[OE-core] [PATCH 2/2] cml1.bbclass: Add fragmentconfig task to cml1

Richard Purdie richard.purdie at linuxfoundation.org
Fri Feb 14 16:44:25 UTC 2014


On Fri, 2014-02-14 at 08:15 -0800, Darren Hart wrote:
> On 2/14/14, 7:46, "Richard Purdie" <richard.purdie at linuxfoundation.org>
> wrote:
> 
> >On Thu, 2014-02-13 at 11:00 -0800, Darren Hart wrote:
> >> On 2/13/14, 3:33, "Richard Purdie" <richard.purdie at linuxfoundation.org>
> >> wrote:
> >> 
> >> >On Wed, 2014-02-12 at 22:20 -0200, João Henrique Ferreira de Freitas
> >> >wrote:
> >> >> fragmentconfig() is a new task that makes a diff between the
> >> >> old and new config files and writes to the fragment.cfg result file.
> >> >> menuconfig() always copy the original config file, so the user
> >> >> doesn't need to copy it.
> >> >> 
> >> >> Signed-off-by: João Henrique Ferreira de Freitas <joaohf at gmail.com>
> >> >> ---
> >> >>  meta/classes/cml1.bbclass | 35 +++++++++++++++++++++++++++++++++++
> >> >>  1 file changed, 35 insertions(+)
> >> >> 
> >> >> diff --git a/meta/classes/cml1.bbclass b/meta/classes/cml1.bbclass
> >> >> index e292ecb..2053361 100644
> >> >> --- a/meta/classes/cml1.bbclass
> >> >> +++ b/meta/classes/cml1.bbclass
> >> 
> >> One more thought on this. As this is currently linux-yocto specific,
> >>does
> >> it belong in cml1.bbclass? It may make sense in busybox as well, so
> >>maybe
> >> this is the right place.
> >> 
> >> RP, do you have a preference?
> >
> >Its starting to head in a direction we could probably use a
> >kern-tools.bbclass to keep the code in a common place.
> 
> I wonder if something like kconfig.bbclass, rather than tying it to the
> linux-yocto tooling (as it doesn't use any of the tooling to create the
> fragment). Busybox could then use this, for example (not that it couldn't
> include kern-tools.bbclass, it just seems a bit less natural to do so).

Well, busybox does DEPEND on the kern-tools-native package. cml1.bbclass
is effectively a kconfig.bbclass. I'd be ok with this code there, the
fragment is kern-tools specific at this point but its not a big deal...

Cheers,

Richard




More information about the Openembedded-core mailing list