10b57cec5SDimitry Andric //===- Relocations.h -------------------------------------------*- C++ -*-===// 20b57cec5SDimitry Andric // 30b57cec5SDimitry Andric // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 40b57cec5SDimitry Andric // See https://llvm.org/LICENSE.txt for license information. 50b57cec5SDimitry Andric // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 60b57cec5SDimitry Andric // 70b57cec5SDimitry Andric //===----------------------------------------------------------------------===// 80b57cec5SDimitry Andric 90b57cec5SDimitry Andric #ifndef LLD_ELF_RELOCATIONS_H 100b57cec5SDimitry Andric #define LLD_ELF_RELOCATIONS_H 110b57cec5SDimitry Andric 120b57cec5SDimitry Andric #include "lld/Common/LLVM.h" 130b57cec5SDimitry Andric #include "llvm/ADT/DenseMap.h" 140b57cec5SDimitry Andric #include <map> 150b57cec5SDimitry Andric #include <vector> 160b57cec5SDimitry Andric 170b57cec5SDimitry Andric namespace lld { 180b57cec5SDimitry Andric namespace elf { 190b57cec5SDimitry Andric class Symbol; 200b57cec5SDimitry Andric class InputSection; 210b57cec5SDimitry Andric class InputSectionBase; 220b57cec5SDimitry Andric class OutputSection; 230b57cec5SDimitry Andric class SectionBase; 240b57cec5SDimitry Andric 250b57cec5SDimitry Andric // Represents a relocation type, such as R_X86_64_PC32 or R_ARM_THM_CALL. 260b57cec5SDimitry Andric using RelType = uint32_t; 275ffd83dbSDimitry Andric using JumpModType = uint32_t; 280b57cec5SDimitry Andric 290b57cec5SDimitry Andric // List of target-independent relocation types. Relocations read 300b57cec5SDimitry Andric // from files are converted to these types so that the main code 310b57cec5SDimitry Andric // doesn't have to know about architecture-specific details. 320b57cec5SDimitry Andric enum RelExpr { 330b57cec5SDimitry Andric R_ABS, 340b57cec5SDimitry Andric R_ADDEND, 350b57cec5SDimitry Andric R_DTPREL, 360b57cec5SDimitry Andric R_GOT, 370b57cec5SDimitry Andric R_GOT_OFF, 380b57cec5SDimitry Andric R_GOT_PC, 390b57cec5SDimitry Andric R_GOTONLY_PC, 400b57cec5SDimitry Andric R_GOTPLTONLY_PC, 410b57cec5SDimitry Andric R_GOTPLT, 420b57cec5SDimitry Andric R_GOTPLTREL, 430b57cec5SDimitry Andric R_GOTREL, 440b57cec5SDimitry Andric R_NONE, 450b57cec5SDimitry Andric R_PC, 460b57cec5SDimitry Andric R_PLT, 470b57cec5SDimitry Andric R_PLT_PC, 48349cc55cSDimitry Andric R_PLT_GOTPLT, 490b57cec5SDimitry Andric R_RELAX_GOT_PC, 500b57cec5SDimitry Andric R_RELAX_GOT_PC_NOPIC, 510b57cec5SDimitry Andric R_RELAX_TLS_GD_TO_IE, 520b57cec5SDimitry Andric R_RELAX_TLS_GD_TO_IE_ABS, 530b57cec5SDimitry Andric R_RELAX_TLS_GD_TO_IE_GOT_OFF, 540b57cec5SDimitry Andric R_RELAX_TLS_GD_TO_IE_GOTPLT, 550b57cec5SDimitry Andric R_RELAX_TLS_GD_TO_LE, 560b57cec5SDimitry Andric R_RELAX_TLS_GD_TO_LE_NEG, 570b57cec5SDimitry Andric R_RELAX_TLS_IE_TO_LE, 580b57cec5SDimitry Andric R_RELAX_TLS_LD_TO_LE, 590b57cec5SDimitry Andric R_RELAX_TLS_LD_TO_LE_ABS, 600b57cec5SDimitry Andric R_SIZE, 61e8d8bef9SDimitry