1 //===- InputChunks.cpp ----------------------------------------------------===//
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 #include "InputChunks.h"
10 #include "Config.h"
11 #include "OutputSegment.h"
12 #include "WriterUtils.h"
13 #include "lld/Common/ErrorHandler.h"
14 #include "lld/Common/LLVM.h"
15 #include "llvm/Support/LEB128.h"
16 
17 #define DEBUG_TYPE "lld"
18 
19 using namespace llvm;
20 using namespace llvm::wasm;
21 using namespace llvm::support::endian;
22 using namespace lld;
23 using namespace lld::wasm;
24 
25 StringRef lld::relocTypeToString(uint8_t RelocType) {
26   switch (RelocType) {
27 #define WASM_RELOC(NAME, REL)                                                  \
28   case REL:                                                                    \
29     return #NAME;
30 #include "llvm/BinaryFormat/WasmRelocs.def"
31 #undef WASM_RELOC
32   }
33   llvm_unreachable("unknown reloc type");
34 }
35 
36 std::string lld::toString(const InputChunk *C) {
37   return (toString(C->File) + ":(" + C->getName() + ")").str();
38 }
39 
40 StringRef InputChunk::getComdatName() const {
41   uint32_t Index = getComdat();
42   if (Index == UINT32_MAX)
43     return StringRef();
44   return File->getWasmObj()->linkingData().Comdats[Index];
45 }
46 
47 void InputChunk::verifyRelocTargets() const {
48   for (const WasmRelocation &Rel : Relocations) {
49     uint32_t ExistingValue;
50     unsigned BytesRead = 0;
51     uint32_t Offset = Rel.Offset - getInputSectionOffset();
52     const uint8_t *Loc = data().data() + Offset;
53     switch (Rel.Type) {
54     case R_WASM_TYPE_INDEX_LEB:
55     case R_WASM_FUNCTION_INDEX_LEB:
56     case R_WASM_GLOBAL_INDEX_LEB:
57     case R_WASM_EVENT_INDEX_LEB:
58     case R_WASM_MEMORY_ADDR_LEB:
59       ExistingValue = decodeULEB128(Loc, &BytesRead);
60       break;
61     case R_WASM_TABLE_INDEX_SLEB:
62     case R_WASM_TABLE_INDEX_REL_SLEB:
63     case R_WASM_MEMORY_ADDR_SLEB:
64     case R_WASM_MEMORY_ADDR_REL_SLEB:
65       ExistingValue = static_cast<uint32_t>(decodeSLEB128(Loc, &BytesRead));
66       break;
67     case R_WASM_TABLE_INDEX_I32:
68     case R_WASM_MEMORY_ADDR_I32:
69     case R_WASM_FUNCTION_OFFSET_I32:
70     case R_WASM_SECTION_OFFSET_I32:
71       ExistingValue = static_cast<uint32_t>(read32le(Loc));
72       break;
73     default:
74       llvm_unreachable("unknown relocation type");
75     }
76 
77     if (BytesRead && BytesRead != 5)
78       warn("expected LEB at relocation site be 5-byte padded");
79 
80     if (Rel.Type != R_WASM_GLOBAL_INDEX_LEB) {
81       uint32_t ExpectedValue = File->calcExpectedValue(Rel);
82       if (ExpectedValue != ExistingValue)
83         warn("unexpected existing value for " + relocTypeToString(Rel.Type) +
84              ": existing=" + Twine(ExistingValue) +
85              " expected=" + Twine(ExpectedValue));
86     }
87   }
88 }
89 
90 // Copy this input chunk to an mmap'ed output file and apply relocations.
