Re: [PATCH 2/4] gdb-cross-canadian: use NATIVESDK paths as it happens to be here

[Richard Purdie <richard.purdie@linuxfoundation.org>]

On Sat, 2012-02-18 at 12:56 +0100, Martin Jansa wrote:
> On Fri, Feb 17, 2012 at 02:36:48PM -0800, Saul Wold wrote:
> > On 02/13/2012 07:40 AM, Martin Jansa wrote:
> > > * seems like config/config in -L was also wrong
> > >
> > > Signed-off-by: Martin Jansa<Martin.Jansa@gmail.com>
> > > ---
> > >   meta/recipes-devtools/gdb/gdb-cross-canadian.inc |   10 ++++++++--
> > >   1 files changed, 8 insertions(+), 2 deletions(-)
> > >
> > > diff --git a/meta/recipes-devtools/gdb/gdb-cross-canadian.inc b/meta/recipes-devtools/gdb/gdb-cross-canadian.inc
> > > index b5746ce..bac63b7 100644
> > > --- a/meta/recipes-devtools/gdb/gdb-cross-canadian.inc
> > > +++ b/meta/recipes-devtools/gdb/gdb-cross-canadian.inc
> > > @@ -10,12 +10,18 @@ RDEPENDS += "python-nativesdk-core python-nativesdk-lang python-nativesdk-re \
> > >
> > >   EXTRA_OECONF_append = "--with-python=${WORKDIR}/python"
> > >
> > > +NATIVESDK_NAME = "oecore-${SDK_ARCH}-${SDK_ARCH}"
> > > +NATIVESDK_PATH = "/usr/local/${NATIVESDK_NAME}"
> 
> Ah, I was reusing definition from bitbake.conf:
> meta/conf/bitbake.conf:SDK_NAME = "oecore-${SDK_ARCH}-${TARGET_ARCH}"
> 
> with only TARGET_ARCH replaced with SDK_ARCH.
> 
> Can we move "oecore" to extra varible like:
> 
> bitbake.conf:
> -SDK_NAME = "oecore-${SDK_ARCH}-${TARGET_ARCH}"
> +SDK_NAME_PREFIX = "oecore"
> +SDK_NAME = "${SDK_NAME_PREFIX}-${SDK_ARCH}-${TARGET_ARCH}"
> 
> meta-yocto/conf/distro/poky.conf:
> -SDK_NAME = "${DISTRO}-${TCLIBC}-${SDK_ARCH}-${TARGET_ARCH}"
> +SDK_NAME_PREFIX = "${DISTRO}-${TCLIBC}"
> 
> SDK_PATH is worse:
> meta/conf/bitbake.conf:SDKPATH = "/usr/local/${SDK_NAME}"
> meta-yocto/conf/distro/poky.conf:SDKPATH = "/opt/${DISTRO}/${SDK_VERSION}"
> 
> Maybe this patch is not going in right direction, but nobody replied in
> "SDK confusion Was: [oe] [meta-oe] gdb-cross-canadian: build gdb with python support"
> thread so I have sent this version which works in default (distroless)
> oe-core (and also for SHR as we don't override default SDK* variables.

This last week has been a bit problematic with conference activities
occupying many of the people who would normally have replied to that
kind of thread.

I'm very worried that you're having to copy the base bitbake definitions
as needing to do that is usually a sign there is some underlying more
fundamental problem. We really need to find a better alternative.

I therefore don't think the patch as it stands can go in but I do want
to understand the problem and try and fix it in a way that addresses the
issues. I'm probably not going to get a chance to look at things until
next week though.

Cheers,

Richard