[obj2yaml] - Fix BB after r373315.
[llvm-complete.git] / lib / Target / AMDGPU / AMDGPURewriteOutArguments.cpp
blob4f095087a57f4cc9f8390fc0915e11cb29fa2914
1 //===- AMDGPURewriteOutArgumentsPass.cpp - Create struct returns ----------===//
2 //
3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 // See https://llvm.org/LICENSE.txt for license information.
5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 //
7 //===----------------------------------------------------------------------===//
8 //
9 /// \file This pass attempts to replace out argument usage with a return of a
10 /// struct.
11 ///
12 /// We can support returning a lot of values directly in registers, but
13 /// idiomatic C code frequently uses a pointer argument to return a second value
14 /// rather than returning a struct by value. GPU stack access is also quite
15 /// painful, so we want to avoid that if possible. Passing a stack object
16 /// pointer to a function also requires an additional address expansion code
17 /// sequence to convert the pointer to be relative to the kernel's scratch wave
18 /// offset register since the callee doesn't know what stack frame the incoming
19 /// pointer is relative to.
20 ///
21 /// The goal is to try rewriting code that looks like this:
22 ///
23 /// int foo(int a, int b, int* out) {
24 /// *out = bar();
25 /// return a + b;
26 /// }
27 ///
28 /// into something like this:
29 ///
30 /// std::pair<int, int> foo(int a, int b) {
31 /// return std::make_pair(a + b, bar());
32 /// }
33 ///
34 /// Typically the incoming pointer is a simple alloca for a temporary variable
35 /// to use the API, which if replaced with a struct return will be easily SROA'd
36 /// out when the stub function we create is inlined
37 ///
38 /// This pass introduces the struct return, but leaves the unused pointer
39 /// arguments and introduces a new stub function calling the struct returning
40 /// body. DeadArgumentElimination should be run after this to clean these up.
42 //===----------------------------------------------------------------------===//
44 #include "AMDGPU.h"
45 #include "Utils/AMDGPUBaseInfo.h"
46 #include "llvm/Analysis/MemoryDependenceAnalysis.h"
47 #include "llvm/ADT/DenseMap.h"
48 #include "llvm/ADT/STLExtras.h"
49 #include "llvm/ADT/SmallSet.h"
50 #include "llvm/ADT/SmallVector.h"
51 #include "llvm/ADT/Statistic.h"
52 #include "llvm/Analysis/MemoryLocation.h"
53 #include "llvm/IR/Argument.h"
54 #include "llvm/IR/Attributes.h"
55 #include "llvm/IR/BasicBlock.h"
56 #include "llvm/IR/Constants.h"
57 #include "llvm/IR/DataLayout.h"
58 #include "llvm/IR/DerivedTypes.h"
59 #include "llvm/IR/Function.h"
60 #include "llvm/IR/IRBuilder.h"
61 #include "llvm/IR/Instructions.h"
62 #include "llvm/IR/Module.h"
63 #include "llvm/IR/Type.h"
64 #include "llvm/IR/Use.h"
65 #include "llvm/IR/User.h"
66 #include "llvm/IR/Value.h"
67 #include "llvm/Pass.h"
68 #include "llvm/Support/Casting.h"
69 #include "llvm/Support/CommandLine.h"
70 #include "llvm/Support/Debug.h"
71 #include "llvm/Support/raw_ostream.h"
72 #include <cassert>
73 #include <utility>
75 #define DEBUG_TYPE "amdgpu-rewrite-out-arguments"
77 using namespace llvm;
79 static cl::opt<bool> AnyAddressSpace(
80 "amdgpu-any-address-space-out-arguments",