Andric R_TPREL, 62e8d8bef9SDimitry Andric R_TPREL_NEG, 630b57cec5SDimitry Andric R_TLSDESC, 640b57cec5SDimitry Andric R_TLSDESC_CALL, 650b57cec5SDimitry Andric R_TLSDESC_PC, 66349cc55cSDimitry Andric R_TLSDESC_GOTPLT, 670b57cec5SDimitry Andric R_TLSGD_GOT, 680b57cec5SDimitry Andric R_TLSGD_GOTPLT, 690b57cec5SDimitry Andric R_TLSGD_PC, 700b57cec5SDimitry Andric R_TLSIE_HINT, 710b57cec5SDimitry Andric R_TLSLD_GOT, 720b57cec5SDimitry Andric R_TLSLD_GOTPLT, 730b57cec5SDimitry Andric R_TLSLD_GOT_OFF, 740b57cec5SDimitry Andric R_TLSLD_HINT, 750b57cec5SDimitry Andric R_TLSLD_PC, 760b57cec5SDimitry Andric 770b57cec5SDimitry Andric // The following is abstract relocation types used for only one target. 780b57cec5SDimitry Andric // 790b57cec5SDimitry Andric // Even though RelExpr is intended to be a target-neutral representation 800b57cec5SDimitry Andric // of a relocation type, there are some relocations whose semantics are 810b57cec5SDimitry Andric // unique to a target. Such relocation are marked with R_<TARGET_NAME>. 820b57cec5SDimitry Andric R_AARCH64_GOT_PAGE_PC, 83e8d8bef9SDimitry Andric R_AARCH64_GOT_PAGE, 840b57cec5SDimitry Andric R_AARCH64_PAGE_PC, 850b57cec5SDimitry Andric R_AARCH64_RELAX_TLS_GD_TO_IE_PAGE_PC, 860b57cec5SDimitry Andric R_AARCH64_TLSDESC_PAGE, 875ffd83dbSDimitry Andric R_ARM_PCA, 880b57cec5SDimitry Andric R_ARM_SBREL, 890b57cec5SDimitry Andric R_MIPS_GOTREL, 900b57cec5SDimitry Andric R_MIPS_GOT_GP, 910b57cec5SDimitry Andric R_MIPS_GOT_GP_PC, 920b57cec5SDimitry Andric R_MIPS_GOT_LOCAL_PAGE, 930b57cec5SDimitry Andric R_MIPS_GOT_OFF, 940b57cec5SDimitry Andric R_MIPS_GOT_OFF32, 950b57cec5SDimitry Andric R_MIPS_TLSGD, 960b57cec5SDimitry Andric R_MIPS_TLSLD, 970b57cec5SDimitry Andric R_PPC32_PLTREL, 980b57cec5SDimitry Andric R_PPC64_CALL, 990b57cec5SDimitry Andric R_PPC64_CALL_PLT, 1000b57cec5SDimitry Andric R_PPC64_RELAX_TOC, 1010b57cec5SDimitry Andric R_PPC64_TOCBASE, 102e8d8bef9SDimitry Andric R_PPC64_RELAX_GOT_PC, 1030b57cec5SDimitry Andric R_RISCV_ADD, 1040b57cec5SDimitry Andric R_RISCV_PC_INDIRECT, 1050b57cec5SDimitry Andric }; 1060b57cec5SDimitry Andric 1070b57cec5SDimitry Andric // Architecture-neutral representation of relocation. 1080b57cec5SDimitry Andric struct Relocation { 1090b57cec5SDimitry Andric RelExpr expr; 1100b57cec5SDimitry Andric RelType type; 1110b57cec5SDimitry Andric uint64_t offset; 1120b57cec5SDimitry Andric int64_t addend; 1130b57cec5SDimitry Andric Symbol *sym; 1140b57cec5SDimitry Andric }; 1150b57cec5SDimitry Andric 1165ffd83dbSDimitry Andric // Manipulate jump instructions with these modifiers. These are used to relax 1175ffd83dbSDimitry Andric // jump instruction opcodes at basic block boundaries and are particularly 1185ffd83dbSDimitry Andric // useful when basic block sections are enabled. 