gdb/nat/fork-inferior.c

   1 /* Fork a Unix child process, and set up to debug it, for GDB and GDBserver.
   2
   3    Copyright (C) 1990-2022 Free Software Foundation, Inc.
   4
   5    This file is part of GDB.
   6
   7    This program is free software; you can redistribute it and/or modify
   8    it under the terms of the GNU General Public License as published by
   9    the Free Software Foundation; either version 3 of the License, or
  10    (at your option) any later version.
  11
  12    This program is distributed in the hope that it will be useful,
  13    but WITHOUT ANY WARRANTY; without even the implied warranty of
  14    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15    GNU General Public License for more details.
  16
  17    You should have received a copy of the GNU General Public License
  18    along with this program.  If not, see <http://www.gnu.org/licenses/>.  */
  19
  20 #include "gdbsupport/common-defs.h"
  21 #include "fork-inferior.h"
  22 #include "target/waitstatus.h"
  23 #include "gdbsupport/filestuff.h"
  24 #include "target/target.h"
  25 #include "gdbsupport/common-inferior.h"
  26 #include "gdbsupport/common-gdbthread.h"
  27 #include "gdbsupport/pathstuff.h"
  28 #include "gdbsupport/signals-state-save-restore.h"
  29 #include "gdbsupport/gdb_tilde_expand.h"
  30 #include <vector>
  31
  32 extern char **environ;
  33
  34 /* Build the argument vector for execv(3).  */
  35
  36 class execv_argv
  37 {
  38 public:
  39   /* EXEC_FILE is the file to run.  ALLARGS is a string containing the
  40      arguments to the program.  If starting with a shell, SHELL_FILE
  41      is the shell to run.  Otherwise, SHELL_FILE is NULL.  */
  42   execv_argv (const char *exec_file, const std::string &allargs,
  43               const char *shell_file);
  44
  45   /* Return a pointer to the built argv, in the type expected by
  46      execv.  The result is (only) valid for as long as this execv_argv
  47      object is live.  We return a "char **" because that's the type
  48      that the execv functions expect.  Note that it is guaranteed that
  49      the execv functions do not modify the argv[] array nor the
  50      strings to which the array point.  */
  51   char **argv ()
  52   {
  53     return const_cast<char **> (&m_argv[0]);
  54   }
  55
  56 private:
  57   DISABLE_COPY_AND_ASSIGN (execv_argv);
  58
  59   /* Helper methods for constructing the argument vector.  */
  60
  61   /* Used when building an argv for a straight execv call, without
  62      going via the shell.  */
  63   void init_for_no_shell (const char *exec_file,
  64                           const std::string &allargs);
  65
  66   /* Used when building an argv for execing a shell that execs the
  67      child program.  */
  68   void init_for_shell (const char *exec_file,
  69                        const std::string &allargs,
  70                        const char *shell_file);
  71
  72   /* The argument vector built.  Holds non-owning pointers.  Elements
  73      either point to the strings passed to the execv_argv ctor, or
  74      inside M_STORAGE.  */
  75   std::vector<const char *> m_argv;
  76
  77   /* Storage.  In the no-shell case, this contains a copy of the
  78      arguments passed to the ctor, split by '\0'.  In the shell case,
  79      this contains the quoted shell command.  I.e., SHELL_COMMAND in
  80      {"$SHELL" "-c", SHELL_COMMAND, NULL}.  */
  81   std::string m_storage;
  82 };
  83
  84 /* Create argument vector for straight call to execvp.  Breaks up
  85    ALLARGS into an argument vector suitable for passing to execvp and
  86    stores it in M_ARGV.  E.g., on "run a b c d" this routine would get
  87    as input the string "a b c d", and as output it would fill in
  88    M_ARGV with the four arguments "a", "b", "c", "d".  Each argument
  89    in M_ARGV points to a substring of a copy of ALLARGS stored in
  90    M_STORAGE.  */
  91
  92 void
  93 execv_argv::init_for_no_shell (const char *exec_file,
  94                                const std::string &allargs)
  95 {
  96
  97   /* Save/work with a copy stored in our storage.  The pointers pushed
  98      to M_ARGV point directly into M_STORAGE, which is modified in
  99      place with the necessary NULL terminators.  This avoids N heap
 100      allocations and string dups when 1 is sufficient.  */
 101   std::string &args_copy = m_storage = allargs;
 102
 103   m_argv.push_back (exec_file);
 104
