Hi,

thanks for adding the EMAC routines!)
I'm appending a patch (RFD) which tries to accomplish four things:

a) align labels to the size of the cache line
b) clears the accumulator before use
c) uses the fetched data to calculate the current
   and the next data. (for .order8 only and assuming an
   even number of input data)
d) reduce rounding error

Comment:
a) probably very little effect
b) no effect if this is the only routine using the MAC
c) the mac rx,ry instruction uses only one instead of two cycles.
   (if input data is not fetched from zero wait state ram
   there will be an additional benefit)
d) probably very little effect except on very low volume

Not having the toolchain installed and not having any
hardware I did not test the appended patch (I most likely
introduced some errors).

Be warned, I do not know wether this assembles/performs better/works at all!

Greetings,
Friede
__________________________________________________________________________
Erweitern Sie FreeMail zu einem noch leistungsstärkeren E-Mail-Postfach!
Mehr Infos unter http://freemail.web.de/home/landingpad/?mc=021131