1//===- TypeBasedAliasAnalysis.cpp - Type-Based Alias Analysis -------------===//
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// This file defines the TypeBasedAliasAnalysis pass, which implements
10// metadata-based TBAA.
11//
12// In LLVM IR, memory does not have types, so LLVM's own type system is not
13// suitable for doing TBAA. Instead, metadata is added to the IR to describe
14// a type system of a higher level language. This can be used to implement
15// typical C/C++ TBAA, but it can also be used to implement custom alias
16// analysis behavior for other languages.
17//
18// Scalar type nodes have up to three fields, e.g.:
19// !0 = !{ !"an example type tree" }
20// !1 = !{ !"int", !0 }
21// !2 = !{ !"float", !0 }
22// !3 = !{ !"const float", !2, i64 1 }
23//
24// The first field is an identity field. It can be any value, usually
25// an MDString, which uniquely identifies the type. The most important
26// name in the tree is the name of the root node. Two trees with
27// different root node names are entirely disjoint, even if they
28// have leaves with common names.
29//
30// The second field identifies the type's parent node in the tree, or
31// is null or omitted for a root node. A type is considered to alias
32// all of its descendants and all of its ancestors in the tree. Also,
33// a type is considered to alias all types in other trees, so that
34// bitcode produced from multiple front-ends is handled conservatively.
35//
36// If the third field is present, it's an integer which if equal to 1
37// indicates that the type is "constant" (meaning pointsToConstantMemory
38// should return true; see
39// http://llvm.org/docs/AliasAnalysis.html#OtherItfs).
40//
41// The MDNodes attached to an instruction using "!tbaa" are not plain type
42// nodes, but path tag nodes.
43//
44// The path tag node has 4 fields with the last field being optional.
45//
46// The first field is the base type node, it can be a struct type node
47// or a scalar type node. The second field is the access type node, it
48// must be a scalar type node. The third field is the offset into the base type.
49// The last field has the same meaning as the last field of our scalar TBAA:
50// it's an integer which if equal to 1 indicates that the access is "constant".
51//
52// The struct type node has a name and a list of pairs, one pair for each member
53// of the struct. The first element of each pair is a type node (a struct type
54// node or a scalar type node), specifying the type of the member, the second
55// element of each pair is the offset of the member.
56//
57// Given an example
58// typedef struct {
59// short s;
60// } A;
61// typedef struct {
62// uint16_t s;
63// A a;
64// } B;
65//
66// For an access to B.a.s, we attach !5 (a path tag node) to the load/store
67// instruction. The base type is !4 (struct B), the access type is !2 (scalar
68// type short) and the offset is 4.
69//
70// !0 = !{!"Simple C/C++ TBAA"}
71// !1 = !{!"omnipotent char", !0} // Scalar type node
72// !2 = !{!"short", !1} // Scalar type node
73// !3 = !{!"A", !2, i64 0} // Struct type node
74// !4 = !{!"B", !2, i64 0, !3, i64 4}
75// // Struct type node
76// !5 = !{!4, !2, i64 4} // Path tag node
77//
78// The struct type nodes and the scalar type nodes form a type DAG.
79// Root (!0)
80// char (!1) -- edge to Root
81// short (!2) -- edge to char
82// A (!3) -- edge with offset 0 to short
83// B (!4) -- edge with offset 0 to short and edge with offset 4 to A
84//
85// To check if two tags (tagX and tagY) can alias, we start from the base type
86// of tagX, follow the edge with the correct offset in the type DAG and adjust
87// the offset until we reach the base type of tagY or until we reach the Root
88// node.
89// If we reach the base type of tagY, compare the adjusted offset with
90// offset of tagY, return Alias if the offsets are the same, return NoAlias
91// otherwise.
92// If we reach the Root node, perform the above starting from base type of tagY
93// to see if we reach base type of tagX.
94//
95// If they have different roots, they're part of different potentially
96// unrelated type systems, so we return Alias to be conservative.
97// If neither node is an ancestor of the other and they have the same root,
98// then we say NoAlias.
