Re: [PATCH] Add rdc321x defconfig file

[Adrian Bunk <bunk@kernel.org>]

On Mon, Feb 25, 2008 at 01:57:07PM +0100, Ingo Molnar wrote:
> 
> * Adrian Bunk <bunk@kernel.org> wrote:
> 
> > > the existing 32-bit and 64-bit defconfigs should be enough for that. 
> > > For better/full coverage, randconfig should be used.
> > 
> > The two big problems with randconfigs are:
> > - either you build each .config both with and without your patch or you
> >   have to manually check which of the failures are caused by your patch
> > - you require at least an order of magnitude more builds for having the
> >   same amount of common configurations covered
> > 
> > And any solution that only works on x86 (e.g. based on the expectation 
> > that all randconfig configurations normally build) is of zero value 
> > for me since x86 is only one out of 23 architectures.
> 
> so if an arguably sane testing method "only" works on x86 then the right 
> solution is to fix the other architectures to be sanely testable too. 

If you want to fix them I won't stop you...

Until they are fixed I'm staying at using the defconfigs.

But then there's still the other problem that at least I simply don't 
want to wait two weeks for having the test compiles of a patch finish.

> I've seen architectures that were build-tested for the _first time_ at 
> around 2.6.24-rc8...

That can't be true.

Can you name what architectures you think of and why you think noone 
tried to compile them before?

> 	Ingo

cu
Adrian

-- 

       "Is there not promise of rain?" Ling Tan asked suddenly out
        of the darkness. There had been need of rain for many days.
       "Only a promise," Lao Er said.
                                       Pearl S. Buck - Dragon Seed