[PATCH] Convert the value of pwd_exp_warning to seconds
by Jakub Hrozek
https://fedorahosted.org/sssd/ticket/1773
Commit bf8cce77a35cb0a3cdb0d21fb9c39b7b6372bc11 broke
pam_pwd_expiration_warning because it started treating its value as if
it was seconds, but in fact it's days.
I tested the path that hits parse_krb5_child_response() and the one that
goes through check_pwexpire_ldap() but I'm not entirely sure how to test
the LDAP authentication that uses check_pwexpire_kerberos(). Could
anyone educate me (or test the patch as part of review maybe?).
I also don't like that the Kerberos password expiration is checked in
the provider and the LDAP password expiration in the responder. I think
this kind of logic belongs to the responder only. I filed #1774 to get
it fixed, but since this patch is intended for the 1.9.4 release that is
going to happen after Wednesday, I think we can do the refactor
separately.
11 years, 3 months
[PATCH] Fix code style
by Jakub Hrozek
Hi,
I pushed the attached patch to master as a one-liner. No functional
change, just code style fix.
11 years, 3 months
[PATCHES] Second refactoring wave for #1747
by Simo Sorce
Attached are 25 patches that are a second wave on top of my previous
refactoring patchset for #1747.
These new patches are a start for the removal of the sysdb argument from
some of the functions, leaving just the domain argument. This part is
not complete, but this subset is self contained.
On top of that there is also a complete refactoring of the code around
the struct be_req structure. This structure has now been made private to
the data_provider_be.c file so that it cannot be abused in future.
I will soon propose a third wave with more sysdb args elimination and
hopefully, finally a change in how subdomains are managed, which is the
end-goal of this long refactoring journey.
This patchset is also compressed as uncompresse dit amounts to ~ 250k
Simo.
--
Simo Sorce * Red Hat, Inc * New York
11 years, 3 months
[PATCH] set struct bet_info->bet_type
by Pavel Březina
I noticed that struct bet_info->bet_type is unused.
Instead of removing it, this patch makes sure that the field is set.
It might be helpful during debugging to get the idea of what be_info
index belong to what backend.
e.g. gdb> p be_ctx->be_info
..., {bet_type = BET_ID, bet_ops = 0x7fbb91ea97a0, pvt_bet_data =
0x1ed00e0, mod_name = 0x1ed21b0 "ipa"}, ...
11 years, 3 months