[SSSD] AUTOFS: Store entry objects below map objects

Pavel Březina pbrezina at redhat.com
Mon Sep 24 14:53:46 UTC 2012


On 09/24/2012 04:52 PM, Jakub Hrozek wrote:
> On Mon, Sep 24, 2012 at 04:41:51PM +0200, Pavel Březina wrote:
>> On 09/24/2012 03:44 PM, Jakub Hrozek wrote:
>>> On Mon, Sep 24, 2012 at 01:12:51PM +0200, Pavel Březina wrote:
>>>> On 09/23/2012 08:47 PM, Jakub Hrozek wrote:
>>>>> [PATCH 1/5] AUTOFS: Do not fail if search base is not provided
>>>>> We need to fall back to rootDSE provided search base instead of failing
>>>>
>>>> Ack.
>>>>
>>>>> [PATCH 2/5] AUTOFS: Add sysdb tests
>>>>> I knew I would be changing the sysdb interface so I coded up tests to
>>>>> avoid regressions
>>>>
>>>> Nack.
>>>> You don't free test_ctx in the three last tests.
>>>>
>>>
>>> Fixed.
>>>
>>>>> [PATCH 3/5] AUTOFS: Add entry objects below map objects
>>>>> https://fedorahosted.org/sssd/ticket/1506
>>>>
>>>> Nack.
>>>> /home/pavka/git/sssd/.git/rebase-apply/patch:273: new blank line at EOF.
>>>> +
>>>> warning: 1 line adds whitespace errors.
>>>>
>>>
>>> Fixed. Sorry, I squashed that hunk into a wrong patch.
>>>
>>>> SSSD cannot be compiled with this patch as HEAD:
>>>> src/db/sysdb_autofs.c:266:56: error: 'name' undeclared (first use in
>>>> this function)
>>>>
>>>>>
>>>>> Changes how the new autofs entry objects are handled. Instead of
>>>>> creating the entry on the cn=autofs,cn=custom level, the entry is
>>>>> created below the map it belongs to.
>>>>>
>>>>> [PATCH 4/5] AUTOFS: Use both key and value in entry RDN
>>>>> This patch switches from using just key in the RDN to using both key and
>>>>> value. That is neccessary to allow multiple direct mounts in a single
>>>>> map.
>>>>
>>>> Ack,
>>>>
>>>>> [PATCH 5/5] AUTOFS: convert the existing autofs entries during a sysdb upgrade
>>>>> The sysdb is upgraded to a new version to convert the existing entries
>>>>> to the new format.
>>>>
>>>> Ack.
>>>
>>> New patches are attached. Thank you for the review.
>>
>> Nack for patch 3.
>> mod_autofs_entry() don't use default branch for enum in the switch
>> statement. If the default branch is present, compiler cannot warn
>> you that you have forgotten some value.
>
> New patches attached again.

Ack. Good job.





More information about the sssd-devel mailing list