[microblaze-uclinux] Re: [uClinux-dev] System Ace and uclinux+microblaze, is there a "reliable" driver to mount and use a Compact Flash file system?

[John Williams <jwilliams@xxxxxxxxxxxxxx>]

Hi Marco,

Marco Cavallini wrote:

> Claudio Lanconelli ha scritto:
> 
>> What is exactly the error (any message)?
>> Does it work without uclinux? Polling mode vs interrupt mode access?
>> If you mount the CF card read only under uclinux, does it work
>> reliably? If so I posted a patch some months ago on the mb-uclinux
>> mail list to fix a write related interrupt bug, you can give it a try.
>>
> Stefano is facing to the same problem on the same board :-)
> The problem is before r/w
> Every mount fails.
> I tried your patch too but without success.
> I suspect the problem is into the bitstream :'(

I'd be looking very closely at the schematics of your board, and the systemAce
connections to the FPGA - and compare them to known-good boards such as ML40x
and XUP2VP.  Do this on a systematic, wire-by-wire basis!

The thing is, that the uClinux systemACE drivers work fine on these boards,
subject to the constraints that have been discussed on the mb-uclinux list.
Specifically the CF card *must* contain a valid systemace image (see previously
recommended and general Xilinx doco for details).

I initially saw a problem on the ML40x whereby if you didn't have the board's
FPGA config mode switch set to "SystemACE", then a required PCB connection
between the FPGA and the SysAce controller wasn't made.  This caused similar
problems to what you are reporting where the driver would just hang.

Similarly, check that you are instantiating the opb_systemace controller IP in
an exactly analogous way to how it is done in the known-good ML40x and XUP2VP
designs.

Regards,

John
___________________________
microblaze-uclinux mailing list
microblaze-uclinux@xxxxxxxxxxxxxx
Project Home Page : http://www.itee.uq.edu.au/~jwilliams/mblaze-uclinux
Mailing List Archive : http://www.itee.uq.edu.au/~listarch/microblaze-uclinux/