99//
100//===----------------------------------------------------------------------===//
101
102#include "llvm/Analysis/TypeBasedAliasAnalysis.h"
103#include "llvm/ADT/SetVector.h"
104#include "llvm/Analysis/AliasAnalysis.h"
105#include "llvm/Analysis/MemoryLocation.h"
106#include "llvm/IR/Constants.h"
107#include "llvm/IR/DataLayout.h"
108#include "llvm/IR/DerivedTypes.h"
109#include "llvm/IR/InstrTypes.h"
110#include "llvm/IR/LLVMContext.h"
111#include "llvm/IR/Metadata.h"
112#include "llvm/IR/Module.h"
113#include "llvm/InitializePasses.h"
114#include "llvm/Pass.h"
115#include "llvm/Support/Casting.h"
116#include "llvm/Support/CommandLine.h"
117#include "llvm/Support/ErrorHandling.h"
118#include <cassert>
119#include <cstdint>
120
121using namespace llvm;
122
123// A handy option for disabling TBAA functionality. The same effect can also be
124// achieved by stripping the !tbaa tags from IR, but this option is sometimes
125// more convenient.
126static cl::opt<bool> EnableTBAA("enable-tbaa", cl::init(Val: true), cl::Hidden);
127
128namespace {
129
130/// isNewFormatTypeNode - Return true iff the given type node is in the new
131/// size-aware format.
132static bool isNewFormatTypeNode(const MDNode *N) {
133 if (N->getNumOperands() < 3)
134 return false;
135 // In the old format the first operand is a string.
136 if (!isa<MDNode>(Val: N->getOperand(I: 0)))
137 return false;
138 return true;
139}
140
141/// This is a simple wrapper around an MDNode which provides a higher-level
142/// interface by hiding the details of how alias analysis information is encoded
143/// in its operands.
144template<typename MDNodeTy>
145class TBAANodeImpl {
146 MDNodeTy *Node = nullptr;
147
148public:
149 TBAANodeImpl() = default;
150 explicit TBAANodeImpl(MDNodeTy *N) : Node(N) {}
151
152 /// getNode - Get the MDNode for this TBAANode.
153 MDNodeTy *getNode() const { return Node; }
154
155 /// isNewFormat - Return true iff the wrapped type node is in the new
156 /// size-aware format.
157 bool isNewFormat() const { return isNewFormatTypeNode(Node); }
158
159 /// getParent - Get this TBAANode's Alias tree parent.
160 TBAANodeImpl<MDNodeTy> getParent() const {
161 if (isNewFormat())
162 return TBAANodeImpl(cast<MDNodeTy>(Node->getOperand(0)));
163
164 if (Node->getNumOperands() < 2)
165 return TBAANodeImpl<MDNodeTy>();
166 MDNodeTy *P = dyn_cast_or_null<MDNodeTy>(Node->getOperand(1));
167 if (!P)
168 return TBAANodeImpl<MDNodeTy>();
169 // Ok, this node has a valid parent. Return it.
170 return TBAANodeImpl<MDNodeTy>(P);
171 }
172
173 /// Test if this TBAANode represents a type for objects which are
174 /// not modified (by any means) in the context where this
175 /// AliasAnalysis is relevant.
176 bool isTypeImmutable() const {
177 if (Node->getNumOperands() < 3)
178 return false;
179 ConstantInt *CI = mdconst::dyn_extract<ConstantInt>(Node->getOperand(2));
180 if (!CI)
181 return false;
182 return CI->getValue()[0];
183 }
184};
185
186/// \name Specializations of \c TBAANodeImpl for const and non const qualified
187/// \c MDNode.
188/// @{
189using TBAANode = TBAANodeImpl<const MDNode>;
190using MutableTBAANode = TBAANodeImpl<MDNode>;
191/// @}
192
193/// This is a simple wrapper around an MDNode which provides a
194/// higher-level interface by hiding the details of how alias analysis
195/// information is encoded in its operands.
196template<typename MDNodeTy>
197class TBAAStructTagNodeImpl {
198 /// This node should be created with createTBAAAccessTag().
199 MDNodeTy *Node;
200
201public:
202 explicit TBAAStructTagNodeImpl(MDNodeTy *N) : Node(N) {}
203
204 /// Get the MDNode for this TBAAStructTagNode.
205 MDNodeTy *getNode() const { return Node; }
206
207 /// isNewFormat - Return true iff the wrapped access tag is in the new
208 /// size-aware format.
