Re: A first sign of life from LEDs

On Wed Feb 2 11:54:34 2005 Michael Opdenacker wrote:
> . . .
>
> Does this make sense? I'm glad you found an explanation which is
> consistent with what we know so far...

I think it makes sense, but I might have it backwards. It could be that
the hx4700 is hooked up directly and the h4000 (and others?) has it's
address shifted *right* one bit.

Another interesting point: the PXA270 register MSC1, which controls
accesses to CS3 where ASIC3 is, says that ASIC3 is set up as a 16-bit
device. This means a 32-bit read at 0x0c000100 is actually two 16-bit
reads at 0x0c000100 and 0x0c000102, and ASIC 3 supposedly sees offsets
0x80 and 0x81 on its address bus. This is documented on page 6-33 of
the PXA manual in the "note". This is why I see two registers with a
single 32-bit read. When I change the CS3 region to 32-bit, I can only
access every other register, and the upper 16 bits are always zero,
which is what you'd expect.

Aric.
Received on Wed Feb 02 2005 - 15:53:42 EST