[meta-freescale] [PATCH v2 2/5] ls-debug: add recipe

Luo Zhenhua zhenhua.luo at freescale.com
Wed Aug 19 20:16:58 PDT 2015

Hello all, 

Thanks for your comments. V2 has been sent out.

Best Regards,


> -----Original Message-----
> From: Swarthout Edward L-SWARTHOU
> Sent: Thursday, August 20, 2015 4:57 AM
> To: Luo Zhenhua-B19537 <zhenhua.luo at freescale.com>; Otavio Salvador
> <otavio.salvador at ossystems.com.br>
> Cc: meta-freescale at yoctoproject.org; Schmitt Richard-B43082
> <richard.schmitt at freescale.com>; Weng White-B18292
> <White.Weng at freescale.com>
> Subject: RE: [meta-freescale] [PATCH v2 2/5] ls-debug: add recipe
> From: Luo Zhenhua
> >> From: Otavio Salvador [mailto:otavio.salvador at ossystems.com.br]
> >> On Wed, Aug 19, 2015 at 6:46 AM, Luo Zhenhua wrote:
> >> >> From: Otavio Salvador [mailto:otavio.salvador at ossystems.com.br]
> >> >> I disagree. It should be ls1021a SoCs not the TWR board only.
> >> > [Luo Zhenhua-B19537] ls1021a family only has one supported
> >> board(ls1021atwr) currently, it is identical to use ls1021a SOCs and
> >> ls1021atwr at this time. We can change to ls1021a SOCs when multiple
> >> boards of ls1021a are supported by this package.
> >>
> >> That is a very serious misconception on the BSP design. We do the
> >> layer development for custom board to be easier to be supported and
> >> Freescale having one board does not mean others cannot reuse the
> >> components for their boards development.  Every BSP component should
> >> be designed with reusability in mind and this is reduces the time to
> >> market development and speeds the development being critical for
> customers.
> >
> >I agree with your point on components reusability.
> > We can't guarantee that this package can work well on 3rd party's
> >boards of same SOC,  so it is better to add the new board in the
> COMPATIBLE_MACHINE after verification.
> How can a different board break this SOC kernel feature?
> Why should I have to touch any recipe to enable the ls-debug feature on my
> ls1021aqds board?
> Ed

More information about the meta-freescale mailing list