Experimental Physics and
Industrial Control System

[email protected] (Bill Brown) · Wed, 25 Feb 1998 07:21:55 -0800

You don't say what your vxWorks host processor is, so I'll take a stab in
the dark here.

The way WRS handles the issue of making processor memory visable from the
vme bus is a bit strange, at least to our way of thinking.  You may recall
that John W. patched the kernel to provide a fix for the memory mapping,
which is necessary because the NI-1014 board does DMA to the cpu memory.
If the memory<->vme bus mapping is not correct you will get errors similar
to what you are describing.  I don't know if there are other examples of
device/driver support that also use DMA.

AFAIK this may only apply to mv162 processors, but it seems reasonable that
there are others that may suffer from the same problem.  I know at one point
I had to kludge other Motorola cards because of similar problems.

You might compare Johns' modified kernel against the old and new "stock
kernels; perhaps the results may point to a solution.

Disclaimer:  Any opinions are my own and have	    |  -bill
    nothing to do with the official policy or the   |   [email protected]
    management of L.B.N.L, who probably couldn't    |   Berkeley, CA
    care less about employees who play with trains. |   aka [email protected]

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System