| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
| |
The current request already returned the SID, we do not need to request
it separately.
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
(cherry picked from commit e6046d23b3e90102fb3c796737ced03fb5a60fea)
|
|
|
|
|
|
|
|
|
|
| |
The call protected by the check does not only expect the version 1 of
the extdom plugin is used but a specific response type as well. Since
version 1 can return older response types as well we want to be on the
safe side.
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
(cherry picked from commit 2fc12875f7d51248799016c19c1298b85e06a286)
|
|
|
|
|
|
|
|
|
| |
The IPA extdom plugin returns the data with the default view already
applied hence it is on needed to look up the override data if the client
has the default view assigned.
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
(cherry picked from commit d8ceb194023a2cdc8bc183acc322e9a7fb6fe2b1)
|
|
|
|
|
|
|
|
| |
Currently ipa_resolve_user_list_send() only looks up the related user
objects but do not check for overrides. This patch tries to fix this.
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
(cherry picked from commit eab17959df71341073f946c533f59fc5e593b35c)
|
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
(cherry picked from commit fbcdc08722aa8ed17c4b114e01fbb37c02cfb2fe)
|
|
|
|
|
|
|
| |
Related to https://fedorahosted.org/sssd/ticket/2481
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
(cherry picked from commit 765d9075bb1e10ae0f09b6c2701bfd50aeb423d4)
|
|
|
|
|
|
|
| |
Related to https://fedorahosted.org/sssd/ticket/2481
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
(cherry picked from commit d32b165fad7b89462f49c82349e1df5a2343afa2)
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://fedorahosted.org/sssd/ticket/2543
The LDAP URI is not valid prior to connecting to LDAP. Moreover,
reconnecting to a different server might invalidate the URI.
Move reading the URI after the connection has been established.
Reviewed-by: Sumit Bose <sbose@redhat.com>
(cherry picked from commit ccff8e75940963a0f68f86efcddc37133318abfa)
|
|
|
|
|
|
|
|
|
|
|
| |
Resolves:
https://fedorahosted.org/sssd/ticket/2544
Use a dedicated fd instead to work around
https://bugzilla.samba.org/show_bug.cgi?id=11036
Reviewed-by: Sumit Bose <sbose@redhat.com>
(cherry picked from commit f00a61b6079d8de81432077a59daf015d85800d2)
|
|
|
|
|
|
|
|
|
|
|
| |
Related to:
https://fedorahosted.org/sssd/ticket/2544
Adds a new function exec_child_ex and moves setting the extra_argv[]
to exec_child_ex() along with specifying the input and output fds.
Reviewed-by: Sumit Bose <sbose@redhat.com>
(cherry picked from commit 16cb0969f0a9ea71524d852077d6a480740d4f12)
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
libsmb logs to stdout by default. It's much more reasonable to log to
stderr by default.
Please also note:
https://bugzilla.samba.org/show_bug.cgi?id=11036
and:
https://fedorahosted.org/sssd/ticket/2544
Reviewed-by: Sumit Bose <sbose@redhat.com>
(cherry picked from commit bb7ddd2be9847bfb07395341c7623da1b104b8a6)
|
|
|
|
|
|
|
|
|
|
|
| |
https://fedorahosted.org/sssd/ticket/2542
If the GPO result object was missing completely, we would error out with
a fatal error code. It's more user-friendly to treat the missing object
as if the requested attribute was missing on the provider level.
Reviewed-by: Pavel Reichl <preichl@redhat.com>
(cherry picked from commit fc2cc91a5b645180e53d46436b0d08011aac8d74)
|
|
|
|
|
|
|
|
|
|
| |
Some callers of libwbclient functions expects the return values are
initialized even it the functions returns an error. This patch adds some
initializations to meet this requirement.
Resolves https://fedorahosted.org/sssd/ticket/2537
Reviewed-by: Pavel Reichl <preichl@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Resolves:
https://fedorahosted.org/sssd/ticket/2519
Not existing user/group in simple_allow_users/simple_allow_groups should not
imply access denied.
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
| |
Resolves https://fedorahosted.org/sssd/ticket/2514
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
By default user and group overrides use the same attribute name for the
GID and this cause SSSD machinery to add the same value twice which
cause an error in ldb_add() or ldm_modify().
Related to https://fedorahosted.org/sssd/ticket/2514
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
When groups are resolved on IPA clients as part of a user lookup not all
groups have to be from the same domain as the used. This has to be
checked to store the group object properly in the cache.
Related to https://fedorahosted.org/sssd/ticket/2529
and https://fedorahosted.org/sssd/ticket/2524
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Depending on the state of the cache group object a freshly created or
updates user entry for a trusted domain user might already be a member
of the group or not. This cache makes sure the requested user is a
member of all groups returned from the extdom request. Special care has
to be taken to cover cross-domain group-memberships properly.
Resolves https://fedorahosted.org/sssd/ticket/2529
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
KRB5KRB_ERR_GENERIC is a generic error and we cannot make any
assumptions about the cause. If there are cases where
KRB5KRB_ERR_GENERIC is returned and SSSD should behave differently this
must be solved by other means.
Resolves https://fedorahosted.org/sssd/ticket/2535
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
sysdb_search_object_by_sid returns ENOENT if no results are found.
Part od solution for:
https://fedorahosted.org/sssd/ticket/1991
Fixes:
https://fedorahosted.org/sssd/ticket/2520
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
| |
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Check that return value of sysdb_delete_by_sid() is not changed as
called SYSDB functions have changed the return value.
Part of patches for:
https://fedorahosted.org/sssd/ticket/1991
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Environment variable SSSD_KRB5_REALM was used to late for initialisation
realm. and therefore default value NULL was used.