209 bool isNewFormat() const {
210 if (Node->getNumOperands() < 4)
211 return false;
212 if (MDNodeTy *AccessType = getAccessType())
213 if (!TBAANodeImpl<MDNodeTy>(AccessType).isNewFormat())
214 return false;
215 return true;
216 }
217
218 MDNodeTy *getBaseType() const {
219 return dyn_cast_or_null<MDNode>(Node->getOperand(0));
220 }
221
222 MDNodeTy *getAccessType() const {
223 return dyn_cast_or_null<MDNode>(Node->getOperand(1));
224 }
225
226 uint64_t getOffset() const {
227 return mdconst::extract<ConstantInt>(Node->getOperand(2))->getZExtValue();
228 }
229
230 uint64_t getSize() const {
231 if (!isNewFormat())
232 return UINT64_MAX;
233 return mdconst::extract<ConstantInt>(Node->getOperand(3))->getZExtValue();
234 }
235
236 /// Test if this TBAAStructTagNode represents a type for objects
237 /// which are not modified (by any means) in the context where this
238 /// AliasAnalysis is relevant.
239 bool isTypeImmutable() const {
240 unsigned OpNo = isNewFormat() ? 4 : 3;
241 if (Node->getNumOperands() < OpNo + 1)
242 return false;
243 ConstantInt *CI = mdconst::dyn_extract<ConstantInt>(Node->getOperand(OpNo));
244 if (!CI)
245 return false;
246 return CI->getValue()[0];
247 }
248};
249
250/// \name Specializations of \c TBAAStructTagNodeImpl for const and non const
251/// qualified \c MDNods.
252/// @{
253using TBAAStructTagNode = TBAAStructTagNodeImpl<const MDNode>;
254using MutableTBAAStructTagNode = TBAAStructTagNodeImpl<MDNode>;
255/// @}
256
257/// This is a simple wrapper around an MDNode which provides a
258/// higher-level interface by hiding the details of how alias analysis
259/// information is encoded in its operands.
260class TBAAStructTypeNode {
261 /// This node should be created with createTBAATypeNode().
262 const MDNode *Node = nullptr;
263
264public:
265 TBAAStructTypeNode() = default;
266 explicit TBAAStructTypeNode(const MDNode *N) : Node(N) {}
267
268 /// Get the MDNode for this TBAAStructTypeNode.
269 const MDNode *getNode() const { return Node; }
270
271 /// isNewFormat - Return true iff the wrapped type node is in the new
272 /// size-aware format.
273 bool isNewFormat() const { return isNewFormatTypeNode(N: Node); }
274
275 bool operator==(const TBAAStructTypeNode &Other) const {
276 return getNode() == Other.getNode();
277 }
278
279 /// getId - Return type identifier.
280 Metadata *getId() const {
281 return Node->getOperand(I: isNewFormat() ? 2 : 0);
282 }
283
284 unsigned getNumFields() const {
285 unsigned FirstFieldOpNo = isNewFormat() ? 3 : 1;
286 unsigned NumOpsPerField = isNewFormat() ? 3 : 2;
287 return (getNode()->getNumOperands() - FirstFieldOpNo) / NumOpsPerField;
288 }
289
290 TBAAStructTypeNode getFieldType(unsigned FieldIndex) const {
291 unsigned FirstFieldOpNo = isNewFormat() ? 3 : 1;
292 unsigned NumOpsPerField = isNewFormat() ? 3 : 2;
293 unsigned OpIndex = FirstFieldOpNo + FieldIndex * NumOpsPerField;
294 auto *TypeNode = cast<MDNode>(Val: getNode()->getOperand(I: OpIndex));
295 return TBAAStructTypeNode(TypeNode);
296 }
297
298 /// Get this TBAAStructTypeNode's field in the type DAG with
299 /// given offset. Update the offset to be relative to the field type.
300 TBAAStructTypeNode getField(uint64_t &Offset) const {
301 bool NewFormat = isNewFormat();
302 const ArrayRef<MDOperand> Operands = Node->operands();
303 const unsigned NumOperands = Operands.size();
304
305 if (NewFormat) {
306 // New-format root and scalar type nodes have no fields.
307 if (NumOperands < 6)
308 return TBAAStructTypeNode();
309 } else {
310 // Parent can be omitted for the root node.
311 if (NumOperands < 2)
312 return TBAAStructTypeNode();
313
314 // Fast path for a scalar type node and a struct type node with a single
315 // field.
316 if (NumOperands <= 3) {
317 uint64_t Cur =
318 NumOperands == 2
319 ? 0
320 : mdconst::extract<ConstantInt>(MD: Operands[2])->getZExtValue();
321 Offset -= Cur;
322 MDNode *P = dyn_cast_or_null<MDNode>(Val: Operands[1]);
323 if (!P)
324 return TBAAStructTypeNode();
325 return TBAAStructTypeNode(P);
326 }
327 }
328
329 // Assume the offsets are in order. We return the previous field if
330 // the current offset is bigger than the given offset.