 105   for (size_t cur_pos = 0; cur_pos < args_copy.size ();)
 106     {
 107       /* Skip whitespace-like chars.  */
 108       std::size_t pos = args_copy.find_first_not_of (" \t\n", cur_pos);
 109
 110       if (pos != std::string::npos)
 111         cur_pos = pos;
 112
 113       /* Find the position of the next separator.  */
 114       std::size_t next_sep = args_copy.find_first_of (" \t\n", cur_pos);
 115
 116       if (next_sep == std::string::npos)
 117         {
 118           /* No separator found, which means this is the last
 119              argument.  */
 120           next_sep = args_copy.size ();
 121         }
 122       else
 123         {
 124           /* Replace the separator with a terminator.  */
 125           args_copy[next_sep++] = '\0';
 126         }
 127
 128       m_argv.push_back (&args_copy[cur_pos]);
 129
 130       cur_pos = next_sep;
 131     }
 132
 133   /* NULL-terminate the vector.  */
 134   m_argv.push_back (NULL);
 135 }
 136
 137 /* When executing a command under the given shell, return true if the
 138    '!' character should be escaped when embedded in a quoted
 139    command-line argument.  */
 140
 141 static bool
 142 escape_bang_in_quoted_argument (const char *shell_file)
 143 {
 144   size_t shell_file_len = strlen (shell_file);
 145
 146   /* Bang should be escaped only in C Shells.  For now, simply check
 147      that the shell name ends with 'csh', which covers at least csh
 148      and tcsh.  This should be good enough for now.  */
 149
 150   if (shell_file_len < 3)
 151     return false;
 152
 153   if (shell_file[shell_file_len - 3] == 'c'
 154       && shell_file[shell_file_len - 2] == 's'
 155       && shell_file[shell_file_len - 1] == 'h')
 156     return true;
 157
 158   return false;
 159 }
 160
 161 /* See declaration.  */
 162
 163 execv_argv::execv_argv (const char *exec_file,
 164                         const std::string &allargs,
 165                         const char *shell_file)
 166 {
 167   if (shell_file == NULL)
 168     init_for_no_shell (exec_file, allargs);
 169   else
 170     init_for_shell (exec_file, allargs, shell_file);
 171 }
 172
 173 /* See declaration.  */
 174
 175 void
 176 execv_argv::init_for_shell (const char *exec_file,
 177                             const std::string &allargs,
 178                             const char *shell_file)
 179 {
 180   const char *exec_wrapper = get_exec_wrapper ();
 181
 182   /* We're going to call a shell.  */
 183   bool escape_bang = escape_bang_in_quoted_argument (shell_file);
 184
 185   /* We need to build a new shell command string, and make argv point
 186      to it.  So build it in the storage.  */
 187   std::string &shell_command = m_storage;
 188
 189   shell_command = "exec ";
 190
 191   /* Add any exec wrapper.  That may be a program name with arguments,
 192      so the user must handle quoting.  */
 193   if (exec_wrapper != NULL)
 194     {
 195       shell_command += exec_wrapper;
 196       shell_command += ' ';
 197     }
 198
 199   /* Now add exec_file, quoting as necessary.  */
 200
 201   /* Quoting in this style is said to work with all shells.  But csh
 202      on IRIX 4.0.1 can't deal with it.  So we only quote it if we need
 203      to.  */
 204   bool need_to_quote;
 205   const char *p = exec_file;
 206   while (1)
 207     {
 208       switch (*p)
 209         {
 210         case '\'':
 211         case '!':
 212         case '"':
 213         case '(':
 214         case ')':
 215         case '$':
 216         case '&':
 217         case ';':
 218         case '<':
 219         case '>':
 220         case ' ':
 221         case '\n':
 222         case '\t':
 223           need_to_quote = true;
 224           goto end_scan;
 225
 226         case '\0':
 227           need_to_quote = false;
 228           goto end_scan;
 229
 230         default:
 231           break;
 232         }
 233       ++p;
 234     }
 235  end_scan:
 236   if (need_to_quote)
 237     {
 238       shell_command += '\'';
 239       for (p = exec_file; *p != '\0'; ++p)
 240         {
 241           if (*p == '\'')
 242             shell_command += "'\\''";
 243           else if (*p == '!' && escape_bang)
 244             shell_command += "\\!";
 245           else
 246             shell_command += *p;
 247         }
 248       shell_command += '\'';
 249     }
 250   else
 251     shell_command += exec_file;
 252
 253   shell_command += ' ' + allargs;
 254
 255   /* If we decided above to start up with a shell, we exec the shell.
