Re: kernel config

Hi Erich;

Am Donnerstag, 16. April 2015, 10:02:58 schrieb Erich Titl:
> Hi KP
> 
> I had a quick look in the kernel config files and came up with a few
> questions you may be able to answer. All my findings are based on the
> latest git master version.
> 
> Bering-3.14.20.config-versatile     Bering-3.14.36.config-i486.patch
> Bering-3.14.36.config               Bering-3.14.36.config-i686
> Bering-3.14.36.config-bcmrpi        Bering-3.14.36.config-i686.patch
> Bering-3.14.36.config-bcmrpi.patch  Bering-3.14.36.config-versatile.patch
> Bering-3.14.36.config-geode         Bering-3.14.36.config-x86_64
> Bering-3.14.36.config-geode.patch   Bering-3.14.36.config-x86_64.patch
> Bering-3.14.36.config-i486
> 
> 1) Here we see the variants of the kernel config files. What is the
> versatile variant for and why is it not updated?
> 
> 2) We are keeping config files and patch files for each variant. Why is
> this so? We could either keep patch files or config files, as they can
> be created from each other. This is confusing. I am tending towards
> keeping the patch files because then we can have a base variant where
> all the common config goes.
> 
> So I would like to see the Berning-xx.yy.zz.config-variant files
> disappear from the git repository. They are just redundant.

See Andrews response; it has been useful to keep the files around. If 
somethings breaks, we can run "make oldconfig" based on these files.

> I did some work on the config files and decided to use i486 as a base
> variant. It was quite easy to create patches for the other variants from
> the i486 and it reduced patch size by more than 90%. I believe the short
> patches are easier to read and understand. Config size is not really a
> concern, config maintenance may be though.

Agreed.

> for example this is the geode variant
> -rw-rw-r-- 1 mega mega 74192 Apr 15 18:23
> Bering-3.14.36.config-geode.patch.old
> -rw-rw-r-- 1 mega mega 355 Apr 15 18:35 Bering-3.14.36.config-geode.patch
> 
> So the size of the geode patch can be reduced to 0.4% of the original
> patch size.
> 
> for the i486 config
> -rw-rw-r-- 1 mega mega 62107 Apr 15 18:23
> Bering-3.14.36.config-i486.patch.old
> -rw-rw-r-- 1 mega mega 355 Apr 15 18:24 Bering-3.14.36.config-i486.patch
> 
> The size of the i486 patch is shrunk to 0.5% of its original size
> 
> for the i686 config the difference is not that big
> -rw-rw-r-- 1 mega mega 62200 Apr 15 18:23
> Bering-3.14.36.config-i686.patch.old
> -rw-rw-r-- 1 mega mega 593 Apr 15 18:25 Bering-3.14.36.config-i686.patch
> 
> Here we just see a reduction to 1% of the original patch file.
> 
> The size of the redundant config files
> -rw------- 1 mega mega  97286 Apr 15 18:18 Bering-3.14.36.config-geode
> -rw------- 1 mega mega 101777 Apr 15 18:37 Bering-3.14.36.config-i486
> -rw------- 1 mega mega 101778 Apr 15 18:38 Bering-3.14.36.config-i686
> -rw------- 1 mega mega 101758 Apr 15 23:07
> Bering-3.14.36.config-x86_64.config
> 
> On an estimate we are keeping about 1MB of redundant information in the
> config files alone (per kernel version). One MB of redundancy is not
> really much, one MB of confusion may be :-)

True :)

I suggest you commit your changes to branch next (I think we need to merge 
master into next before?) or a a topic branch. Either way we can test your 
work before make it into master.

kp

------------------------------------------------------------------------------
BPM Camp - Free Virtual Workshop May 6th at 10am PDT/1PM EDT
Develop your own process in accordance with the BPMN 2 standard
Learn Process modeling best practices with Bonita BPM through live exercises
http://www.bonitasoft.com/be-part-of-it/events/bpm-camp-virtual- event?utm_
source=Sourceforge_BPM_Camp_5_6_15&utm_medium=email&utm_campaign=VA_SF