update-rc.d initscript shows up as S00... rather than S09...

update-rc.d initscript shows up as S00... rather than S09...

[Darcy Watkins]

Hello,

In one of my bitbake recipes, I have...


inherit update-rc.d
...
INITSCRIPT_NAME = "watchdog-progressd"
INITSCRIPT_PARAMS = "start 09 S . stop 90 0 6 ."


...and then I install my initscript at
${D}${sysconfdir}/init.d/watchdog-progressd during do_install()

According to examples I have seen (and some of my other packages/recipes
that worked with no problem), this should result
in /etc/rcS.d/S09watchdog-progressd symlinking
to /etc/init.d/watchdog-progressd

What sort of error would cause it to show up in /etc/rcS.d/ as
S00watchdog-progressd instead?

That is 'S00' rather than 'S09'.


I am building for an x86 target using 'daisy' branch.



-- 

Regards,

Darcy

---

Darcy Watkins
Staff Engineer, Firmware
In Motion Technology, a
Sierra Wireless company
350-625 Agnes St. New Westminster, BC
Canada, V3M 5Y4
[P1]

Re: update-rc.d initscript shows up as S00... rather than S09...

[Darcy Watkins]

On Mon, 2014-05-26 at 11:21 -0700, Darcy Watkins wrote:
> Hello,
> 
> In one of my bitbake recipes, I have...
> 
> 
> inherit update-rc.d
> ...
> INITSCRIPT_NAME = "watchdog-progressd"
> INITSCRIPT_PARAMS = "start 09 S . stop 90 0 6 ."

I found what I think the cause is.  The '09' must be parsed as a number
rather than a sort-key string of digits so it is handled as an illegal
octal literal.

'07' and less was fine.  '08 and '09' failed.  I am guessing that the
literal parser returned a zero value, hence the S00.

So I switched to NOT use leading zero and it was handled OK.  It also
padded the '8' or '9' with a leading zero in the output.

     INITSCRIPT_PARAMS = "start 9 S . stop 90 0 6 ."

This is one to add to a "...for dummies" series.  ;-)


> 
> 
> ...and then I install my initscript at
> ${D}${sysconfdir}/init.d/watchdog-progressd during do_install()
> 
> According to examples I have seen (and some of my other packages/recipes
> that worked with no problem), this should result
> in /etc/rcS.d/S09watchdog-progressd symlinking
> to /etc/init.d/watchdog-progressd
> 
> What sort of error would cause it to show up in /etc/rcS.d/ as
> S00watchdog-progressd instead?
> 
> That is 'S00' rather than 'S09'.
> 
> 
> I am building for an x86 target using 'daisy' branch.
> 
> 
> 

-- 

Regards,

Darcy

---

Darcy Watkins
Staff Engineer, Firmware
In Motion Technology, a
Sierra Wireless company
350-625 Agnes St. New Westminster, BC
Canada, V3M 5Y4
[P1]

Re: update-rc.d initscript shows up as S00... rather than S09...

[ChenQi]

Hi Darcy,

Thank you for your feedback.
I think this is a bug which needs to be fixed.
I've filed a bug in bugzilla with information in your emails.
https://bugzilla.yoctoproject.org/show_bug.cgi?id=6376

Best Regards,
Chen Qi


On 05/27/2014 03:54 AM, Darcy Watkins wrote:
> On Mon, 2014-05-26 at 11:21 -0700, Darcy Watkins wrote:
>> Hello,
>>
>> In one of my bitbake recipes, I have...
>>
>>
>> inherit update-rc.d
>> ...
>> INITSCRIPT_NAME = "watchdog-progressd"
>> INITSCRIPT_PARAMS = "start 09 S . stop 90 0 6 ."
> I found what I think the cause is.  The '09' must be parsed as a number
> rather than a sort-key string of digits so it is handled as an illegal
> octal literal.
>
> '07' and less was fine.  '08 and '09' failed.  I am guessing that the
> literal parser returned a zero value, hence the S00.
>
> So I switched to NOT use leading zero and it was handled OK.  It also
> padded the '8' or '9' with a leading zero in the output.
>
>       INITSCRIPT_PARAMS = "start 9 S . stop 90 0 6 ."
>
> This is one to add to a "...for dummies" series.  ;-)
>
>
>>
>> ...and then I install my initscript at
>> ${D}${sysconfdir}/init.d/watchdog-progressd during do_install()
>>
>> According to examples I have seen (and some of my other packages/recipes
>> that worked with no problem), this should result
>> in /etc/rcS.d/S09watchdog-progressd symlinking
>> to /etc/init.d/watchdog-progressd
>>
>> What sort of error would cause it to show up in /etc/rcS.d/ as
>> S00watchdog-progressd instead?
>>
>> That is 'S00' rather than 'S09'.
>>
>>
>> I am building for an x86 target using 'daisy' branch.
>>
>>
>>