 256      "-c" says to interpret the next arg as a shell command to
 257      execute, and this command is "exec <target-program> <args>".  */
 258   m_argv.reserve (4);
 259   m_argv.push_back (shell_file);
 260   m_argv.push_back ("-c");
 261   m_argv.push_back (shell_command.c_str ());
 262   m_argv.push_back (NULL);
 263 }
 264
 265 /* See nat/fork-inferior.h.  */
 266
 267 pid_t
 268 fork_inferior (const char *exec_file_arg, const std::string &allargs,
 269                char **env, void (*traceme_fun) (),
 270                gdb::function_view<void (int)> init_trace_fun,
 271                void (*pre_trace_fun) (),
 272                const char *shell_file_arg,
 273                void (*exec_fun)(const char *file, char * const *argv,
 274                                 char * const *env))
 275 {
 276   pid_t pid;
 277   /* Set debug_fork then attach to the child while it sleeps, to debug.  */
 278   int debug_fork = 0;
 279   const char *shell_file;
 280   const char *exec_file;
 281   char **save_our_env;
 282   int i;
 283   int save_errno;
 284
 285   /* If no exec file handed to us, get it from the exec-file command
 286      -- with a good, common error message if none is specified.  */
 287   if (exec_file_arg == NULL)
 288     exec_file = get_exec_file (1);
 289   else
 290     exec_file = exec_file_arg;
 291
 292   /* 'startup_with_shell' is declared in inferior.h and bound to the
 293      "set startup-with-shell" option.  If 0, we'll just do a
 294      fork/exec, no shell, so don't bother figuring out what shell.  */
 295   if (startup_with_shell)
 296     {
 297       shell_file = shell_file_arg;
 298
 299       /* Figure out what shell to start up the user program under.  */
 300       if (shell_file == NULL)
 301         shell_file = get_shell ();
 302
 303       gdb_assert (shell_file != NULL);
 304     }
 305   else
 306     shell_file = NULL;
 307
 308   /* Build the argument vector.  */
 309   execv_argv child_argv (exec_file, allargs, shell_file);
 310
 311   /* Retain a copy of our environment variables, since the child will
 312      replace the value of environ and if we're vforked, we have to
 313      restore it.  */
 314   save_our_env = environ;
 315
 316   /* Perform any necessary actions regarding to TTY before the
 317      fork/vfork call.  */
 318   prefork_hook (allargs.c_str ());
 319
 320   /* It is generally good practice to flush any possible pending stdio
 321      output prior to doing a fork, to avoid the possibility of both
 322      the parent and child flushing the same data after the fork.  */
 323   gdb_flush_out_err ();
 324
 325   /* Check if the user wants to set a different working directory for
 326      the inferior.  */
 327   std::string inferior_cwd = get_inferior_cwd ();
 328
 329   if (!inferior_cwd.empty ())
 330     {
 331       /* Expand before forking because between fork and exec, the child
 332          process may only execute async-signal-safe operations.  */
 333       inferior_cwd = gdb_tilde_expand (inferior_cwd.c_str ());
 334     }
 335
 336   /* If there's any initialization of the target layers that must
 337      happen to prepare to handle the child we're about fork, do it
 338      now...  */
 339   if (pre_trace_fun != NULL)
 340     (*pre_trace_fun) ();
 341
 342   /* Create the child process.  Since the child process is going to
 343      exec(3) shortly afterwards, try to reduce the overhead by
 344      calling vfork(2).  However, if PRE_TRACE_FUN is non-null, it's
 345      likely that this optimization won't work since there's too much
 346      work to do between the vfork(2) and the exec(3).  This is known
 347      to be the case on ttrace(2)-based HP-UX, where some handshaking
 348      between parent and child needs to happen between fork(2) and
 349      exec(2).  However, since the parent is suspended in the vforked
 350      state, this doesn't work.  Also note that the vfork(2) call might
 351      actually be a call to fork(2) due to the fact that autoconf will
 352      ``#define vfork fork'' on certain platforms.  */
 353 #if !(defined(__UCLIBC__) && defined(HAS_NOMMU))
 354   if (pre_trace_fun || debug_fork)
