraseFromParent();
167   }
168 
169   if (LPI->getNumUses() == 0)
170     return;
171 
172   // There are still some uses of LPI. Construct an aggregate with the exception
173   // values and replace the LPI with that aggregate.
174   Type *LPadType = LPI->getType();
175   Value *LPadVal = UndefValue::get(LPadType);
176   IRBuilder<> Builder(
177       std::next(BasicBlock::iterator(cast<Instruction>(SelVal))));
178   LPadVal = Builder.CreateInsertValue(LPadVal, ExnVal, 0, "lpad.val");
179   LPadVal = Builder.CreateInsertValue(LPadVal, SelVal, 1, "lpad.val");
180 
181   LPI->replaceAllUsesWith(LPadVal);
182 }
183 
184 /// setupFunctionContext - Allocate the function context on the stack and fill
185 /// it with all of the data that we know at this point.
186 Value *SjLjEHPrepare::setupFunctionContext(Function &F,
187                                            ArrayRef<LandingPadInst *> LPads) {
188   BasicBlock *EntryBB = F.begin();
189 
190   // Create an alloca for the incoming jump buffer ptr and the new jump buffer
191   // that needs to be restored on all exits from the function. This is an alloca
192   // because the value needs to be added to the global context list.
193   const TargetLowering *TLI = TM->getTargetLowering();
194   unsigned Align =
195       TLI->getDataLayout()->getPrefTypeAlignment(FunctionContextTy);
196   FuncCtx = new AllocaInst(FunctionContextTy, nullptr, Align, "fn_context",
197                            EntryBB->begin());
198 
199   // Fill in the function context structure.
200   for (unsigned I = 0, E = LPads.size(); I != E; ++I) {
201     LandingPadInst *LPI = LPads[I];
202     IRBuilder<> Builder(LPI->getParent()->getFirstInsertionPt());
203 
204     // Reference the __data field.
205     Value *FCData = Builder.CreateConstGEP2_32(FuncCtx, 0, 2, "__data");
206 
207     // The exception values come back in context->__data[0].
208     Value *ExceptionAddr =
209         Builder.CreateConstGEP2_32(FCData, 0, 0, "exception_gep");
210     Value *ExnVal = Builder.CreateLoad(ExceptionAddr, true, "exn_val");
211     ExnVal = Builder.CreateIntToPtr(ExnVal, Builder.getInt8PtrTy());
212 
213     Value *SelectorAddr =
214         Builder.CreateConstGEP2_32(FCData, 0, 1, "exn_selector_gep");
215     Value *SelVal = Builder.CreateLoad(SelectorAddr, true, "exn_selector_val");
216 
217     substituteLPadValues(LPI, ExnVal, SelVal);
218   }
219 
220   // Personality function
221   IRBuilder<> Builder(EntryBB->getTerminator());
222   if (!PersonalityFn)
223     PersonalityFn = LPads[0]->getPersonalityFn();
224   Value *PersonalityFieldPtr =
225       Builder.CreateConstGEP2_32(FuncCtx, 0, 3, "pers_fn_gep");
226   Builder.CreateStore(
227       Builder.CreateBitCast(PersonalityFn, Builder.getInt8PtrTy()),
228       PersonalityFieldPtr, /*isVolatile=*/true);
229 
230   // LSDA address
231   Value *LSDA = Builder.CreateCall(LSDAAddrFn, "lsda_addr");
232   Value *LSDAFieldPtr = Builder.CreateConstGEP2_32(FuncCtx, 0, 4, "lsda_gep");
233   Builder.CreateStore(LSDA, LSDAFieldPtr, /*isVolatile=*/true);
234 
235   return FuncCtx;
236 }
237 
238 /// lowerIncomingArguments - To avoid having to handle incoming arguments
239 /// specially, we lower each arg to a copy instruction in the entry block. This
240 /// ensures that the argument value itself cannot be live out of the entry
241 /// block.
