On 01/19/2013 04:20 PM, Gregg Levine wrote:
For some embedded environments the kernel (or any programming) is
rarely built on the native platform. The builders, (you Stuart,
yourself Stanley, or myself) would use a properly setup cross
compiling environment.
The big problem is setting the appropriate switches.
Cross compiling the kernel is the most trivial part of the cross
compilation. You need only to set the ARCH= and CROSS_COMPILE= variables
and have the cross toolchain installed and accessible in the path.
I have never had troubles with this part and it is really, really handy
to build the kernel in the host machine with cross toolchain when you
develop some driver or want to try out different kernel configs. And
especially if you have your target (native) machine set to boot the
kernel over NFS, the "config, build and try" procedure is as simple as
configuring and building the kernel and then rebooting the target machine.
It's a real time saver.
--
Best regards,
Andrzej Telszewski
_______________________________________________
ARMedslack mailing list
[email protected]
http://lists.armedslack.org/mailman/listinfo/armedslack