 355     pid = fork ();
 356   else
 357 #endif
 358     pid = vfork ();
 359
 360   if (pid < 0)
 361     perror_with_name (("vfork"));
 362
 363   if (pid == 0)
 364     {
 365       /* Close all file descriptors except those that gdb inherited
 366          (usually 0/1/2), so they don't leak to the inferior.  Note
 367          that this closes the file descriptors of all secondary
 368          UIs.  */
 369       close_most_fds ();
 370
 371       /* Change to the requested working directory if the user
 372          requested it.  */
 373       if (!inferior_cwd.empty ())
 374         {
 375           if (chdir (inferior_cwd.c_str ()) < 0)
 376             trace_start_error_with_name (inferior_cwd.c_str ());
 377         }
 378
 379       if (debug_fork)
 380         sleep (debug_fork);
 381
 382       /* Execute any necessary post-fork actions before we exec.  */
 383       postfork_child_hook ();
 384
 385       /* Changing the signal handlers for the inferior after
 386          a vfork can also change them for the superior, so we don't mess
 387          with signals here.  See comments in
 388          initialize_signals for how we get the right signal handlers
 389          for the inferior.  */
 390
 391       /* "Trace me, Dr. Memory!"  */
 392       (*traceme_fun) ();
 393
 394       /* The call above set this process (the "child") as debuggable
 395         by the original gdb process (the "parent").  Since processes
 396         (unlike people) can have only one parent, if you are debugging
 397         gdb itself (and your debugger is thus _already_ the
 398         controller/parent for this child), code from here on out is
 399         undebuggable.  Indeed, you probably got an error message
 400         saying "not parent".  Sorry; you'll have to use print
 401         statements!  */
 402
 403       restore_original_signals_state ();
 404
 405       /* There is no execlpe call, so we have to set the environment
 406          for our child in the global variable.  If we've vforked, this
 407          clobbers the parent, but environ is restored a few lines down
 408          in the parent.  By the way, yes we do need to look down the
 409          path to find $SHELL.  Rich Pixley says so, and I agree.  */
 410       environ = env;
 411
 412       char **argv = child_argv.argv ();
 413
 414       if (exec_fun != NULL)
 415         (*exec_fun) (argv[0], &argv[0], env);
 416       else
 417         execvp (argv[0], &argv[0]);
 418
 419       /* If we get here, it's an error.  */
 420       save_errno = errno;
 421       warning ("Cannot exec %s", argv[0]);
 422
 423       for (i = 1; argv[i] != NULL; i++)
 424         warning (" %s", argv[i]);
 425
 426       warning ("Error: %s", safe_strerror (save_errno));
 427
 428       _exit (0177);
 429     }
 430
 431   /* Restore our environment in case a vforked child clob'd it.  */
 432   environ = save_our_env;
 433
 434   postfork_hook (pid);
 435
 436   /* Now that we have a child process, make it our target, and
 437      initialize anything target-vector-specific that needs
 438      initializing.  */
 439   if (init_trace_fun)
 440     init_trace_fun (pid);
 441
 442   /* We are now in the child process of interest, having exec'd the
 443      correct program, and are poised at the first instruction of the
 444      new program.  */
 445   return pid;
 446 }
 447
 448 /* See nat/fork-inferior.h.  */
 449
 450 ptid_t
 451 startup_inferior (process_stratum_target *proc_target, pid_t pid, int ntraps,
 452                   struct target_waitstatus *last_waitstatus,
 453                   ptid_t *last_ptid)
 454 {
 455   int pending_execs = ntraps;
 456   int terminal_initted = 0;
 457   ptid_t resume_ptid;
 458
 459   if (startup_with_shell)
 460     {
 461       /* One trap extra for exec'ing the shell.  */
 462       pending_execs++;
 463     }
 464
 465   if (target_supports_multi_process ())
 466     resume_ptid = ptid_t (pid);
 467   else
 468     resume_ptid = minus_one_ptid;
 469
 470   /* The process was started by the fork that created it, but it will
 471      have stopped one instruction after execing the shell.  Here we