242 void SjLjEHPrepare::lowerIncomingArguments(Function &F) {
243   BasicBlock::iterator AfterAllocaInsPt = F.begin()->begin();
244   while (isa<AllocaInst>(AfterAllocaInsPt) &&
245          isa<ConstantInt>(cast<AllocaInst>(AfterAllocaInsPt)->getArraySize()))
246     ++AfterAllocaInsPt;
247 
248   for (Function::arg_iterator AI = F.arg_begin(), AE = F.arg_end(); AI != AE;
249        ++AI) {
250     Type *Ty = AI->getType();
251 
252     // Use 'select i8 true, %arg, undef' to simulate a 'no-op' instruction.
253     Value *TrueValue = ConstantInt::getTrue(F.getContext());
254     Value *UndefValue = UndefValue::get(Ty);
255     Instruction *SI = SelectInst::Create(TrueValue, AI, UndefValue,
256                                          AI->getName() + ".tmp",
257                                          AfterAllocaInsPt);
258     AI->replaceAllUsesWith(SI);
259 
260     // Reset the operand, because it  was clobbered by the RAUW above.
261     SI->setOperand(1, AI);
262   }
263 }
264 
265 /// lowerAcrossUnwindEdges - Find all variables which are alive across an unwind
266 /// edge and spill them.
267 void SjLjEHPrepare::lowerAcrossUnwindEdges(Function &F,
268                                            ArrayRef<InvokeInst *> Invokes) {
269   // Finally, scan the code looking for instructions with bad live ranges.
270   for (Function::iterator BB = F.begin(), BBE = F.end(); BB != BBE; ++BB) {
271     for (BasicBlock::iterator II = BB->begin(), IIE = BB->end(); II != IIE;
272          ++II) {
273       // Ignore obvious cases we don't have to handle. In particular, most
274       // instructions either have no uses or only have a single use inside the
275       // current block. Ignore them quickly.
276       Instruction *Inst = II;
277       if (Inst->use_empty())
278         continue;
279       if (Inst->hasOneUse() &&
280           cast<Instruction>(Inst->user_back())->getParent() == BB &&
281           !isa<PHINode>(Inst->user_back()))
282         continue;
283 
284       // If this is an alloca in the entry block, it's not a real register
285       // value.
286       if (AllocaInst *AI = dyn_cast<AllocaInst>(Inst))
287         if (isa<ConstantInt>(AI->getArraySize()) && BB == F.begin())
288           continue;
289 
290       // Avoid iterator invalidation by copying users to a temporary vector.
291       SmallVector<Instruction *, 16> Users;
292       for (User *U : Inst->users()) {
293         Instruction *UI = cast<Instruction>(U);
294         if (UI->getParent() != BB || isa<PHINode>(UI))
295           Users.push_back(UI);
296       }
297 
298       // Find all of the blocks that this value is live in.
299       SmallPtrSet<BasicBlock *, 64> LiveBBs;
300       LiveBBs.insert(Inst->getParent());
301       while (!Users.empty()) {
302         Instruction *U = Users.back();
303         Users.pop_back();
304 
305         if (!isa<PHINode>(U)) {
306           MarkBlocksLiveIn(U->getParent(), LiveBBs);
307         } else {
308           // Uses for a PHI node occur in their predecessor block.
309           PHINode *PN = cast<PHINode>(U);
310           for (unsigned i = 0, e = PN->getNumIncomingValues(); i != e; ++i)
311             if (PN->getIncomingValue(i) == Inst)
312               MarkBlocksLiveIn(PN->getIncomingBlock(i), LiveBBs);
313         }
314       }
315 
316       // Now that we know all of the blocks that this thing is live in, see if
317       // it includes any of the unwind locations.
318       bool NeedsSpill = false;
319       for (unsigned i = 0, e = Invokes.size(); i != e; ++i) {
320         BasicBlock *UnwindBlock = Invokes[i]->getUnwindDest();
321         if (UnwindBlock != BB && LiveBBs.count(UnwindBlock)) {
322           DEBUG(dbgs() << "SJLJ Spill: " << *Inst << " around "
323                        << UnwindBlock->getName() << "\n");
324           NeedsSpill = true;
325           break;
326         }
327       }
328 
329       // If we decided we need a spill, do it.
