1//===- Chunks.h -------------------------------------------------*- C++ -*-===//
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#ifndef LLD_COFF_CHUNKS_H
10#define LLD_COFF_CHUNKS_H
11
12#include "Config.h"
13#include "InputFiles.h"
14#include "lld/Common/LLVM.h"
15#include "llvm/ADT/ArrayRef.h"
16#include "llvm/ADT/PointerIntPair.h"
17#include "llvm/ADT/iterator.h"
18#include "llvm/ADT/iterator_range.h"
19#include "llvm/MC/StringTableBuilder.h"
20#include "llvm/Object/COFF.h"
21#include "llvm/Object/WindowsMachineFlag.h"
22#include <utility>
23#include <vector>
24
25namespace lld::coff {
26
27using llvm::COFF::ImportDirectoryTableEntry;
28using llvm::object::chpe_range_type;
29using llvm::object::coff_relocation;
30using llvm::object::coff_section;
31using llvm::object::COFFSymbolRef;
32using llvm::object::SectionRef;
33
34class Baserel;
35class Defined;
36class DefinedImportData;
37class DefinedRegular;
38class ObjFile;
39class OutputSection;
40class RuntimePseudoReloc;
41class Symbol;
42
43// Mask for permissions (discardable, writable, readable, executable, etc).
44const uint32_t permMask = 0xFE000000;
45
46// Mask for section types (code, data, bss).
47const uint32_t typeMask = 0x000000E0;
48
49// The log base 2 of the largest section alignment, which is log2(8192), or 13.
50enum : unsigned { Log2MaxSectionAlignment = 13 };
51
52// A Chunk represents a chunk of data that will occupy space in the
53// output (if the resolver chose that). It may or may not be backed by
54// a section of an input file. It could be linker-created data, or
55// doesn't even have actual data (if common or bss).
56class Chunk {
57public:
58 enum Kind : uint8_t {
59 SectionKind,
60 SectionECKind,
61 OtherKind,
62 ImportThunkKind,
63 ECExportThunkKind
64 };
65 Kind kind() const { return chunkKind; }
66
67 // Returns the size of this chunk (even if this is a common or BSS.)
68 size_t getSize() const;
69
70 // Returns chunk alignment in power of two form. Value values are powers of
71 // two from 1 to 8192.
72 uint32_t getAlignment() const { return 1U << p2Align; }
73
74 // Update the chunk section alignment measured in bytes. Internally alignment
75 // is stored in log2.
76 void setAlignment(uint32_t align) {
77 // Treat zero byte alignment as 1 byte alignment.
78 align = align ? align : 1;
79 assert(llvm::isPowerOf2_32(align) && "alignment is not a power of 2");
80 p2Align = llvm::Log2_32(Value: align);
81 assert(p2Align <= Log2MaxSectionAlignment &&
82 "impossible requested alignment");
83 }
84
85 // Write this chunk to a mmap'ed file, assuming Buf is pointing to
86 // beginning of the file. Because this function may use RVA values
87 // of other chunks for relocations, you need to set them properly
88 // before calling this function.
89 void writeTo(uint8_t *buf) const;
90
91 // The writer sets and uses the addresses. In practice, PE images cannot be
92 // larger than 2GB. Chunks are always laid as part of the image, so Chunk RVAs
93 // can be stored with 32 bits.
94 uint32_t getRVA() const { return rva; }
95 void setRVA(uint64_t v) {
96 // This may truncate. The writer checks for overflow later.
97 rva = (uint32_t)v;
98 }
99
100 // Returns readable/writable/executable bits.
101 uint32_t getOutputCharacteristics() const;
102
103 // Returns the section name if this is a section chunk.
104 // It is illegal to call this function on non-section chunks.
105 StringRef getSectionName() const;
106
107 // An output section has pointers to chunks in the section, and each
108 // chunk has a back pointer to an output section.
109 void setOutputSectionIdx(uint16_t o) { osidx = o; }
110 uint16_t getOutputSectionIdx() const { return osidx; }
111
112 // Windows-specific.
113 // Collect all locations that contain absolute addresses for base relocations.
114 void getBaserels(std::vector<Baserel> *res);
115
116 // Returns a human-readable name of this chunk. Chunks are unnamed chunks of
117 // bytes, so this is used only for logging or debugging.
118 StringRef getDebugName() const;
119
120 // Return true if this file has the hotpatch flag set to true in the
121 // S_COMPILE3 record in codeview debug info. Also returns true for some thunks
122 // synthesized by the linker.
123 bool isHotPatchable() const;
124
125 MachineTypes getMachine() const;
126 llvm::Triple::ArchType getArch() const;
127 std::optional<chpe_range_type> getArm64ECRangeType() const;
128
129 // ARM64EC entry thunk associated with the chunk.
