freebsd stable - Apr 2012 - -DNO_CLEAN compiling & recent changes as of ~two weeks.

Somewhere is the last two weeks it seems something has changed on
8-STABLE that causes a complete rebuild of the kernel every time it is
built.

While this is not too much of a problem, it is quite a neusense when
attempting to make a quick change that would only require a recompile of
the effected code and relink that would only amount to about 5 or 10
minutes.

I am running FreeBSD 8.3-STABLE #0 r234692 i386 where this has been an
issue. The earliest part of the kernel I always see compile is ACPI code
if that helps to determine the problem.

A few committs that I picked out that might cause it but only a guess
is... (I could be wrong about these)

After 8.3-RELEASE
r234634
r234631
r234604

Before 8.3-RELEASE
r234534
r234512
r234509

Maybe just a text or data relocation issue ? Has anyone else noticed this ?


To test after a full build of kernel #0.

1) Change config ROUTETABLES={ROUTETABLES+1}
2) cd /usr/src
3) time -h make -DNO_CLEAN buildkernel

This should only take a few minutes as an example only recompiling what
only is needed.

-- 

 - (2^(N-1))
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 455 bytes
Desc: not available
Url :
http://lists.freebsd.org/pipermail/freebsd-stable/attachments/20120429/aa8f3fd8/attachment.pgp

On 29 April 2012 13:31, Jason Hellenthal <jhellenthal@dataix.net>
wrote:
>
> Somewhere is the last two weeks it seems something has changed on
> 8-STABLE that causes a complete rebuild of the kernel every time it is
> built.
>
> While this is not too much of a problem, it is quite a neusense when
> attempting to make a quick change that would only require a recompile of
> the effected code and relink that would only amount to about 5 or 10
> minutes.
>
> I am running FreeBSD 8.3-STABLE #0 r234692 i386 where this has been an
> issue. The earliest part of the kernel I always see compile is ACPI code
> if that helps to determine the problem.
>
> A few committs that I picked out that might cause it but only a guess
> is... (I could be wrong about these)
>
> After 8.3-RELEASE
> r234634
> r234631
> r234604
>
> Before 8.3-RELEASE
> r234534
> r234512
> r234509
>
> Maybe just a text or data relocation issue ? Has anyone else noticed this ?
>
>
> To test after a full build of kernel #0.
>
> 1) Change config ROUTETABLES={ROUTETABLES+1}
> 2) cd /usr/src
> 3) time -h make -DNO_CLEAN buildkernel
use -DKERNFAST or -DNO_KERNELCLEAN

-- 
Eitan Adler

On Sun, Apr 29, 2012 at 01:35:20PM -0400, Eitan Adler
wrote:
> On 29 April 2012 13:31, Jason Hellenthal <jhellenthal@dataix.net>
wrote:
> >
> > Somewhere is the last two weeks it seems something has changed on
> > 8-STABLE that causes a complete rebuild of the kernel every time it is
> > built.
> >
> > While this is not too much of a problem, it is quite a neusense when
> > attempting to make a quick change that would only require a recompile
of
> > the effected code and relink that would only amount to about 5 or 10
> > minutes.
> >
> > I am running FreeBSD 8.3-STABLE #0 r234692 i386 where this has been an
> > issue. The earliest part of the kernel I always see compile is ACPI
code
> > if that helps to determine the problem.
> >
> > A few committs that I picked out that might cause it but only a guess
> > is... (I could be wrong about these)
> >
> > After 8.3-RELEASE
> > r234634
> > r234631
> > r234604
> >
> > Before 8.3-RELEASE
> > r234534
> > r234512
> > r234509
> >
> > Maybe just a text or data relocation issue ? Has anyone else noticed
this ?
> >
> >
> > To test after a full build of kernel #0.
> >
> > 1) Change config ROUTETABLES={ROUTETABLES+1}
> > 2) cd /usr/src
> > 3) time -h make -DNO_CLEAN buildkernel
> 
> use -DKERNFAST or -DNO_KERNELCLEAN
> 
Sorry but NO_CLEAN has been doing fine for a long time now and has not
changed recently. Like I said this happened about 2 weeks ago and was
not because of these defines.

-- 

 - (2^(N-1))