As shadow_logfd variable is not set at the beginning of the program if
something fails and fprintf() is called a segmentation fault happens.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=2021339
Signed-off-by: Iker Pedrosa <ipedrosa@redhat.com>
Always set SIGCHLD handler to default, even if the caller of vipw has
set SIGCHLD to ignore. If SIGCHLD is ignored no zombie processes would
be created, which in turn could mean that kill is called with an already
recycled pid.
Proof of Concept:
1. Compile nochld:
--
#include <signal.h>
#include <unistd.h>
int main(void) {
char *argv[] = { "vipw", NULL };
signal(SIGCHLD, SIG_IGN);
execvp("vipw", argv);
return 1;
}
--
2. Run nochld
3. Suspend child vi, which suspends vipw too:
`kill -STOP childpid`
4. Kill vi:
`kill -9 childpid`
5. You can see with ps that childpid is no zombie but disappeared
6. Bring vipw back into foreground
`fg`
The kill call sends SIGCONT to "childpid" which in turn could have been
already recycled for another process.
This is definitely not a vulnerability. It would take super user
operations, at which point an attacker would have already elevated
permissions.
Signed-off-by: Tobias Stoeckmann <tobias@stoeckmann.org>
Change SELinux labels for files copied from the skeleton directory to
the home directory.
This could cause gnome's graphical user adding to fail without copying
the full skeleton files.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=2022658
Signed-off-by: Iker Pedrosa <ipedrosa@redhat.com>
Fix segmentation fault in newgrp when xgetspnam() returns a NULL value
that is immediately freed.
The error was committed in
e65cc6aebc
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=2019553
Signed-off-by: Iker Pedrosa <ipedrosa@redhat.com>
When using groupdel with a prefix, groupdel will attempt to read a
passwd file to look for any user in the group. When the file does not
exist it cores with segmentation fault.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1986111
If a line in hushlogins file, e.g. /etc/hushlogins, starts with
'\0', then current code performs an out of boundary write.
If the line lacks a newline at the end, then another character is
overridden.
With strcspn both cases are solved.
Signed-off-by: Tobias Stoeckmann <tobias@stoeckmann.org>
During shadowtcb_move() the directory is temporarily changed to be
owned by root:root with permissions 0700. After the change is done,
the ownership and permissions were supposed to be restored. The
call for chown() was there, but the chmod() call was missing. This
resulted in the broken TCB functionality. The added chmod() fixes
the issue.
Close the selabel handle to update the file_context. This means that the
file_context will be remmaped and used by selabel_lookup() to return
the appropriate context to label the home folder.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1993081
Signed-off-by: Iker Pedrosa <ipedrosa@redhat.com>
Create the home and mail folders after the SELinux user has been set for
the added user. This will allow the folders to be created with the
SELinux user label.
Signed-off-by: Iker Pedrosa <ipedrosa@redhat.com>
The buggy code was introduced nearly 5 years ago at the
commit 08fd4b69e8. The
desired behavior is that SIGKILL will be sent to the
child if it does not exit within 2 seconds after it
receives SIGTERM. However, SIGALRM is masked while
waiting for the child so it cannot wake the program
up after 2 seconds to send SIGKILL.
An example shows the buggy behavior, which exists in
Ubuntu 18.04 LTS (with login 1:4.5-1ubuntu2).
```bash
user1@localhost:~$ su user2 -c '
_term() {
echo SIGTERM received
}
trap _term TERM
while true; do
sleep 1
echo still alive
done'
Password:
still alive
Session terminated, terminating shell...Terminated
SIGTERM received
still alive
still alive
still alive
still alive
```
(SIGTERM is sent in another user1's terminal by
executing `killall su`.)
Here is the desired behavior, which shows what the
commit fixes.
```bash
user1@localhost:~$ su user2 -c '
_term() {
echo SIGTERM received
}
trap _term TERM
while true; do
sleep 1
echo still alive
done'
Password:
still alive
Session terminated, terminating shell...Terminated
SIGTERM received
still alive
still alive
...killed.
user1@localhost:~$ echo $?
255
```
In some circumstances I want the default behaviour of useradd to
not add user entries to the lastlog and faillog databases. Allowing
this options behaviour to be controlled by the config file
/etc/default/useradd.
`sgent` is only initialized in `get_group()` if `is_shadowgrp` is true.
So we should also only attempt to free it if this is actually the case.
Can otherwise lead to:
```
free() double free detected in tcache 2 (gpasswd)
```
The GitHub and Debian permanently moved to HTTPS URLs and redirect
there. The Gentoo URL does not redirect to HTTPS, but still use it to
address certain kinds of attacks. Lastly, the NetBSD URL is only
available using HTTP.
subuid_count won't get used by usr_update(), but since we're passing it
as an argument we have to make sure it's always defined. So just define
it as pre-set to 0.
Closes#402
Signed-off-by: Serge Hallyn <serge@hallyn.com>
xml2po is deprecated. We've previously replaced xml2po with
itstool in man/generate_translations.mak, but there was still
an instance of it that only is exercised for 'make dist'.
Update that one. Now 'make dist' succeeds on a ubuntu focal
or newer host where xml2po is not available.
Signed-off-by: Serge Hallyn <serge@hallyn.com>
If SHA_CRYPT_MIN_ROUNDS and SHA_CRYPT_MAX_ROUNDS are both unspecified,
use SHA_ROUNDS_DEFAULT.
Previously, the code fell through, calling shadow_random(-1, -1). This
ultimately set rounds = (unsigned long) -1, which ends up being a very
large number! This then got capped to SHA_ROUNDS_MAX later in the
function.
The new behavior matches BCRYPT_get_salt_rounds().
Bug: https://bugs.gentoo.org/808195
Fixes: https://github.com/shadow-maint/shadow/issues/393
Serge Hallyn
Iker Pedrosa
Balint Reczey
Tobias Stoeckmann
François Rigault
(GalaxyMaster)
Ruihan Li
Andy Zaugg
Michael Vetter
Frans Spiesschaert
Luís Ferreira
a1346054
Paul Menzel
Mike Gilbert