81 cl::desc("Replace pointer out arguments with "
82 "struct returns for non-private address space"),
83 cl::Hidden,
84 cl::init(false));
86 static cl::opt<unsigned> MaxNumRetRegs(
87 "amdgpu-max-return-arg-num-regs",
88 cl::desc("Approximately limit number of return registers for replacing out arguments"),
89 cl::Hidden,
90 cl::init(16));
92 STATISTIC(NumOutArgumentsReplaced,
93 "Number out arguments moved to struct return values");
94 STATISTIC(NumOutArgumentFunctionsReplaced,
95 "Number of functions with out arguments moved to struct return values");
97 namespace {
99 class AMDGPURewriteOutArguments : public FunctionPass {
100 private:
101 const DataLayout *DL = nullptr;
102 MemoryDependenceResults *MDA = nullptr;
104 bool checkArgumentUses(Value &Arg) const;
105 bool isOutArgumentCandidate(Argument &Arg) const;
107 #ifndef NDEBUG
108 bool isVec3ToVec4Shuffle(Type *Ty0, Type* Ty1) const;
109 #endif
111 public:
112 static char ID;
114 AMDGPURewriteOutArguments() : FunctionPass(ID) {}
116 void getAnalysisUsage(AnalysisUsage &AU) const override {
117 AU.addRequired<MemoryDependenceWrapperPass>();
118 FunctionPass::getAnalysisUsage(AU);
121 bool doInitialization(Module &M) override;
122 bool runOnFunction(Function &F) override;
125 } // end anonymous namespace
127 INITIALIZE_PASS_BEGIN(AMDGPURewriteOutArguments, DEBUG_TYPE,
128 "AMDGPU Rewrite Out Arguments", false, false)
129 INITIALIZE_PASS_DEPENDENCY(MemoryDependenceWrapperPass)
130 INITIALIZE_PASS_END(AMDGPURewriteOutArguments, DEBUG_TYPE,
131 "AMDGPU Rewrite Out Arguments", false, false)
133 char AMDGPURewriteOutArguments::ID = 0;
135 bool AMDGPURewriteOutArguments::checkArgumentUses(Value &Arg) const {
136 const int MaxUses = 10;
137 int UseCount = 0;
139 for (Use &U : Arg.uses()) {
140 StoreInst *SI = dyn_cast<StoreInst>(U.getUser());
141 if (UseCount > MaxUses)
142 return false;
144 if (!SI) {
145 auto *BCI = dyn_cast<BitCastInst>(U.getUser());
146 if (!BCI || !BCI->hasOneUse())
147 return false;
149 // We don't handle multiple stores currently, so stores to aggregate
150 // pointers aren't worth the trouble since they are canonically split up.
151 Type *DestEltTy = BCI->getType()->getPointerElementType();
152 if (DestEltTy->isAggregateType())
153 return false;
155 // We could handle these if we had a convenient way to bitcast between
156 // them.
157 Type *SrcEltTy = Arg.getType()->getPointerElementType();
158 if (SrcEltTy->isArrayTy())
159 return false;
161 // Special case handle structs with single members. It is useful to handle
162 // some casts between structs and non-structs, but we can't bitcast
163 // directly between them. directly bitcast between them. Blender uses
164 // some casts that look like { <3 x float> }* to <4 x float>*
165 if ((SrcEltTy->isStructTy() && (SrcEltTy->getStructNumElements() != 1)))
166 return false;
168 // Clang emits OpenCL 3-vector type accesses with a bitcast to the
169 // equivalent 4-element vector and accesses that, and we're looking for
170 // this pointer cast.
171 if (DL->getTypeAllocSize(SrcEltTy) != DL->getTypeAllocSize(DestEltTy))
172 return false;
174 return checkArgumentUses(*BCI);
177 if (!SI->isSimple() ||
178 U.getOperandNo() != StoreInst::getPointerOperandIndex())
179 return false;
181 ++UseCount;
184 // Skip unused arguments.