1195ffd83dbSDimitry Andric struct JumpInstrMod { 1205ffd83dbSDimitry Andric JumpModType original; 1215ffd83dbSDimitry Andric uint64_t offset; 1225ffd83dbSDimitry Andric unsigned size; 1235ffd83dbSDimitry Andric }; 1245ffd83dbSDimitry Andric 1250b57cec5SDimitry Andric // This function writes undefined symbol diagnostics to an internal buffer. 1260b57cec5SDimitry Andric // Call reportUndefinedSymbols() after calling scanRelocations() to emit 1270b57cec5SDimitry Andric // the diagnostics. 1280b57cec5SDimitry Andric template <class ELFT> void scanRelocations(InputSectionBase &); 129*0eae32dcSDimitry Andric void postScanRelocations(); 1300b57cec5SDimitry Andric 1310b57cec5SDimitry Andric template <class ELFT> void reportUndefinedSymbols(); 1320b57cec5SDimitry Andric 1335ffd83dbSDimitry Andric void hexagonTLSSymbolUpdate(ArrayRef<OutputSection *> outputSections); 1345ffd83dbSDimitry Andric bool hexagonNeedsTLSSymbol(ArrayRef<OutputSection *> outputSections); 1355ffd83dbSDimitry Andric 1360b57cec5SDimitry Andric class ThunkSection; 1370b57cec5SDimitry Andric class Thunk; 138e8d8bef9SDimitry Andric class InputSectionDescription; 1390b57cec5SDimitry Andric 1400b57cec5SDimitry Andric class ThunkCreator { 1410b57cec5SDimitry Andric public: 1420b57cec5SDimitry Andric // Return true if Thunks have been added to OutputSections 1430b57cec5SDimitry Andric bool createThunks(ArrayRef<OutputSection *> outputSections); 1440b57cec5SDimitry Andric 1450b57cec5SDimitry Andric // The number of completed passes of createThunks this permits us 1460b57cec5SDimitry Andric // to do one time initialization on Pass 0 and put a limit on the 1470b57cec5SDimitry Andric // number of times it can be called to prevent infinite loops. 1480b57cec5SDimitry Andric uint32_t pass = 0; 1490b57cec5SDimitry Andric 1500b57cec5SDimitry Andric private: 1510b57cec5SDimitry Andric void mergeThunks(ArrayRef<OutputSection *> outputSections); 1520b57cec5SDimitry Andric 1530b57cec5SDimitry Andric ThunkSection *getISDThunkSec(OutputSection *os, InputSection *isec, 154fe6060f1SDimitry Andric InputSectionDescription *isd, 155fe6060f1SDimitry Andric const Relocation &rel, uint64_t src); 1560b57cec5SDimitry Andric 1570b57cec5SDimitry Andric ThunkSection *getISThunkSec(InputSection *isec); 1580b57cec5SDimitry Andric 1590b57cec5SDimitry Andric void createInitialThunkSections(ArrayRef<OutputSection *> outputSections); 1600b57cec5SDimitry Andric 1610b57cec5SDimitry Andric std::pair<Thunk *, bool> getThunk(InputSection *isec, Relocation &rel, 1620b57cec5SDimitry Andric uint64_t src); 1630b57cec5SDimitry Andric 1640b57cec5SDimitry Andric ThunkSection *addThunkSection(OutputSection *os, InputSectionDescription *, 1650b57cec5SDimitry Andric uint64_t off); 1660b57cec5SDimitry Andric 1670b57cec5SDimitry Andric bool normalizeExistingThunk(Relocation &rel, uint64_t src); 1680b57cec5SDimitry Andric 169480093f4SDimitry Andric // Record all the available Thunks for a (Symbol, addend) pair, where Symbol 170480093f4SDimitry Andric // is represented as a (section, offset) pair. There may be multiple 171480093f4SDimitry Andric // relocations sharing the same (section, offset + addend) pair. We may revert 172480093f4SDimitry Andric // a relocation back to its original non-Thunk target, and restore the 173480093f4SDimitry Andric // original addend, so we cannot fold offset + addend. A nested pair is used 174480093f4SDimitry Andric // because DenseMapInfo is not specialized for std::tuple. 