331 unsigned FirstFieldOpNo = NewFormat ? 3 : 1;
332 unsigned NumOpsPerField = NewFormat ? 3 : 2;
333 unsigned TheIdx = 0;
334
335 for (unsigned Idx = FirstFieldOpNo; Idx < NumOperands;
336 Idx += NumOpsPerField) {
337 uint64_t Cur =
338 mdconst::extract<ConstantInt>(MD: Operands[Idx + 1])->getZExtValue();
339 if (Cur > Offset) {
340 assert(Idx >= FirstFieldOpNo + NumOpsPerField &&
341 "TBAAStructTypeNode::getField should have an offset match!");
342 TheIdx = Idx - NumOpsPerField;
343 break;
344 }
345 }
346 // Move along the last field.
347 if (TheIdx == 0)
348 TheIdx = NumOperands - NumOpsPerField;
349 uint64_t Cur =
350 mdconst::extract<ConstantInt>(MD: Operands[TheIdx + 1])->getZExtValue();
351 Offset -= Cur;
352 MDNode *P = dyn_cast_or_null<MDNode>(Val: Operands[TheIdx]);
353 if (!P)
354 return TBAAStructTypeNode();
355 return TBAAStructTypeNode(P);
356 }
357};
358
359} // end anonymous namespace
360
361AliasResult TypeBasedAAResult::alias(const MemoryLocation &LocA,
362 const MemoryLocation &LocB,
363 AAQueryInfo &AAQI, const Instruction *) {
364 if (!shouldUseTBAA())
365 return AliasResult::MayAlias;
366
367 if (Aliases(A: LocA.AATags.TBAA, B: LocB.AATags.TBAA))
368 return AliasResult::MayAlias;
369
370 // Otherwise return a definitive result.
371 return AliasResult::NoAlias;
372}
373
374AliasResult TypeBasedAAResult::aliasErrno(const MemoryLocation &Loc,
375 const Instruction *CtxI) {
376 if (!shouldUseTBAA())
377 return AliasResult::MayAlias;
378
379 const auto *N = Loc.AATags.TBAA;
380 if (!N)
381 return AliasResult::MayAlias;
382
383 // There cannot be any alias with errno if TBAA proves the given memory
384 // location does not alias errno.
385 const auto *ErrnoTBAAMD =
386 CtxI->getModule()->getNamedMetadata(Name: "llvm.errno.tbaa");
387 if (!ErrnoTBAAMD || any_of(Range: ErrnoTBAAMD->operands(), P: [&](const auto *Node) {
388 return Aliases(A: N, B: Node);
389 }))
390 return AliasResult::MayAlias;
391 return AliasResult::NoAlias;
392}
393
394ModRefInfo TypeBasedAAResult::getModRefInfoMask(const MemoryLocation &Loc,
395 AAQueryInfo &AAQI,
396 bool IgnoreLocals) {
397 if (!shouldUseTBAA())
398 return ModRefInfo::ModRef;
399
400 const MDNode *M = Loc.AATags.TBAA;
401 if (!M)
402 return ModRefInfo::ModRef;
403
404 // If this is an "immutable" type, we can assume the pointer is pointing
405 // to constant memory.
406 if (TBAAStructTagNode(M).isTypeImmutable())
407 return ModRefInfo::NoModRef;
408
409 return ModRefInfo::ModRef;
410}
411
412MemoryEffects TypeBasedAAResult::getMemoryEffects(const CallBase *Call,
413 AAQueryInfo &AAQI) {
414 if (!shouldUseTBAA())
415 return MemoryEffects::unknown();
416
417 // If this is an "immutable" type, the access is not observable.
418 if (const MDNode *M = Call->getMetadata(KindID: LLVMContext::MD_tbaa))
419 if (TBAAStructTagNode(M).isTypeImmutable())
420 return MemoryEffects::none();
421
422 return MemoryEffects::unknown();
423}
424
425MemoryEffects TypeBasedAAResult::getMemoryEffects(const Function *F) {
426 // Functions don't have metadata.
