src/dyldInitialization.cpp

   1 /* -*- mode: C++; c-basic-offset: 4; tab-width: 4 -*-
   2  *
   3  * Copyright (c) 2004-2005 Apple Computer, Inc. All rights reserved.
   4  *
   5  * @APPLE_LICENSE_HEADER_START@
   6  *
   7  * This file contains Original Code and/or Modifications of Original Code
   8  * as defined in and that are subject to the Apple Public Source License
   9  * Version 2.0 (the 'License'). You may not use this file except in
  10  * compliance with the License. Please obtain a copy of the License at
  11  * http://www.opensource.apple.com/apsl/ and read it before using this
  12  * file.
  13  *
  14  * The Original Code and all software distributed under the License are
  15  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  16  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  17  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  18  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  19  * Please see the License for the specific language governing rights and
  20  * limitations under the License.
  21  *
  22  * @APPLE_LICENSE_HEADER_END@
  23  */
  24
  25 #include <stddef.h>
  26 #include <string.h>
  27 #include <mach/mach.h>
  28 #include <mach-o/loader.h>
  29 #include <mach-o/ldsyms.h>
  30 #include <mach-o/reloc.h>
  31 #if __ppc__ || __ppc64__
  32         #include <mach-o/ppc/reloc.h>
  33 #endif
  34 #if __x86_64__
  35         #include <mach-o/x86_64/reloc.h>
  36 #endif
  37 #include "dyld.h"
  38
  39 #if __LP64__
  40         #define macho_header                    mach_header_64
  41         #define LC_SEGMENT_COMMAND              LC_SEGMENT_64
  42         #define macho_segment_command   segment_command_64
  43         #define macho_section                   section_64
  44         #define RELOC_SIZE                              3
  45 #else
  46         #define macho_header                    mach_header
  47         #define LC_SEGMENT_COMMAND              LC_SEGMENT
  48         #define macho_segment_command   segment_command
  49         #define macho_section                   section
  50         #define RELOC_SIZE                              2
  51 #endif
  52
  53 #if __x86_64__
  54         #define POINTER_RELOC X86_64_RELOC_UNSIGNED
  55 #else
  56         #define POINTER_RELOC GENERIC_RELOC_VANILLA
  57 #endif
  58
  59 //
  60 //  Code to bootstrap dyld into a runnable state
  61 //
  62 //
  63
  64 namespace dyldbootstrap {
  65
  66
  67 typedef void (*Initializer)(int argc, const char* argv[], const char* envp[], const char* apple[]);
  68
  69 //
  70 // For a regular executable, the crt code calls dyld to run the executables initializers.
  71 // For a static executable, crt directly runs the initializers.
  72 // dyld (should be static) but is a dynamic executable and needs this hack to run its own initializers.
  73 // We pass argc, argv, etc in case libc.a uses those arguments
  74 //
  75 static void runDyldInitializers(const struct macho_header* mh, intptr_t slide, int argc, const char* argv[], const char* envp[], const char* apple[])
  76 {
  77         const uint32_t cmd_count = mh->ncmds;
  78         const struct load_command* const cmds = (struct load_command*)(((char*)mh)+sizeof(macho_header));
  79         const struct load_command* cmd = cmds;
  80         for (uint32_t i = 0; i < cmd_count; ++i) {
  81                 switch (cmd->cmd) {
  82                         case LC_SEGMENT_COMMAND:
  83                                 {
  84                                         const struct macho_segment_command* seg = (struct macho_segment_command*)cmd;
  85                                         const struct macho_section* const sectionsStart = (struct macho_section*)((char*)seg + sizeof(struct macho_segment_command));
  86                                         const struct macho_section* const sectionsEnd = &sectionsStart[seg->nsects];
  87                                         for (const struct macho_section* sect=sectionsStart; sect < sectionsEnd; ++sect) {
  88                                                 const uint8_t type = sect->flags & SECTION_TYPE;
  89                                                 if ( type == S_MOD_INIT_FUNC_POINTERS ){
  90                                                         Initializer* inits = (Initializer*)(sect->addr + slide);
  91                                                         const uint32_t count = sect->size / sizeof(uintptr_t);
