markgc.c

   1 /*
   2  * Copyright (c) 2007-2009 Apple Inc.  All Rights Reserved.
   3  *
   4  * @APPLE_LICENSE_HEADER_START@
   5  *
   6  * This file contains Original Code and/or Modifications of Original Code
   7  * as defined in and that are subject to the Apple Public Source License
   8  * Version 2.0 (the 'License'). You may not use this file except in
   9  * compliance with the License. Please obtain a copy of the License at
  10  * http://www.opensource.apple.com/apsl/ and read it before using this
  11  * file.
  12  *
  13  * The Original Code and all software distributed under the License are
  14  * distributed on an 'AS IS' basis, WITHOUT WARRANTY OF ANY KIND, EITHER
  15  * EXPRESS OR IMPLIED, AND APPLE HEREBY DISCLAIMS ALL SUCH WARRANTIES,
  16  * INCLUDING WITHOUT LIMITATION, ANY WARRANTIES OF MERCHANTABILITY,
  17  * FITNESS FOR A PARTICULAR PURPOSE, QUIET ENJOYMENT OR NON-INFRINGEMENT.
  18  * Please see the License for the specific language governing rights and
  19  * limitations under the License.
  20  *
  21  * @APPLE_LICENSE_HEADER_END@
  22  */
  23
  24 #include <stdlib.h>
  25 #include <unistd.h>
  26 #include <string.h>
  27 #include <stdio.h>
  28 #include <stdbool.h>
  29 #include <fcntl.h>
  30 #include <limits.h>
  31 #include <sys/stat.h>
  32 #include <mach-o/fat.h>
  33 #include <mach-o/arch.h>
  34 #include <mach-o/loader.h>
  35
  36 // from "objc-private.h"
  37 // masks for objc_image_info.flags
  38 #define OBJC_IMAGE_IS_REPLACEMENT (1<<0)
  39 #define OBJC_IMAGE_SUPPORTS_GC (1<<1)
  40 #define OBJC_IMAGE_REQUIRES_GC (1<<2)
  41 #define OBJC_IMAGE_OPTIMIZED_BY_DYLD (1<<3)
  42 #define OBJC_IMAGE_SUPPORTS_COMPACTION (1<<4)
  43
  44 bool debug;
  45 bool verbose;
  46 bool quiet;
  47 bool rrOnly;
  48 bool patch = true;
  49 bool unpatch = false;
  50
  51 struct gcinfo {
  52         bool hasObjC;
  53         bool hasInfo;
  54         uint32_t flags;
  55         char *arch;
  56 } GCInfo[4];
  57
  58 void dumpinfo(char *filename);
  59
  60 int Errors = 0;
  61 char *FileBase;
  62 size_t FileSize;
  63 const char *FileName;
  64
  65 int main(int argc, char *argv[]) {
  66     //NSAutoreleasePool *pool = [NSAutoreleasePool new];
  67     int i;
  68     //dumpinfo("/System/Library/Frameworks/AppKit.framework/AppKit");
  69     if (argc == 1) {
  70         printf("Usage: markgc [-v] [-r] [--] library_or_executable_image [image2 ...]\n");
  71         printf(" changes Garbage Collection readiness of named images, ignoring those without ObjC segments\n");
  72         printf("  -p        - patch RR binary to (apparently) support GC (default)\n");
  73         printf("  -u        - unpatch GC binary to RR only\n");
  74         printf("\nAuthor: blaine@apple.com\n");
  75         exit(0);
  76     }
  77     for (i = 1; i < argc; ++i) {
  78         if (!strcmp(argv[i], "-v")) {
  79             verbose = true;
  80             continue;
  81         }
  82         if (!strcmp(argv[i], "-d")) {
  83             debug = true;
  84             continue;
  85         }
  86         if (!strcmp(argv[i], "-q")) {
  87             quiet = true;
  88             continue;
  89         }
  90         if (!strcmp(argv[i], "-p")) {
  91             patch = true;
  92             continue;
  93         }
  94         if (!strcmp(argv[i], "-u")) {
  95             unpatch = true;
  96             patch = false;