130 Defined *getEntryThunk() const;
131 void setEntryThunk(Defined *entryThunk);
132
133protected:
134 Chunk(Kind k = OtherKind) : chunkKind(k), hasData(true), p2Align(0) {}
135
136 const Kind chunkKind;
137
138public:
139 // Returns true if this has non-zero data. BSS chunks return
140 // false. If false is returned, the space occupied by this chunk
141 // will be filled with zeros. Corresponds to the
142 // IMAGE_SCN_CNT_UNINITIALIZED_DATA section characteristic bit.
143 uint8_t hasData : 1;
144
145public:
146 // The alignment of this chunk, stored in log2 form. The writer uses the
147 // value.
148 uint8_t p2Align : 7;
149
150 // The output section index for this chunk. The first valid section number is
151 // one.
152 uint16_t osidx = 0;
153
154 // The RVA of this chunk in the output. The writer sets a value.
155 uint32_t rva = 0;
156};
157
158class NonSectionChunk : public Chunk {
159public:
160 virtual ~NonSectionChunk() = default;
161
162 // Returns the size of this chunk (even if this is a common or BSS.)
163 virtual size_t getSize() const = 0;
164
165 virtual uint32_t getOutputCharacteristics() const { return 0; }
166
167 // Write this chunk to a mmap'ed file, assuming Buf is pointing to
168 // beginning of the file. Because this function may use RVA values
169 // of other chunks for relocations, you need to set them properly
170 // before calling this function.
171 virtual void writeTo(uint8_t *buf) const {}
172
173 // Returns the section name if this is a section chunk.
174 // It is illegal to call this function on non-section chunks.
175 virtual StringRef getSectionName() const {
176 llvm_unreachable("unimplemented getSectionName");
177 }
178
179 // Windows-specific.
180 // Collect all locations that contain absolute addresses for base relocations.
181 virtual void getBaserels(std::vector<Baserel> *res) {}
182
183 virtual MachineTypes getMachine() const { return IMAGE_FILE_MACHINE_UNKNOWN; }
184
185 // Returns a human-readable name of this chunk. Chunks are unnamed chunks of
186 // bytes, so this is used only for logging or debugging.
187 virtual StringRef getDebugName() const { return ""; }
188
189 // Verify that chunk relocations are within their ranges.
190 virtual bool verifyRanges() { return true; };
191
192 // If needed, extend the chunk to ensure all relocations are within the
193 // allowed ranges. Return the additional space required for the extension.
194 virtual uint32_t extendRanges() { return 0; };
195
196 static bool classof(const Chunk *c) { return c->kind() >= OtherKind; }
197
198protected:
199 NonSectionChunk(Kind k = OtherKind) : Chunk(k) {}
200};
201
202class NonSectionCodeChunk : public NonSectionChunk {
203public:
204 virtual uint32_t getOutputCharacteristics() const override {
205 return llvm::COFF::IMAGE_SCN_MEM_READ | llvm::COFF::IMAGE_SCN_MEM_EXECUTE;
206 }
207
208protected:
209 NonSectionCodeChunk(Kind k = OtherKind) : NonSectionChunk(k) {}
210};
211
212// MinGW specific; information about one individual location in the image
213// that needs to be fixed up at runtime after loading. This represents
214// one individual element in the PseudoRelocTableChunk table.
215class RuntimePseudoReloc {
216public:
217 RuntimePseudoReloc(Defined *sym, SectionChunk *target, uint32_t targetOffset,
218 int flags)
219 : sym(sym), target(target), targetOffset(targetOffset), flags(flags) {}
220
221 Defined *sym;
222 SectionChunk *target;
223 uint32_t targetOffset;
224 // The Flags field contains the size of the relocation, in bits. No other
225 // flags are currently defined.
226 int flags;
227};
228
229// A chunk corresponding a section of an input file.
230class SectionChunk : public Chunk {
231 // Identical COMDAT Folding feature accesses section internal data.
232 friend class ICF;
233
234public:
235 class symbol_iterator : public llvm::iterator_adaptor_base<
236 symbol_iterator, const coff_relocation *,
237 std::random_access_iterator_tag, Symbol *> {
238 friend SectionChunk;
239
240 ObjFile *file;
241
242 symbol_iterator(ObjFile *file, const coff_relocation *i)
243 : symbol_iterator::iterator_adaptor_base(i), file(file) {}
244
245 public:
246 symbol_iterator() = default;
247
248 Symbol *operator*() const { return file->getSymbol(symbolIndex: I->SymbolTableIndex); }
249 };
250
251 SectionChunk(ObjFile *file, const coff_section *header, Kind k = SectionKind);
252 static bool classof(const Chunk *c) { return c->kind() <= SectionECKind; }
253 size_t getSize() const { return header->SizeOfRawData; }
254 ArrayRef<uint8_t> getContents() const;
255 void writeTo(uint8_t *buf) const;
256 MachineTypes getMachine() const;
257
258 // Defend against unsorted relocations. This may be overly conservative.
