man5/core.5

   1 .\" Copyright (c) 2006, 2008 by Michael Kerrisk <mtk.manpages@gmail.com>
   2 .\"
   3 .\" Permission is granted to make and distribute verbatim copies of this
   4 .\" manual provided the copyright notice and this permission notice are
   5 .\" preserved on all copies.
   6 .\"
   7 .\" Permission is granted to copy and distribute modified versions of this
   8 .\" manual under the conditions for verbatim copying, provided that the
   9 .\" entire resulting derived work is distributed under the terms of a
  10 .\" permission notice identical to this one.
  11 .\"
  12 .\" Since the Linux kernel and libraries are constantly changing, this
  13 .\" manual page may be incorrect or out-of-date.  The author(s) assume no
  14 .\" responsibility for errors or omissions, or for damages resulting from
  15 .\" the use of the information contained herein.  The author(s) may not
  16 .\" have taken the same level of care in the production of this manual,
  17 .\" which is licensed free of charge, as they might when working
  18 .\" professionally.
  19 .\"
  20 .\" Formatted or processed versions of this manual, if unaccompanied by
  21 .\" the source, must acknowledge the copyright and authors of this work.
  22 .\"
  23 .TH CORE 5 2008-08-26 "Linux" "Linux Programmer's Manual"
  24 .SH NAME
  25 core \- core dump file
  26 .SH DESCRIPTION
  27 The default action of certain signals is to cause a process to terminate
  28 and produce a
  29 .IR "core dump file" ,
  30 a disk file containing an image of the process's memory at
  31 the time of termination.
  32 This image can be used in a debugger (e.g.,
  33 .BR gdb (1))
  34 to inspect the state of the program at the time that it terminated.
  35 A list of the signals which cause a process to dump core can be found in
  36 .BR signal (7).
  37
  38 A process can set its soft
  39 .B RLIMIT_CORE
  40 resource limit to place an upper limit on the size of the core dump file
  41 that will be produced if it receives a "core dump" signal; see
  42 .BR getrlimit (2)
  43 for details.
  44
  45 There are various circumstances in which a core dump file is
  46 not produced:
  47 .IP * 3
  48 The process does not have permission to write the core file.
  49 (By default the core file is called
  50 .IR core ,
  51 and is created in the current working directory.
  52 See below for details on naming.)
  53 Writing the core file will fail if the directory in which
  54 it is to be created is non-writable,
  55 or if a file with the same name exists and
  56 is not writable
  57 or is not a regular file
  58 (e.g., it is a directory or a symbolic link).
  59 .IP *
  60 A (writable, regular) file with the same name as would be used for the
  61 core dump already exists, but there is more than one hard link to that
  62 file.
  63 .IP *
  64 The file system where the core dump file would be created is full;
  65 or has run out of inodes; or is mounted read-only;
  66 or the user has reached their quota for the file system.
  67 .IP *
  68 The directory in which the core dump file is to be created does
  69 not exist.
  70 .IP *
  71 The
  72 .B RLIMIT_CORE
  73 (core file size) or
  74 .B RLIMIT_FSIZE
  75 (file size) resource limits for the process are set to zero; see
  76 .BR getrlimit (2)
  77 and the documentation of the shell's
  78 .I ulimit
  79 command
  80 .RI ( limit
  81 in
  82 .BR csh (1)).
  83 .IP *
  84 The binary being executed by the process does not have read
  85 permission enabled.
  86 .IP *
  87 The process is executing a set-user-ID (set-group-ID) program
  88 that is owned by a user (group) other than the real user (group)
  89 ID of the process.
  90 (However, see the description of the
  91 .BR prctl (2)
  92 .B PR_SET_DUMPABLE
  93 operation, and the description of the
  94 .I /proc/sys/fs/suid_dumpable
  95 .\" FIXME . Perhaps relocate discussion of /proc/sys/fs/suid_dumpable
  96 .\" and PR_SET_DUMPABLE to this page?
  97 file in
  98 .BR proc (5).)
  99 .SS Naming of core dump files
 100 By default, a core dump file is named
 101 .IR core ,
 102 but the
 103 .I /proc/sys/kernel/core_pattern
 104 file (since Linux 2.6 and 2.4.21)
 105 can be set to define a template that is used to name core dump files.
