Re: [PATCH 1/4] mtd: introduce the mtd_pairing_scheme concept

From: Brian Norris
Date: Mon Jun 13 2016 - 02:22:55 EST


On Sun, Jun 12, 2016 at 10:55:32PM -0700, Brian Norris wrote:
> On Sat, Jun 11, 2016 at 08:54:08AM +0200, Boris Brezillon wrote:
> > On Fri, 10 Jun 2016 19:17:15 -0700
> > Brian Norris <computersforpeace@xxxxxxxxx> wrote:
[...]
> > > Also, the "pair" term (and examples you use) seem to imply 2-cell MLC,
> > > whereas I believe you're trying to handle TLC too. I don't know if we
> > > should drop the "pair" term, or just explain it better.
> >
> > I clearly have some problems with the words I've chosen, but those terms
> > were extracted from NAND datasheets (group and pair), and I think
> > keeping the same wording help people converting datasheet specs into
> > pairing scheme implementation.
> >
> > Any suggestions to replace those 2 words?
>
> I'm not sure we should replace the words (esp. if those are used by
> multiple vendors). [...]

I see that George highlighted a Micron datasheet in other parts of this
thread, and I noticed it uses the term "shared page." That explains why
I couldn't find the word "pair" in my quick search of Micron datasheets!
So I guess "shared page" would be a nomination, though I'm certainly not
forcing it, if you think pair is better.

Brian