In the current data sheets corresponding to Toshiba and Samsung Nand
Flash device, invalid blocks are detected by verifying the content of
the address 517 on the first page of each block, as specified in SSFDC
standard. In the old NAND data sheets, the way to detect these was
checking the whole content on the first page of each block by being
different to 0xff.
What I am writing is an independent generic device driver. Can I
assume that the invalid blocks detection is the first one described
above? If not, do you have any suggestion?

Thanks in advance !

If you look at the data sheets from both manufacturers there seem to be
different techniques.

ST Micro's flash apparenty guarantees that the 6the byte of the C area of
either page 1 or 2 are the indicators. however, when I test these parts, ALL
of the C area bytes are set to Zero. So I stick with the MFG recommendation.
From what I saw, the Samsung and Toshiba parts are slightly different so I
use the MFG ID to vector to the proper routines.. Or at least I plan to..
When I get samples of the other parts..

In general, you only need to test those bytes.. not the entire page.


"Ricardo Malerba" <r_malerba@hotmail.com> wrote in message
news:1f4618e.0407300650.7c230e82@posting.google.co m...