 106 The template can contain % specifiers which are substituted
 107 by the following values when a core file is created:
 108 .PP
 109 .RS 4
 110 .PD 0
 111 .TP 4
 112 %%
 113 a single % character
 114 .TP
 115 %p
 116 PID of dumped process
 117 .TP
 118 %u
 119 (numeric) real UID of dumped process
 120 .TP
 121 %g
 122 (numeric) real GID of dumped process
 123 .TP
 124 %s
 125 number of signal causing dump
 126 .TP
 127 %t
 128 time of dump, expressed as seconds since the Epoch (00:00h, 1\ Jan 1970, UTC)
 129 .TP
 130 %h
 131 hostname (same as \fInodename\fP returned by \fBuname\fP(2))
 132 .TP
 133 %e
 134 executable filename (without path prefix)
 135 .TP
 136 %c
 137 core file size soft resource limit of crashing process (since Linux 2.6.24)
 138 .PD
 139 .RE
 140 .PP
 141 A single % at the end of the template is dropped from the
 142 core filename, as is the combination of a % followed by any
 143 character other than those listed above.
 144 All other characters in the template become a literal
 145 part of the core filename.
 146 The template may include \(aq/\(aq characters, which are interpreted
 147 as delimiters for directory names.
 148 The maximum size of the resulting core filename is 128 bytes (64 bytes
 149 in kernels before 2.6.19).
 150 The default value in this file is "core".
 151 For backward compatibility, if
 152 .I /proc/sys/kernel/core_pattern
 153 does not include "%p" and
 154 .I /proc/sys/kernel/core_uses_pid
 155 (see below)
 156 is non-zero, then .PID will be appended to the core filename.
 157
 158 Since version 2.4, Linux has also provided
 159 a more primitive method of controlling
 160 the name of the core dump file.
 161 If the
 162 .I /proc/sys/kernel/core_uses_pid
 163 file contains the value 0, then a core dump file is simply named
 164 .IR core .
 165 If this file contains a non-zero value, then the core dump file includes
 166 the process ID in a name of the form
 167 .IR core.PID .
 168 .SS Piping core dumps to a program
 169 Since kernel 2.6.19, Linux supports an alternate syntax for the
 170 .I /proc/sys/kernel/core_pattern
 171 file.
 172 If the first character of this file is a pipe symbol (\fB|\fP),
 173 then the remainder of the line is interpreted as a program to be
 174 executed.
 175 Instead of being written to a disk file, the core dump is given as
 176 standard input to the program.
 177 Note the following points:
 178 .IP * 3
 179 The program must be specified using an absolute pathname (or a
 180 pathname relative to the root directory, \fI/\fP),
 181 and must immediately follow the '|' character.
 182 .IP *
 183 The process created to run the program runs as user and group
 184 .IR root .
 185 .IP *
 186 Command-line arguments can be supplied to the
 187 program (since kernel 2.6.24),
 188 delimited by white space (up to a total line length of 128 bytes).
 189 .IP *
 190 The command-line arguments can include any of
 191 the % specifiers listed above.
 192 For example, to pass the PID of the process that is being dumped, specify
 193 .I %p
 194 in an argument.
 195 .SS Controlling which mappings are written to the core dump
 196 Since kernel 2.6.23, the Linux-specific
 197 .IR /proc/PID/coredump_filter
 198 file can be used to control which memory segments are written to the
 199 core dump file in the event that a core dump is performed for the
 200 process with the corresponding process ID.
 201
 202 The value in the file is a bit mask of memory mapping types (see
 203 .BR mmap (2)).
 204 If a bit is set in the mask, then memory mappings of the
 205 corresponding type are dumped; otherwise they are not dumped.
 206 The bits in this file have the following meanings:
 207 .PP
 208 .PD 0
 209 .RS 4
 210 .TP
 211 bit 0
 212 Dump anonymous private mappings.
 213 .TP
 214 bit 1
 215 Dump anonymous shared mappings.
 216 .TP
 217 bit 2
 218 Dump file-backed private mappings.
 219 .TP
 220 bit 3
 221 Dump file-backed shared mappings.
 222 .\" file-backed shared mappings of course also update the underlying
 223 .\" mapped file.
 224 .RE
 225 .PD
 226 .PP
 227 The default value of
 228 .I coredump_filter
 229 is 0x3;
 230 this reflects traditional Linux behavior and means that
 231 only anonymous memory segments are dumped.
 232
 233 Memory-mapped I/O pages such as frame buffer are never dumped, and
 234 virtual DSO pages are always dumped, regardless of the
 235 .I coredump_filter
 236 value.
 237
 238 A child process created via
 239 .BR fork (2)
 240 inherits its parents
 241 .I coredump_filter
 242 value;
 243 the
 244 .I coredump_filter
 245 value is preserved across an
 246 .BR execve (2).
 247
 248 It can be useful to set
 249 .I coredump_filter
 250 in the parent shell before running a program, for example:
 251
 252 .in +4n
 253 .nf
 254 .RB "$" " echo 0x7 > /proc/self/coredump_filter"
 255 .RB "$" " ./some_program"
 256 .fi
 257 .in
 258 .PP
 259 This file is only provided if the kernel was built with the
 260 .B CONFIG_ELF_CORE
 261 configuration option.