427 return MemoryEffects::unknown();
428}
429
430ModRefInfo TypeBasedAAResult::getModRefInfo(const CallBase *Call,
431 const MemoryLocation &Loc,
432 AAQueryInfo &AAQI) {
433 if (!shouldUseTBAA())
434 return ModRefInfo::ModRef;
435
436 if (const MDNode *L = Loc.AATags.TBAA)
437 if (const MDNode *M = Call->getMetadata(KindID: LLVMContext::MD_tbaa))
438 if (!Aliases(A: L, B: M))
439 return ModRefInfo::NoModRef;
440
441 return ModRefInfo::ModRef;
442}
443
444ModRefInfo TypeBasedAAResult::getModRefInfo(const CallBase *Call1,
445 const CallBase *Call2,
446 AAQueryInfo &AAQI) {
447 if (!shouldUseTBAA())
448 return ModRefInfo::ModRef;
449
450 if (const MDNode *M1 = Call1->getMetadata(KindID: LLVMContext::MD_tbaa))
451 if (const MDNode *M2 = Call2->getMetadata(KindID: LLVMContext::MD_tbaa))
452 if (!Aliases(A: M1, B: M2))
453 return ModRefInfo::NoModRef;
454
455 return ModRefInfo::ModRef;
456}
457
458bool MDNode::isTBAAVtableAccess() const {
459 // For struct-path aware TBAA, we use the access type of the tag.
460 TBAAStructTagNode Tag(this);
461 TBAAStructTypeNode AccessType(Tag.getAccessType());
462 if(auto *Id = dyn_cast<MDString>(Val: AccessType.getId()))
463 if (Id->getString() == "vtable pointer")
464 return true;
465 return false;
466}
467
468static bool matchAccessTags(const MDNode *A, const MDNode *B,
469 const MDNode **GenericTag = nullptr);
470
471MDNode *MDNode::getMostGenericTBAA(MDNode *A, MDNode *B) {
472 const MDNode *GenericTag;
473 matchAccessTags(A, B, GenericTag: &GenericTag);
474 return const_cast<MDNode*>(GenericTag);
475}
476
477static const MDNode *getLeastCommonType(const MDNode *A, const MDNode *B) {
478 if (!A || !B)
479 return nullptr;
480
481 if (A == B)
482 return A;
483
484 SmallSetVector<const MDNode *, 4> PathA;
485 TBAANode TA(A);
486 while (TA.getNode()) {
487 if (!PathA.insert(X: TA.getNode()))
488 report_fatal_error(reason: "Cycle found in TBAA metadata.");
489 TA = TA.getParent();
490 }
491
492 SmallSetVector<const MDNode *, 4> PathB;
493 TBAANode TB(B);
494 while (TB.getNode()) {
495 if (!PathB.insert(X: TB.getNode()))
496 report_fatal_error(reason: "Cycle found in TBAA metadata.");
497 TB = TB.getParent();
498 }
499
500 int IA = PathA.size() - 1;
501 int IB = PathB.size() - 1;
502
503 const MDNode *Ret = nullptr;
504 while (IA >= 0 && IB >= 0) {
505 if (PathA[IA] == PathB[IB])
506 Ret = PathA[IA];
507 else
508 break;
509 --IA;
510 --IB;
511 }
512
513 return Ret;
514}
515
516AAMDNodes AAMDNodes::merge(const AAMDNodes &Other) const {
517 AAMDNodes Result;
518 Result.TBAA = MDNode::getMostGenericTBAA(A: TBAA, B: Other.TBAA);
519 Result.TBAAStruct = nullptr;
520 Result.Scope = MDNode::getMostGenericAliasScope(A: Scope, B: Other.Scope);
521 Result.NoAlias = MDNode::intersect(A: NoAlias, B: Other.NoAlias);
522 Result.NoAliasAddrSpace = MDNode::getMostGenericNoaliasAddrspace(
523 A: NoAliasAddrSpace, B: Other.NoAliasAddrSpace);
524 return Result;
525}
526
527AAMDNodes AAMDNodes::concat(const AAMDNodes &Other) const {
528 AAMDNodes Result;
529 Result.TBAA = Result.TBAAStruct = nullptr;
530 Result.Scope = MDNode::getMostGenericAliasScope(A: Scope, B: Other.Scope);
531 Result.NoAlias = MDNode::intersect(A: NoAlias, B: Other.NoAlias);
532 Result.NoAliasAddrSpace = MDNode::getMostGenericNoaliasAddrspace(
533 A: NoAliasAddrSpace, B: Other.NoAliasAddrSpace);
534 return Result;
535}
536
537static const MDNode *createAccessTag(const MDNode *AccessType) {
538 // If there is no access type or the access type is the root node, then
539 // we don't have any useful access tag to return.
