| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
| |
FIXME - this removed the internal dp callback
|
|
|
|
| |
vs. autofs_setent_notify
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
* This broke corner cases when used with
default_tkt_types = des-cbc-crc
and DES enabled on an AD domain.
* This is fixed in kerberos instead, in a more correct way
and in a way which we cannot replicate.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
This simplifies configuration by eliminating the need to
specifiy both krb5_keytab and ldap_krb5_keytab if the keytab is
not located at /etc/krb5.keytab
|
| |
|
|
|
|
|
| |
This patch adds support for checking whether a user is expired or
disabled in AD.
|
|
|
|
|
|
| |
These new providers take advantage of existing code for the KRB5
provider, providing sensible defaults for operating against an
Active Directory 2008 R2 or later server.
|
|
|
|
|
|
| |
This new identity provider takes advantage of existing code for
the LDAP provider, but provides sensible defaults for operating
against an Active Directory 2008 R2 or later server.
|
|
|
|
| |
This will eliminate ambiguity for the AD provider
|
|
|
|
|
| |
This will reduce code duplication between the krb5, ipa and ad
providers
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This function is not supposed to return any newly-allocated memory
directly. It was actually leaking the memory for krb5_servers if
krb5_kdcip was being used, though it was undetectable because it
was allocated on the provided memctx.
This patch removes the memctx parameter and allocates krb5_servers
temporarily on NULL and ensures that it is freed on all exit
conditions. It is not necessary to retain this memory, as
dp_opt_set_string() performs a talloc_strdup onto the appropriate
context internally.
It also updates the DEBUG messages for this function to the
appropriate new macro levels.
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
* This is similar to the code in ipa_dyndns_update_send()
|
| |
|
|
|
|
|
| |
This request and attached memory would be freed at the end of
access-check processing, but it's a waste to keep it around.
|
|
|
|
|
| |
Sudo support is optional, when it is not configured sudorules_map is not
initialized and dereferencing it will cause a segmentation fault.
|
|
|
|
|
|
|
|
|
| |
We weren't guaranteeing that the cctype-specific callbacks were
initialized before using them.
This bug only presented itself for users who were logging in
without a ccacheFile attribute in the LDB (for example, first-time
logins).
|
| |
|
| |
|
|
|
|
| |
Removes old options and adds new ones.
|
|
|
|
|
| |
sudoHost attribute may contain IPv4 or IPv6 host/network address.
This patch adds support for autoconfiguration of these information.
|
|
|
|
| |
Add host information to LDAP filters.
|
| |
|
|
|
|
|
|
|
| |
We need to load host information during provider initialization.
Currently it loads only values from configuration files, but it is
implemented as an asynchrounous request as it will later try to
autodetect these settings (which will need to contact DNS).
|
|
|
|
|
|
|
|
|
|
| |
Adds some option that allows to manually configure a host filter.
ldap_sudo_use_host_filter - if false, we will download all rules regardless their sudoHost attribute
ldap_sudo_hostnames - list hostnames and/or fqdn that should be downloaded, separated with spaces
ldap_sudo_ip - list of IPv4/6 address and/or network that should be downloaded, separated with spaces
ldap_sudo_include_netgroups - include rules that contains netgroup in sudoHost
ldap_sudo_include_regexp - include rules that contains regular expression in sudoHost
|
|
|
|
|
| |
I had to create a new context structure to store additional
information such as ip addresses and hostnames.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When an expired rule is not present on the server server during specific rule
refresh, the provider will notify the sudo responder that it has been deleted.
Because there is a high probability that some other rules were deleted from
the server as well, we want to remove them from sysdb as soon as possible.
Once the responder is notified, it will schedule an out of band full refresh.
This is issued by responder, because we already have a mechanism that
prohibits creation of similar request (i.e. once the OOB full refresh is
scheduled, there won't be another).
The notification is done by returning:
DP error = DP_ERR_OK, error = ENOENT
|
|
|
|
| |
sdap_sudo_refresh_recv()
|
| |
|
| |
|
|
|
|
| |
sudosrv_get_sudorules_query_cache()
|
| |
|
|
|
|
|
|
|
|
|
| |
sss_sudo_type represents query type that comes to the responder
sss_dp_sudo_type represents query type to DP that is issued by the responder
I'm leaving current values of sss_dp_sudo_type untouched so the compilation
is not broken. Hovewer, they will be changed to new DP types once the DP
interface is updated.
|
|
|
|
|
|
|
|
|
|
|
|
| |
When SSSD is started, then full refresh is scheduled.
The smart refresh is scheduled after this full refresh,
if USN (or modifyTimestamp) values are available.
If full refresh interval <= smart refresh interval then
full refresh will be disabled.
If both refresh types are 0 then smart refresh interval
is set to default value.
|
| |
|