Re: [PATCH 0/3] MTD: Change meaning of -EUCLEAN return code on reads

[Mike Dunn <mikedunn@newsguy.com>]

On 03/19/2012 12:18 PM, Shmulik Ladkani wrote:
> On Mon, 19 Mar 2012 12:09:02 -0700 Mike Dunn <mikedunn@newsguy.com> wrote:
>>
>> I didn't get Shmulik's original email for some reason, only what was quoted in
>> Matthieu's. 
> 
> http://lists.infradead.org/pipermail/linux-mtd/2012-March/040343.html


Many thanks Shmulik.  I'm glad I didn't miss this.  Don't know why some emails
are getting dropped.  What you describe is exactly what I had in mind, and is
just an extension of the previous patches to the finer granularity of ecc steps.
 Yes it's not too hard, but touching drivers that I don't have the means to test
has to be done very carefully.

Quoting from your post off the list archive...


  [ BTW, at a side note, I failed to understand why you prefer the
  threshold comparison at the generic 'mtd_read' wrapper, and not within
  nand_base.c.
  I guess you didn't want to duplicate the condition into onenand_base.c
  and alauda.c?
  Thing is, I feel mtd->bitflip_threshold is a NAND property, so it
  makes more sense if it is tested within the NAND infrastricture (and
  clones).
  Changing the 'mtd->_read' interface was less elegant IMO. ]

I agree, it is a NAND property and would have preferred to keep it in the nand
interface,  But in addition to alauda, devices/docg3.c also is a nand device
that does not use the nand interface,  After being bandied about, it was decided
to make it part of the mtd driver interface.

Thanks,
Mike