Christoph Hellwig wrote:
>
> On Thu, Jul 27, 2000 at 12:18:42PM -0400, Theodore Y. Ts'o wrote:
> > /lib/include and /lib/include/`uname -r`, if we all can agree on it as a
> > standard, is as good as any.
> >
> > (Note that's not just the /lib/include/linux we would need symlinked; we
> > would also need /lib/include/asm as well.)
>
> It's yust plain ugly.
> a) they dont't have to be in the root filesystem
> b) headers in <foo>/lib are plain ugly, too
>
> /usr/include/$(uname -a)/$(uname -r)
>
> would be much better an support propoer cross-compiling, too.
I hope you don't really mean $(uname -a)! Along with its redundancy
w.r.t. $(uname -r), it's needlessly specific and awful for parsing by
the shell. Perhaps you meant the kernel architecture name, which is not
available from uname(1).
It would be great if people creating external kernel modules could do
something like:
a) assume their module directory is in the kernel's root directory and
use relative paths.
or
b) use a hand-edited define that specifies the location of the kernel's
root directory (as suggested by Theodore).
It would be trivial to set this up for the newbie that can't deal with
the concepts. The rest of us would just do the Right Thing.
-- David N. Lombard MSC.Software- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Mon Jul 31 2000 - 21:00:24 EST