  97             continue;
  98         }
  99         dumpinfo(argv[i]);
 100     }
 101     return Errors;
 102 }
 103
 104 struct imageInfo {
 105     uint32_t version;
 106     uint32_t flags;
 107 };
 108
 109 void patchFile(uint32_t value, size_t offset) {
 110     int fd = open(FileName, 1);
 111     off_t lresult = lseek(fd, offset, SEEK_SET);
 112     if (lresult == -1) {
 113         printf("couldn't seek to 0x%lx position on fd %d\n", offset, fd);
 114         ++Errors;
 115         return;
 116     }
 117     size_t wresult = write(fd, &value, 4);
 118     if (wresult != 4) {
 119         ++Errors;
 120         printf("didn't write new value\n");
 121     }
 122     else {
 123         printf("patched %s at offset 0x%lx\n", FileName, offset);
 124     }
 125     close(fd);
 126 }
 127
 128 uint32_t iiflags(struct imageInfo *ii, size_t size, bool needsFlip) {
 129     if (needsFlip) {
 130         ii->flags = OSSwapInt32(ii->flags);
 131     }
 132     if (debug) printf("flags->%x, nitems %lu\n", ii->flags, size/sizeof(struct imageInfo));
 133     uint32_t support_mask = (OBJC_IMAGE_SUPPORTS_GC | OBJC_IMAGE_SUPPORTS_COMPACTION);
 134     uint32_t flags = ii->flags;
 135     if (patch && (flags & support_mask) != support_mask) {
 136         //printf("will patch %s at offset %p\n", FileName, (char*)(&ii->flags) - FileBase);
 137         uint32_t newvalue = flags | support_mask;
 138         if (needsFlip) newvalue = OSSwapInt32(newvalue);
 139         patchFile(newvalue, (char*)(&ii->flags) - FileBase);
 140     }
 141     if (unpatch && (flags & support_mask) == support_mask) {
 142         uint32_t newvalue = flags & ~support_mask;
 143         if (needsFlip) newvalue = OSSwapInt32(newvalue);
 144         patchFile(newvalue, (char*)(&ii->flags) - FileBase);
 145     }
 146     for(unsigned niis = 1; niis < size/sizeof(struct imageInfo); ++niis) {
 147         if (needsFlip) ii[niis].flags = OSSwapInt32(ii[niis].flags);
 148         if (ii[niis].flags != flags) {
 149             // uh, oh.
 150             printf("XXX ii[%d].flags %x != ii[0].flags %x\n", niis, ii[niis].flags, flags);
 151             ++Errors;
 152         }
 153     }
 154     return flags;
 155 }
 156
 157 void printflags(uint32_t flags) {
 158     if (flags & 0x1) printf(" F&C");
 159     if (flags & 0x2) printf(" GC");
 160     if (flags & 0x4) printf(" GC-only");
 161     else printf(" RR");
 162 }
 163
 164 /*
 165 void doimageinfo(struct imageInfo *ii, uint32_t size, bool needsFlip) {
 166     uint32_t flags = iiflags(ii, size, needsFlip);
 167     printflags(flags);
 168 }
 169 */
 170
 171
 172 void dosect32(void *start, struct section *sect, bool needsFlip, struct gcinfo *gcip) {
 173     if (debug) printf("section %s from segment %s\n", sect->sectname, sect->segname);
 174     if (strcmp(sect->segname, "__OBJC")) return;
 175     gcip->hasObjC = true;
 176     if (strcmp(sect->sectname, "__image_info")) return;
 177     gcip->hasInfo = true;
 178     if (needsFlip) {
 179         sect->offset = OSSwapInt32(sect->offset);
 180         sect->size = OSSwapInt32(sect->size);
 181     }
 182     // these guys aren't inline - they point elsewhere
 183     gcip->flags = iiflags(start + sect->offset, sect->size, needsFlip);
 184 }
 185
 186 void dosect64(void *start, struct section_64 *sect, bool needsFlip, struct gcinfo *gcip) {
 187     if (debug) printf("section %s from segment %s\n", sect->sectname, sect->segname);