259 void sortRelocations();
260
261 // Write and relocate a portion of the section. This is intended to be called
262 // in a loop. Relocations must be sorted first.
263 void writeAndRelocateSubsection(ArrayRef<uint8_t> sec,
264 ArrayRef<uint8_t> subsec,
265 uint32_t &nextRelocIndex, uint8_t *buf) const;
266
267 uint32_t getOutputCharacteristics() const {
268 return header->Characteristics & (permMask | typeMask);
269 }
270 StringRef getSectionName() const {
271 return StringRef(sectionNameData, sectionNameSize);
272 }
273 void getBaserels(std::vector<Baserel> *res);
274 bool isCOMDAT() const;
275 void applyRelocation(uint8_t *off, const coff_relocation &rel) const;
276 void applyRelX64(uint8_t *off, uint16_t type, OutputSection *os, uint64_t s,
277 uint64_t p, uint64_t imageBase) const;
278 void applyRelX86(uint8_t *off, uint16_t type, OutputSection *os, uint64_t s,
279 uint64_t p, uint64_t imageBase) const;
280 void applyRelARM(uint8_t *off, uint16_t type, OutputSection *os, uint64_t s,
281 uint64_t p, uint64_t imageBase) const;
282 void applyRelARM64(uint8_t *off, uint16_t type, OutputSection *os, uint64_t s,
283 uint64_t p, uint64_t imageBase) const;
284
285 void getRuntimePseudoRelocs(std::vector<RuntimePseudoReloc> &res);
286
287 // Called if the garbage collector decides to not include this chunk
288 // in a final output. It's supposed to print out a log message to stdout.
289 void printDiscardedMessage() const;
290
291 // Adds COMDAT associative sections to this COMDAT section. A chunk
292 // and its children are treated as a group by the garbage collector.
293 void addAssociative(SectionChunk *child);
294
295 StringRef getDebugName() const;
296
297 // True if this is a codeview debug info chunk. These will not be laid out in
298 // the image. Instead they will end up in the PDB, if one is requested.
299 bool isCodeView() const {
300 return getSectionName() == ".debug" || getSectionName().starts_with(Prefix: ".debug$");
301 }
302
303 // True if this is a DWARF debug info or exception handling chunk.
304 bool isDWARF() const {
305 return getSectionName().starts_with(Prefix: ".debug_") || getSectionName() == ".eh_frame";
306 }
307
308 // Allow iteration over the bodies of this chunk's relocated symbols.
309 llvm::iterator_range<symbol_iterator> symbols() const {
310 return llvm::make_range(x: symbol_iterator(file, relocsData),
311 y: symbol_iterator(file, relocsData + relocsSize));
312 }
313
314 ArrayRef<coff_relocation> getRelocs() const {
315 return llvm::ArrayRef(relocsData, relocsSize);
316 }
317
318 // Reloc setter used by ARM range extension thunk insertion.
319 void setRelocs(ArrayRef<coff_relocation> newRelocs) {
320 relocsData = newRelocs.data();
321 relocsSize = newRelocs.size();
322 assert(relocsSize == newRelocs.size() && "reloc size truncation");
323 }
324
325 // Single linked list iterator for associated comdat children.
326 class AssociatedIterator
327 : public llvm::iterator_facade_base<
328 AssociatedIterator, std::forward_iterator_tag, SectionChunk> {
329 public:
330 AssociatedIterator() = default;
331 AssociatedIterator(SectionChunk *head) : cur(head) {}
332 bool operator==(const AssociatedIterator &r) const { return cur == r.cur; }
333 // FIXME: Wrong const-ness, but it makes filter ranges work.
334 SectionChunk &operator*() const { return *cur; }
335 SectionChunk &operator*() { return *cur; }
336 AssociatedIterator &operator++() {
337 cur = cur->assocChildren;
338 return *this;
339 }
340
341 private:
342 SectionChunk *cur = nullptr;
343 };
344
345 // Allow iteration over the associated child chunks for this section.
346 llvm::iterator_range<AssociatedIterator> children() const {
347 // Associated sections do not have children. The assocChildren field is
348 // part of the parent's list of children.
349 bool isAssoc = selection == llvm::COFF::IMAGE_COMDAT_SELECT_ASSOCIATIVE;
350 return llvm::make_range(
351 x: AssociatedIterator(isAssoc ? nullptr : assocChildren),
352 y: AssociatedIterator(nullptr));
353 }
354
355 // The section ID this chunk belongs to in its Obj.
356 uint32_t getSectionNumber() const;
357
358 ArrayRef<uint8_t> consumeDebugMagic();
359
360 static ArrayRef<uint8_t> consumeDebugMagic(ArrayRef<uint8_t> data,
361 StringRef sectionName);
362
363 static SectionChunk *findByName(ArrayRef<SectionChunk *> sections,
364 StringRef name);
365
366 // The file that this chunk was created from.