175480093f4SDimitry Andric llvm::DenseMap<std::pair<std::pair<SectionBase *, uint64_t>, int64_t>, 176480093f4SDimitry Andric std::vector<Thunk *>> 177480093f4SDimitry Andric thunkedSymbolsBySectionAndAddend; 178480093f4SDimitry Andric llvm::DenseMap<std::pair<Symbol *, int64_t>, std::vector<Thunk *>> 179480093f4SDimitry Andric thunkedSymbols; 1800b57cec5SDimitry Andric 1810b57cec5SDimitry Andric // Find a Thunk from the Thunks symbol definition, we can use this to find 1820b57cec5SDimitry Andric // the Thunk from a relocation to the Thunks symbol definition. 1830b57cec5SDimitry Andric llvm::DenseMap<Symbol *, Thunk *> thunks; 1840b57cec5SDimitry Andric 1850b57cec5SDimitry Andric // Track InputSections that have an inline ThunkSection placed in front 1860b57cec5SDimitry Andric // an inline ThunkSection may have control fall through to the section below 1870b57cec5SDimitry Andric // so we need to make sure that there is only one of them. 1880b57cec5SDimitry Andric // The Mips LA25 Thunk is an example of an inline ThunkSection. 1890b57cec5SDimitry Andric llvm::DenseMap<InputSection *, ThunkSection *> thunkedSections; 1900b57cec5SDimitry Andric }; 1910b57cec5SDimitry Andric 1920b57cec5SDimitry Andric // Return a int64_t to make sure we get the sign extension out of the way as 1930b57cec5SDimitry Andric // early as possible. 1940b57cec5SDimitry Andric template <class ELFT> 1950b57cec5SDimitry Andric static inline int64_t getAddend(const typename ELFT::Rel &rel) { 1960b57cec5SDimitry Andric return 0; 1970b57cec5SDimitry Andric } 1980b57cec5SDimitry Andric template <class ELFT> 1990b57cec5SDimitry Andric static inline int64_t getAddend(const typename ELFT::Rela &rel) { 2000b57cec5SDimitry Andric return rel.r_addend; 2010b57cec5SDimitry Andric } 202fe6060f1SDimitry Andric 203fe6060f1SDimitry Andric template <typename RelTy> 204fe6060f1SDimitry Andric ArrayRef<RelTy> sortRels(ArrayRef<RelTy> rels, SmallVector<RelTy, 0> &storage) { 205fe6060f1SDimitry Andric auto cmp = [](const RelTy &a, const RelTy &b) { 206fe6060f1SDimitry Andric return a.r_offset < b.r_offset; 207fe6060f1SDimitry Andric }; 208fe6060f1SDimitry Andric if (!llvm::is_sorted(rels, cmp)) { 209fe6060f1SDimitry Andric storage.assign(rels.begin(), rels.end()); 210fe6060f1SDimitry Andric llvm::stable_sort(storage, cmp); 211fe6060f1SDimitry Andric rels = storage; 212fe6060f1SDimitry Andric } 213fe6060f1SDimitry Andric return rels; 214fe6060f1SDimitry Andric } 2150b57cec5SDimitry Andric } // namespace elf 2160b57cec5SDimitry Andric } // namespace lld 2170b57cec5SDimitry Andric 2180b57cec5SDimitry Andric #endif 219