Two issues. One minor, one major.

Andreas Tobler andreast-list at fgznet.ch
Thu Dec 11 03:08:14 PST 2008


--------- Original Message --------
From: Garrett Cooper <yanefbsd at gmail.com>
To: Horst Günther Burkhardt III <horst at sxemacs.org>
Cc: FreeBSD PowerPC ML <freebsd-ppc at freebsd.org>
Subject: Re: Two issues. One minor, one major.
Date: 11/12/08 11:01

> On Wed, Dec 10, 2008 at 9:59 AM, Horst Günther Burkhardt III
> &lt;horst at sxemacs.org&gt; wrote:
> &gt; Another thing I forgot: for those brave enough to look at this
problem :
> &gt; the entire session log can be found at
http://rafb.net/p/MOTbHY26.html
> &gt; relating to the failure of python25 to build. Any help would be
greatly
> &gt; appreciated.
> &gt;
> &gt; I'm completely stumped as to the cause of this odd error.
> &gt;
> &gt; ktrace and kdump links available:
> &gt; ktrace  :
http://midcom.unlogicnetworks/tmpfiles/python-compileall-trace.txt.bz2
> &gt; kdump -f:
http://midcom.unlogicnetworks/tmpfiles/python-compileall-kdump.txt.bz2
> &gt; Hopefully these will help unearth what went wrong to some extent.
> &gt;
> &gt; Is there anything further I can provide to assist with diagnosing the
problem? (It has been suggested that this is an upstream bug perhaps.)
> &gt;
> &gt; Thanks again,
> &gt; -- Horst.
> 
> Why are you using mcpu and mtune directly? You can use CPUTYPE, e.g.:
> 
> CPUTYPE                 := nocona
> 
> (Replace nocona with 7400 8-)...)
> 
> I did a quick google search and while -mcpu=7400 and -mtune=7400
> doesn't necessarily imply -maltivec, I would ask the GCC folks whether
> or not that is truly the case. Or a quicker means to determine if that
> is the problem is to use mcpu=common (or a known working 32-bit ppc
> arch) and see whether or not you run across the same error.

-mcpu=7400 does imply -mppc -maltivec. See the rs6000.h in the
gcc/config/rs6000 directory.
I'd suggest to build w/o special cpu flags.

Andreas




More information about the freebsd-ppc mailing list