185 return UseCount > 0;
188 bool AMDGPURewriteOutArguments::isOutArgumentCandidate(Argument &Arg) const {
189 const unsigned MaxOutArgSizeBytes = 4 * MaxNumRetRegs;
190 PointerType *ArgTy = dyn_cast<PointerType>(Arg.getType());
192 // TODO: It might be useful for any out arguments, not just privates.
193 if (!ArgTy || (ArgTy->getAddressSpace() != DL->getAllocaAddrSpace() &&
194 !AnyAddressSpace) ||
195 Arg.hasByValAttr() || Arg.hasStructRetAttr() ||
196 DL->getTypeStoreSize(ArgTy->getPointerElementType()) > MaxOutArgSizeBytes) {
197 return false;
200 return checkArgumentUses(Arg);
203 bool AMDGPURewriteOutArguments::doInitialization(Module &M) {
204 DL = &M.getDataLayout();
205 return false;
208 #ifndef NDEBUG
209 bool AMDGPURewriteOutArguments::isVec3ToVec4Shuffle(Type *Ty0, Type* Ty1) const {
210 VectorType *VT0 = dyn_cast<VectorType>(Ty0);
211 VectorType *VT1 = dyn_cast<VectorType>(Ty1);
212 if (!VT0 || !VT1)
213 return false;
215 if (VT0->getNumElements() != 3 ||
216 VT1->getNumElements() != 4)
217 return false;
219 return DL->getTypeSizeInBits(VT0->getElementType()) ==
220 DL->getTypeSizeInBits(VT1->getElementType());
222 #endif
224 bool AMDGPURewriteOutArguments::runOnFunction(Function &F) {
225 if (skipFunction(F))
226 return false;
228 // TODO: Could probably handle variadic functions.
229 if (F.isVarArg() || F.hasStructRetAttr() ||
230 AMDGPU::isEntryFunctionCC(F.getCallingConv()))
231 return false;
233 MDA = &getAnalysis<MemoryDependenceWrapperPass>().getMemDep();
235 unsigned ReturnNumRegs = 0;
236 SmallSet<int, 4> OutArgIndexes;
237 SmallVector<Type *, 4> ReturnTypes;
238 Type *RetTy = F.getReturnType();
239 if (!RetTy->isVoidTy()) {
240 ReturnNumRegs = DL->getTypeStoreSize(RetTy) / 4;
242 if (ReturnNumRegs >= MaxNumRetRegs)
243 return false;
245 ReturnTypes.push_back(RetTy);
248 SmallVector<Argument *, 4> OutArgs;
249 for (Argument &Arg : F.args()) {
250 if (isOutArgumentCandidate(Arg)) {
251 LLVM_DEBUG(dbgs() << "Found possible out argument " << Arg
252 << " in function " << F.getName() << '\n');
253 OutArgs.push_back(&Arg);
257 if (OutArgs.empty())
258 return false;
260 using ReplacementVec = SmallVector<std::pair<Argument *, Value *>, 4>;
262 DenseMap<ReturnInst *, ReplacementVec> Replacements;
264 SmallVector<ReturnInst *, 4> Returns;
265 for (BasicBlock &BB : F) {
266 if (ReturnInst *RI = dyn_cast<ReturnInst>(&BB.back()))
267 Returns.push_back(RI);
270 if (Returns.empty())
271 return false;
273 bool Changing;
275 do {
276 Changing = false;
278 // Keep retrying if we are able to successfully eliminate an argument. This
279 // helps with cases with multiple arguments which may alias, such as in a
280 // sincos implemntation. If we have 2 stores to arguments, on the first
281 // attempt the MDA query will succeed for the second store but not the
282 // first. On the second iteration we've removed that out clobbering argument
283 // (by effectively moving it into another function) and will find the second
284 // argument is OK to move.
285 for (Argument *OutArg : OutArgs) {
286 bool ThisReplaceable = true;
287 SmallVector<std::pair<ReturnInst *, StoreInst *>, 4> ReplaceableStores;
289 Type *ArgTy = OutArg->getType()->getPointerElementType();
291 // Skip this argument if converting it will push us over the register
292 // count to return limit.
