Broken git-native in dev

Broken git-native in dev

[Henning Heinold]

Hi,

today I stumbled over some mess up with git-native. I wondered why disk was full. I examined
that my staging dir was about 3 gig's. I examined further and found ot that most of the space comes
from i686-linux/libexec/git-core. Problem is we install git-bins without calling strip
so each 130 executables is about 22 Megs. If no one has a quick solution I will try dig further.

Maybee we should put a test in the native class which checks if the executables are striped.

Bye Henning

Re: Broken git-native in dev

[Henning Heinold]

Hi again,

okay seems the problem is rather angstroem related which set BUILD_OPTIMIZATION = "-O2 -ggdb3" and
so the native packages suffer from it too.

Bye Henning

Re: Broken git-native in dev

[Khem Raj]

On (14/07/09 19:04), Henning Heinold wrote:
> Hi,
> 
> today I stumbled over some mess up with git-native. I wondered why disk was full. I examined
> that my staging dir was about 3 gig's. I examined further and found ot that most of the space comes
> from i686-linux/libexec/git-core. Problem is we install git-bins without calling strip
> so each 130 executables is about 22 Megs. If no one has a quick solution I will try dig further.
> 

I think this is due to -ggdb3 flag being passed to native recipes.
Unless we want to debug native recipes we should avoid excessive debug
info for those packages to save on disk space.

> Maybee we should put a test in the native class which checks if the executables are striped.

yeah that would work too but then why generate debug info at all if we 
do not want to keep it.

Thx
-Khem

Re: Broken git-native in dev

[Florian Boor]

Hi,

Henning Heinold schrieb:
> Maybee we should put a test in the native class which checks if the executables are striped.

+1

Greetings

Florian

-- 
The dream of yesterday                  Florian Boor
is the hope of today                    Tel: +49 271-771091-15
and the reality of tomorrow.            Fax: +49 271-771091-19
[Robert Hutchings Goddard, 1904]        florian.boor@kernelconcepts.de

1D78 2D4D 6C53 1CA4 5588  D07B A8E7 940C 25B7 9A76

Re: Broken git-native in dev

[Richard Purdie]

On Tue, 2009-07-14 at 21:10 +0200, Florian Boor wrote:
> Hi,
> 
> Henning Heinold schrieb:
> > Maybee we should put a test in the native class which checks if the executables are striped.
> 
> +1

We should be stripping native binaries and the optimisation flags sound
wrong for native builds. That should be straightforward to fix.

Sanity checks should go into insane.bbclass though.

Cheers,

Richard

-- 
Richard Purdie
Intel Open Source Technology Centre