src/unexsunos4.c

   1 /* Unexec for Sunos 4 using shared libraries.
   2    Copyright (C) 1990, 1994 Free Software Foundation, Inc.
   3
   4 This file is part of GNU Emacs.
   5
   6 GNU Emacs is free software; you can redistribute it and/or modify
   7 it under the terms of the GNU General Public License as published by
   8 the Free Software Foundation; either version 2, or (at your option)
   9 any later version.
  10
  11 GNU Emacs is distributed in the hope that it will be useful,
  12 but WITHOUT ANY WARRANTY; without even the implied warranty of
  13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  14 GNU General Public License for more details.
  15
  16 You should have received a copy of the GNU General Public License
  17 along with GNU Emacs; see the file COPYING.  If not, write to
  18 the Free Software Foundation, 675 Mass Ave, Cambridge, MA 02139, USA.  */
  19
  20 /* Contributed by Viktor Dukhovni.  */
  21 /*
  22  * Unexec for Berkeley a.out format + SUNOS shared libraries
  23  * The unexeced executable contains the __DYNAMIC area from the
  24  * original text file,  and then the rest of data + bss + malloced area of
  25  * the current process.  (The __DYNAMIC area is at the top of the process
  26  * data segment,  we use "data_start" defined externally to mark the start
  27  * of the "real" data segment.)
  28  *
  29  * For programs that want to remap some of the data segment read only
  30  * a run_time_remap is provided.  This attempts to remap largest area starting
  31  * and ending on page boundaries between "data_start" and "bndry"
  32  * For this it to figure out where the text file is located.  A path search
  33  * is attempted after trying argv[0] and if all fails we simply do not remap
  34  *
  35  * One feature of run_time_remap () is mandatory:  reseting the break.
  36  *
  37  *  Note that we can no longer map data into the text segment,  as this causes
  38  *  the __DYNAMIC struct to become read only,  breaking the runtime loader.
  39  *  Thus we no longer need to mess with a private crt0.c,  the standard one
  40  *  will do just fine,  since environ can live in the writable area between
  41  *  __DYNAMIC and data_start,  just make sure that pre-crt0.o (the name
  42  *  is somewhat abused here) is loaded first!
  43  *
  44  */
  45 #include <sys/param.h>
  46 #include <sys/mman.h>
  47 #include <sys/file.h>
  48 #include <sys/stat.h>
  49 #include <string.h>
  50 #include <stdio.h>
  51 #include <a.out.h>
  52
  53 /* Do this after the above #include's in case a configuration file wants
  54    to define things for this file based on what <a.out.h> defines.  */
  55 #ifdef emacs
  56 #include <config.h>
  57 #endif
  58
  59 #if defined (SUNOS4) || defined (__FreeBSD__)
  60 #define UNDO_RELOCATION
  61 #endif
  62
  63 #ifdef UNDO_RELOCATION
  64 #include <link.h>
  65 #endif
  66
  67 #ifdef HAVE_UNISTD_H
  68 #include <unistd.h>
  69 #endif
  70
  71 /* NetBSD needs this bit, but SunOS does not have it.  */
  72 #ifndef MAP_FILE
  73 #define MAP_FILE 0
  74 #endif
  75
  76
  77 /*
  78  * for programs other than emacs
  79  * define data_start + initialized here,  and make sure
  80  * this object is loaded first!
  81  * emacs will define these elsewhere,  and load the object containing
  82  * data_start (pre-crt0.o or firstfile.o?) first!
  83  * The custom crt0.o *must not* be loaded!
  84  */
  85 #ifndef emacs
  86   static int data_start = 0;
  87   static int initialized = 0;
  88 #else
  89   extern int initialized;
  90   extern unsigned data_start;
  91   extern int pureptr;
  92 #endif
  93
  94 extern char *getenv ();
  95 static unsigned brk_value;
  96 static struct exec nhdr;
  97 static int rd_only_len;
  98 static long cookie;
  99
 100
 101 unexec (new_name, a_name, bndry, bss_start, entry)
 102      char *new_name, *a_name;
 103      unsigned bndry, bss_start, entry;
 104 {
 105   int fd, new;
 106   char *old;
 107   struct exec ohdr;             /* Allocate on the stack,  not needed in the next life */
 108   struct stat stat;
 109
 110   if ((fd = open (a_name, O_RDONLY)) < 0)
 111     {
 112       fprintf (stderr, "%s: open: ", a_name);
 113       perror (a_name);
 114       exit (1);
 115     }
 116   if ((new = open (new_name, O_WRONLY | O_CREAT, 0666)) == -1)
 117     {
 118       fprintf (stderr, "%s: open: ", a_name);
 119       perror (new_name);
 120       exit (1);
 121     }
 122
 123   if ((fstat (fd, &stat) == -1))
 124     {
 125       fprintf (stderr, "%s: ", a_name);
 126       perror ("fstat");
 127       exit (1);
 128     }
 129
 130   old = (char *)mmap (0, stat.st_size, PROT_READ, MAP_FILE|MAP_SHARED, fd, 0);
 131   if (old == (char *)-1)
 132     {
 133       fprintf (stderr, "%s: ", a_name);
 134       perror ("mmap");
 135       exit (1);
 136     }
 137   close (fd);
 138
 139   nhdr = ohdr = (*(struct exec *)old);
 140
 141
 142   /*
 143    * Remember a magic cookie so we know we've got the right binary
 144    * when remapping.