 188     if (strcmp(sect->segname, "__OBJC") && strcmp(sect->segname, "__DATA")) return;
 189     if (strcmp(sect->sectname, "__image_info") && strncmp(sect->sectname, "__objc_imageinfo", 16)) return;
 190     gcip->hasObjC = true;
 191     gcip->hasInfo = true;
 192     if (needsFlip) {
 193         sect->offset = OSSwapInt32(sect->offset);
 194         sect->size = OSSwapInt64(sect->size);
 195     }
 196     // these guys aren't inline - they point elsewhere
 197     gcip->flags = iiflags(start + sect->offset, (size_t)sect->size, needsFlip);
 198 }
 199
 200 void doseg32(void *start, struct segment_command *seg, bool needsFlip, struct gcinfo *gcip) {
 201     // lets do sections
 202     if (needsFlip) {
 203         seg->fileoff = OSSwapInt32(seg->fileoff);
 204         seg->nsects = OSSwapInt32(seg->nsects);
 205     }
 206     if (debug) printf("segment name: %s, nsects %d\n", seg->segname, seg->nsects);
 207     if (seg->segname[0]) {
 208         if (strcmp("__OBJC", seg->segname)) return;
 209     }
 210     struct section *sect = (struct section *)(seg + 1);
 211     for (uint32_t nsects = 0; nsects < seg->nsects; ++nsects) {
 212         // sections directly follow
 213
 214         dosect32(start, sect + nsects, needsFlip, gcip);
 215     }
 216 }
 217 void doseg64(void *start, struct segment_command_64 *seg, bool needsFlip, struct gcinfo *gcip) {
 218     if (debug) printf("segment name: %s\n", seg->segname);
 219     if (seg->segname[0] && strcmp("__OBJC", seg->segname) && strcmp("__DATA", seg->segname)) return;
 220     gcip->hasObjC = true;
 221     // lets do sections
 222     if (needsFlip) {
 223         seg->fileoff = OSSwapInt64(seg->fileoff);
 224         seg->nsects = OSSwapInt32(seg->nsects);
 225     }
 226     struct section_64 *sect = (struct section_64 *)(seg + 1);
 227     for (uint32_t nsects = 0; nsects < seg->nsects; ++nsects) {
 228         // sections directly follow
 229
 230         dosect64(start, sect + nsects, needsFlip, gcip);
 231     }
 232 }
 233
 234 #if 0
 235 /*
 236  * A variable length string in a load command is represented by an lc_str
 237  * union.  The strings are stored just after the load command structure and
 238  * the offset is from the start of the load command structure.  The size
 239  * of the string is reflected in the cmdsize field of the load command.
 240  * Once again any padded bytes to bring the cmdsize field to a multiple
 241  * of 4 bytes must be zero.
 242  */
 243 union lc_str {
 244         uint32_t        offset; /* offset to the string */
 245 #ifndef __LP64__
 246         char            *ptr;   /* pointer to the string */
 247 #endif
 248 };
 249
 250 struct dylib {
 251     union lc_str  name;                 /* library's path name */
 252     uint32_t timestamp;                 /* library's build time stamp */
 253     uint32_t current_version;           /* library's current version number */
 254     uint32_t compatibility_version;     /* library's compatibility vers number*/
 255 };
 256
 257  * A dynamically linked shared library (filetype == MH_DYLIB in the mach header)
 258  * contains a dylib_command (cmd == LC_ID_DYLIB) to identify the library.
 259  * An object that uses a dynamically linked shared library also contains a
 260  * dylib_command (cmd == LC_LOAD_DYLIB, LC_LOAD_WEAK_DYLIB, or
 261  * LC_REEXPORT_DYLIB) for each library it uses.
