Hi Thilo, On Saturday, 10. October 2015 13:30:22 Thilo Schulz wrote: > at this time, libftdi does not seem to feature an explicit mechanism for > reading the user area (except directly accessing the data buffer in ftdi > context structure) and more importantly, it has no direct support for > writing to the user area with a correct EEPROM checksum. > > Is there a good reason for this? One could image extending the ftdi_eeprom > structure (which is opaque to programs using the library anyways) by > appropriate fields and adding library support functions for reading and > setting of user space area data, so the API would remain fully backwards > compatible.
so far probably nobody needed to really store data there. If you need access to it, providing an API that deals with the opaque ftdi_eeprom structure is a good idea. Backwards compatibility is king :) > I'd also be interested to know about the significance of the EEPROM > checksum. My FTDI chip comes preshipped with the checksum that libftdi > would compute, so where does this come from? > Does it have any effect on chip initialisation (e.g. EEPROM checksum > invalid, chip will start up with some default values) or is it solely to > satisfy user- space programs running on the host-side? back in the days, the FT245BM type chips ignored the eeprom content when the checksum was wrong. I guess never chips behave the same. If the checksum does not match -> use default values. > Last but not least I wanted to point something out that might be of > interest to you. > From the discussion: > http://developer.intra2net.com/mailarchive/html/libftdi/2015/msg00010.html > > The descriptive strings begin at byte 0x1C (=28) in the EEPROMs I have > > at > > hand (both written with FTProg and ftdi_eeprom). > > In the EEPROM from my FT232R that is factory-shipped, the descriptive > strings start at 0x18 (=24). See attached EEPROM. sorry, can't investigate atm. Currently in a "release cycle" at $dayjob. Cheers, Thomas -- libftdi - see http://www.intra2net.com/en/developer/libftdi for details. To unsubscribe send a mail to [email protected]
