--qcHopEYAB45HaUaB
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline

Aymeric asked that I send this to the list, so that more people can cast
their eyes over the information - I hope it's useful.

As an extra data point, I got a panic last night whilst doing a build.sh
run: pmap_pinit(): out of segments

Unfortunately, I didn't write down the backtrace - if I can get this
again, I'll try to do so.

Ian.
 

--qcHopEYAB45HaUaB
Content-Type: message/rfc822
Content-Disposition: inline

Date: Tue, 13 Jan 2004 12:27:12 +0000
From: Ian Fry <Ian.Fry@sophos.com>
To: Aymeric Vincent <Aymeric.Vincent@labri.fr>
Subject: Re: macppc port "unstable" in -current
Message-ID: <20040113122712.GA23341@sophos.com>
References: <wnh7k0btwwv.fsf@phantom.labri.fr>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <wnh7k0btwwv.fsf@phantom.labri.fr>

On Thu, Jan 01, 2004 at 06:13:04PM +0100, Aymeric Vincent wrote:
> Does anyone have a hint as to what might have changed in the last
> months that causes this?
> 
> Here are a few thoughts:
> 
> - signal trampolines
> - top down VM
> - PPC64 additions

I managed to find an old kernel at home that is pretty stable, so I've
done some comparisons of the source file revisions (using ident). Attached
is a diff of the idents for a 1.6ZF kernel and a 1.6ZG kernel with the MI
file stuff stripped out. Looking at the commit messages for these files in
cvsweb indicates that the culprit could either be the PPC64 changes or
some signal changes.

I hope this information is useful, I'm afraid I haven't been able to build
a ZG kernel with some of the older revisions to see if I can track down
the exact change that causes the current instability, but if you want me
to test patches, please let me know, and I'll see what I can do.

Ian.


--qcHopEYAB45HaUaB--