 145    */
 146   cookie = time (0);
 147
 148   /* Save the break, it is reset to &_end (by ld.so?).  */
 149   brk_value = (unsigned) sbrk (0);
 150
 151   /*
 152    * Round up data start to a page boundary (Lose if not a 2 power!)
 153    */
 154   data_start = ((((int)&data_start) - 1) & ~(N_PAGSIZ (nhdr) - 1)) + N_PAGSIZ (nhdr);
 155
 156   /*
 157    * Round down read only pages to a multiple of the page size
 158    */
 159   if (bndry)
 160     rd_only_len = ((int)bndry & ~(N_PAGSIZ (nhdr) - 1)) - data_start;
 161
 162 #ifndef emacs
 163   /* Have to do this some time before dumping the data */
 164   initialized = 1;
 165 #endif
 166
 167   /* Handle new data and bss sizes and optional new entry point.
 168      No one actually uses bss_start and entry,  but tradition compels
 169      one to support them.
 170      Could complain if bss_start > brk_value,
 171      but the caller is *supposed* to know what she is doing.  */
 172   nhdr.a_data = (bss_start ? bss_start : brk_value) - N_DATADDR (nhdr);
 173   nhdr.a_bss  = bss_start ? brk_value - bss_start : 0;
 174   if (entry)
 175     nhdr.a_entry = entry;
 176
 177   /*
 178    * Write out the text segment with new header
 179    * Dynamic executables are ZMAGIC with N_TXTOFF==0 and the header
 180    * part of the text segment, but no need to rely on this.
 181    * So write the TEXT first,  then go back replace the header.
 182    * Doing it in the other order is less general!
 183    */
 184   lseek (new, N_TXTOFF (nhdr), L_SET);
 185   write (new, old + N_TXTOFF (ohdr), N_TXTOFF (ohdr) + ohdr.a_text);
 186   lseek (new, 0L, L_SET);
 187   write (new, &nhdr, sizeof (nhdr));
 188
 189   /*
 190    * Write out the head of the old data segment from the file not
 191    * from core, this has the unresolved __DYNAMIC relocation data
 192    * we need to reload
 193    */
 194   lseek (new, N_DATOFF (nhdr), L_SET);
 195   write (new, old + N_DATOFF (ohdr), (int)&data_start - N_DATADDR (ohdr));
 196
 197   /*
 198    * Copy the rest of the data from core
 199    */
 200   write (new, &data_start, N_BSSADDR (nhdr) - (int)&data_start);
 201
 202   /*
 203    * Copy the symbol table and line numbers
 204    */
 205   lseek (new, N_TRELOFF (nhdr), L_SET);
 206   write (new, old + N_TRELOFF (ohdr), stat.st_size - N_TRELOFF (ohdr));
 207
 208   /* Some other BSD systems use this file.
 209      We don't know whether this change is right for them.  */
 210 #ifdef UNDO_RELOCATION
 211   /* Undo the relocations done at startup by ld.so.
