Re: adding specific config files

["JC" <jc@vtkloud.com>]

[-- Attachment #1: Type: text/plain, Size: 2441 bytes --]

On jeu., sept. 5, 2013 at 12:00 PM, Paul Eggleton <paul.eggleton@linux.intel.com="mailto:paul.eggleton@linux.intel.com">> wrote:

    On Thursday 05 September 2013 11:45:26 JC wrote:

> On 05/09/2013 11:32, Paul Eggleton wrote:

> > Unfortunately we have to have a fixed configuration for smart during

> > do_rootfs, so it has to be written to. I think though that rather than

> > trying to install a configuration file for smart using a separate recipe,

> > I would suggest modifying the smart configuration at the end of do_rootfs

> > to add the configuration elements you need. The way to do this would be

> > to define a shell function that you can add a call to in

> > ROOTFS_POSTPROCESS_COMMAND that will run smart:

> > 

> > add_extra_smart_config() {

> > 

> > 	smart --data-dir=${IMAGE_ROOTFS}/var/lib/smart channel --add rpmsys

> > 	type=rpm-sys -y> 

> > }

> > ROOTFS_POSTPROCESS_COMMAND += "add_extra_smart_config; "

> > 

> > You can put the above in the image recipe itself or in a class that your

> > image recipe inherits.

> > 

> > Cheers,

> > Paul

> 

> Thanks, I'm going to try it !!

> 

> Assuming I don't want to create a new image (i still haven't exactly

> figured how to do this in a clean way), where should I append this code ?


To be honest I always recommend people create their own image recipes

whatever they are doing since image recipes are usually trivial and

almost always need to be customised. The easiest way to do this is to

copy an existing image recipe and name it and modify it as you prefer.


> How can I make it dependend on the "package-management" feature?


If "package-management" is not in IMAGE_FEATURES the smart config gets deleted

anyway so this will have no effect. However, if desired you can use an inline

python expression within the shell function to do things conditionally, e.g.


if ${@base_contains("IMAGE_FEATURES", "package-management", "true", "false", d)}; then

...

fi


It's important to realise though that the evaluation of the inline

python happens when the expression is parsed, not when the shell

function it is executed; by the time the shell function executes the

expression will have been converted to its result.


Cheers,

Paul




    ​

    
      


    Thanks ! I'm gonna try and let you know !

    
      


    Jay

[-- Attachment #2: Type: text/html, Size: 3344 bytes --]