libftdi Archives

Subject: Re: [EEPROM branch] "size_check" init value

From: Thomas Jarosch <thomas.jarosch@xxxxxxxxxxxxx>
To: libftdi@xxxxxxxxxxxxxxxxxxxxxxx
Date: Fri, 24 Sep 2010 16:44:28 +0200
On Friday, 24. September 2010 16:38:51 Uwe Bonnes wrote:
>     Thomas> Could you check with MProg if they limit all strings to a
> total Thomas> of 96 bytes?  If yes, we can greatly simplify our
> "size_check" Thomas> code like this:
> I don't have access to modified wine here. So I testet linux ftd2xx.
>     Thomas> if (manfacturer_size*2 + product_size*2 + serial_size*2) >
> 96) Thomas> explode_and_die();
> On a FT232R, FT_EE_Program() failed with error 18 if the string size was
> > 46. So I think, there's a bug in AN_121, and using plain 96 on all
> > chips

Interesting, the new FTProg manual also mentions 46 characters as the upper 
limit, so this seems to be a documentation bug.

> will lead to error. E.g. on a FT2232H, strings start at 0x1a(0x9a) and +
> 0x66 (0x60 bytes and 3 times length and string marker) will bring us to
>  0x7f(0xff), leaving no room for the checksum. And then the PNP marker
>  (0x02030000 or 0x02030100) is left out too.

Please see the updated eeprom map, I've documented the magic bytes
0x2 0x3 0x1. (I noticed they are mentioned on the NDAed FT2232C eeprom 
document at the end)

>     Thomas> AN_121 also mentions the "manufacturer ID" string
> additionally Thomas> to the "manufacturer" string. Any idea what that
> is?
> The manufacturer ID is a tag used for the "automatic" generation  of
> serial numbers. I have never seen it in the real EEPROM, beside
> eventually as the first letters of the generated

So I guess this is something "software only"
and not stored in the eeprom at all, too.


libftdi - see for details.
To unsubscribe send a mail to libftdi+unsubscribe@xxxxxxxxxxxxxxxxxxxxxxx   

Current Thread