330       // FIXME: Spilling this way is overkill, as it forces all uses of
331       // the value to be reloaded from the stack slot, even those that aren't
332       // in the unwind blocks. We should be more selective.
333       if (NeedsSpill) {
334         DemoteRegToStack(*Inst, true);
335         ++NumSpilled;
336       }
337     }
338   }
339 
340   // Go through the landing pads and remove any PHIs there.
341   for (unsigned i = 0, e = Invokes.size(); i != e; ++i) {
342     BasicBlock *UnwindBlock = Invokes[i]->getUnwindDest();
343     LandingPadInst *LPI = UnwindBlock->getLandingPadInst();
344 
345     // Place PHIs into a set to avoid invalidating the iterator.
346     SmallPtrSet<PHINode *, 8> PHIsToDemote;
347     for (BasicBlock::iterator PN = UnwindBlock->begin(); isa<PHINode>(PN); ++PN)
348       PHIsToDemote.insert(cast<PHINode>(PN));
349     if (PHIsToDemote.empty())
350       continue;
351 
352     // Demote the PHIs to the stack.
353     for (SmallPtrSet<PHINode *, 8>::iterator I = PHIsToDemote.begin(),
354                                              E = PHIsToDemote.end();
355          I != E; ++I)
356       DemotePHIToStack(*I);
357 
358     // Move the landingpad instruction back to the top of the landing pad block.
359     LPI->moveBefore(UnwindBlock->begin());
360   }
361 }
362 
363 /// setupEntryBlockAndCallSites - Setup the entry block by creating and filling
364 /// the function context and marking the call sites with the appropriate
365 /// values. These values are used by the DWARF EH emitter.
366 bool SjLjEHPrepare::setupEntryBlockAndCallSites(Function &F) {
367   SmallVector<ReturnInst *, 16> Returns;
368   SmallVector<InvokeInst *, 16> Invokes;
369   SmallSetVector<LandingPadInst *, 16> LPads;
370 
371   // Look through the terminators of the basic blocks to find invokes.
372   for (Function::iterator BB = F.begin(), E = F.end(); BB != E; ++BB)
373     if (InvokeInst *II = dyn_cast<InvokeInst>(BB->getTerminator())) {
374       if (Function *Callee = II->getCalledFunction())
375         if (Callee->isIntrinsic() &&
376             Callee->getIntrinsicID() == Intrinsic::donothing) {
377           // Remove the NOP invoke.
378           BranchInst::Create(II->getNormalDest(), II);
379           II->eraseFromParent();
380           continue;
381         }
382 
383       Invokes.push_back(II);
384       LPads.insert(II->getUnwindDest()->getLandingPadInst());
385     } else if (ReturnInst *RI = dyn_cast<ReturnInst>(BB->getTerminator())) {
386       Returns.push_back(RI);
387     }
388 
389   if (Invokes.empty())
390     return false;
391 
392   NumInvokes += Invokes.size();
393 
394   lowerIncomingArguments(F);
395   lowerAcrossUnwindEdges(F, Invokes);
396 
397   Value *FuncCtx =
398       setupFunctionContext(F, makeArrayRef(LPads.begin(), LPads.end()));
399   BasicBlock *EntryBB = F.begin();
400   IRBuilder<> Builder(EntryBB->getTerminator());
401 
402   // Get a reference to the jump buffer.
403   Value *JBufPtr = Builder.CreateConstGEP2_32(FuncCtx, 0, 5, "jbuf_gep");
404 
405   // Save the frame pointer.
406   Value *FramePtr = Builder.CreateConstGEP2_32(JBufPtr, 0, 0, "jbuf_fp_gep");
407 
408   Value *Val = Builder.CreateCall(FrameAddrFn, Builder.getInt32(0), "fp");
409   Builder.CreateStore(Val, FramePtr, /*isVolatile=*/true);
410 
411   // Save the stack pointer.