91 void InputChunk::writeTo(uint8_t *Buf) const {
92   // Copy contents
93   memcpy(Buf + OutputOffset, data().data(), data().size());
94 
95   // Apply relocations
96   if (Relocations.empty())
97     return;
98 
99 #ifndef NDEBUG
100   verifyRelocTargets();
101 #endif
102 
103   LLVM_DEBUG(dbgs() << "applying relocations: " << getName()
104                     << " count=" << Relocations.size() << "\n");
105   int32_t Off = OutputOffset - getInputSectionOffset();
106 
107   for (const WasmRelocation &Rel : Relocations) {
108     uint8_t *Loc = Buf + Rel.Offset + Off;
109     uint32_t Value = File->calcNewValue(Rel);
110     LLVM_DEBUG(dbgs() << "apply reloc: type=" << relocTypeToString(Rel.Type));
111     if (Rel.Type != R_WASM_TYPE_INDEX_LEB)
112       LLVM_DEBUG(dbgs() << " sym=" << File->getSymbols()[Rel.Index]->getName());
113     LLVM_DEBUG(dbgs() << " addend=" << Rel.Addend << " index=" << Rel.Index
114                       << " value=" << Value << " offset=" << Rel.Offset
115                       << "\n");
116 
117     switch (Rel.Type) {
118     case R_WASM_TYPE_INDEX_LEB:
119     case R_WASM_FUNCTION_INDEX_LEB:
120     case R_WASM_GLOBAL_INDEX_LEB:
121     case R_WASM_EVENT_INDEX_LEB:
122     case R_WASM_MEMORY_ADDR_LEB:
123       encodeULEB128(Value, Loc, 5);
124       break;
125     case R_WASM_TABLE_INDEX_SLEB:
126     case R_WASM_TABLE_INDEX_REL_SLEB:
127     case R_WASM_MEMORY_ADDR_SLEB:
128     case R_WASM_MEMORY_ADDR_REL_SLEB:
129       encodeSLEB128(static_cast<int32_t>(Value), Loc, 5);
130       break;
131     case R_WASM_TABLE_INDEX_I32:
132     case R_WASM_MEMORY_ADDR_I32:
133     case R_WASM_FUNCTION_OFFSET_I32:
134     case R_WASM_SECTION_OFFSET_I32:
135       write32le(Loc, Value);
136       break;
137     default:
138       llvm_unreachable("unknown relocation type");
139     }
140   }
141 }
142 
143 // Copy relocation entries to a given output stream.
144 // This function is used only when a user passes "-r". For a regular link,
145 // we consume relocations instead of copying them to an output file.
146 void InputChunk::writeRelocations(raw_ostream &OS) const {
147   if (Relocations.empty())
148     return;
149 
150   int32_t Off = OutputOffset - getInputSectionOffset();
151   LLVM_DEBUG(dbgs() << "writeRelocations: " << File->getName()
152                     << " offset=" << Twine(Off) << "\n");
153 
154   for (const WasmRelocation &Rel : Relocations) {
155     writeUleb128(OS, Rel.Type, "reloc type");
156     writeUleb128(OS, Rel.Offset + Off, "reloc offset");
157     writeUleb128(OS, File->calcNewIndex(Rel), "reloc index");
158 
159     if (relocTypeHasAddend(Rel.Type))
160       writeSleb128(OS, File->calcNewAddend(Rel), "reloc addend");
161   }
162 }
163 
164 void InputFunction::setFunctionIndex(uint32_t Index) {
165   LLVM_DEBUG(dbgs() << "InputFunction::setFunctionIndex: " << getName()
166                     << " -> " << Index << "\n");
167   assert(!hasFunctionIndex());
168   FunctionIndex = Index;
169 }
170 
171 void InputFunction::setTableIndex(uint32_t Index) {
172   LLVM_DEBUG(dbgs() << "InputFunction::setTableIndex: " << getName() << " -> "
173                     << Index << "\n");
174   assert(!hasTableIndex());
175   TableIndex = Index;
176 }
177 
178 // Write a relocation value without padding and return the number of bytes
179 // witten.
