Why it would be good to have AC interdiff patches

Why it would be good to have AC interdiff patches

[Gregoire Favre]

Hello,

I know I could do interdiff myself using the patchutils.
I have done it for example for patch-2.4.20-pre4-ac1-ac2.bz2 and it gave
the first reason:

1) patch-2.4.20-pre4-ac1-ac2.bz2 is only 26K (patch-2.4.20-pre4-ac1.bz2
   and patch-2.4.20-pre4-ac2.bz2 are both 1.1M), so as AC kernels are
   very popular, it would reduce bandwight
2) One could more easyly watch what has changed
3) The time for compilation would be smaller as you modify less files.
4) All script on kernel.org are already written for vanilla kernels and
   I am quiete sure it wouldn't be that hard to include the AC...
(5) With my ISDN connection it would be much faster to test those news
   kernels...).

Thank you very much, and in case of response, don't cc to me, receiving
email once is enough ;-)

Have a great day,

	Grégoire
________________________________________________________________
http://ulima.unil.ch/greg ICQ:16624071 mailto:greg@ulima.unil.ch

Re: Why it would be good to have AC interdiff patches

[Axel Siebenwirth]

Hi Gregoire!

Gregoire Favre schrieb am Thursday, den 29. August 2002:

> 3) The time for compilation would be smaller as you modify less files.

Somebody wrote me one time that it is always a good thing to make an
"mrproper" after I have applied a patch to kernel sources.

Is that right?

Thank you,
Axel

Re: Why it would be good to have AC interdiff patches

[Gregoire Favre]

On Thu, Aug 29, 2002 at 10:15:33AM +0200, Axel Siebenwirth wrote:

Hello list and Axel ;-)

> > 3) The time for compilation would be smaller as you modify less files.
> 
> Somebody wrote me one time that it is always a good thing to make an
> "mrproper" after I have applied a patch to kernel sources.
> 
> Is that right?

Well, I never did...

Why CC to me even if I ask not to?

Thank you,

	Grégoire
________________________________________________________________
http://ulima.unil.ch/greg ICQ:16624071 mailto:greg@ulima.unil.ch

Re: Why it would be good to have AC interdiff patches

[Tomas Szepe]

> > > 3) The time for compilation would be smaller as you modify less files.
> > 
> > Somebody wrote me one time that it is always a good thing to make an
> > "mrproper" after I have applied a patch to kernel sources.
> > 
> > Is that right?
> 
> Well, I never did...

You're asking for serious problems. ld will often not fail when
the ABI changes and evil things *will* happen if a fc ends up
popping nonsense from stack.