294 // TODO: This is an approximation. When legalized this could be more. We
295 // can ask TLI for exactly how many.
296 unsigned ArgNumRegs = DL->getTypeStoreSize(ArgTy) / 4;
297 if (ArgNumRegs + ReturnNumRegs > MaxNumRetRegs)
298 continue;
300 // An argument is convertible only if all exit blocks are able to replace
301 // it.
302 for (ReturnInst *RI : Returns) {
303 BasicBlock *BB = RI->getParent();
305 MemDepResult Q = MDA->getPointerDependencyFrom(MemoryLocation(OutArg),
306 true, BB->end(), BB, RI);
307 StoreInst *SI = nullptr;
308 if (Q.isDef())
309 SI = dyn_cast<StoreInst>(Q.getInst());
311 if (SI) {
312 LLVM_DEBUG(dbgs() << "Found out argument store: " << *SI << '\n');
313 ReplaceableStores.emplace_back(RI, SI);
314 } else {
315 ThisReplaceable = false;
316 break;
320 if (!ThisReplaceable)
321 continue; // Try the next argument candidate.
323 for (std::pair<ReturnInst *, StoreInst *> Store : ReplaceableStores) {
324 Value *ReplVal = Store.second->getValueOperand();
326 auto &ValVec = Replacements[Store.first];
327 if (llvm::find_if(ValVec,
328 [OutArg](const std::pair<Argument *, Value *> &Entry) {
329 return Entry.first == OutArg;}) != ValVec.end()) {
330 LLVM_DEBUG(dbgs()
331 << "Saw multiple out arg stores" << *OutArg << '\n');
332 // It is possible to see stores to the same argument multiple times,
333 // but we expect these would have been optimized out already.
334 ThisReplaceable = false;
335 break;
338 ValVec.emplace_back(OutArg, ReplVal);
339 Store.second->eraseFromParent();
342 if (ThisReplaceable) {
343 ReturnTypes.push_back(ArgTy);
344 OutArgIndexes.insert(OutArg->getArgNo());
345 ++NumOutArgumentsReplaced;
346 Changing = true;
349 } while (Changing);
351 if (Replacements.empty())
352 return false;
354 LLVMContext &Ctx = F.getParent()->getContext();
355 StructType *NewRetTy = StructType::create(Ctx, ReturnTypes, F.getName());
357 FunctionType *NewFuncTy = FunctionType::get(NewRetTy,
358 F.getFunctionType()->params(),
359 F.isVarArg());
361 LLVM_DEBUG(dbgs() << "Computed new return type: " << *NewRetTy << '\n');
363 Function *NewFunc = Function::Create(NewFuncTy, Function::PrivateLinkage,
364 F.getName() + ".body");
365 F.getParent()->getFunctionList().insert(F.getIterator(), NewFunc);
366 NewFunc->copyAttributesFrom(&F);
367 NewFunc->setComdat(F.getComdat());
369 // We want to preserve the function and param attributes, but need to strip
370 // off any return attributes, e.g. zeroext doesn't make sense with a struct.
371 NewFunc->stealArgumentListFrom(F);
373 AttrBuilder RetAttrs;
374 RetAttrs.addAttribute(Attribute::SExt);
375 RetAttrs.addAttribute(Attribute::ZExt);
376 RetAttrs.addAttribute(Attribute::NoAlias);
377 NewFunc->removeAttributes(AttributeList::ReturnIndex, RetAttrs);
378 // TODO: How to preserve metadata?
380 // Move the body of the function into the new rewritten function, and replace
381 // this function with a stub.