367 ObjFile *file;
368
369 // Pointer to the COFF section header in the input file.
370 const coff_section *header;
371
372 // The COMDAT leader symbol if this is a COMDAT chunk.
373 DefinedRegular *sym = nullptr;
374
375 // The CRC of the contents as described in the COFF spec 4.5.5.
376 // Auxiliary Format 5: Section Definitions. Used for ICF.
377 uint32_t checksum = 0;
378
379 // Used by the garbage collector.
380 bool live;
381
382 // Whether this section needs to be kept distinct from other sections during
383 // ICF. This is set by the driver using address-significance tables.
384 bool keepUnique = false;
385
386 // The COMDAT selection if this is a COMDAT chunk.
387 llvm::COFF::COMDATType selection = (llvm::COFF::COMDATType)0;
388
389 // A pointer pointing to a replacement for this chunk.
390 // Initially it points to "this" object. If this chunk is merged
391 // with other chunk by ICF, it points to another chunk,
392 // and this chunk is considered as dead.
393 SectionChunk *repl;
394
395private:
396 SectionChunk *assocChildren = nullptr;
397
398 // Used for ICF (Identical COMDAT Folding)
399 void replace(SectionChunk *other);
400 uint32_t eqClass[2] = {0, 0};
401
402 // Relocations for this section. Size is stored below.
403 const coff_relocation *relocsData;
404
405 // Section name string. Size is stored below.
406 const char *sectionNameData;
407
408 uint32_t relocsSize = 0;
409 uint32_t sectionNameSize = 0;
410};
411
412// A section chunk corresponding a section of an EC input file.
413class SectionChunkEC final : public SectionChunk {
414public:
415 static bool classof(const Chunk *c) { return c->kind() == SectionECKind; }
416
417 SectionChunkEC(ObjFile *file, const coff_section *header)
418 : SectionChunk(file, header, SectionECKind) {}
419 Defined *entryThunk = nullptr;
420};
421
422// Inline methods to implement faux-virtual dispatch for SectionChunk.
423
424inline size_t Chunk::getSize() const {
425 if (isa<SectionChunk>(Val: this))
426 return static_cast<const SectionChunk *>(this)->getSize();
427 return static_cast<const NonSectionChunk *>(this)->getSize();
428}
429
430inline uint32_t Chunk::getOutputCharacteristics() const {
431 if (isa<SectionChunk>(Val: this))
432 return static_cast<const SectionChunk *>(this)->getOutputCharacteristics();
433 return static_cast<const NonSectionChunk *>(this)->getOutputCharacteristics();
434}
435
436inline void Chunk::writeTo(uint8_t *buf) const {
437 if (isa<SectionChunk>(Val: this))
438 static_cast<const SectionChunk *>(this)->writeTo(buf);
439 else
440 static_cast<const NonSectionChunk *>(this)->writeTo(buf);
441}
442
443inline StringRef Chunk::getSectionName() const {
444 if (isa<SectionChunk>(Val: this))
445 return static_cast<const SectionChunk *>(this)->getSectionName();
446 return static_cast<const NonSectionChunk *>(this)->getSectionName();
447}
448
449inline void Chunk::getBaserels(std::vector<Baserel> *res) {
450 if (isa<SectionChunk>(Val: this))
451 static_cast<SectionChunk *>(this)->getBaserels(res);
452 else
453 static_cast<NonSectionChunk *>(this)->getBaserels(res);
454}
455
456inline StringRef Chunk::getDebugName() const {
457 if (isa<SectionChunk>(Val: this))
458 return static_cast<const SectionChunk *>(this)->getDebugName();
459 return static_cast<const NonSectionChunk *>(this)->getDebugName();
460}
461
462inline MachineTypes Chunk::getMachine() const {
463 if (isa<SectionChunk>(Val: this))
464 return static_cast<const SectionChunk *>(this)->getMachine();
465 return static_cast<const NonSectionChunk *>(this)->getMachine();
466}
467
468inline llvm::Triple::ArchType Chunk::getArch() const {
469 return llvm::getMachineArchType(machine: getMachine());
470}
471
472inline std::optional<chpe_range_type> Chunk::getArm64ECRangeType() const {
473 // Data sections don't need codemap entries.
474 if (!(getOutputCharacteristics() & llvm::COFF::IMAGE_SCN_MEM_EXECUTE))
475 return std::nullopt;
476
477 switch (getMachine()) {
478 case AMD64:
479 return chpe_range_type::Amd64;
480 case ARM64EC:
481 return chpe_range_type::Arm64EC;
482 default:
483 return chpe_range_type::Arm64;
484 }
485}
486
487// This class is used to implement an lld-specific feature (not implemented in
488// MSVC) that minimizes the output size by finding string literals sharing tail
489// parts and merging them.
