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