382 NewFunc->getBasicBlockList().splice(NewFunc->begin(), F.getBasicBlockList());
384 for (std::pair<ReturnInst *, ReplacementVec> &Replacement : Replacements) {
385 ReturnInst *RI = Replacement.first;
386 IRBuilder<> B(RI);
387 B.SetCurrentDebugLocation(RI->getDebugLoc());
389 int RetIdx = 0;
390 Value *NewRetVal = UndefValue::get(NewRetTy);
392 Value *RetVal = RI->getReturnValue();
393 if (RetVal)
394 NewRetVal = B.CreateInsertValue(NewRetVal, RetVal, RetIdx++);
396 for (std::pair<Argument *, Value *> ReturnPoint : Replacement.second) {
397 Argument *Arg = ReturnPoint.first;
398 Value *Val = ReturnPoint.second;
399 Type *EltTy = Arg->getType()->getPointerElementType();
400 if (Val->getType() != EltTy) {
401 Type *EffectiveEltTy = EltTy;
402 if (StructType *CT = dyn_cast<StructType>(EltTy)) {
403 assert(CT->getNumElements() == 1);
404 EffectiveEltTy = CT->getElementType(0);
407 if (DL->getTypeSizeInBits(EffectiveEltTy) !=
408 DL->getTypeSizeInBits(Val->getType())) {
409 assert(isVec3ToVec4Shuffle(EffectiveEltTy, Val->getType()));
410 Val = B.CreateShuffleVector(Val, UndefValue::get(Val->getType()),
411 { 0, 1, 2 });
414 Val = B.CreateBitCast(Val, EffectiveEltTy);
416 // Re-create single element composite.
417 if (EltTy != EffectiveEltTy)
418 Val = B.CreateInsertValue(UndefValue::get(EltTy), Val, 0);
421 NewRetVal = B.CreateInsertValue(NewRetVal, Val, RetIdx++);
424 if (RetVal)
425 RI->setOperand(0, NewRetVal);
426 else {
427 B.CreateRet(NewRetVal);
428 RI->eraseFromParent();
432 SmallVector<Value *, 16> StubCallArgs;
433 for (Argument &Arg : F.args()) {
434 if (OutArgIndexes.count(Arg.getArgNo())) {
435 // It's easier to preserve the type of the argument list. We rely on
436 // DeadArgumentElimination to take care of these.
437 StubCallArgs.push_back(UndefValue::get(Arg.getType()));
438 } else {
439 StubCallArgs.push_back(&Arg);
443 BasicBlock *StubBB = BasicBlock::Create(Ctx, "", &F);
444 IRBuilder<> B(StubBB);
445 CallInst *StubCall = B.CreateCall(NewFunc, StubCallArgs);
447 int RetIdx = RetTy->isVoidTy() ? 0 : 1;
448 for (Argument &Arg : F.args()) {
449 if (!OutArgIndexes.count(Arg.getArgNo()))
450 continue;
452 PointerType *ArgType = cast<PointerType>(Arg.getType());
454 auto *EltTy = ArgType->getElementType();
455 unsigned Align = Arg.getParamAlignment();
456 if (Align == 0)
457 Align = DL->getABITypeAlignment(EltTy);
459 Value *Val = B.CreateExtractValue(StubCall, RetIdx++);
460 Type *PtrTy = Val->getType()->getPointerTo(ArgType->getAddressSpace());
462 // We can peek through bitcasts, so the type may not match.
463 Value *PtrVal = B.CreateBitCast(&Arg, PtrTy);
465 B.CreateAlignedStore(Val, PtrVal, Align);
468 if (!RetTy->isVoidTy()) {
469 B.CreateRet(B.CreateExtractValue(StubCall, 0));
470 } else {
471 B.CreateRetVoid();
474 // The function is now a stub we want to inline.
475 F.addFnAttr(Attribute::AlwaysInline);
477 ++NumOutArgumentFunctionsReplaced;
478 return true;
481 FunctionPass *llvm::createAMDGPURewriteOutArgumentsPass() {
482 return new AMDGPURewriteOutArguments();