 262 .SH NOTES
 263 The
 264 .BR gdb (1)
 265 .I gcore
 266 command can be used to obtain a core dump of a running process.
 267
 268 If a multithreaded process (or, more precisely, a process that
 269 shares its memory with another process by being created with the
 270 .B CLONE_VM
 271 flag of
 272 .BR clone (2))
 273 dumps core, then the process ID is always appended to the core filename,
 274 unless the process ID was already included elsewhere in the
 275 filename via a %p specification in
 276 .IR /proc/sys/kernel/core_pattern .
 277 (This is primarily useful when employing the LinuxThreads implementation,
 278 where each thread of a process has a different PID.)
 279 .\" Always including the PID in the name of the core file made
 280 .\" sense for LinuxThreads, where each thread had a unique PID,
 281 .\" but doesn't seem to serve any purpose with NPTL, where all the
 282 .\" threads in a process share the same PID (as POSIX.1 requires).
 283 .\" Probably the behavior is maintained so that applications using
 284 .\" LinuxThreads continue appending the PID (the kernel has no easy
 285 .\" way of telling which threading implementation the userspace
 286 .\" application is using). -- mtk, April 2006
 287 .SH EXAMPLE
 288 The program below can be used to demonstrate the use of the
 289 pipe syntax in the
 290 .I /proc/sys/kernel/core_pattern
 291 file.
 292 The following shell session demonstrates the use of this program
 293 (compiled to create an executable named
 294 .IR core_pattern_pipe_test ):
 295 .PP
 296 .in +4n
 297 .nf
 298 .RB "$" " cc \-o core_pattern_pipe_test core_pattern_pipe_test.c"
 299 .RB "$" " su"
 300 Password:
 301 .RB "#" " echo \(aq|$PWD/core_pattern_pipe_test %p \
 302 UID=%u GID=%g sig=%s\(aq > \e"
 303 .B "    /proc/sys/kernel/core_pattern"
 304 .RB "#" " exit"
 305 .RB "$" " sleep 100"
 306 .BR "^\e" "                     # type control-backslash"
 307 Quit (core dumped)
 308 .RB "$" " cat core.info"
 309 argc=5
 310 argc[0]=</home/mtk/core_pattern_pipe_test>
 311 argc[1]=<20575>
 312 argc[2]=<UID=1000>
 313 argc[3]=<GID=100>
 314 argc[4]=<sig=3>
 315 Total bytes in core dump: 282624
 316 .fi
 317 .in
 318 .SS Program source
 319 .R " "
 320 .nf
 321 /* core_pattern_pipe_test.c */
 322
 323 #define _GNU_SOURCE
 324 #include <sys/stat.h>
 325 #include <fcntl.h>
 326 #include <limits.h>
 327 #include <stdio.h>
 328 #include <stdlib.h>
 329 #include <unistd.h>
 330
 331 #define BUF_SIZE 1024
 332
 333 int
 334 main(int argc, char *argv[])
 335 {
 336     int tot, j;
 337     ssize_t nread;
 338     char buf[BUF_SIZE];
 339     FILE *fp;
 340     char cwd[PATH_MAX];
 341
 342     /* Change our current working directory to that of the
 343        crashing process */
 344
 345     snprintf(cwd, PATH_MAX, "/proc/%s/cwd", argv[1]);
 346     chdir(cwd);
 347
 348     /* Write output to file "core.info" in that directory */
 349
 350     fp = fopen("core.info", "w+");
 351     if (fp == NULL)
 352         exit(EXIT_FAILURE);
 353
 354     /* Display command\-line arguments given to core_pattern
 355        pipe program */
 356
 357     fprintf(fp, "argc=%d\\n", argc);
 358     for (j = 0; j < argc; j++)
 359         fprintf(fp, "argc[%d]=<%s>\\n", j, argv[j]);
 360
 361     /* Count bytes in standard input (the core dump) */
 362
 363     tot = 0;
 364     while ((nread = read(STDIN_FILENO, buf, BUF_SIZE)) > 0)
 365         tot += nread;
 366     fprintf(fp, "Total bytes in core dump: %d\\n", tot);
 367
 368     exit(EXIT_SUCCESS);
 369 }
 370 .fi
 371 .SH SEE ALSO
 372 .BR bash (1),
 373 .BR gdb (1),
 374 .BR getrlimit (2),
 375 .BR mmap (2),
 376 .BR prctl (2),
 377 .BR sigaction (2),
 378 .BR elf (5),
 379 .BR proc (5),
 380 .BR pthreads (7),
 381 .BR signal (7)