On Thu, Nov 19, 2015 at 09:18:46AM +0100, Lukas Slebodnik wrote:
On (14/11/15 13:56), Jakub Hrozek wrote:
>On Fri, Nov 13, 2015 at 09:42:15PM +0200, Nikolai Kondrashov wrote:
>> On 11/13/2015 07:06 PM, Pavel Reichl wrote:
>> >On 11/13/2015 05:44 PM, Nikolai Kondrashov wrote:
>> >>On 11/13/2015 06:19 PM, Pavel Reichl wrote:
>> >>>Thanks Nick, updated patch attached.
>> >>
>> >>Thanks, Pavel, this looks really neat now! I have only two comments.
>> >>
>> >>>+def test_simple_user_override(ldap_conn,
env_simple_user_override):
>> >>>+ """Test entries are overriden"""
>> >>>+
>> >>>+
>> >>
>> >>Shouldn't there be an actual test here?
>> >Sorry :-)
>>
>> No problem :)
>>
>> >>
>> >>>+# Regression test for bug #2802
>> >>>+# sss_override segfaults when accidentally adding --help flag to
some commands
>> >>>+
>> >>>+
>> >>>+def test_regr_2802_override(request, ldap_conn):
>> >>>+ prepare_sssd(request, ldap_conn)
>> >>
>> >>I had no idea this could actually work, I mean getting
"request" in a test.
>> >>Can you give a link to documentation where this is explained?
>> >
>> >I don't have any, I was in a hurry and I did a mistake.
>>
>> Ah, fine.
>>
>> >>Otherwise, could you please move the setup to a proper fixture? Tests
are not
>> >>supposed to have setup/teardown inside, AFAIK, unless this is really
official.
>> >>
>> >>>+
>> >>>+ subprocess.check_call(["sss_override",
"user-del", "--help"])
>> >
>> >Sure.
>>
>> Thanks a lot!
>>
>> Looks fine to me now. Tests passed on my machine both all together, and each
>> separately. Great job!
>>
>> ACK
>>
>> Nick
>
>Pushed to master so far: 8d1dcb6af723f2968410c4b088d06d63d02b4fea
>
>We can push all these to sssd-1-13 later, when CI ressurects.
We can "ressurects" CI after fixing failure introduced by this commit.
Patch is attached.
LS
OK, thanks. I submitted the patch to CI.