540 if (!AccessType || AccessType->getNumOperands() < 2)
541 return nullptr;
542
543 Type *Int64 = IntegerType::get(C&: AccessType->getContext(), NumBits: 64);
544 auto *OffsetNode = ConstantAsMetadata::get(C: ConstantInt::get(Ty: Int64, V: 0));
545
546 if (TBAAStructTypeNode(AccessType).isNewFormat()) {
547 // TODO: Take access ranges into account when matching access tags and
548 // fix this code to generate actual access sizes for generic tags.
549 uint64_t AccessSize = UINT64_MAX;
550 auto *SizeNode =
551 ConstantAsMetadata::get(C: ConstantInt::get(Ty: Int64, V: AccessSize));
552 Metadata *Ops[] = {const_cast<MDNode*>(AccessType),
553 const_cast<MDNode*>(AccessType),
554 OffsetNode, SizeNode};
555 return MDNode::get(Context&: AccessType->getContext(), MDs: Ops);
556 }
557
558 Metadata *Ops[] = {const_cast<MDNode*>(AccessType),
559 const_cast<MDNode*>(AccessType),
560 OffsetNode};
561 return MDNode::get(Context&: AccessType->getContext(), MDs: Ops);
562}
563
564static bool hasField(TBAAStructTypeNode BaseType,
565 TBAAStructTypeNode FieldType) {
566 for (unsigned I = 0, E = BaseType.getNumFields(); I != E; ++I) {
567 TBAAStructTypeNode T = BaseType.getFieldType(FieldIndex: I);
568 if (T == FieldType || hasField(BaseType: T, FieldType))
569 return true;
570 }
571 return false;
572}
573
574/// Return true if for two given accesses, one of the accessed objects may be a
575/// subobject of the other. The \p BaseTag and \p SubobjectTag parameters
576/// describe the accesses to the base object and the subobject respectively.
577/// \p CommonType must be the metadata node describing the common type of the
578/// accessed objects. On return, \p MayAlias is set to true iff these accesses
579/// may alias and \p Generic, if not null, points to the most generic access
580/// tag for the given two.
581static bool mayBeAccessToSubobjectOf(TBAAStructTagNode BaseTag,
582 TBAAStructTagNode SubobjectTag,
583 const MDNode *CommonType,
584 const MDNode **GenericTag,
585 bool &MayAlias) {
586 // If the base object is of the least common type, then this may be an access
587 // to its subobject.
588 if (BaseTag.getAccessType() == BaseTag.getBaseType() &&
589 BaseTag.getAccessType() == CommonType) {
590 if (GenericTag)
591 *GenericTag = createAccessTag(AccessType: CommonType);
592 MayAlias = true;
593 return true;
594 }
595
596 // If the access to the base object is through a field of the subobject's
597 // type, then this may be an access to that field. To check for that we start
598 // from the base type, follow the edge with the correct offset in the type DAG
599 // and adjust the offset until we reach the field type or until we reach the
600 // access type.
601 bool NewFormat = BaseTag.isNewFormat();
602 TBAAStructTypeNode BaseType(BaseTag.getBaseType());
603 uint64_t OffsetInBase = BaseTag.getOffset();
604
605 for (;;) {
606 // In the old format there is no distinction between fields and parent
607 // types, so in this case we consider all nodes up to the root.
608 if (!BaseType.getNode()) {
609 assert(!NewFormat && "Did not see access type in access path!");
610 break;
611 }
612
613 if (BaseType.getNode() == SubobjectTag.getBaseType()) {
614 MayAlias = OffsetInBase == SubobjectTag.getOffset() ||
615 BaseType.getNode() == BaseTag.getAccessType() ||
616 SubobjectTag.getBaseType() == SubobjectTag.getAccessType();
617 if (GenericTag) {
618 *GenericTag =
619 MayAlias ? SubobjectTag.getNode() : createAccessTag(AccessType: CommonType);
620 }
621 return true;
622 }
623
624 // With new-format nodes we stop at the access type.
625 if (NewFormat && BaseType.getNode() == BaseTag.getAccessType())
626 break;
627
628 // Follow the edge with the correct offset. Offset will be adjusted to
629 // be relative to the field type.
630 BaseType = BaseType.getField(Offset&: OffsetInBase);
631 }
632
633 // If the base object has a direct or indirect field of the subobject's type,
634 // then this may be an access to that field. We need this to check now that
635 // we support aggregates as access types.
