Hi Alain, Bernd, Johnson, Tom,

> >                 FreeDOS        MS-DOS7.10
> > UDMA2, no cache        R=6.3 W=4.4        R=54.2 W=16.0
> > UDMA2, cache        R=5.5 W=4.4        R=39.5 W=15.7

Bernd is probably right: EMM386, in particular Pentium VME-unaware
versions, can be the main cause of bad FreeDOS performance here.
Please try the same test without any EMM386. You should also test
with DR-DOS EMM386, which I remember to be very fast, and with
older MS DOS EMM386, which will probably be a lot slower than the
MS DOS 7 (probably Pentium-aware) one. Johnson is right, too, of
course :-).

> testing with RAWSPEED 2.2 (from older udma versions) writes/reads a 
> 256Mb file to disk (figures in Mb/s)

So you only read the file once - Tom is right in telling that a cache
cannot help with that, but on the other hand, a cache should be fast
in detecting that it cannot help, and waste only little CPU time in
such cases. You should try if the TUNA option of LBACACHE makes any
difference here. And of course: Compare to other caches. Your test
changed kernel, himem, emm386, shell and cache at the same time, so it
is a bit hard to tell which of them has how much performance impact.

> caching is useful only, if the same data are accessed twice (at least)
> else it gets only in the way.

Eric



-------------------------------------------------------
This SF.Net email is sponsored by: NEC IT Guy Games.  How far can you shotput
a projector? How fast can you ride your desk chair down the office luge track?
If you want to score the big prize, get to know the little guy.  
Play to win an NEC 61" plasma display: http://www.necitguy.com/?r=20
_______________________________________________
Freedos-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/freedos-devel

Reply via email to