[OE-core] more pedantry -- looking to clarify layers definition, BBPATH, BBFILES, etc.

[Martin Jansa]

On Sun, Jul 08, 2012 at 01:59:48PM -0400, Robert P. J. Day wrote:
> 
>   (most of this is sort of self-evident but it's not documented as
> well as it could be so i just want to make sure i have it exactly
> right.)
> 
>   snippets of some layer.conf files:
> 
> oe-core:
> 
> BBPATH .= ":${LAYERDIR}"
> BBFILES += "${LAYERDIR}/recipes-*/*/*.bb"
> 
> meta-hob:
> 
> BBPATH := "${BBPATH}:${LAYERDIR}"
> BBFILES := "${BBFILES} ${LAYERDIR}/recipes-*/*/*.bb"
> 
> meta-yocto:
> 
> BBPATH := "${LAYERDIR}:${BBPATH}"
> BBFILES := "${BBFILES} ${LAYERDIR}/recipes-*/*/*.bb \
>             ${LAYERDIR}/recipes-*/*/*.bbappend"
> 
>   note how meta-hob *appends* itself to BBPATH, while meta-yocto
> *prepends* itself.  is there no possibility that this will cause some
> confusion based on how BBPATH is used to resolve the location of class
> or conf files?  i realize you should try to avoid that sort of
> conflict but it still seems possible that someone might choose the
> names of some files badly and the unpredictable ordering in BBPATH
> will cause grief, no?

Yes, order is important and prepending makes it harder for person
writting bblayers.conf to influence it:

http://lists.linuxtogo.org/pipermail/openembedded-core/2012-July/024914.html

http://lists.linuxtogo.org/pipermail/openembedded-core/2012-February/018115.html:
BBFILE_PRIORITY controls recipe priority. BBLAYERS order controls
BBPATH order which in turn controls bbclass/config priority, as things
stand today.

See also this thread for more details:
http://lists.linuxtogo.org/pipermail/openembedded-devel/2011-May/032387.html

Cheers,

-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa at gmail.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://lists.openembedded.org/pipermail/openembedded-core/attachments/20120709/c26862df/attachment-0002.sig>