[oe] [PATCH] base.bbclass: fix soc-family test

Leon Woestenberg leon.woestenberg at gmail.com
Thu Sep 9 10:07:05 UTC 2010


Hello,

On Thu, Sep 9, 2010 at 11:30 AM, Phil Blundell <philb at gnu.org> wrote:
> On Thu, 2010-09-09 at 09:11 +0200, Frans Meulenbroeks wrote:
>> 2010/9/2 Frans Meulenbroeks <fransmeulenbroeks at gmail.com>:
>> > -                if this_soc_family and not re.match(need_machine, this_soc_family):
>> > +                if (this_soc_family and not re.match(need_machine, this_soc_family)) or not this_soc_family:
>> >                     raise bb.parse.SkipPackage("incompatible with machine %s" % this_machine)
>
> I am still far from convinced that this hunk of SOC_FAMILY code is
> desirable to have in the first place but, if it's going to stay there,
> clearly it should be fixed so that it doesn't cause problems for
> non-users.
>
Same feeling here.

I don't think we should clutter OpenEmbedded with fixes for one rare
case and leave it unattended / non-understood / unmaintained for all
other cases.

Hands up for those who can maintain SOC_FAMILY support. :-)

Regards,
-- 
Leon




More information about the Openembedded-devel mailing list