FAT binaries and CONFIG_SMP

Teunis Peters (teunis@usa.net)
Tue, 16 Sep 1997 09:52:52 -0600 (MDT)


On Wed, 10 Sep 1997, [ISO-8859-1] Kristian Köhntopp wrote:

> In netuse.lists.linux-kernel you write:
> >Instead of the suggested 'EXPORT_SYMBOL(i_am_SMP)' solution, IMO a better
> >way is to implement some 'powermodule' thing, single .o file, both SMP and
> >UP version of all functions. Generated automagically by "make modules",
> >interpreted by modutils. Could even have more 'styles', eg: P5,i386
> >optimizations, etc
>
> >we could even put several platforms into a single .o module file ... this
> >makes sense for distribution makers i guess.
>
> We call this "FAT binaries" in Nextstep. Our GUI frontend to
> make generates production code for Motorola, Intel, Sun and HP
> platforms in a single executeable file.

I have curiousity : Does such capability exist for Linux?
[ELF doesn't have a problem with FAT binaries...]

Hmm - this could be setup via 'interpreter', yes?
Does having an interpreter mean that the interpreter handles issues such
as platform [eg. SCO; Linux; Linux-GLIBC2; etc], relocations (does Linux
support ELF-style relocation?), and other specifics?
[relocations aren't in the kernel - and it's better that way :]

And could modules be autoloaded using this method? (as in - make them
runnable binaries :) [that would prolly be a security risc IMHO]

hmmm.....
(But I'll keep checking out this thread :)

G'day, eh?
- Teunis