636 if (NewFormat) {
637 // TBAAStructTypeNode BaseAccessType(BaseTag.getAccessType());
638 TBAAStructTypeNode FieldType(SubobjectTag.getBaseType());
639 if (hasField(BaseType, FieldType)) {
640 if (GenericTag)
641 *GenericTag = createAccessTag(AccessType: CommonType);
642 MayAlias = true;
643 return true;
644 }
645 }
646
647 return false;
648}
649
650/// matchTags - Return true if the given couple of accesses are allowed to
651/// overlap. If \arg GenericTag is not null, then on return it points to the
652/// most generic access descriptor for the given two.
653static bool matchAccessTags(const MDNode *A, const MDNode *B,
654 const MDNode **GenericTag) {
655 if (A == B) {
656 if (GenericTag)
657 *GenericTag = A;
658 return true;
659 }
660
661 // Accesses with no TBAA information may alias with any other accesses.
662 if (!A || !B) {
663 if (GenericTag)
664 *GenericTag = nullptr;
665 return true;
666 }
667
668 TBAAStructTagNode TagA(A), TagB(B);
669 const MDNode *CommonType = getLeastCommonType(A: TagA.getAccessType(),
670 B: TagB.getAccessType());
671
672 // If the final access types have different roots, they're part of different
673 // potentially unrelated type systems, so we must be conservative.
674 if (!CommonType) {
675 if (GenericTag)
676 *GenericTag = nullptr;
677 return true;
678 }
679
680 // If one of the accessed objects may be a subobject of the other, then such
681 // accesses may alias.
682 bool MayAlias;
683 if (mayBeAccessToSubobjectOf(/* BaseTag= */ TagA, /* SubobjectTag= */ TagB,
684 CommonType, GenericTag, MayAlias) ||
685 mayBeAccessToSubobjectOf(/* BaseTag= */ TagB, /* SubobjectTag= */ TagA,
686 CommonType, GenericTag, MayAlias))
687 return MayAlias;
688
689 // Otherwise, we've proved there's no alias.
690 if (GenericTag)
691 *GenericTag = createAccessTag(AccessType: CommonType);
692 return false;
693}
694
695/// Aliases - Test whether the access represented by tag A may alias the
696/// access represented by tag B.
697bool TypeBasedAAResult::Aliases(const MDNode *A, const MDNode *B) const {
698 return matchAccessTags(A, B);
699}
700
701bool TypeBasedAAResult::shouldUseTBAA() const {
702 return EnableTBAA && !UsingTypeSanitizer;
703}
704
705AnalysisKey TypeBasedAA::Key;
706
707TypeBasedAAResult TypeBasedAA::run(Function &F, FunctionAnalysisManager &AM) {
708 return TypeBasedAAResult(F.hasFnAttribute(Kind: Attribute::SanitizeType));
709}
710
711char TypeBasedAAWrapperPass::ID = 0;
712INITIALIZE_PASS(TypeBasedAAWrapperPass, "tbaa", "Type-Based Alias Analysis",
713 false, true)
714
715ImmutablePass *llvm::createTypeBasedAAWrapperPass() {
716 return new TypeBasedAAWrapperPass();
717}
718
719TypeBasedAAWrapperPass::TypeBasedAAWrapperPass() : ImmutablePass(ID) {}
720
721bool TypeBasedAAWrapperPass::doInitialization(Module &M) {
722 Result.reset(p: new TypeBasedAAResult(/*UsingTypeSanitizer=*/false));
723 return false;
724}
725
726bool TypeBasedAAWrapperPass::doFinalization(Module &M) {
727 Result.reset();
728 return false;
729}
730
731void TypeBasedAAWrapperPass::getAnalysisUsage(AnalysisUsage &AU) const {
732 AU.setPreservesAll();
733}
734
735MDNode *AAMDNodes::shiftTBAA(MDNode *MD, size_t Offset) {
736 // Fast path if there's no offset
737 if (Offset == 0)
738 return MD;
739
740 // The correct behavior here is to add the offset into the TBAA
741 // struct node offset. The base type, however may not have defined
742 // a type at this additional offset, resulting in errors. Since
743 // this method is only used within a given load/store access
744 // the offset provided is only used to subdivide the previous load
745 // maintaining the validity of the previous TBAA.
746 //
747 // This, however, should be revisited in the future.
