Just a thought that popped into my head

Steve Bougerolle steveb at creek-and-cowley.com
Wed May 8 16:04:08 PDT 2002


On Wed, 8 May 2002 17:44:18 +0100
Ian Molton <spyro at armlinux.org> wrote:

> Look, Im not trying to be rude, but 'compiling something' ISNT a test.

It would be nice if somebody in this debate would post (and start
arguing about) what exactly IS a good test.

At a minimum, compiled software should pass its own tests - that is,
"make check" shouldn't produce grievous error messages.  That was enough
to rule out an earlier version of gcc 3.0.x for me.

We've already mentioned that the system should be able to rebuild itself
(and I would insist that the rebuilt system should also pass its
checks).  That also ruled out an earlier version of gcc 3.0.x.

What are some other tests we could be doing?  

I have a new system I built using gcc 3.0.4 that's passed its first
round of checks - I haven't rebuilt it and done said checks again but
will.  However, if Gerard says he's done this and it all works then I
won't doubt him at this point.

On the other hand, I've seen one message giving an example of how gcc
3.0.x can't compile the kernel properly.  If true then my new system is
no good (because I build it for other people to use on other computers
and it MUST be able to recompile kernels).  A nice kernel testing suite
would be invaluable.

-- 
Steve Bougerolle
Creek & Cowley Consulting

http://www.creek-and-cowley.com
-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message



More information about the lfs-dev mailing list