On Tuesday 29 April 2003 19:49, Stephan Linke wrote:
>
>
> Hi Charles,
>
> We recently did some tests to verify the number of possible erase/write
> cycles on a NAND. There was one interresting detail related to error
> correction. At some point we detected an error when reading back the data
> (several single bit errors, same bit in every corrupted byte). After
> continuing the test by erasing the block this bit error disappeared until
> the end of the test... So maybe YAFFS is a bit too aggressive in dealing
> with single bit errors.
>
> Stephan

Stepahn 

I hope you don't mind me CCing the yaffs list.

My understanding of the way NAND decays with use/age is that the blocks will 
start reporting intermittent errors (most likely single bit errors) and 
slowly get worse until the block becomes unpredictable and unsafe to use.

Developing a strategy for dealing with these single-bit errors is quite 
tricky:
* Too aggressive and you retire blocks early.
* Not aggressive enough and you let the problem worsen to the stage where 
single bit errors become unrecoverable (2-bit errors).

Getting good information on how NAND fails is a challenge. If anyone has good 
info on this, I'd like to hear.

Personally, I'd rather sacrifice a few blocks to preserve of data integrity, 
until I have information that can show it is safe to do otherwise.

-- Charles






---------------------------------------------------------------------------------------
This mailing list is hosted by Toby Churchill open software (www.toby-churchill.org).
If mailing list membership is no longer wanted you can remove yourself from the list by 
sending an email to yaffs-request@toby-churchill.org with the text "unsubscribe" 
(without the quotes) as the subject.