 262
 263 struct dylib_command {
 264         uint32_t        cmd;            /* LC_ID_DYLIB, LC_LOAD_{,WEAK_}DYLIB,
 265                                            LC_REEXPORT_DYLIB */
 266         uint32_t        cmdsize;        /* includes pathname string */
 267         struct dylib    dylib;          /* the library identification */
 268 };
 269 #endif
 270
 271 void dodylib(void *start, struct dylib_command *dylibCmd, bool needsFlip) {
 272     if (!verbose) return;
 273     if (needsFlip) {
 274     }
 275     size_t count = dylibCmd->cmdsize - sizeof(struct dylib_command);
 276     //printf("offset is %d, count is %d\n", dylibCmd->dylib.name.offset, count);
 277     if (dylibCmd->dylib.name.offset > count) return;
 278     //printf("-->%.*s<---", count, ((void *)dylibCmd)+dylibCmd->dylib.name.offset);
 279     if (verbose) printf("load %s\n", ((char *)dylibCmd)+dylibCmd->dylib.name.offset);
 280 }
 281
 282 struct load_command *doloadcommand(void *start, struct load_command *lc, bool needsFlip, bool is32, struct gcinfo *gcip) {
 283     if (needsFlip) {
 284         lc->cmd = OSSwapInt32(lc->cmd);
 285         lc->cmdsize = OSSwapInt32(lc->cmdsize);
 286     }
 287
 288     switch(lc->cmd) {
 289     case LC_SEGMENT_64:
 290         if (debug) printf("...segment64\n");
 291         if (is32) printf("XXX we have a 64-bit segment in a 32-bit mach-o\n");
 292         doseg64(start, (struct segment_command_64 *)lc, needsFlip, gcip);
 293         break;
 294     case LC_SEGMENT:
 295         if (debug) printf("...segment32\n");
 296         doseg32(start, (struct segment_command *)lc, needsFlip, gcip);
 297         break;
 298     case LC_SYMTAB: if (debug) printf("...dynamic symtab\n"); break;
 299     case LC_DYSYMTAB: if (debug) printf("...symtab\n"); break;
 300     case LC_LOAD_DYLIB:
 301         dodylib(start, (struct dylib_command *)lc, needsFlip);
 302         break;
 303     case LC_SUB_UMBRELLA: if (debug) printf("...load subumbrella\n"); break;
 304     default:    if (debug) printf("cmd is %x\n", lc->cmd); break;
 305     }
 306
 307     return (struct load_command *)((void *)lc + lc->cmdsize);
 308 }
 309
 310 void doofile(void *start, size_t size, struct gcinfo *gcip) {
 311     struct mach_header *mh = (struct mach_header *)start;
 312     bool isFlipped = false;
 313     if (mh->magic == MH_CIGAM || mh->magic == MH_CIGAM_64) {
 314         if (debug) printf("(flipping)\n");
 315         mh->magic = OSSwapInt32(mh->magic);
 316         mh->cputype = OSSwapInt32(mh->cputype);
 317         mh->cpusubtype = OSSwapInt32(mh->cpusubtype);
 318         mh->filetype = OSSwapInt32(mh->filetype);
 319         mh->ncmds = OSSwapInt32(mh->ncmds);
 320         mh->sizeofcmds = OSSwapInt32(mh->sizeofcmds);
 321         mh->flags = OSSwapInt32(mh->flags);
 322         isFlipped = true;
 323     }
 324     if (rrOnly && mh->filetype != MH_DYLIB) return; // ignore executables
 325     NXArchInfo *info = (NXArchInfo *)NXGetArchInfoFromCpuType(mh->cputype, mh->cpusubtype);
 326     //printf("%s:", info->description);
 327     gcip->arch = (char *)info->description;
 328     //if (debug) printf("...description is %s\n", info->description);
 329     bool is32 = !(mh->cputype & CPU_ARCH_ABI64);
 330     if (debug) printf("is 32? %d\n", is32);
 331     if (debug) printf("filetype -> %d\n", mh->filetype);
 332     if (debug) printf("ncmds -> %d\n", mh->ncmds);
 333     struct load_command *lc = (is32 ? (struct load_command *)(mh + 1) : (struct load_command *)((struct mach_header_64 *)start + 1));
 334     unsigned ncmds;
 335     for (ncmds = 0; ncmds < mh->ncmds; ++ncmds) {
 336         lc = doloadcommand(start, lc, isFlipped, is32, gcip);
 337     }
 338     //printf("\n");
 339 }
 340
 341 void initGCInfo() {
 342     bzero((void *)GCInfo, sizeof(GCInfo));