490//
491// If string tail merging is enabled and a section is identified as containing a
492// string literal, it is added to a MergeChunk with an appropriate alignment.
493// The MergeChunk then tail merges the strings using the StringTableBuilder
494// class and assigns RVAs and section offsets to each of the member chunks based
495// on the offsets assigned by the StringTableBuilder.
496class MergeChunk : public NonSectionChunk {
497public:
498 MergeChunk(uint32_t alignment);
499 static void addSection(COFFLinkerContext &ctx, SectionChunk *c);
500 void finalizeContents();
501 void assignSubsectionRVAs();
502
503 uint32_t getOutputCharacteristics() const override;
504 StringRef getSectionName() const override { return ".rdata"; }
505 size_t getSize() const override;
506 void writeTo(uint8_t *buf) const override;
507
508 std::vector<SectionChunk *> sections;
509
510private:
511 llvm::StringTableBuilder builder;
512 bool finalized = false;
513};
514
515// A chunk for common symbols. Common chunks don't have actual data.
516class CommonChunk : public NonSectionChunk {
517public:
518 CommonChunk(const COFFSymbolRef sym);
519 size_t getSize() const override { return sym.getValue(); }
520 uint32_t getOutputCharacteristics() const override;
521 StringRef getSectionName() const override { return ".bss"; }
522
523private:
524 const COFFSymbolRef sym;
525};
526
527// A chunk for linker-created strings.
528class StringChunk : public NonSectionChunk {
529public:
530 explicit StringChunk(StringRef s) : str(s) {}
531 size_t getSize() const override { return str.size() + 1; }
532 void writeTo(uint8_t *buf) const override;
533
534private:
535 StringRef str;
536};
537
538static const uint8_t importThunkX86[] = {
539 0xff, 0x25, 0x00, 0x00, 0x00, 0x00, // JMP *0x0
540};
541
542static const uint8_t importThunkARM[] = {
543 0x40, 0xf2, 0x00, 0x0c, // mov.w ip, #0
544 0xc0, 0xf2, 0x00, 0x0c, // mov.t ip, #0
545 0xdc, 0xf8, 0x00, 0xf0, // ldr.w pc, [ip]
546};
547
548static const uint8_t importThunkARM64[] = {
549 0x10, 0x00, 0x00, 0x90, // adrp x16, #0
550 0x10, 0x02, 0x40, 0xf9, // ldr x16, [x16]
551 0x00, 0x02, 0x1f, 0xd6, // br x16
552};
553
554static const uint8_t importThunkARM64EC[] = {
555 0x0b, 0x00, 0x00, 0x90, // adrp x11, 0x0
556 0x6b, 0x01, 0x40, 0xf9, // ldr x11, [x11]
557 0x0a, 0x00, 0x00, 0x90, // adrp x10, 0x0
558 0x4a, 0x01, 0x00, 0x91, // add x10, x10, #0x0
559 0x00, 0x00, 0x00, 0x14 // b 0x0
560};
561
562// Windows-specific.
563// A chunk for DLL import jump table entry. In a final output, its
564// contents will be a JMP instruction to some __imp_ symbol.
565class ImportThunkChunk : public NonSectionCodeChunk {
566public:
567 ImportThunkChunk(COFFLinkerContext &ctx, Defined *s);
568 static bool classof(const Chunk *c) { return c->kind() == ImportThunkKind; }
569
570 // We track the usage of the thunk symbol separately from the import file
571 // to avoid generating unnecessary thunks.
572 bool live;
573
574protected:
575 Defined *impSymbol;
576 COFFLinkerContext &ctx;
577};
578
579class ImportThunkChunkX64 : public ImportThunkChunk {
580public:
581 explicit ImportThunkChunkX64(COFFLinkerContext &ctx, Defined *s);
582 size_t getSize() const override { return sizeof(importThunkX86); }
583 void writeTo(uint8_t *buf) const override;
584 MachineTypes getMachine() const override { return AMD64; }
585};
586
587class ImportThunkChunkX86 : public ImportThunkChunk {
588public:
589 explicit ImportThunkChunkX86(COFFLinkerContext &ctx, Defined *s)
590 : ImportThunkChunk(ctx, s) {}
591 size_t getSize() const override { return sizeof(importThunkX86); }
592 void getBaserels(std::vector<Baserel> *res) override;
593 void writeTo(uint8_t *buf) const override;
594 MachineTypes getMachine() const override { return I386; }
595};
596
597class ImportThunkChunkARM : public ImportThunkChunk {
598public:
599 explicit ImportThunkChunkARM(COFFLinkerContext &ctx, Defined *s)
600 : ImportThunkChunk(ctx, s) {
601 setAlignment(2);
602 }
603 size_t getSize() const override { return sizeof(importThunkARM); }
604 void getBaserels(std::vector<Baserel> *res) override;
605 void writeTo(uint8_t *buf) const override;
606 MachineTypes getMachine() const override { return ARMNT; }
607};
608
609class ImportThunkChunkARM64 : public ImportThunkChunk {
610public:
611 explicit ImportThunkChunkARM64(COFFLinkerContext &ctx, Defined *s,
612 MachineTypes machine)
613 : ImportThunkChunk(ctx, s), machine(machine) {
614 setAlignment(4);
615 }
616 size_t getSize() const override { return sizeof(importThunkARM64); }
617 void writeTo(uint8_t *buf) const override;
618 MachineTypes getMachine() const override { return machine; }
619
620private:
621 MachineTypes machine;
622};
623
624// ARM64EC __impchk_* thunk implementation.
