On Tue, 3 Feb 2004, Ralf Baechle wrote:
> > Perhaps it wasn't really tested. Have you ever run the code on a PC
> > variant? Has anyone else?
>
> Yes, it has. Olivetti M700-10, around 2.2 or so. The code used at that
> time in arch/mips/mm/r4xx0.c was not much different from what pg-r4k.c is
> generating now that is it violates this hazard.
OK then.
> > I suppose so -- without the "mips-pg-r4k-scache" patch my system is very
> > unstable and the difference is essentially that in addition to
> > Create_Dirty_Excl_SD there are additional Create_Dirty_Excl_D ones, that,
> > apart from being a performance hit, shouldn't have any effect. I have to
> > investigate it further yet.
>
> Interesting, I was expecting somewhat better performance from the
> combination of both. Anyway, what is now in CVS is tested on R4400SC V4.0.
After rereading the description I agree with you, so I am going to get at
the code again soon.
--
+ Maciej W. Rozycki, Technical University of Gdansk, Poland +
+--------------------------------------------------------------+
+ e-mail: macro@ds2.pg.gda.pl, PGP key available +
|