[Mageia-dev] Help: Please find me initscripts without LSB headers!

[Colin Guthrie]

'Twas brillig, and Guillaume Rousse at 18/04/12 12:11 did gyre and gimble:
> Le 18/04/2012 00:48, Colin Guthrie a écrit :
>> 'Twas brillig, and Pascal Terjan at 11/04/12 11:55 did gyre and gimble:
>>> On Wed, Apr 11, 2012 at 11:30, Colin Guthrie<mageia at colin.guthr.ie> 
>>> wrote:
>>>> 'Twas brillig, and Pascal Terjan at 11/04/12 11:24 did gyre and gimble:
>>>>> On Wed, Apr 11, 2012 at 11:07, Colin
>>>>> Guthrie<mageia at colin.guthr.ie>  wrote:
>>>>>> Hi,
>>>>>>
>>>>>> I need to find all initscripts we ship that do not contain LSB header
>>>>>> information!
>>>>>>
>>>>>> Ideally you should find them and fix the scripts, but if in doubt
>>>>>> ask me!
>>>>>>
>>>>>> This is the result of debug work on:
>>>>>> https://bugs.mageia.org/show_bug.cgi?id=5262#c19
>>>>>>
>>>>>> Where overzealous use of priorities from non-LSB compatible
>>>>>> scripts has
>>>>>> resulted in ordering cycle deadlocks. We need to nuke these problems.
>>>>>>
>>>>>> This can be fixed by providing proper LSB headers in initscripts
>>>>>> (likely
>>>>>> easier) or native systemd units (better if you can do it).
>>>>>>
>>>>>> Here is a document that tracks all packages we provide which provide
>>>>>> sysvinit scripts. Please can people go through the list and fill
>>>>>> in the
>>>>>> blanks. I've started it off. The only cells to free edit are
>>>>>> either your
>>>>>> name, or notes about things. The others have data validation  -
>>>>>> i.e. a
>>>>>> drop down. I've provided links to subversion which in many cases is
>>>>>> sufficient to verify the initscripts as they are often just
>>>>>> included in
>>>>>> the SOURCES folder, not packaged upstream. For some however you will
>>>>>> need to check a bit.
>>>>>
>>>>> Can't this be automated?
>>>>> Get the list of binary rpms containing initscripts with urpmf (I guess
>>>>> that's what you did), extract the initiscripts from a local mirror,
>>>>> grep to see if they have LSB Headers?
>>>>
>>>> That part can be automated I guess. I don't have a handy mirror tho'.
>>>> Also if the script has native systemd it might be a false negative, but
>>>> I guess they would be suitably small to be checked by hand.
>>>>
>>>> That said, the list isn't super long :)
>>>
>>> Just did a quick try, packages with initscripts not containing LSB
>>> headers.
>>> 92 initscripts in 67 packages, but there a few false positives like
>>> initscripts containing /etc/init.d/functions
>>>
>>> afbackup
>>> arpwatch
>>> bip
>> ...
>>> yum
>>> zoneminder
>>
>> I've incorporated this list into the spreadsheet. I've fixed a few of
>> them.
>>
>> Again, if any packagers are able to help out here, Please do!!!
>>
>> Here is the link again:
>> https://docs.google.com/spreadsheet/ccc?key=0AjumKqYMRG9WdHdkRWM3dUhEYms5T1BkR05KRnp6V1E
>>
>>
>> You only need to look at the ones marked "Broken" in bold now. If you
>> fix it, please change it to "Has LSB" or "Native" depending on your fix.

> Well, there is another list of migrated packages:
> https://wiki.mageia.org/en/Features/Systemd
> It's a bit painful to maintain both lists simultaneously.

Well this isn't about migrating to native systemd specifically, but
rather ensuring it doesn't break with system. The wiki page is more for
the former, although, yes there could easily be massive overlap.

I figured using the google doc would be generally easier than editing a
wiki page, but if we prefer I can just migrate the extra data in the
google doc over to the wiki?

Col


-- 

Colin Guthrie
colin(at)mageia.org
http://colin.guthr.ie/

Day Job:
  Tribalogic Limited http://www.tribalogic.net/
Open Source:
  Mageia Contributor http://www.mageia.org/
  PulseAudio Hacker http://www.pulseaudio.org/
  Trac Hacker http://trac.edgewall.org/