Source Mage code commit list

[Jeremy Blosser <jblosser-smgl AT firinn.org>]

On Feb 24, Eric Sandall [eric AT sandall.us] wrote:
> On Saturday 24 February 2007 01:10:19 Juuso Alasuutari wrote:
> > On Friday 23 February 2007 23:19, Karsten Behrmann wrote:
> > > GIT changes to master cauldron by Karsten Behrmann
> > > <BearPerson AT sourcemage.org>:
> > >
> > >  iso/data/config-2.6 |  979
> > > ++++++++++++++++++++++++++++++++++++++-------------- 1 files changed, 
> > > 720
> > > insertions(+), 259 deletions(-)
> > >
> > > New commits:
> > > commit c8826657025d00ccd780bbf03ac4862b815d4c62
> > > Author: Karsten Behrmann <BearPerson AT sourcemage.org>
> > > Commit: Karsten Behrmann <BearPerson AT sourcemage.org>
> > >
> > >     Check in new kernel config, thanks muchly to sobukus for making one
> >
> > <snip>
> >
> > > +# CONFIG_SMP is not set
> >
> > My guesstimate based on Debian's kernel config is that SMP can be enabled
> > by default. The Debian linux-image-2.6-686-smp package is a meta package
> > for linux-image-2.6-686, which in turn is a meta package for
> > linux-image-2.6.18-4-686. Multi-core CPUs are becoming commonplace and the
> > SMP setting apparently causes no harm on single-core systems. But we'd
> > still need to test this in practice too just to make sure.
> 
> FYI,
> 
> SMP breaks on my single-CPU Athlon-XP with an nForce2 board. :(

There was a huge amount of debate on this in -cauldron when the kernel
config was being written, the end result was basically leaving it without
SMP, but people really wanting details on what it would still break to turn
it on.  So thanks for at least one example.  Do you have any more info on
why it breaks there?

How about if someone who really wants the SMP choice from the installer
files a bug to have both kernels included in 0.9.7 or 0.9.8?

Attachment: pgpR4uRIF4hxo.pgp
Description: PGP signature