  92                                                         for (uint32_t i=0; i < count; ++i) {
  93                                                                 Initializer func = inits[i];
  94                                                                 func(argc, argv, envp, apple);
  95                                                         }
  96                                                 }
  97                                         }
  98                                 }
  99                                 break;
 100                 }
 101                 cmd = (const struct load_command*)(((char*)cmd)+cmd->cmdsize);
 102         }
 103 }
 104
 105 //
 106 // If the kernel does not load dyld at its preferred address, we need to apply
 107 // fixups to various initialized parts of the __DATA segment
 108 //
 109 static void rebaseDyld(const struct macho_header* mh, intptr_t slide)
 110 {
 111         // rebase non-lazy pointers (which all point internal to dyld, since dyld uses no shared libraries)
 112         // and get interesting pointers into dyld
 113         const uint32_t cmd_count = mh->ncmds;
 114         const struct load_command* const cmds = (struct load_command*)(((char*)mh)+sizeof(macho_header));
 115         const struct load_command* cmd = cmds;
 116         const struct macho_segment_command* linkEditSeg = NULL;
 117 #if __x86_64__
 118         const struct macho_segment_command* firstWritableSeg = NULL;
 119 #endif
 120         const struct dysymtab_command* dynamicSymbolTable = NULL;
 121         for (uint32_t i = 0; i < cmd_count; ++i) {
 122                 switch (cmd->cmd) {
 123                         case LC_SEGMENT_COMMAND:
 124                                 {
 125                                         const struct macho_segment_command* seg = (struct macho_segment_command*)cmd;
 126                                         if ( strcmp(seg->segname, "__LINKEDIT") == 0 )
 127                                                 linkEditSeg = seg;
 128                                         const struct macho_section* const sectionsStart = (struct macho_section*)((char*)seg + sizeof(struct macho_segment_command));
 129                                         const struct macho_section* const sectionsEnd = &sectionsStart[seg->nsects];
 130                                         for (const struct macho_section* sect=sectionsStart; sect < sectionsEnd; ++sect) {
 131                                                 const uint8_t type = sect->flags & SECTION_TYPE;
 132                                                 if ( type == S_NON_LAZY_SYMBOL_POINTERS ) {
 133                                                         // rebase non-lazy pointers (which all point internal to dyld, since dyld uses no shared libraries)
 134                                                         const uint32_t pointerCount = sect->size / sizeof(uintptr_t);
 135                                                         uintptr_t* const symbolPointers = (uintptr_t*)(sect->addr + slide);
 136                                                         for (uint32_t j=0; j < pointerCount; ++j) {
 137                                                                 symbolPointers[j] += slide;
 138                                                         }
 139                                                 }
 140                                         }
 141 #if __x86_64__
 142                                         if ( (firstWritableSeg == NULL) && (seg->initprot & VM_PROT_WRITE) )
 143                                                 firstWritableSeg = seg;
 144 #endif
 145                                 }
 146                                 break;
 147                         case LC_DYSYMTAB:
 148                                 dynamicSymbolTable = (struct dysymtab_command *)cmd;
 149                                 break;
 150                 }
 151                 cmd = (const struct load_command*)(((char*)cmd)+cmd->cmdsize);
 152         }
 153
 154         // use reloc's to rebase all random data pointers
 155 #if __x86_64__
 156         const uintptr_t relocBase = firstWritableSeg->vmaddr + slide;
 157 #else
 158         const uintptr_t relocBase = (uintptr_t)mh;
 159 #endif