412   Value *StackPtr = Builder.CreateConstGEP2_32(JBufPtr, 0, 2, "jbuf_sp_gep");
413 
414   Val = Builder.CreateCall(StackAddrFn, "sp");
415   Builder.CreateStore(Val, StackPtr, /*isVolatile=*/true);
416 
417   // Call the setjmp instrinsic. It fills in the rest of the jmpbuf.
418   Value *SetjmpArg = Builder.CreateBitCast(JBufPtr, Builder.getInt8PtrTy());
419   Builder.CreateCall(BuiltinSetjmpFn, SetjmpArg);
420 
421   // Store a pointer to the function context so that the back-end will know
422   // where to look for it.
423   Value *FuncCtxArg = Builder.CreateBitCast(FuncCtx, Builder.getInt8PtrTy());
424   Builder.CreateCall(FuncCtxFn, FuncCtxArg);
425 
426   // At this point, we are all set up, update the invoke instructions to mark
427   // their call_site values.
428   for (unsigned I = 0, E = Invokes.size(); I != E; ++I) {
429     insertCallSiteStore(Invokes[I], I + 1);
430 
431     ConstantInt *CallSiteNum =
432         ConstantInt::get(Type::getInt32Ty(F.getContext()), I + 1);
433 
434     // Record the call site value for the back end so it stays associated with
435     // the invoke.
436     CallInst::Create(CallSiteFn, CallSiteNum, "", Invokes[I]);
437   }
438 
439   // Mark call instructions that aren't nounwind as no-action (call_site ==
440   // -1). Skip the entry block, as prior to then, no function context has been
441   // created for this function and any unexpected exceptions thrown will go
442   // directly to the caller's context, which is what we want anyway, so no need
443   // to do anything here.
444   for (Function::iterator BB = F.begin(), E = F.end(); ++BB != E;)
445     for (BasicBlock::iterator I = BB->begin(), end = BB->end(); I != end; ++I)
446       if (CallInst *CI = dyn_cast<CallInst>(I)) {
447         if (!CI->doesNotThrow())
448           insertCallSiteStore(CI, -1);
449       } else if (ResumeInst *RI = dyn_cast<ResumeInst>(I)) {
450         insertCallSiteStore(RI, -1);
451       }
452 
453   // Register the function context and make sure it's known to not throw
454   CallInst *Register =
455       CallInst::Create(RegisterFn, FuncCtx, "", EntryBB->getTerminator());
456   Register->setDoesNotThrow();
457 
458   // Following any allocas not in the entry block, update the saved SP in the
459   // jmpbuf to the new value.
460   for (Function::iterator BB = F.begin(), E = F.end(); BB != E; ++BB) {
461     if (BB == F.begin())
462       continue;
463     for (BasicBlock::iterator I = BB->begin(), E = BB->end(); I != E; ++I) {
464       if (CallInst *CI = dyn_cast<CallInst>(I)) {
465         if (CI->getCalledFunction() != StackRestoreFn)
466           continue;
467       } else if (!isa<AllocaInst>(I)) {
468         continue;
469       }
470       Instruction *StackAddr = CallInst::Create(StackAddrFn, "sp");
471       StackAddr->insertAfter(I);
472       Instruction *StoreStackAddr = new StoreInst(StackAddr, StackPtr, true);
473       StoreStackAddr->insertAfter(StackAddr);
474     }
475   }
476 
477   // Finally, for any returns from this function, if this function contains an
478   // invoke, add a call to unregister the function context.
479   for (unsigned I = 0, E = Returns.size(); I != E; ++I)
480     CallInst::Create(UnregisterFn, FuncCtx, "", Returns[I]);
481 
482   return true;
483 }
484 
485 bool SjLjEHPrepare::runOnFunction(Function &F) {
486   bool Res = setupEntryBlockAndCallSites(F);
487   return Res;
488 }
489