180 static unsigned writeCompressedReloc(uint8_t *Buf, const WasmRelocation &Rel,
181                                      uint32_t Value) {
182   switch (Rel.Type) {
183   case R_WASM_TYPE_INDEX_LEB:
184   case R_WASM_FUNCTION_INDEX_LEB:
185   case R_WASM_GLOBAL_INDEX_LEB:
186   case R_WASM_EVENT_INDEX_LEB:
187   case R_WASM_MEMORY_ADDR_LEB:
188     return encodeULEB128(Value, Buf);
189   case R_WASM_TABLE_INDEX_SLEB:
190   case R_WASM_MEMORY_ADDR_SLEB:
191     return encodeSLEB128(static_cast<int32_t>(Value), Buf);
192   default:
193     llvm_unreachable("unexpected relocation type");
194   }
195 }
196 
197 static unsigned getRelocWidthPadded(const WasmRelocation &Rel) {
198   switch (Rel.Type) {
199   case R_WASM_TYPE_INDEX_LEB:
200   case R_WASM_FUNCTION_INDEX_LEB:
201   case R_WASM_GLOBAL_INDEX_LEB:
202   case R_WASM_EVENT_INDEX_LEB:
203   case R_WASM_MEMORY_ADDR_LEB:
204   case R_WASM_TABLE_INDEX_SLEB:
205   case R_WASM_MEMORY_ADDR_SLEB:
206     return 5;
207   default:
208     llvm_unreachable("unexpected relocation type");
209   }
210 }
211 
212 static unsigned getRelocWidth(const WasmRelocation &Rel, uint32_t Value) {
213   uint8_t Buf[5];
214   return writeCompressedReloc(Buf, Rel, Value);
215 }
216 
217 // Relocations of type LEB and SLEB in the code section are padded to 5 bytes
218 // so that a fast linker can blindly overwrite them without needing to worry
219 // about the number of bytes needed to encode the values.
220 // However, for optimal output the code section can be compressed to remove
221 // the padding then outputting non-relocatable files.
222 // In this case we need to perform a size calculation based on the value at each
223 // relocation.  At best we end up saving 4 bytes for each relocation entry.
224 //
225 // This function only computes the final output size.  It must be called
226 // before getSize() is used to calculate of layout of the code section.
227 void InputFunction::calculateSize() {
228   if (!File || !Config->CompressRelocations)
229     return;
230 
231   LLVM_DEBUG(dbgs() << "calculateSize: " << getName() << "\n");
232 
233   const uint8_t *SecStart = File->CodeSection->Content.data();
234   const uint8_t *FuncStart = SecStart + getInputSectionOffset();
235   uint32_t FunctionSizeLength;
236   decodeULEB128(FuncStart, &FunctionSizeLength);
237 
238   uint32_t Start = getInputSectionOffset();
239   uint32_t End = Start + Function->Size;
240 
241   uint32_t LastRelocEnd = Start + FunctionSizeLength;
242   for (const WasmRelocation &Rel : Relocations) {
243     LLVM_DEBUG(dbgs() << "  region: " << (Rel.Offset - LastRelocEnd) << "\n");
244     CompressedFuncSize += Rel.Offset - LastRelocEnd;
245     CompressedFuncSize += getRelocWidth(Rel, File->calcNewValue(Rel));
246     LastRelocEnd = Rel.Offset + getRelocWidthPadded(Rel);
247   }
248   LLVM_DEBUG(dbgs() << "  final region: " << (End - LastRelocEnd) << "\n");
249   CompressedFuncSize += End - LastRelocEnd;
250 
251   // Now we know how long the resulting function is we can add the encoding
252   // of its length
253   uint8_t Buf[5];
254   CompressedSize = CompressedFuncSize + encodeULEB128(CompressedFuncSize, Buf);
255 
256   LLVM_DEBUG(dbgs() << "  calculateSize orig: " << Function->Size << "\n");
257   LLVM_DEBUG(dbgs() << "  calculateSize  new: " << CompressedSize << "\n");
258 }
259 
260 // Override the default writeTo method so that we can (optionally) write the
261 // compressed version of the function.