 343 }
 344
 345 void printGCInfo(char *filename) {
 346     if (!GCInfo[0].hasObjC) return; // don't bother
 347     // verify that flags are all the same
 348     uint32_t flags = GCInfo[0].flags;
 349     bool allSame = true;
 350     for (int i = 1; i < 4 && GCInfo[i].arch; ++i) {
 351         if (flags != GCInfo[i].flags) {
 352             allSame = false;
 353         }
 354     }
 355     if (rrOnly) {
 356         if (allSame && (flags & 0x2))
 357             return;
 358         printf("*** not all GC in %s:\n", filename);
 359     }
 360     if (allSame && !verbose) {
 361         printf("%s:", filename);
 362         printflags(flags);
 363         printf("\n");
 364     }
 365     else {
 366         printf("%s:\n", filename);
 367         for (int i = 0; i < 4 && GCInfo[i].arch; ++i) {
 368             printf("%s:", GCInfo[i].arch);
 369             printflags(GCInfo[i].flags);
 370             printf("\n");
 371         }
 372         printf("\n");
 373     }
 374 }
 375
 376 void dofat(void *start) {
 377     struct fat_header *fh = start;
 378     bool needsFlip = false;
 379     if (fh->magic == FAT_CIGAM) {
 380         fh->nfat_arch = OSSwapInt32(fh->nfat_arch);
 381         needsFlip = true;
 382     }
 383     if (debug) printf("%d architectures\n", fh->nfat_arch);
 384     unsigned narchs;
 385     struct fat_arch *arch_ptr = (struct fat_arch *)(fh + 1);
 386     for (narchs = 0; narchs < fh->nfat_arch; ++narchs) {
 387         if (debug) printf("doing arch %d\n", narchs);
 388         if (needsFlip) {
 389             arch_ptr->offset = OSSwapInt32(arch_ptr->offset);
 390             arch_ptr->size = OSSwapInt32(arch_ptr->size);
 391         }
 392         doofile(start+arch_ptr->offset, arch_ptr->size, &GCInfo[narchs]);
 393         arch_ptr++;
 394     }
 395 }
 396
 397 bool openFile(const char *filename) {
 398     FileName = filename;
 399     // get size
 400     struct stat statb;
 401     int fd = open(filename, 0);
 402     if (fd < 0) {
 403         printf("couldn't open %s for reading\n", filename);
 404         return false;
 405     }
 406     int osresult = fstat(fd, &statb);
 407     if (osresult != 0) {
 408         printf("couldn't get size of %s\n", filename);
 409         close(fd);
 410         return false;
 411     }
 412         if ((sizeof(size_t) == 4) && ((size_t)statb.st_size > SIZE_T_MAX)) {
 413         printf("couldn't malloc %llu bytes\n", statb.st_size);
 414         close(fd);
 415         return false;
 416         }
 417     FileSize = (size_t)statb.st_size;
 418     FileBase = malloc(FileSize);
 419     if (!FileBase) {
 420         printf("couldn't malloc %lu bytes\n", FileSize);
 421         close(fd);
 422         return false;
 423     }
 424     ssize_t readsize = read(fd, FileBase, FileSize);
 425     if ((readsize == -1) || ((size_t)readsize != FileSize)) {
 426         printf("read %ld bytes, wanted %ld\n", (size_t)readsize, FileSize);
 427         close(fd);
 428         return false;
 429     }
 430     close(fd);
 431     return true;
 432 }
 433
 434 void closeFile() {
 435     free(FileBase);
 436 }
 437
 438 void dumpinfo(char *filename) {
 439     initGCInfo();
 440     if (!openFile(filename)) exit(1);
 441     struct fat_header *fh = (struct fat_header *)FileBase;
 442     if (fh->magic == FAT_MAGIC || fh->magic == FAT_CIGAM) {
 443         dofat((void *)FileBase);
 444         //printGCInfo(filename);
 445     }
 446     else if (fh->magic == MH_MAGIC || fh->magic == MH_CIGAM || fh->magic == MH_MAGIC_64 || fh->magic == MH_CIGAM_64) {
 447         doofile((void *)FileBase, FileSize, &GCInfo[0]);
 448         //printGCInfo(filename);
 449     }
 450     else if (!quiet) {
 451         printf("don't understand %s!\n", filename);
 452     }
 453     closeFile();
 454  }
 455