[RFC PATCH v2] mtd: nand: add erased-page bitflip correction

Peter van Vugt pwvanvugt at gmail.com
Thu Mar 19 12:37:45 PDT 2015


Gupta, Pekon <pekon <at> ti.com> writes:
 
> I'll wait for your next revision of patch,  but please re-think on my 
proposal
> to leave out ecc.layout->oobfree[], and just reset ecc.laout->ecc.pos[].
> as it keeps compatibility with other file-systems which use OOB to store
> some metadata.
> 
> (may be you can put resetting complete OOB, as separate patch later).
> 
> >> - Therefore in my earlier patch [2] I was just counting bit-flips in 
read_ecc[]
> >>   *not* read_oob[].
> >
> >You can't have your cake and eat it too. You're trying to support
> >intentionally programming bitflips (i.e., the 'cleanmarker') to spare
> >area, but you're also trying to correct bits in such a page. Pick one.
> >Or improve the JFFS2 / MTD API so that JFFS2 can communicate when it
> >does and doesn't want error correction, rather than making assumptions.
> >

For the record, has this patch been abandoned?

I have encountered issues with one (and occasionally more) non-1 bits in 
erased pages on several vendors' MLC NAND devices when using the GPMI 
controller, which causes significant issues when using UBI. There was an 
earlier GPMI-specific, not platform-agnostic patch 
(http://lists.infradead.org/pipermail/linux-mtd/2014-
January/thread.html#51356) that appears it may have been abandoned due to 
silicon bugs that may have been fixed in my silicon revision, and a OMAP-
specific patch that made it into the mainline in commit 62116e51 .

The GPMI patch appears to have a few small issues in our use case that 
might be fixable, but if I've missed something and some version of this 
patch is still alive in some other form, I'd prefer to use/support it.

Regards,

Peter





More information about the linux-mtd mailing list