Naming of core dump files¶

By default, a core dump file is named core, but the /proc/sys/kernel/core_pattern file (since Linux 2.6 and 2.4.21) can be set to define a template that is used to name core dump files. The template can contain % specifiers which are substituted by the following values when a core file is created:

A single % at the end of the template is dropped from the core filename, as is the combination of a % followed by any character other than those listed above. All other characters in the template become a literal part of the core filename. The template may include '/' characters, which are interpreted as delimiters for directory names. The maximum size of the resulting core filename is 128 bytes (64 bytes in kernels before 2.6.19). The default value in this file is "core". For backward compatibility, if /proc/sys/kernel/core_pattern does not include %p and /proc/sys/kernel/core_uses_pid (see below) is nonzero, then .PID will be appended to the core filename.

Paths are interpreted according to the settings that are active for the crashing process. That means the crashing process's mount namespace (see mount_namespaces(7)), its current working directory (found via getcwd(2)), and its root directory (see chroot(2)).

Since version 2.4, Linux has also provided a more primitive method of controlling the name of the core dump file. If the /proc/sys/kernel/core_uses_pid file contains the value 0, then a core dump file is simply named core. If this file contains a nonzero value, then the core dump file includes the process ID in a name of the form core.PID.

Since Linux 3.6, if /proc/sys/fs/suid_dumpable is set to 2 ("suidsafe"), the pattern must be either an absolute pathname (starting with a leading '/' character) or a pipe, as defined below.

Piping core dumps to a program¶

Since kernel 2.6.19, Linux supports an alternate syntax for the /proc/sys/kernel/core_pattern file. If the first character of this file is a pipe symbol (|), then the remainder of the line is interpreted as the command-line for a user-space program (or script) that is to be executed. Instead of being written to a disk file, the core dump is given as standard input to the program. Note the following points:

• The program must be specified using an absolute pathname (or a pathname relative to the root directory, /), and must immediately follow the '|' character.
• The command-line arguments can include any of the % specifiers listed above. For example, to pass the PID of the process that is being dumped, specify %p in an argument.
• The process created to run the program runs as user and group root.
• Running as root does not confer any exceptional security bypasses. Namely, LSMs (e.g., SELinux) are still active and may prevent the handler from accessing details about the crashed process via /proc/[pid].
• The program pathname is interpreted with respect to the initial mount namespace as it is always executed there. It is not affected by the settings (e.g., root directory, mount namespace, current working directory) of the crashing process.
• The process runs in the initial namespaces (PID, mount, user, and so on) and not in the namespaces of the crashing process. One can utilize specifiers such as %P to find the right /proc/[pid] directory and probe/enter the crashing process's namespaces if needed.
• The process starts with its current working directory as the root directory. If desired, it is possible change to the working directory of the dumping process by employing the value provided by the %P specifier to change to the location of the dumping process via /proc/[pid]/cwd.
• Command-line arguments can be supplied to the program (since Linux 2.6.24), delimited by white space (up to a total line length of 128 bytes).
• The RLIMIT_CORE limit is not enforced for core dumps that are piped to a program via this mechanism.

/proc/sys/kernel/core_pipe_limit¶

When collecting core dumps via a pipe to a user-space program, it can be useful for the collecting program to gather data about the crashing process from that process's /proc/[pid] directory. In order to do this safely, the kernel must wait for the program collecting the core dump to exit, so as not to remove the crashing process's /proc/[pid] files prematurely. This in turn creates the possibility that a misbehaving collecting program can block the reaping of a crashed process by simply never exiting.

Since Linux 2.6.32, the /proc/sys/kernel/core_pipe_limit can be used to defend against this possibility. The value in this file defines how many concurrent crashing processes may be piped to user-space programs in parallel. If this value is exceeded, then those crashing processes above this value are noted in the kernel log and their core dumps are skipped.

A value of 0 in this file is special. It indicates that unlimited processes may be captured in parallel, but that no waiting will take place (i.e., the collecting program is not guaranteed access to /proc/<crashing-PID>). The default value for this file is 0.

Controlling which mappings are written to the core dump¶

Since kernel 2.6.23, the Linux-specific /proc/[pid]/coredump_filter file can be used to control which memory segments are written to the core dump file in the event that a core dump is performed for the process with the corresponding process ID.

