Commit Graph

43081 Commits

Author SHA1 Message Date
Zbigniew Jędrzejewski-Szmek
b2ae4d9eb8 sysctl: move hashmap allocation out of main function
This allocation is a low level detail, and it seems nicer to keep it
out of run().
2020-01-16 14:45:37 +01:00
Zbigniew Jędrzejewski-Szmek
e76c60bf2a man: rework section about configuration file precedence
This section is loaded in a bunch of places, so this affects many
man pages.

1. point the reader to the synopsis section, which has the exact paths
that are used to load files.
2. put the "reference" part first, and recommendations later, in separate
paragraphs.
3. describe how individual settings and whole files are replaces.

Closes #12791.
2020-01-16 14:45:37 +01:00
Zbigniew Jędrzejewski-Szmek
251d3d20c3 Merge pull request #14581 from poettering/setcred-alternative-fix
alternative pam_setcred() fix
2020-01-16 09:53:26 +01:00
Lennart Poettering
4bb68f2fee core: on each iteration processing /proc/self/mountinfo merge all discovery flags for each path
This extends on d253a45e1c, and instead of
merging just a single flag from previous mount entries of
/proc/self/mountinfo for the same path we merge all three.

This shouldn't change behaviour, but I think make things more readable.

Previously we'd set MOUNT_PROC_IS_MOUNTED unconditionally, we still do.

Previously we'd inherit MOUNT_PROC_JUST_MOUNTED from a previous entry on
the same line, we still do.

MOUNT_PROC_JUST_CHANGED should generally stay set too. Why that? If we
have two mount entries on the same mount point we'd first process one
and then the other, and the almost certainly different mount parameters
of the two would mean we'd set MOUNT_PROC_JUST_CHANGED for the second.
And with this we'll definitely do that still.

This also adds a comment explaining the situation a bit, and why we get
into this situation.
2020-01-15 17:42:12 +01:00
Lennart Poettering
dd1b23a313 Merge pull request #14375 from poettering/userdb
New varlink API for user and group management
2020-01-15 17:41:29 +01:00
Lennart Poettering
46d7c6afbf execute: allow pam_setcred() to fail, ignore errors
Fixes: #14567
Alternative-To: #14569
2020-01-15 17:10:43 +01:00
Lennart Poettering
5b8d1f6b77 execute: add const to array parameters, where possible 2020-01-15 17:10:28 +01:00
Lennart Poettering
c903ee8976 docs: add documentation for the varlink user/group APIs 2020-01-15 15:34:09 +01:00
Lennart Poettering
32eb3c4229 docs: add documentation for JSON group records, too 2020-01-15 15:34:09 +01:00
Lennart Poettering
812862db71 docs: add documentation for JSON user records 2020-01-15 15:34:09 +01:00
Lennart Poettering
0ba56d3657 man: document the new nss-systemd behaviour
(This also changes the suggested /etc/nsswitch.conf line to use for
hooking up nss-system to use glibc's [SUCCESS=merge] feature so that we
can properly merge group membership lists).
2020-01-15 15:31:00 +01:00
Lennart Poettering
7d9ad0e5e5 man: document systemd-userdbd.service 2020-01-15 15:30:40 +01:00
Lennart Poettering
3b2db6f110 man: document userdbctl(1) 2020-01-15 15:30:20 +01:00
Lennart Poettering
fc89f88e56 man: document new pam_systemd features in man page
This also updates the suggested PAM snippet in a number of way:

1. Be closer to the logic nowadays implemented in Fedora where the
   auth/account/password stacks are all finished off with
   pam_{deny|permit}.so

2. Make pam_unix.so just "sufficient" instead of "required" (paving
   ground for pam_systemd_home.so being hooked in as additional
   sufficient module.

3. Only do pam_nologin in the "account" stack, since it's about account
   validity really.

4. Use modern parameters to pam_unix when changing passwords, i.e.
   sha512 and shadow, and use already set up passwords (preparing ground
   for pam_systemd_home again)
2020-01-15 15:30:06 +01:00
Lennart Poettering
f9c1f4e193 pam-systemd: apply user record properties to session
This way any component providing us with JSON user record data can use
this for automatic resource management and other session properties.
2020-01-15 15:30:02 +01:00
Lennart Poettering
7bfbf6cc92 pam-systemd: normalize return values of append_session_xyz()
Let's propagate the PAM errors we got.
2020-01-15 15:29:59 +01:00
Lennart Poettering
9ab0d3ebe5 pam-systemd: port over to use a UserRecord structure
Later on this allows us to set various session properties from user
record.
2020-01-15 15:29:55 +01:00
Lennart Poettering
355c9966c2 pam-systemd: share bus connection with pam_systemd_home if we can
Let's use the pam-util.h provided helpers to acquire them.
2020-01-15 15:29:52 +01:00
Lennart Poettering
d750dde2a6 pam-systemd: port to pam_bus_log_{create|parse}_error() and pam_log_oom() 2020-01-15 15:29:48 +01:00
Lennart Poettering
cef9f2a647 shared: add pam utility helpers 2020-01-15 15:29:31 +01:00
Lennart Poettering
d510589fd0 logind: honour per-user stopDelayUSec property 2020-01-15 15:29:27 +01:00
Lennart Poettering
156a363750 logind: honour killProcesses field of user record 2020-01-15 15:29:24 +01:00
Lennart Poettering
e8e4b7a0b6 logind: enforce user record resource settings when user logs in 2020-01-15 15:29:21 +01:00
Lennart Poettering
22c902facc logind: port to UserRecord object
This changes the user tracking of logind to use the new-style UserRecord
object.

In a later commit this enables us to do per-user resource management.
2020-01-15 15:29:17 +01:00
Lennart Poettering
1684c56f40 nss: hook up nss-systemd with userdb varlink bits
This changes nss-systemd to use the new varlink user/group APIs for
looking up everything.

(This also changes the factory /etc/nsswitch.conf line to use for
hooking up nss-system to use glibc's [SUCCESS=merge] feature so that we
can properly merge group membership lists).

Fixes: #12492
2020-01-15 15:29:07 +01:00