Re: [Fx2lib-devel] eeprom reset program

[Dennis M. <de...@ub...>]

Gerry Rozema wrote:
> Ok, this did the trick.  I fixed my git installation first, so I could 
> check out the whole lib, instead of copy/paste from the viewer online.  
> Built it all, seemed to build ok, then I wrote up a quick little thingy 
> to read the eprom instead of write it.  Two of my devices it read back 
> fine, and I found exactly what I expected, a C0 format, and to my 
> surprise, only 8 bytes there.  The third one surprised me, it's got a C2 
> format and appears to have some firmware embedded into a larger eprom.  
> All of these are in sealed packages which will entail breaking the 
> plastic to open up and look at the parts on board.  I am _hoping_ I wont 
> have to resort to that, because one of them will definitely not go back 
> together if I do that.
>   
c2 is valid too.  It is the code for embedded firmware like you said:

c0 = just boot up with a default vendor_id/product_id and default usb 
descriptors for the chip but the firmware doesn't do anything.
c2= load the firmware from the eeprom.

> Just another 'point of interest', the third one did give me a bit of 
> trouble initially, kept reading C3 back from the first byte in the 
> eeprom, and that's not valid.  Then for giggles, I commented out the one 
> line that sets up cpu clock, unplugged and replugged the device so that 
> it comes up 'stock out of the box', and voila, now when I put it into 
> reset, load my own little bit (using cycfx2prog for now), take it out of 
> reset, let it run the program that dumps eprom contents into ram, reset 
> again, then dump dram, there it was, a properly formatted C2 header.  It 
> wont take me very long to poke at all the registers and find out what 
> setup the on board firmware is putting in place.
>   
I'm not sure why you got c3 at any time.
> I think I'm away to the races at this point.  Next step, gonna write 
> something up that dumps the entire eprom out of all of them, then store 
> that away for safe keeping, and _then_ start reprogramming eproms.
>
>   
This is exactly what I first did.  I have a program that can dump eeprom 
contents from an endpoint.  It dynamically reads the eeprom as it's 
dumping to know who much to read.  In the TRM for the ez-usb, you can 
see the iic from of the c2 data and you just need to read the length 
field and then read that many bytes in a loop until you read the entire 
iic file.  There end block stop condition is in there too.

-Dennis