625// Performs an indirect call to an imported function pointer
626// using the __icall_helper_arm64ec helper function.
627class ImportThunkChunkARM64EC : public ImportThunkChunk {
628public:
629 explicit ImportThunkChunkARM64EC(ImportFile *file);
630 size_t getSize() const override;
631 MachineTypes getMachine() const override { return ARM64EC; }
632 void writeTo(uint8_t *buf) const override;
633 bool verifyRanges() override;
634 uint32_t extendRanges() override;
635
636 Defined *exitThunk;
637 Defined *sym = nullptr;
638 bool extended = false;
639
640private:
641 ImportFile *file;
642};
643
644class RangeExtensionThunkARM : public NonSectionCodeChunk {
645public:
646 explicit RangeExtensionThunkARM(COFFLinkerContext &ctx, Defined *t)
647 : target(t), ctx(ctx) {
648 setAlignment(2);
649 }
650 size_t getSize() const override;
651 void writeTo(uint8_t *buf) const override;
652 MachineTypes getMachine() const override { return ARMNT; }
653
654 Defined *target;
655
656private:
657 COFFLinkerContext &ctx;
658};
659
660// A ragnge extension thunk used for both ARM64EC and ARM64 machine types.
661class RangeExtensionThunkARM64 : public NonSectionCodeChunk {
662public:
663 explicit RangeExtensionThunkARM64(MachineTypes machine, Defined *t)
664 : target(t), machine(machine) {
665 setAlignment(4);
666 assert(llvm::COFF::isAnyArm64(machine));
667 }
668 size_t getSize() const override;
669 void writeTo(uint8_t *buf) const override;
670 MachineTypes getMachine() const override { return machine; }
671
672 Defined *target;
673
674private:
675 MachineTypes machine;
676};
677
678// Windows-specific.
679// See comments for DefinedLocalImport class.
680class LocalImportChunk : public NonSectionChunk {
681public:
682 explicit LocalImportChunk(COFFLinkerContext &ctx, Defined *s);
683 size_t getSize() const override;
684 void getBaserels(std::vector<Baserel> *res) override;
685 void writeTo(uint8_t *buf) const override;
686
687private:
688 Defined *sym;
689 COFFLinkerContext &ctx;
690};
691
692// Duplicate RVAs are not allowed in RVA tables, so unique symbols by chunk and
693// offset into the chunk. Order does not matter as the RVA table will be sorted
694// later.
695struct ChunkAndOffset {
696 Chunk *inputChunk;
697 uint32_t offset;
698
699 struct DenseMapInfo {
700 static ChunkAndOffset getEmptyKey() {
701 return {.inputChunk: llvm::DenseMapInfo<Chunk *>::getEmptyKey(), .offset: 0};
702 }
703 static ChunkAndOffset getTombstoneKey() {
704 return {.inputChunk: llvm::DenseMapInfo<Chunk *>::getTombstoneKey(), .offset: 0};
705 }
706 static unsigned getHashValue(const ChunkAndOffset &co) {
707 return llvm::DenseMapInfo<std::pair<Chunk *, uint32_t>>::getHashValue(
708 PairVal: {co.inputChunk, co.offset});
709 }
710 static bool isEqual(const ChunkAndOffset &lhs, const ChunkAndOffset &rhs) {
711 return lhs.inputChunk == rhs.inputChunk && lhs.offset == rhs.offset;
712 }
713 };
714};
715
716using SymbolRVASet = llvm::DenseSet<ChunkAndOffset>;
717
718// Table which contains symbol RVAs. Used for /safeseh and /guard:cf.
719class RVATableChunk : public NonSectionChunk {
720public:
721 explicit RVATableChunk(SymbolRVASet s) : syms(std::move(s)) {}
722 size_t getSize() const override { return syms.size() * 4; }
723 void writeTo(uint8_t *buf) const override;
724
725private:
726 SymbolRVASet syms;
727};
728
729// Table which contains symbol RVAs with flags. Used for /guard:ehcont.
