Loading PRI into cards with serial EEPROM only

[Pavel Roskin]

On Fri, 24 Oct 2003, Bjorn Wesen wrote:

> On Wed, 22 Oct 2003, Pavel Roskin wrote:
> > I made a Google search for these numbers and found this page:
> > http://web.utk.edu/~rmahurin/inspiron5100/node4.html, which links to a
> > driver for Orinoco Proxim cards: http://greenblaze.com/proxim.html
> >
> > The driver is released by Agere under GPL.  I cannot find it on the Agere
> > website, but I didn't try hard.
>
> Yup I have the latest RC of this driver actually. But since I rather use the
> hostap drivers, I have now started testing my other cards instead, real
> PRISM3 cards this time :)

Good to know.

> This time I actually managed to get the aux interface to work - I can read
> and write data to the SRAM. Great. One problem though - I can't find the PDA
> anywhere (but the CIS is correct) so I guess it either is not necessary for
> plugging the primary (but I guess I have to put some default values there if
> so) or they have to be read out from the EEPROM manually in some magic way
> (since there is no primary fw, the EEPROM reading commands probably don't
> work - I haven't tried though.. perhaps I should).

I can tell you what happens in Symbol Spectrum24 Trilogy.  It may be
similar to Prism 3.  Maybe it's even called Trilogy because it's a clone
of Prism 3 :-)

Primary firmware is loaded without plugging.  Then a command 0x30 is run
by the firmware.  It reads serial EEPROM, which contains CIS and PDA.
PDA and CIS cannot be found in the memory - I scanned all memory, both RAM
and NVRAM.  So it's the only way to get the PDA.  The secondary firmware
is plugged by the PDA.

I haven't seen any problems with it, but if you care about plugging the
primary firmware, you can plug it and write again.  Or you can plug the
adapter memory using the PDRs from primary firmware.

See http://www.red-bean.com/~proski/symbol/ for the code.

> Actually I am a bit unsure on what commands are available at all before
> running with a firmware (if any).

Try ReadMIF (0x30).

> BTW it looks like the structure of the hostap driver currently is not at
> all designed to cope with cards with no firmware. Is anybody working on
> that or should I simply try to make it as good as I can and mail a patch
> somewhere ?

I'm not aware of any similar efforts.

-- 
Regards,
Pavel Roskin