 160         const relocation_info* const relocsStart = (struct relocation_info*)(linkEditSeg->vmaddr + slide + dynamicSymbolTable->locreloff - linkEditSeg->fileoff);
 161         const relocation_info* const relocsEnd = &relocsStart[dynamicSymbolTable->nlocrel];
 162         for (const relocation_info* reloc=relocsStart; reloc < relocsEnd; ++reloc) {
 163         #if __ppc__ || __ppc64__ || __i36__
 164                 if ( (reloc->r_address & R_SCATTERED) != 0 )
 165                         throw "scattered relocation in dyld";
 166         #endif
 167                 if ( reloc->r_length != RELOC_SIZE )
 168                         throw "relocation in dyld has wrong size";
 169
 170                 if ( reloc->r_type != POINTER_RELOC )
 171                         throw "relocation in dyld has wrong type";
 172
 173                 // update pointer by amount dyld slid
 174                 *((uintptr_t*)(reloc->r_address + relocBase)) += slide;
 175         }
 176 }
 177
 178
 179 //
 180 // For some reason the kernel loads dyld with __TEXT and __LINKEDIT writable
 181 // rdar://problem/3702311
 182 //
 183 static void segmentProtectDyld(const struct macho_header* mh, intptr_t slide)
 184 {
 185         const uint32_t cmd_count = mh->ncmds;
 186         const struct load_command* const cmds = (struct load_command*)(((char*)mh)+sizeof(macho_header));
 187         const struct load_command* cmd = cmds;
 188         for (uint32_t i = 0; i < cmd_count; ++i) {
 189                 switch (cmd->cmd) {
 190                         case LC_SEGMENT_COMMAND:
 191                                 {
 192                                         const struct macho_segment_command* seg = (struct macho_segment_command*)cmd;
 193                                         vm_address_t addr = seg->vmaddr + slide;
 194                                         vm_size_t size = seg->vmsize;
 195                                         const bool setCurrentPermissions = false;
 196                                         vm_protect(mach_task_self(), addr, size, setCurrentPermissions, seg->initprot);
 197                                         //fprintf(stderr, "dyld: segment %s, 0x%08X -> 0x%08X, set to %d\n", seg->segname, addr, addr+size-1, seg->initprot);
 198                                 }
 199                                 break;
 200                 }
 201                 cmd = (const struct load_command*)(((char*)cmd)+cmd->cmdsize);
 202         }
 203
 204 }
 205
 206 extern "C" void dyld_exceptions_init(const struct macho_header*, uintptr_t slide); // in dyldExceptions.cpp
 207 extern "C" void mach_init();
 208
 209 //
 210 //  This is code to bootstrap dyld.  This work in normally done for a program by dyld and crt.
 211 //  In dyld we have to do this manually.
 212 //
 213 uintptr_t start(const struct mach_header* appsMachHeader, int argc, const char* argv[], intptr_t slide)
 214 {
 215         // _mh_dylinker_header is magic symbol defined by static linker (ld), see <mach-o/ldsyms.h>
 216         const struct macho_header* dyldsMachHeader =  (const struct macho_header*)(((char*)&_mh_dylinker_header)+slide);
 217
 218         // if kernel had to slide dyld, we need to fix up load sensitive locations
 219         // we have to do this before using any global variables
 220         if ( slide != 0 ) {
 221                 rebaseDyld(dyldsMachHeader, slide);
 222         }
 223
 224         // enable C++ exceptions to work inside dyld
 225         dyld_exceptions_init(dyldsMachHeader, slide);
 226
 227         // allow dyld to use mach messaging
 228         mach_init();
 229
 230         // set protection on segments (has to be done after mach_init)
 231         segmentProtectDyld(dyldsMachHeader, slide);
 232
 233         // kernel sets up env pointer to be just past end of agv array
 234         const char** envp = &argv[argc+1];
 235
 236         // kernel sets up apple pointer to be just past end of envp array
 237         const char** apple = envp;
 238         while(*apple != NULL) { ++apple; }
 239         ++apple;
 240
 241         // run all C++ initializers inside dyld
 242         runDyldInitializers(dyldsMachHeader, slide, argc, argv, envp, apple);
 243
 244         // now that we are done bootstrapping dyld, call dyld's main
 245         return dyld::_main(appsMachHeader, argc, argv, envp, apple);
 246 }
 247
 248
 249
 250
 251 } // end of namespace
 252
 253
 254
 255