730class RVAFlagTableChunk : public NonSectionChunk {
731public:
732 explicit RVAFlagTableChunk(SymbolRVASet s) : syms(std::move(s)) {}
733 size_t getSize() const override { return syms.size() * 5; }
734 void writeTo(uint8_t *buf) const override;
735
736private:
737 SymbolRVASet syms;
738};
739
740// Windows-specific.
741// This class represents a block in .reloc section.
742// See the PE/COFF spec 5.6 for details.
743class BaserelChunk : public NonSectionChunk {
744public:
745 BaserelChunk(uint32_t page, Baserel *begin, Baserel *end);
746 size_t getSize() const override { return data.size(); }
747 void writeTo(uint8_t *buf) const override;
748
749private:
750 std::vector<uint8_t> data;
751};
752
753class Baserel {
754public:
755 Baserel(uint32_t v, uint8_t ty) : rva(v), type(ty) {}
756 explicit Baserel(uint32_t v, llvm::COFF::MachineTypes machine)
757 : Baserel(v, getDefaultType(machine)) {}
758 static uint8_t getDefaultType(llvm::COFF::MachineTypes machine);
759
760 uint32_t rva;
761 uint8_t type;
762};
763
764// This is a placeholder Chunk, to allow attaching a DefinedSynthetic to a
765// specific place in a section, without any data. This is used for the MinGW
766// specific symbol __RUNTIME_PSEUDO_RELOC_LIST_END__, even though the concept
767// of an empty chunk isn't MinGW specific.
768class EmptyChunk : public NonSectionChunk {
769public:
770 EmptyChunk() {}
771 size_t getSize() const override { return 0; }
772 void writeTo(uint8_t *buf) const override {}
773};
774
775class ECCodeMapEntry {
776public:
777 ECCodeMapEntry(Chunk *first, Chunk *last, chpe_range_type type)
778 : first(first), last(last), type(type) {}
779 Chunk *first;
780 Chunk *last;
781 chpe_range_type type;
782};
783
784// This is a chunk containing CHPE code map on EC targets. It's a table
785// of address ranges and their types.
786class ECCodeMapChunk : public NonSectionChunk {
787public:
788 ECCodeMapChunk(std::vector<ECCodeMapEntry> &map) : map(map) {}
789 size_t getSize() const override;
790 void writeTo(uint8_t *buf) const override;
791
792private:
793 std::vector<ECCodeMapEntry> &map;
794};
795
796class CHPECodeRangesChunk : public NonSectionChunk {
797public:
798 CHPECodeRangesChunk(std::vector<std::pair<Chunk *, Defined *>> &exportThunks)
799 : exportThunks(exportThunks) {}
800 size_t getSize() const override;
801 void writeTo(uint8_t *buf) const override;
802
803private:
804 std::vector<std::pair<Chunk *, Defined *>> &exportThunks;
805};
806
807class CHPERedirectionChunk : public NonSectionChunk {
808public:
809 CHPERedirectionChunk(std::vector<std::pair<Chunk *, Defined *>> &exportThunks)
810 : exportThunks(exportThunks) {}
811 size_t getSize() const override;
812 void writeTo(uint8_t *buf) const override;
813
814private:
815 std::vector<std::pair<Chunk *, Defined *>> &exportThunks;
816};
817
818static const uint8_t ECExportThunkCode[] = {
819 0x48, 0x8b, 0xc4, // movq %rsp, %rax
820 0x48, 0x89, 0x58, 0x20, // movq %rbx, 0x20(%rax)
821 0x55, // pushq %rbp
822 0x5d, // popq %rbp
823 0xe9, 0, 0, 0, 0, // jmp *0x0
824 0xcc, // int3
825 0xcc // int3
826};
827
828class ECExportThunkChunk : public NonSectionCodeChunk {
829public:
830 explicit ECExportThunkChunk(Defined *targetSym)
831 : NonSectionCodeChunk(ECExportThunkKind), target(targetSym) {}
832 static bool classof(const Chunk *c) { return c->kind() == ECExportThunkKind; }
833
834 size_t getSize() const override { return sizeof(ECExportThunkCode); };
835 void writeTo(uint8_t *buf) const override;
836 MachineTypes getMachine() const override { return AMD64; }
837
838 Defined *target;
839};
840
841// ARM64X relocation value, potentially relative to a symbol.
842class Arm64XRelocVal {
843public:
844 Arm64XRelocVal(uint64_t value = 0) : value(value) {}
845 Arm64XRelocVal(Defined *sym, int32_t offset = 0) : sym(sym), value(offset) {}
846 Arm64XRelocVal(Chunk *chunk, int32_t offset = 0)
847 : chunk(chunk), value(offset) {}
848 uint64_t get() const;
849
850private:
851 Defined *sym = nullptr;
852 Chunk *chunk = nullptr;
853 uint64_t value;
854};
855
856// ARM64X entry for dynamic relocations.
