2 title: systemd Coredump Handling
5 SPDX-License-Identifier: LGPL-2.1-or-later
8 # systemd Coredump Handling
10 ## Support in the Service Manager (PID 1)
12 The systemd service manager natively provides coredump handling functionality,
13 as implemented by the Linux kernel.
14 Specifically, PID 1 provides the following functionality:
16 1. During very early boot it will raise the
17 [`LIMIT_CORE`](https://man7.org/linux/man-pages/man2/getrlimit.2.html)
18 resource limit for itself to infinity (and thus implicitly also all its children).
19 This removes any limits on the size of generated coredumps,
20 for all invoked processes, from earliest boot on.
21 (The Linux kernel sets the limit to 0 by default.)
23 2. At the same time it will turn off coredump handling in the kernel by writing
24 `|/bin/false` into `/proc/sys/kernel/core_pattern` (also known as the
25 "`kernel.core_pattern` sysctl"; see
26 [core(5)](https://man7.org/linux/man-pages/man5/core.5.html) for
28 This means that coredumps are not actually processed.
29 (The Linux kernel sets the pattern to `core` by default, so that coredumps are written
30 to the current working directory of the crashing process.)
32 Net effect: after PID1 has started and performed this setup coredumps are
33 disabled, but by means of the the `kernel.core_pattern` sysctl rather than by
35 This is generally preferable, since the pattern can be updated trivially at the right time to enable coredumping once the system is ready, taking comprehensive effect on all userspace.
36 (Or to say this differently: disabling coredumps via the size limit is problematic, since it cannot easily
37 be undone without iterating through all already running processes once the system is ready for coredump handling.)
39 Processing of core dumps may be enabled at the appropriate time by updating the
40 `kernel.core_pattern` sysctl.
41 Only coredumps that happen later will be processed.
43 During the final shutdown phase the `kernel.core_pattern` sysctl is updated
44 again to `|/bin/false`, disabling coredump support again, should it have been
45 enabled in the meantime.
47 This means coredump handling is generally not available during earliest boot
48 and latest shutdown, reflecting the fact that storage is typically not
49 available in these environments, and many other facilities are missing too that
50 are required to collect and process a coredump successfully.
52 ## `systemd-coredump` Handler
54 The systemd suite provides a coredump handler
55 [`systemd-coredump`](https://www.freedesktop.org/software/systemd/man/systemd-coredump.html)
56 which can be enabled at build-time. It is activated during boot via the
57 `/usr/lib/sysctl.d/50-coredump.conf` drop-in file for
58 `systemd-sysctl.service`. It registers the `systemd-coredump` tool as
59 `kernel.core_pattern` sysctl.
61 `systemd-coredump` is implemented as socket activated service: when the kernel
62 invokes the userspace coredump handler, the received coredump file descriptor
63 is immediately handed off to the socket activated service
64 `systemd-coredump@.service` via the `systemd-coredump.socket` socket unit. This
65 means the coredump handler runs for a very short time only, and the potentially
66 *heavy* and security sensitive coredump processing work is done as part of the
67 specified service unit, and thus can take benefit of regular service resource
68 management and sandboxing.
70 The `systemd-coredump` handler will extract a backtrace and
71 [ELF packaging metadata](/ELF_PACKAGE_METADATA) from any coredumps it
72 receives and log both.
73 The information about coredumps stored in the journal can be enumerated and queried with the
74 [`coredumpctl`](https://www.freedesktop.org/software/systemd/man/coredumpctl.html)
75 tool, for example for directly invoking a debugger such as `gdb` on a collected
78 The handler writes coredump files to `/var/lib/systemd/coredump/`.
79 Old files are cleaned up periodically by
80 [`systemd-tmpfiles(8)`](https://www.freedesktop.org/software/systemd/man/systemd-tmpfiles.html).
84 With the above, any coredumps generated on the system are by default collected
85 and turned into logged events — except during very early boot and late
87 Individual services, processes or users can opt-out of coredump collection,
88 by setting `LIMIT_CORE` to 0 (or alternatively invoke
89 [`PR_SET_DUMPABLE`](https://man7.org/linux/man-pages/man2/prctl.2.html)).
90 The resource limit can be set freely by daemons/processes/users to arbitrary
91 values, which the coredump handler will respect.
92 The `coredumpctl` tool may be used to further analyze/debug coredumps.
94 ## Alternative Coredump Handlers
96 While we recommend usage of the `systemd-coredump` handler, it's fully
97 supported to use alternative coredump handlers instead.
98 A similar implementation pattern is recommended.
101 1. Use a `sysctl.d/` drop-in to register your handler with the kernel.
102 Make sure to include the `%c` specifier in the pattern (which reflects the
103 crashing process' `RLIMIT_CORE`) and act on it:
104 limit the stored coredump file to the specified limit.
106 2. Do not do heavy processing directly in the coredump handler.
107 Instead, quickly pass off the kernel's coredump file descriptor to an
108 auxiliary service running as service under the service manager,
109 so that it can be done under supervision, sandboxing and resource management.
111 Note that at any given time only a single handler can be enabled, i.e. the
112 `kernel.core_pattern` sysctl cannot reference multiple executables.
116 It might make sense to split `systemd-coredump` into a separate distribution
118 If doing so, make sure that `/usr/lib/sysctl.d/50-coredump.conf` and
119 the associated service and socket units are also added to the split off package.
121 Note that in a scenario where `systemd-coredump` is split out and not
122 installed, coredumping is turned off during the entire runtime of the system —
123 unless an alternative handler is installed, or behaviour is manually reverted
124 to legacy style handling (see below).
126 ## Restoring Legacy Coredump Handling
128 The default policy of the kernel to write coredumps into the current working
129 directory of the crashing process is considered highly problematic by many,
130 including by the systemd maintainers.
131 Nonetheless, if users locally want to return to this behaviour, two changes must be made (followed by a reboot):
134 $ mkdir -p /etc/sysctl.d
135 $ cat >/etc/sysctl.d/50-coredump.conf <<EOF
136 # Party like it's 1995!
137 kernel.core_pattern=core
144 $ mkdir -p /etc/systemd/system.conf.d
145 $ cat >/etc/systemd/system.conf.d/50-coredump.conf <<EOF
147 DefaultLimitCORE=0:infinity