Re: 2.6.13-mm3

[Nish Aravamudan <nish.aravamudan@gmail.com>]

On 9/12/05, Paolo Ciarrocchi <paolo.ciarrocchi@gmail.com> wrote:
> On 9/12/05, Nish Aravamudan <nish.aravamudan@gmail.com> wrote:
> > On 9/12/05, Paolo Ciarrocchi <paolo.ciarrocchi@gmail.com> wrote:
> > > On 9/12/05, Andrew Morton <akpm@osdl.org> wrote:
> [...]
> > > >
> > > > - There are several performance tuning patches here which need careful
> > > >  attention and testing.  (Does anyone do performance testing any more?)
> > >
> > > How about the tool announced months ago by Martin J. Bligh ?
> > >
> > > http://ftp.kernel.org/pub/linux/kernel/people/mbligh/abat/regression_matrix.html
> >
> > Preferred location is: test.kernel.org (much shorter too!)
> 
> I wasn't aware of that, thank you! Now I won't forget anymore that URL ;-)

That was the idea, I think :)
 
> > Also, the problem for -mm3 is that -mm2 did not build on most
> > machines. -mm1 did on 4/6. Probably some determination could be made
> > from those.
> 
> I see. But I still think that automated testing is a great opportunity
> for the community to pinpoint problems.

Wasn't arguing that point by any means.

> Is there anything we can do to make thinks work better ?

See why the builds failed (the logs should say), e.g. for elm3b6, the
x86-64 box:

arch/x86_64/pci/built-in.o(.init.text+0xa88): In function `pci_acpi_scan_root':
: undefined reference to `pxm_to_node'

and try to fix 'em. Buildable kernels (which has been a problem for
-mm lately, I guess, with certain .configs at least) mean testable
kernels.

Might be fixed in -mm3, I dunno (those jobs haven't been spawned yet,
it would seem).

Thanks,
Nish