Date: Thu, 4 Mar 2004 14:53:02 -0600 (CST) From: Quincey Koziol <koziol@ncsa.uiuc.edu> To: James Van Artsdalen <james-freebsd-amd64@jrv.org> Cc: freebsd-amd64@freebsd.org Subject: Re: freebsd-amd64 Digest, Vol 39, Issue 3 Message-ID: <200403042053.i24Kr2ne021252@sleipnir.ncsa.uiuc.edu> In-Reply-To: <200403040046.i240kgLn001980@bigtex.jrv.org>
next in thread | previous in thread | raw e-mail | index | archive | help
> > From: Quincey Koziol <koziol@ncsa.uiuc.edu>
> > Date: Wed, 3 Mar 2004 15:01:35 -0600 (CST)
> >
> > Ok, I managed to copy the source code off of the 5.2.1-RELEASE CD-ROM and
> > (after hacking the makefiles a bit) build a new world & kernel. However, after
> > installing the new kernel (and rebooting in single user, with ACPI disabled),
> > I'm back to the same behavior as when I boot the 5.2.1 CD-ROM itself - the
> > system is _glacially_ slow, taking many, many minutes to make progress in
> > booting and making interactive use (at the console, since network connections
> > always time-out) extremely difficult.
>
> This is not normal. Does the i386 CD do this too? Can you disable ACPI in the
> system ROM setup?
Hmm, I hadn't thought about the i386 CD, I'll see about trying that. I
can't disable ACPI in the BIOS setup, unfortunately...
> > Is this a general symptom of amd64 machines that others have seen? Would
> > disabling "INVARIANTS" or "WITNESS" in the GENERIC kernel build help? Or, has
> > this been addressed in 5-CURRENT and I just need the most recent source?
> > Could anyone help me out by providing a .tgz of a recent /usr/src?
>
> Does your GENERIC enable WITNESS? I don't think it ships that way.
When I re-built the kernel, it had WITNESS enabled, but I've since re-built
without INVARIANT and WITNESS and that didn't help either. :-(
Quincey
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200403042053.i24Kr2ne021252>