Re: linux-next: Tree for March 13

[Stephen Rothwell]

Hi Geert,

On Thu, 13 Mar 2008 11:31:39 +0100 (CET) Geert Uytterhoeven <geert@xxxxxxxxxxxxxx> wrote:
>
> I noticed you added `m68k-allnoconfig', which fails because of the following

This was an experiment.  I disabled it again.

> reasons:
>   - There's no single CPU type enabled, hence we get:
> 
> 	#define FPSTATESIZE error no_cpu_type_configured
> 
>   - There's no single platform type enabled, hence we get no definition of
>     NR_IRQS:
> 
> 	#error unknown nr of irqs
> 
> Well yes, these are all `features' of allnoconfig :-)
> Or should we make sure you cannot build invalid configs like this?

I think that for architectures that do allow allnoconfig to build, they
have a fallback for these situations.  allnoconfig is a good quick way to
see if the core code builds ok for an architecture.

So maybe you could choose (even at random) defaults for the above two
things rather than erroring.  That would also help with randconfig builds.

If you make it work, let me know and I will reenable the build.
-- 
Cheers,
Stephen Rothwell                    sfr@xxxxxxxxxxxxxxxx
http://www.canb.auug.org.au/~sfr/

Attachment: pgp00000.pgp
Description: PGP signature