748 return MD;
749}
750
751MDNode *AAMDNodes::shiftTBAAStruct(MDNode *MD, size_t Offset) {
752 // Fast path if there's no offset
753 if (Offset == 0)
754 return MD;
755 SmallVector<Metadata *, 3> Sub;
756 for (size_t i = 0, size = MD->getNumOperands(); i < size; i += 3) {
757 ConstantInt *InnerOffset = mdconst::extract<ConstantInt>(MD: MD->getOperand(I: i));
758 ConstantInt *InnerSize =
759 mdconst::extract<ConstantInt>(MD: MD->getOperand(I: i + 1));
760 // Don't include any triples that aren't in bounds
761 if (InnerOffset->getZExtValue() + InnerSize->getZExtValue() <= Offset)
762 continue;
763
764 uint64_t NewSize = InnerSize->getZExtValue();
765 uint64_t NewOffset = InnerOffset->getZExtValue() - Offset;
766 if (InnerOffset->getZExtValue() < Offset) {
767 NewOffset = 0;
768 NewSize -= Offset - InnerOffset->getZExtValue();
769 }
770
771 // Shift the offset of the triple
772 Sub.push_back(Elt: ConstantAsMetadata::get(
773 C: ConstantInt::get(Ty: InnerOffset->getType(), V: NewOffset)));
774 Sub.push_back(Elt: ConstantAsMetadata::get(
775 C: ConstantInt::get(Ty: InnerSize->getType(), V: NewSize)));
776 Sub.push_back(Elt: MD->getOperand(I: i + 2));
777 }
778 return MDNode::get(Context&: MD->getContext(), MDs: Sub);
779}
780
781MDNode *AAMDNodes::extendToTBAA(MDNode *MD, ssize_t Len) {
782 // Fast path if 0-length
783 if (Len == 0)
784 return nullptr;
785
786 TBAAStructTagNode Tag(MD);
787
788 // Only new format TBAA has a size
789 if (!Tag.isNewFormat())
790 return MD;
791
792 // If unknown size, drop the TBAA.
793 if (Len == -1)
794 return nullptr;
795
796 // Otherwise, create TBAA with the new Len
797 ArrayRef<MDOperand> MDOperands = MD->operands();
798 SmallVector<Metadata *, 4> NextNodes(MDOperands);
799 ConstantInt *PreviousSize = mdconst::extract<ConstantInt>(MD&: NextNodes[3]);
800
801 // Don't create a new MDNode if it is the same length.
802 if (PreviousSize->equalsInt(V: Len))
803 return MD;
804
805 NextNodes[3] =
806 ConstantAsMetadata::get(C: ConstantInt::get(Ty: PreviousSize->getType(), V: Len));
807 return MDNode::get(Context&: MD->getContext(), MDs: NextNodes);
808}
809
810AAMDNodes AAMDNodes::adjustForAccess(unsigned AccessSize) {
811 AAMDNodes New = *this;
812 MDNode *M = New.TBAAStruct;
813 if (!New.TBAA && M && M->getNumOperands() >= 3 && M->getOperand(I: 0) &&
814 mdconst::hasa<ConstantInt>(MD: M->getOperand(I: 0)) &&
815 mdconst::extract<ConstantInt>(MD: M->getOperand(I: 0))->isZero() &&
816 M->getOperand(I: 1) && mdconst::hasa<ConstantInt>(MD: M->getOperand(I: 1)) &&
817 mdconst::extract<ConstantInt>(MD: M->getOperand(I: 1))->getValue() ==
818 AccessSize &&
819 M->getOperand(I: 2) && isa<MDNode>(Val: M->getOperand(I: 2)))
820 New.TBAA = cast<MDNode>(Val: M->getOperand(I: 2));
821
822 New.TBAAStruct = nullptr;
823 return New;
824}
825
826AAMDNodes AAMDNodes::adjustForAccess(size_t Offset, Type *AccessTy,
827 const DataLayout &DL) {
828 AAMDNodes New = shift(Offset);
829 if (!DL.typeSizeEqualsStoreSize(Ty: AccessTy))
830 return New;
831 TypeSize Size = DL.getTypeStoreSize(Ty: AccessTy);
832 if (Size.isScalable())
833 return New;
834
835 return New.adjustForAccess(AccessSize: Size.getKnownMinValue());
836}
837
838AAMDNodes AAMDNodes::adjustForAccess(size_t Offset, unsigned AccessSize) {
839 AAMDNodes New = shift(Offset);
840 return New.adjustForAccess(AccessSize);
841}
842