[OE-core] Architecture mismatch QA check not fatal?

[Mark Hatle]

On 6/20/11 2:41 AM, Koen Kooi wrote:
> Hi,
> 
> I was building qt this weekend and I noticed this one:
> 
> WARNING: QA Issue: Architecture did not match (40 to 62) on /work/armv7a-angstrom-linux-gnueabi/qt4-embedded-4.7.3-r26.1/packages-split/qt4-embedded-tools/usr/bin/moc
> WARNING: QA Issue: Architecture did not match (40 to 62) on /work/armv7a-angstrom-linux-gnueabi/qt4-embedded-4.7.3-r26.1/packages-split/qt4-embedded-tools/usr/bin/uic
> WARNING: QA Issue: Architecture did not match (40 to 62) on /work/armv7a-angstrom-linux-gnueabi/qt4-embedded-4.7.3-r26.1/packages-split/qt4-embedded-tools/usr/bin/rcc
> 
> Shouldn't that be a fatal error, shipping x86 binaries in arm packages?

There are a couple of very minor use cases where this may be necessary.  So I'm
wondering if we can put in an override mechanism that tells the system to make
the QA a warning instead of an error for select packages.  Otherwise, yes.. this
should be an error.

(The minor cases are primarily firmware loaded into off-board cards in PCI
chasis and such.. occasionally these firmware are ELF and of an architecture not
the same as the host..  it's rare, but I have seen it before.)

--Mark

> regards,
> 
> Koen
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core at lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core