Age | Commit message (Collapse) | Author | Files | Lines |
|
Coverity: 11922
|
|
c-ares timeout to wait for response from DNS server
before moving to next DNS server is lowered from 5s
to 2s.
Partially solves https://fedorahosted.org/sssd/ticket/1966
|
|
Partially solves ticket:
https://fedorahosted.org/sssd/ticket/1966
To avoid the problem mentioned in the ticket above, option
dns_discovery_domain must be set properly.
|
|
The subdomain users user FQDN in their name attribute. However, handling
of whether to use FQDN in the LDAP code was not really good. This patch
introduces a utility function and converts code that was relying on
user/group names matching to this utility function.
This is a temporary fix until we can refactor the sysdb API in #2011.
|
|
We check if function ipa_get_ad_id_ctx returns NULL,
but function ipa_get_ad_id_ctx could never return NULL.
This issue was found by scan-build.
|
|
In unlikely case that an NSS module returns a reference to a group
and we are unable to resolve it shortly after that, make sure these
groups are skipped.
|
|
|
|
The krb5 child contacts the PAC responder for any user except for the
IPA native users if the PAC is configured. This works fine for the
general case but the ipa_server_mode is a special one. The PAC responder
is there, but since in the server mode we should be operating as AD
provider default, the PAC shouldn't be analyzed either in this case.
|
|
|
|
getgrouplist(3) call is missing from Python older than Python 3.3
Introduce supplementary binding to provide getgrouplist as part of pysss
interface. Since getgrouplist() can be run against any domain, place
it at top module rather than pysss.local namespace.
pysss.getgrouplist(username) -> tuple(group list as strings)
|
|
https://fedorahosted.org/sssd/ticket/2025
|
|
|
|
|
|
|
|
Commit 52ae806bd17c3c00d70bd1aed437f10f5ae51a1c changed the default FQDN
format by accident to the one we only ever user internally. This commit
fixes the mistake.
|
|
https://fedorahosted.org/sssd/ticket/2007
PAC, SSH, SUDO and AUTOFS services' debug levels weren't modified when
sss_debuglevel tool was used.
|
|
https://fedorahosted.org/sssd/ticket/2023
When the option values are copied using dp_opt_copy_map, the .val member
is used if it's not NULL. At the same time, the bool options are never
NULL, unlike integers or strings that can have special NULL-like values
such as NULL_STRING. This effectively means that when copying a bool
option, the .val member is always used.
But in the AD maps, some .val fields were set differently from the
.def_val fields. The effect was that when the AD subdomain provider was
initialized from IPA subdomain provider using only the defaults, some
options (notably referral chasing) were set to a value that didn't make
sense for the AD provider.
This patch makes sure that for all boolean option, the .val is always
the same as .def_val.
|
|
https://fedorahosted.org/sssd/ticket/2009
If the IPA server mode is on and the SSSD is running on the IPA server,
then the server's extdom plugin calls getpwnam_r to read info about trusted
users from the AD server and return them to the clients that called the
extended operation.
The SSSD returns the subdomain users fully-qualified, ie "user@domain"
by default. The format of the fully qualified name is configurable.
However, the extdom plugin returns the user name without the domain
component.
With this patch, when ipa_server_mode is on, warn if the full_name_format
is set to a non-default value. That would prompt the admin to change the
format if he changed it to something exotic.
|
|
|
|
|
|
warning: format string is not a string literal (potentially insecure)
[-Wformat-security]
|
|
|
|
Instead of multiple calls of sss_authtok_get_type, perform the call just
once and store into variable.
|
|
https://fedorahosted.org/sssd/ticket/1814
Return authentication error when empty password is passed.
|
|
https://fedorahosted.org/sssd/ticket/1814
When the authtok_length is zero, it shouldn't call
sss_authtok_set_password, because it tries to determine lenght of passed
string by itself and would read parts of DBus message behind boundaries
of authtok.
|
|
https://fedorahosted.org/sssd/ticket/2005
Some users were confused by our description of min_id/max_id and thought
the limits only applied to returning entries from the NSS responder.
However, the limits are actually enforced on the back end side, so the
entries are not even saved to cache.
|
|
https://fedorahosted.org/sssd/ticket/1998
Currently using IP address as value of ad_server is not supported, so
the man pages should not mention that as an option.
|
|
https://fedorahosted.org/sssd/ticket/1965
After we added a section that clarified what access_provider=ad did,
some users were confused and thought that "ad" was also the default
access provider if "id_provider=ad" was specified.
|
|
https://fedorahosted.org/sssd/ticket/1992
|
|
https://fedorahosted.org/sssd/ticket/1992
dyndns test will crash with this patch
|
|
The NSS responder recently started using libsss_idmap in the getbysid
functions. The bug itself was spotted by one of our automated QA tools.
|
|
Storing cyclic groups into sysdb can cause adding ghost members,
which has already been stored. Function ldb_modify will fail
with error [Attribute or value exists].
With permisive control, duplicated attributes will be skipped
as if it was never added.
https://fedorahosted.org/sssd/ticket/1846
|
|
https://fedorahosted.org/sssd/ticket/1827
|
|
https://fedorahosted.org/sssd/ticket/1997
Most of the time, the SIDs are resolved as a call coming from the PAC
responder during initgroups. In that case at least, it makes sense to
search for group matching that SID first, then users.
We may consider making this behaviour configurable ie for the server
mode where typically the users should be queried first.
|
|
|
|
|
|
Some krb5 functions needn't be available for retrieving ccache
with principal. Therefore ifdef is used to solve this situation with older
version of libkrb5. There were two functions with similar functionality
in krb5_child and krb5_utils. They were merged to one universal function, which
was moved to file src/util/sss_krb5.c
|
|
|
|
Dircache can be either file or directory. Wrong value was used
when dircache was itself directory.
https://fedorahosted.org/sssd/ticket/2002
|
|
We did not set port status for metaservers (srv servers)
in fo_reset_services().
Fixes:
https://fedorahosted.org/sssd/ticket/1933
|
|
The switch statement was dead code due to missing case/default.
|
|
Option -E/--everething was added to invalide all types of entries.
https://fedorahosted.org/sssd/ticket/1988
|
|
https://fedorahosted.org/sssd/ticket/1778
When trying to copy special file, only message is logged now.
|
|
|
|
This will ensure that we aren't pulling in extra samba4
dependencies for the Kerberos provider.
|
|
|
|
The PAC responder is now used by both IPA and AD providers.
|
|
There are no longer any Fedora platforms running SSSD with SYSV
init scripts. We don't need the upgrade logic any more.
|
|
|
|
https://fedorahosted.org/sssd/ticket/1999
ldap_auth.c code which was added to SSSD for updating the
shadowLastChange when "ldap_chpass_update_last_change" option is
enabled updates shadowLastChange even when the PAM password change
status reports failure.
We should only update shadowLastChange on PAM password change success or
we open up a work around for users to avoid changing their passwords
periodically as required by policy. The user simply attempts to change
password, fails by trying to set new password which invalid (denied due
to password history check) yet shadowLastChange is updated, avoiding
their need to actually change the password they are using.
|