Re: Prevent Nand page writes on Power failure

[Sascha Hauer <s.hauer@xxxxxxxxxxxxxx>]

On Thu, Feb 21, 2019 at 01:05:04AM +0100, Richard Weinberger wrote:
> On Wed, Feb 20, 2019 at 2:58 PM Sascha Hauer <s.hauer@xxxxxxxxxxxxxx> wrote:
> >
> >
> > Hi All,
> >
> > I have hardware here for which the normal way to turn off is just to cut
> > the power. When the powercut happens during a NAND page write then we
> > get more or less completely written pages during next boot. Very rarely
> > it seems to happen that such a half written page with only very few
> > flipped bits is erroneously detected as empty and written again which
> > then results in ECC errors when reading the data.
> >
> > The Nand in question is a Micron MT29F4G08ABADAH4 and in TN2917 Micron
> > clearly states:
> >
> > | Power loss during NAND array operations (especially Program/Erase) is a
> > | violation of the NAND voltage specifications, which is not supported and
> > | should be avoided
> >
> > Micron suggests to make the capacitors on the Nand chips supply input
> > big enough that every started operation will be finished before the
> > power goes down. Now we don't have that situation here, what I have
> > though is a power good status GPIO, so my job is to wire that up to the
> > Nand write operations.
> >
> > Now my question is how could that be done? I assume for some people a
> > power good failure means that we should write all important data away,
> > rather than preventing any Nand access. Given it's a policy decision I
> > assume user space should be involved, right?  An option might be to
> > introcude some sysfs entry to switch mtd devices to readonly mode. Would
> > that be fine? Other suggestions?
> 
> Well, we need to make sure that no new write/erase command is issued and
> the running one can complete. So, in nand_base.c you can (ab)use a lock
> to ensure that.

Nice idea.

> Regarding important data, users that care need to use fsync() anyway, so
> there is no need to trigger whatever writeback upon power failure.
> Or what else important data do you have in mind?

What you want to do on a power failure probably depends on the time
you have left. When you have a second left then you get other ideas than
when you have just a few milliseconds. If you have enough time then you
maybe even want to do an ordered shutdown which would include writing to
the nand.

Sascha

-- 
Pengutronix e.K.                           |                             |
Industrial Linux Solutions                 | http://www.pengutronix.de/  |
Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0    |
Amtsgericht Hildesheim, HRA 2686           | Fax:   +49-5121-206917-5555 |

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/