262 void InputFunction::writeTo(uint8_t *Buf) const {
263   if (!File || !Config->CompressRelocations)
264     return InputChunk::writeTo(Buf);
265 
266   Buf += OutputOffset;
267   uint8_t *Orig = Buf;
268   (void)Orig;
269 
270   const uint8_t *SecStart = File->CodeSection->Content.data();
271   const uint8_t *FuncStart = SecStart + getInputSectionOffset();
272   const uint8_t *End = FuncStart + Function->Size;
273   uint32_t Count;
274   decodeULEB128(FuncStart, &Count);
275   FuncStart += Count;
276 
277   LLVM_DEBUG(dbgs() << "write func: " << getName() << "\n");
278   Buf += encodeULEB128(CompressedFuncSize, Buf);
279   const uint8_t *LastRelocEnd = FuncStart;
280   for (const WasmRelocation &Rel : Relocations) {
281     unsigned ChunkSize = (SecStart + Rel.Offset) - LastRelocEnd;
282     LLVM_DEBUG(dbgs() << "  write chunk: " << ChunkSize << "\n");
283     memcpy(Buf, LastRelocEnd, ChunkSize);
284     Buf += ChunkSize;
285     Buf += writeCompressedReloc(Buf, Rel, File->calcNewValue(Rel));
286     LastRelocEnd = SecStart + Rel.Offset + getRelocWidthPadded(Rel);
287   }
288 
289   unsigned ChunkSize = End - LastRelocEnd;
290   LLVM_DEBUG(dbgs() << "  write final chunk: " << ChunkSize << "\n");
291   memcpy(Buf, LastRelocEnd, ChunkSize);
292   LLVM_DEBUG(dbgs() << "  total: " << (Buf + ChunkSize - Orig) << "\n");
293 }
294 
295 // Generate code to apply relocations to the data section at runtime.
296 // This is only called when generating shared libaries (PIC) where address are
297 // not known at static link time.
298 void InputSegment::generateRelocationCode(raw_ostream &OS) const {
299   LLVM_DEBUG(dbgs() << "generating runtime relocations: " << getName()
300                     << " count=" << Relocations.size() << "\n");
301 
302   // TODO(sbc): Encode the relocations in the data section and write a loop
303   // here to apply them.
304   uint32_t SegmentVA = OutputSeg->StartVA + OutputSegmentOffset;
305   for (const WasmRelocation &Rel : Relocations) {
306     uint32_t Offset = Rel.Offset - getInputSectionOffset();
307     uint32_t OutputOffset = SegmentVA + Offset;
308 
309     LLVM_DEBUG(dbgs() << "gen reloc: type=" << relocTypeToString(Rel.Type)
310                       << " addend=" << Rel.Addend << " index=" << Rel.Index
311                       << " output offset=" << OutputOffset << "\n");
312 
313     // Get __memory_base
314     writeU8(OS, WASM_OPCODE_GLOBAL_GET, "GLOBAL_GET");
315     writeUleb128(OS, WasmSym::MemoryBase->getGlobalIndex(), "memory_base");
316 
317     // Add the offset of the relocation
318     writeU8(OS, WASM_OPCODE_I32_CONST, "I32_CONST");
319     writeSleb128(OS, OutputOffset, "offset");
320     writeU8(OS, WASM_OPCODE_I32_ADD, "ADD");
321 
322     Symbol *Sym = File->getSymbol(Rel);
323     // Now figure out what we want to store
324     if (Sym->hasGOTIndex()) {
325       writeU8(OS, WASM_OPCODE_GLOBAL_GET, "GLOBAL_GET");
326       writeUleb128(OS, Sym->getGOTIndex(), "global index");
327       if (Rel.Addend) {
328         writeU8(OS, WASM_OPCODE_I32_CONST, "CONST");
329         writeSleb128(OS, Rel.Addend, "addend");
330         writeU8(OS, WASM_OPCODE_I32_ADD, "ADD");
331       }
332     } else {
333       const GlobalSymbol* BaseSymbol = WasmSym::MemoryBase;
334       if (Rel.Type == R_WASM_TABLE_INDEX_I32)
335         BaseSymbol = WasmSym::TableBase;
336       writeU8(OS, WASM_OPCODE_GLOBAL_GET, "GLOBAL_GET");
337       writeUleb128(OS, BaseSymbol->getGlobalIndex(), "base");
338       writeU8(OS, WASM_OPCODE_I32_CONST, "CONST");
339       writeSleb128(OS, File->calcNewValue(Rel), "offset");
340       writeU8(OS, WASM_OPCODE_I32_ADD, "ADD");
341     }
342 
343     // Store that value at the virtual address
344     writeU8(OS, WASM_OPCODE_I32_STORE, "I32_STORE");
345     writeUleb128(OS, 2, "align");
346     writeUleb128(OS, 0, "offset");
347   }
348 }
349