The SSSD_KRB5_REALM (kr->realm) was used as fast_principal_realm for checking
fast cache: privileged_krb5_setup -> k5c_setup_fast -> check_fast_ccache
And therefore wrong principal was used when the option krb5_fast_principal is
empty.
[find_principal_in_keytab] (0x4000): Trying to find principal (null)@(null) in keytab.
[match_principal] (0x1000): Principal matched to the sample ((null)@(null)).
[get_tgt_times] (0x1000): FAST ccache must be recreated
[get_tgt_times] (0x0020): krb5_cc_retrieve_cred failed
[get_tgt_times] (0x0020): 1688: [-1765328243][Matching credential not found]
[check_fast_ccache] (0x0040): Valid FAST TGT not found after attempting to renew it
[k5c_setup_fast] (0x0020): check_fast_ccache failed.
[k5c_setup_fast] (0x0020): 1956: [1432158213][Unknown code UUz 5]
[privileged_krb5_setup] (0x0040): Cannot set up FAST
[main] (0x0020): privileged_krb5_setup failed.
[main] (0x0020): krb5_child failed!
As a result of this user was not able to authenticate.
Resolves:
https://fedorahosted.org/sssd/ticket/2526
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
| |
Resolves:
https://fedorahosted.org/sssd/ticket/2515
|
|
|
|
|
|
|
|
|
|
|
|
| |
When processing group membership check sysdb for group members from
extern domain and include them in newly processed group membership as
extern members are curently found only when initgroups() is called.
Resolves:
https://fedorahosted.org/sssd/ticket/2492
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
| |
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
| |
This patch makes it more discoverable for the admin to find typos in the
various user lists. Typically, the user lists are used to add access to
some feature and printing a syslog message would make sure the admin
sees the mistake.
Reviewed-by: Pavel Reichl <preichl@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
| |
The man page claimed that failing to resolve an user name results in
failure to start SSSD, but it's not the case and shouldn't be, because
marking a user as trusted only elevates privileges, so it's safe to
ignore that failure.
https://fedorahosted.org/sssd/ticket/2530
Reviewed-by: Pavel Reichl <preichl@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
When the administrator sets the domains= list, he usually wants to
restrict the set of domains. An empty list is an undefined configuration
and it's safer to fail then.
https://fedorahosted.org/sssd/ticket/2516
Reviewed-by: Pavel Reichl <preichl@redhat.com>
|
|
|
|
|
|
|
|
| |
The pam_public_domains option and matching the domain requested by a
trusted process was done in a case-sensitive manner which is different
from how we match domain names in SSSD normally.
Reviewed-by: Pavel Reichl <preichl@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://fedorahosted.org/sssd/ticket/2484
When OTPs are used, we can only used each authtoken at most once. When
it comes to Kerberos password changes, this was only working previously
by accident, because the old authtoken was first used to verify the old
password is valid and not expired and then also to acquire a chpass
principal.
This patch looks at the user object in LDAP to check if the user has any
OTPs enabled. If he does, the CHAUTHTOK_PRELIM step is skipped
completely so that the OTP can be used to acquire the chpass ticket
later.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
glibc is inconsistent with how it treats and returns NSS_STATUS_UNAVAIL.
The sss nss plugin is present in nsswitch by default on some platforms
due to glibc caching and problem with long living applications (e.g. GNOME).
But sssd needn't be configuread and it cause problems in some programs.
In this situation, the SSSD nss plugin should behave as if it was functioning
but had no data even thought sssd is not running. The errors have to be passed
from nss plugin up to the user with minimal moidiffication.
Thanks to Stephen Gallagher for initial patch.
Resolves:
https://fedorahosted.org/sssd/ticket/2439
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Lukáš Slebodník <lslebodn@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Usernames from AD subdomains are already in fqdn we should not append
domain name in this case.
Resolves:
https://fedorahosted.org/sssd/ticket/2512
Reviewed-by: Michal Židek <mzidek@redhat.com>
|
|
|
|
| |
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Related:
https://fedorahosted.org/sssd/ticket/2503
In order to avoid calling Kerberos library calls as root, the krb5_child
forks itself and recreates the FAST ccache as the SSSD user.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
Related:
https://fedorahosted.org/sssd/ticket/2503
Currently all child processes use the same arguments, the construction
of argv[] is even hardcoded in exec_child(). Add an extra_args[] array
that extends the common set of argvs so that we can have child-specific
arguments. Also adds a unit test.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
| |
The child_common.c module had no unit tests, yet we need to amend it.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
| |
The host keytab and the FAST credential cache are copied into memory
early at startup to allow to drop privileges earlier.
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
| |
https://fedorahosted.org/sssd/ticket/2510
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
| |
https://fedorahosted.org/sssd/ticket/2510
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
| |
https://fedorahosted.org/sssd/ticket/2510
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
|
|
| |
Keeping a per-request flag in a global structure is really dangerous.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
https://fedorahosted.org/sssd/ticket/2501
Moving the checks to one place has the advantage of not duplicating
security decisions. Previously, the checks were scattered all over the
responder code, making testing hard.
The disadvantage is that we actually check for the presence of the user,
which might trigger some back end lookups. But I think the benefits
overweight the disadvantage.
Also only check the requested domains from a trusted client. An untrusted
client should simply have no say in what domains he wants to talk to, it
should ignore the 'domains' option.
Reviewed-by: Sumit Bose <sbose@redhat.com>
|
|
|
|
|
|
|
|
|
|
| |
Currently some manual steps are needed on a FreeIPA to switch from one
view to another. With this patch the IPA provider checks at startup if
the view name changed and does the needed steps automatically. Besides
saving the new view name this includes removing the old view data and
marking the user and group entries as invalid.
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|
|
|
|
| |
Reviewed-by: Jakub Hrozek <jhrozek@redhat.com>
|