7i90 in EPP mode

Sorry to bother you Pete.

ATM I have 7i90 EPP firmware running on a xc6slx16-csg324 spartan chip. Running from a halrun session the boards is recognised, running from a config with just a basic just 72 GPIO file it appears to be fine. Linuxcnc starts and I can see the board in the config menu of AXIS.

Now the issue I have, I've tried on 2 different motherboards with built Parallel Ports set to EPP 1.9, is that mesaflash can't detect the board. So to change the firmware I have to go through the JTAG interface.
Couple of questions.
1. Is this an issue with Mesa built boards or just on my end ? In the i90_x9card I changed "constant FPGAPins: integer := 144;" to "constant FPGAPins: integer := 324;" in TopEPPSHostmot2 I changed the card & pin file sections to suit and did the usual mod to the clocks.

2. Is there a way to write the boot block, fallback and user bit files via JTAG ? If not I can just go about the usual procedure of loading firmware via JTAG.

If the FPGA part is the issue I have a couple of dev boards with 144 pin devices that I can use. But I'd have to opt for the 7c81 EPP firmware as not enough pins are broken out.

Only trying the EPP route until the RP5 drivers are sorted.

Thanks
Rob

I haven't seen an issue (we initialize the 7I90s,7C80s, and 7C81s
via EPP with mesaflash)

I will try a current version of mesaflash with a 7I90 oi EPP mode sometime
next week to make sure some bug has not crept into mesaflashes's EPP
mode support.

It might be some obscure hardware mismatch like on the EPP interfaces.

What is the exact mesaflash error message?

No 7i90 board found.
I get a "epp wide is broken" message if I uncomment that line.

I have some more debugging messages but I have to setup the other machine.
Just finished watching the final round of the MotoGP and it's about 2am.

I have a couple of boxes with Parallel Ports I can give a try. So I'll see if any of those work. Just a little confusing that the hm2_7i90 driver has no issues.

Once again thanks for your help.

OK it seemed too simple

changed:to
and becuase I'm using a different FPGA
to
So I can do all the basic, --write, --fallback, --fix-boot-block, --reahmid functions

I wondered if there was a FPGA check that failed.
(it should have a better error message in this case)

Is this something you want added to mesaflash?
(add new card type without messing with an existing one)

I think this is very isolated case, but it would be a useful feature when embarking on this kind of endeavour, would this also have to trickle down to the hostmot2 driver ? Tho I do wonder how it does affect Genuine 7i90 cards as that is the string they would report ?

Once again cheers

Probably better to use another 4 character card name.

hm2 will work with an unknown name but will complain
and not print the pin list correctly at startup, though its
pretty trivial to add a new card there also.