857class Arm64XDynamicRelocEntry {
858public:
859 Arm64XDynamicRelocEntry(llvm::COFF::Arm64XFixupType type, uint8_t size,
860 Arm64XRelocVal offset, Arm64XRelocVal value)
861 : offset(offset), value(value), type(type), size(size) {}
862
863 size_t getSize() const;
864 void writeTo(uint8_t *buf) const;
865
866 Arm64XRelocVal offset;
867 Arm64XRelocVal value;
868
869private:
870 llvm::COFF::Arm64XFixupType type;
871 uint8_t size;
872};
873
874// Dynamic relocation chunk containing ARM64X relocations for the hybrid image.
875class DynamicRelocsChunk : public NonSectionChunk {
876public:
877 DynamicRelocsChunk() {}
878 size_t getSize() const override { return size; }
879 void writeTo(uint8_t *buf) const override;
880 void finalize();
881
882 void add(llvm::COFF::Arm64XFixupType type, uint8_t size,
883 Arm64XRelocVal offset, Arm64XRelocVal value = Arm64XRelocVal()) {
884 arm64xRelocs.emplace_back(args&: type, args&: size, args&: offset, args&: value);
885 }
886
887 void set(uint32_t rva, Arm64XRelocVal value);
888
889private:
890 std::vector<Arm64XDynamicRelocEntry> arm64xRelocs;
891 size_t size;
892};
893
894// MinGW specific, for the "automatic import of variables from DLLs" feature.
895// This provides the table of runtime pseudo relocations, for variable
896// references that turned out to need to be imported from a DLL even though
897// the reference didn't use the dllimport attribute. The MinGW runtime will
898// process this table after loading, before handling control over to user
899// code.
900class PseudoRelocTableChunk : public NonSectionChunk {
901public:
902 PseudoRelocTableChunk(std::vector<RuntimePseudoReloc> &relocs)
903 : relocs(std::move(relocs)) {
904 setAlignment(4);
905 }
906 size_t getSize() const override;
907 void writeTo(uint8_t *buf) const override;
908
909private:
910 std::vector<RuntimePseudoReloc> relocs;
911};
912
913// MinGW specific. A Chunk that contains one pointer-sized absolute value.
914class AbsolutePointerChunk : public NonSectionChunk {
915public:
916 AbsolutePointerChunk(SymbolTable &symtab, uint64_t value)
917 : value(value), symtab(symtab) {
918 setAlignment(getSize());
919 }
920 size_t getSize() const override;
921 void writeTo(uint8_t *buf) const override;
922 MachineTypes getMachine() const override;
923
924private:
925 uint64_t value;
926 SymbolTable &symtab;
927};
928
929// Return true if this file has the hotpatch flag set to true in the S_COMPILE3
930// record in codeview debug info. Also returns true for some thunks synthesized
931// by the linker.
932inline bool Chunk::isHotPatchable() const {
933 if (auto *sc = dyn_cast<SectionChunk>(Val: this))
934 return sc->file->hotPatchable;
935 else if (isa<ImportThunkChunk>(Val: this))
936 return true;
937 return false;
938}
939
940inline Defined *Chunk::getEntryThunk() const {
941 if (auto *c = dyn_cast<const SectionChunkEC>(Val: this))
942 return c->entryThunk;
943 return nullptr;
944}
945
946inline void Chunk::setEntryThunk(Defined *entryThunk) {
947 if (auto c = dyn_cast<SectionChunkEC>(Val: this))
948 c->entryThunk = entryThunk;
949}
950
951void applyMOV32T(uint8_t *off, uint32_t v);
952void applyBranch24T(uint8_t *off, int32_t v);
953
954void applyArm64Addr(uint8_t *off, uint64_t s, uint64_t p, int shift);
955void applyArm64Imm(uint8_t *off, uint64_t imm, uint32_t rangeLimit);
956void applyArm64Branch26(uint8_t *off, int64_t v);
957
958// Convenience class for initializing a coff_section with specific flags.
959class FakeSection {
960public:
961 FakeSection(int c) { section.Characteristics = c; }
962
963 coff_section section;
964};
965
966// Convenience class for initializing a SectionChunk with specific flags.
967class FakeSectionChunk {
968public:
969 FakeSectionChunk(const coff_section *section) : chunk(nullptr, section) {
970 // Comdats from LTO files can't be fully treated as regular comdats
971 // at this point; we don't know what size or contents they are going to
972 // have, so we can't do proper checking of such aspects of them.
973 chunk.selection = llvm::COFF::IMAGE_COMDAT_SELECT_ANY;
974 }
975
976 SectionChunk chunk;
977};
978
979} // namespace lld::coff
980
981namespace llvm {
982template <>
983struct DenseMapInfo<lld::coff::ChunkAndOffset>
984 : lld::coff::ChunkAndOffset::DenseMapInfo {};
985}
986
987#endif
988