 472      must get it up to actual execution of the real program.  */
 473   if (get_exec_wrapper () != NULL)
 474     pending_execs++;
 475
 476   while (1)
 477     {
 478       enum gdb_signal resume_signal = GDB_SIGNAL_0;
 479       ptid_t event_ptid;
 480
 481       struct target_waitstatus ws;
 482       event_ptid = target_wait (resume_ptid, &ws, 0);
 483
 484       if (last_waitstatus != NULL)
 485         *last_waitstatus = ws;
 486       if (last_ptid != NULL)
 487         *last_ptid = event_ptid;
 488
 489       if (ws.kind () == TARGET_WAITKIND_IGNORE)
 490         /* The inferior didn't really stop, keep waiting.  */
 491         continue;
 492
 493       switch (ws.kind ())
 494         {
 495           case TARGET_WAITKIND_SPURIOUS:
 496           case TARGET_WAITKIND_LOADED:
 497           case TARGET_WAITKIND_FORKED:
 498           case TARGET_WAITKIND_VFORKED:
 499           case TARGET_WAITKIND_SYSCALL_ENTRY:
 500           case TARGET_WAITKIND_SYSCALL_RETURN:
 501             /* Ignore gracefully during startup of the inferior.  */
 502             switch_to_thread (proc_target, event_ptid);
 503             break;
 504
 505           case TARGET_WAITKIND_SIGNALLED:
 506             target_terminal::ours ();
 507             target_mourn_inferior (event_ptid);
 508             error (_("During startup program terminated with signal %s, %s."),
 509                    gdb_signal_to_name (ws.sig ()),
 510                    gdb_signal_to_string (ws.sig ()));
 511             return resume_ptid;
 512
 513           case TARGET_WAITKIND_EXITED:
 514             target_terminal::ours ();
 515             target_mourn_inferior (event_ptid);
 516             if (ws.exit_status ())
 517               error (_("During startup program exited with code %d."),
 518                      ws.exit_status ());
 519             else
 520               error (_("During startup program exited normally."));
 521             return resume_ptid;
 522
 523           case TARGET_WAITKIND_EXECD:
 524             /* Handle EXEC signals as if they were SIGTRAP signals.  */
 525             resume_signal = GDB_SIGNAL_TRAP;
 526             switch_to_thread (proc_target, event_ptid);
 527             break;
 528
 529           case TARGET_WAITKIND_STOPPED:
 530             resume_signal = ws.sig ();
 531             switch_to_thread (proc_target, event_ptid);
 532             break;
 533         }
 534
 535       if (resume_signal != GDB_SIGNAL_TRAP)
 536         {
 537           /* Let shell child handle its own signals in its own way.  */
 538           target_continue (resume_ptid, resume_signal);
 539         }
 540       else
 541         {
 542           /* We handle SIGTRAP, however; it means child did an exec.  */
 543           if (!terminal_initted)
 544             {
 545               /* Now that the child has exec'd we know it has already
 546                  set its process group.  On POSIX systems, tcsetpgrp
 547                  will fail with EPERM if we try it before the child's
 548                  setpgid.  */
 549
 550               /* Set up the "saved terminal modes" of the inferior
 551                  based on what modes we are starting it with.  */
 552               target_terminal::init ();
 553
 554               /* Install inferior's terminal modes.  */
 555               target_terminal::inferior ();
 556
 557               terminal_initted = 1;
 558             }
 559
 560           if (--pending_execs == 0)
 561             break;
 562
 563           /* Just make it go on.  */
 564           target_continue_no_signal (resume_ptid);
 565         }
 566     }
 567
 568   return resume_ptid;
 569 }
 570
 571 /* See nat/fork-inferior.h.  */
 572
 573 void
 574 trace_start_error (const char *fmt, ...)
 575 {
 576   va_list ap;
 577
 578   va_start (ap, fmt);
 579   warning ("Could not trace the inferior process.");
 580   vwarning (fmt, ap);
 581   va_end (ap);
 582
 583   gdb_flush_out_err ();
 584   _exit (0177);
 585 }
 586
 587 /* See nat/fork-inferior.h.  */
 588
 589 void
 590 trace_start_error_with_name (const char *string)
 591 {
 592   trace_start_error ("%s: %s", string, safe_strerror (errno));
 593 }