On Thu, 21 Dec 2000, Paul Slootman wrote:
> On Thu 21 Dec 2000, Rich Payne wrote:
> > On Thu, 21 Dec 2000, Paul Slootman wrote:
> > > On Wed 20 Dec 2000, Naohiko Shimizu wrote:
> > > > In article <200012182050.eBIKo3Y28204@sirppi.helsinki.fi>
> > > > Topi.Junkkari@iki.fi writes:
> > > >
> > > > >> The system is a 166 MHz UDB (noname) with 128 MB RAM and enough disk
> > > > >> space. I am running plain vanilla potato with kernel 2.2.17.
> > > >
> > > > I think 2.2.17 kernel needs a patch to use with EV4 processors.
> > >
> > > A UDB doesn't have an EV4 processor:
> > LCA4 was the "Low Cost Alpha", it's still based upon the EV4.
> You learn something every day :-)
> However, mine is running the 2.2.18pre version without any additional
> patches just fine.
One of the better tests for the mentioned bug is either
ghostscript or blender. Both programs loop forever using loads of
kerneltime. Even with the bug normal integer work is not affected, so not
everybody triggers this one. (Especially since ghostscript can be
compiled, for floating-point-weak pattforms, so there might be some
work-arounds.) On my Jensen this problem with traps.c had to be patched
from 2.2.14 (where I learned, that this is a kernel problem) to 2.2.17
(I have not yet tried 2.2.18).
> Paul Slootman
> home: email@example.com http://www.wurtel.demon.nl/
> work: firstname.lastname@example.org http://www.murphy.nl/
> debian: email@example.com http://www.debian.org/
> isdn4linux: firstname.lastname@example.org http://www.isdn4linux.org/
> To UNSUBSCRIBE, email to email@example.com
> with a subject of "unsubscribe". Trouble? Contact firstname.lastname@example.org