The value in the file is a bit mask of memory mapping types (see mmap(2)). If a bit is set in the mask, then memory mappings of the corresponding type are dumped; otherwise they are not dumped. The bits in this file have the following meanings:

By default, the following bits are set: 0, 1, 4 (if the CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS kernel configuration option is enabled), and 5. This default can be modified at boot time using the coredump_filter boot option.

The value of this file is displayed in hexadecimal. (The default value is thus displayed as 33.)

Memory-mapped I/O pages such as frame buffer are never dumped, and virtual DSO pages are always dumped, regardless of the coredump_filter value.

A child process created via fork(2) inherits its parent's coredump_filter value; the coredump_filter value is preserved across an execve(2).

It can be useful to set coredump_filter in the parent shell before running a program, for example:

$ echo 0x7 > /proc/self/coredump_filter
$ ./some_program

This file is provided only if the kernel was built with the CONFIG_ELF_CORE configuration option.

Core dumps and systemd¶

On systems using the systemd(1) init framework, core dumps may be placed in a location determined by systemd(1). To do this, systemd(1) employs the core_pattern feature that allows piping core dumps to a program. One can verify this by checking whether core dumps are being piped to the systemd-coredump(8) program:

$ cat /proc/sys/kernel/core_pattern
|/usr/lib/systemd/systemd-coredump %P %u %g %s %t %c %e

In this case, core dumps will be placed in the location configured for systemd-coredump(8), typically as lz4(1) compressed files in the directory /var/lib/systemd/coredump/. One can list the core dumps that have been recorded by systemd-coredump(8) using coredumpctl(1):

$ coredumpctl list | tail -5
Wed 2017-10-11 22:25:30 CEST  2748 1000 1000 3 present  /usr/bin/sleep
Thu 2017-10-12 06:29:10 CEST  2716 1000 1000 3 present  /usr/bin/sleep
Thu 2017-10-12 06:30:50 CEST  2767 1000 1000 3 present  /usr/bin/sleep
Thu 2017-10-12 06:37:40 CEST  2918 1000 1000 3 present  /usr/bin/cat
Thu 2017-10-12 08:13:07 CEST  2955 1000 1000 3 present  /usr/bin/cat

The information shown for each core dump includes the date and time of the dump, the PID, UID, and GID of the dumping process, the signal number that caused the core dump, and the pathname of the executable that was being run by the dumped process. Various options to coredumpctl(1) allow a specified coredump file to be pulled from the systemd(1) location into a specified file. For example, to extract the core dump for PID 2955 shown above to a file named core in the current directory, one could use:

$ coredumpctl dump 2955 -o core

For more extensive details, see the coredumpctl(1) manual page.

To disable the systemd(1) mechanism that archives core dumps, restoring to something more like traditional Linux behavior, one can set an override for the systemd(1) mechanism, using something like:

# echo "kernel.core_pattern=core.%p" > /etc/sysctl.d/50-coredump.conf
# /lib/systemd/systemd-sysctl

Program source¶

/* core_pattern_pipe_test.c */
#define _GNU_SOURCE
#include <sys/stat.h>
#include <fcntl.h>
#include <limits.h>
#include <stdio.h>
#include <stdlib.h>
#include <unistd.h>
#define BUF_SIZE 1024
int
main(int argc, char *argv[])
{


    int tot, j;


    ssize_t nread;


    char buf[BUF_SIZE];


    FILE *fp;


    char cwd[PATH_MAX];


    /* Change our current working directory to that of the


       crashing process */


    snprintf(cwd, PATH_MAX, "/proc/%s/cwd", argv[1]);


    chdir(cwd);


    /* Write output to file "core.info" in that directory */


    fp = fopen("core.info", "w+");


    if (fp == NULL)


        exit(EXIT_FAILURE);


    /* Display command-line arguments given to core_pattern


       pipe program */


    fprintf(fp, "argc=%d\n", argc);


    for (j = 0; j < argc; j++)


        fprintf(fp, "argc[%d]=<%s>\n", j, argv[j]);


    /* Count bytes in standard input (the core dump) */


    tot = 0;


    while ((nread = read(STDIN_FILENO, buf, BUF_SIZE)) > 0)


        tot += nread;


    fprintf(fp, "Total bytes in core dump: %d\n", tot);


    fclose(fp);


    exit(EXIT_SUCCESS);
}