standard for kernel version calculation?

Robert P. J. Day rpjday at mindspring.com
Thu Mar 30 11:19:05 UTC 2006


On Wed, 29 Mar 2006, Rob Landley wrote:

> On Wednesday 29 March 2006 2:55 pm, Robert P. J. Day wrote:
> >   um ... would someone like to suggest a single standard for
> > kernel version comparison?  the (abridged) result of a recursive
> > grep:
> >
> > 	if (get_kernel_revision() <= 2*65536+6*256)
> > 	if (get_kernel_revision() < KRELEASE(2, 1, 0))
> > 	if (get_kernel_revision() < MAKE_VERSION(2, 1, 117))
> >         if (get_kernel_revision() <= 66314
> >
> > i'm thinking that perhaps just one of these variations would be
> > sufficient. :-)
>
> It would be nice if anything that actually needed to care could be
> moved to platform.h or platform.c somehow, but I doubt it's likely
> in the short term...

sure, but in the short term, we can at least standardize on the
comparison.  the kernel source uses the macro:

#define KERNEL_VERSION(a,b,c) (((a) << 16) + ((b) << 8) + (c))


any reason not to add that definition to something like platform.h or
libbb.h and change all those tests to use that?

rday



More information about the busybox mailing list