In article <199604270255.TAA00411@lestat.nas.nasa.gov> Jason Thorpe <thorpej@nas.nasa.gov> writes:
>I see two major things that have changed between my working sources and 
>broken sources:
>
>	- SBus DMA burst size is now checked
>
>	- Calibrated delay was introduced.  This changed delay()
>	  significantly.
>
>Don't have time to work on this anymore, tonight ... but that's my 
>findings, diffing my own private cvs tree.
>
>That's only sparc-specific code ... I haven't looked at MI code, yet.
>

Funny thing is that those exact changes fixed my ELC problem... There may be
some timing calculation in the esp driver that makes the difference.

christos