 212      It will do these relocations again when we start the dumped Emacs.
 213      Doing them twice gives incorrect results.  */
 214   {
 215     unsigned long daddr = N_DATADDR (ohdr);
 216     unsigned long rel, erel;
 217 #ifdef SUNOS4
 218     extern struct link_dynamic _DYNAMIC;
 219
 220     /*  SunOS4.x's ld_rel is relative to N_TXTADDR. */
 221     if (_DYNAMIC.ld_version < 2)
 222       {
 223         rel = _DYNAMIC.ld_un.ld_1->ld_rel + N_TXTADDR (ohdr);
 224         erel = _DYNAMIC.ld_un.ld_1->ld_hash + N_TXTADDR (ohdr);
 225       }
 226     else
 227       {
 228         rel = _DYNAMIC.ld_un.ld_2->ld_rel + N_TXTADDR (ohdr);
 229         erel = _DYNAMIC.ld_un.ld_2->ld_hash + N_TXTADDR (ohdr);
 230       }
 231 #ifdef sparc
 232 #define REL_INFO_TYPE           struct reloc_info_sparc
 233 #else
 234 #define REL_INFO_TYPE           struct relocation_info
 235 #endif /* sparc */
 236 #define REL_TARGET_ADDRESS(r)   (((REL_INFO_TYPE *)(r))->r_address)
 237 #endif /* SUNOS4 */
 238 #ifdef __FreeBSD__
 239     extern struct _dynamic _DYNAMIC;
 240
 241     /*  FreeBSD's LD_REL is a virtual address itself. */
 242     rel = LD_REL (&_DYNAMIC);
 243     erel = rel + LD_RELSZ (&_DYNAMIC);
 244 #define REL_INFO_TYPE           struct relocation_info
 245 #define REL_TARGET_ADDRESS(r)   (((REL_INFO_TYPE *)(r))->r_address)
 246 #endif
 247
 248     for (; rel < erel; rel += sizeof (REL_INFO_TYPE))
 249       {
 250         /*  This is the virtual address where ld.so will do relocation.  */
 251         unsigned long target = REL_TARGET_ADDRESS (rel);
 252         /*  This is the offset in the data segment.  */
 253         unsigned long segoffset = target - daddr;
 254
 255         /*  If it is located below data_start, we have to do nothing here,
 256             because the old data has been already written to the location. */
 257         if (target < (unsigned long)&data_start)
 258             continue;
 259
 260         lseek (new, N_DATOFF (nhdr) + segoffset, L_SET);
 261         write (new, old + N_DATOFF (ohdr) + segoffset, sizeof (unsigned long));
 262       }
 263   }
 264 #endif /* UNDO_RELOCATION */
 265
 266   fchmod (new, 0755);
 267 }
 268
 269 void
 270 run_time_remap (progname)
 271      char *progname;
 272 {
 273   char aout[MAXPATHLEN];
 274   register char *path, *p;
 275
 276   /* Just in case */
 277   if (!initialized)
 278     return;
 279
 280   /* Restore the break */
 281   brk ((char *) brk_value);
 282
 283   /*  If nothing to remap:  we are done! */
 284   if (rd_only_len == 0)
 285     return;
 286
 287   /*
 288    * Attempt to find the executable
 289    * First try argv[0],  will almost always succeed as shells tend to give
 290    * the full path from the hash list rather than using execvp ()
 291    */
 292   if (is_it (progname))
 293     return;
 294
 295   /*
 296    * If argv[0] is a full path and does not exist,  not much sense in
 297    * searching further
 298    */
 299   if (strchr (progname, '/'))
 300     return;
 301
 302   /*
 303    * Try to search for  argv[0] on the PATH
 304    */
 305   path = getenv ("PATH");
 306   if (path == NULL)
 307     return;
 308
 309   while (*path)
 310     {
 311       /* copy through ':' or end */
 312       for (p = aout; *p = *path; ++p, ++path)
 313         if (*p == ':')
 314           {
 315             ++path;             /* move past ':' */
 316             break;
 317           }
 318       *p++ = '/';
 319       strcpy (p, progname);
 320       /*
 321        * aout is a candidate full path name
 322        */
 323       if (is_it (aout))
 324         return;
 325     }
 326 }
 327
 328 is_it (filename)
 329   char *filename;
 330 {
 331   int fd;
 332   long filenames_cookie;
 333   struct exec hdr;
 334
 335   /*
 336    * Open an executable  and check for a valid header!
 337    * Can't bcmp the header with what we had,  it may have been stripped!
 338    * so we may save looking at non executables with the same name, mostly
 339    * directories.
 340    */
 341   fd = open (filename, O_RDONLY);
 342   if (fd != -1)
 343     {
 344       if (read (fd, &hdr, sizeof (hdr)) == sizeof (hdr)
 345           && !N_BADMAG (hdr) && N_DATOFF (hdr) == N_DATOFF (nhdr)
 346           && N_TRELOFF (hdr) == N_TRELOFF (nhdr))
 347         {
 348           /* compare cookies */
 349           lseek (fd, N_DATOFF (hdr) + (int)&cookie - N_DATADDR (hdr), L_SET);
 350           read (fd, &filenames_cookie, sizeof (filenames_cookie));
 351           if (filenames_cookie == cookie)
 352             {                   /* Eureka */
 353
 354               /*
 355                * Do the mapping
 356                * The PROT_EXEC may not be needed,  but it is safer this way.
 357                * should the shared library decide to indirect through
 358                * addresses in the data segment not part of __DYNAMIC
 359                */
 360               mmap ((char *) data_start, rd_only_len, PROT_READ | PROT_EXEC,
 361                     MAP_FILE | MAP_SHARED | MAP_FIXED, fd,
 362                     N_DATOFF (hdr) + data_start - N_DATADDR (hdr));
 363               close (fd);
 364               return 1;
 365             }
 366         }
 367       close (fd);
 368     }
 369   return 0;
 370 }