Re: Harddisk transferrates

[Philip Blundell <Philip.Blundell@pobox.com>]

>Nothing special, I changed only the Makefile to compile with -O4
>-mtune=strongarm.

You almost certainly don't want to use -O4.  That will usually generate worse 
code for the kernel than -O2.

>But I investigated this a  bit more and at least the Arm610 transferrate is
>totally wrong (reported by "hdparm -t /dev/hda") because the time reported for
>the measurement is is only about half of that I measure with my own watch and
>this does for instance not happen on a x86 linux box.

Ah.  Maybe you are losing clock ticks; try `hdparm -u'.

p.