Re: [PATCH 00/15] mtd: rawnand: 3rd batch of cleanup

[Miquel Raynal <miquel.raynal@bootlin.com>]

Hi Boris,

Boris Brezillon <boris.brezillon@bootlin.com> wrote on Tue, 23 Oct 2018
20:49:56 +0200:

> Hello,
> 
> After the sanitization of the raw NAND API, here is a smaller
> patchset continuing in this direction.
> 
> Patch 1 is just a follow-up on the "stop passing mtd_info objects
> around" patches, except this times it takes care of internal functions.
> There's still a lot of sanitization to do in drivers in order to get
> rid of those mtd params, but I'll leave that to driver maintainers.
> 
> Patch 2 is trying to get rid of 3 forward declarations that are not
> really needed.
> 
> Patch 3 is getting rid of useless test in nand_legacy_set_defaults().
> 
> Patch 4 is making nand_exec_op() private, so that NAND controller
> drivers are not tempted to use it.
> 
> Patch 5 is getting rid of an unused flag that has been around for quite
> some time.
> 
> Patches 6 to 13 are deprecating the ->select_chip() hook, which should
> now be replaced by internal target/die selection logic directly called
> from the ->exec_op() and ecc.read/write_xxx() implementations. The
> rational behind this change is that modern controller now take care of
> asserting/deasserting the CS pin on a per-operation basis, and if we
> ever want to implement advanced queuing logic (instead of the basic
> serialization we currently have) we'll have to attach a CS line to the
> operation we want to execute.

Would you mind rebasing this series on nand/next (small conflict
because of the recent introduction of gpio helpers in the NAND core)
and also add a patch to do the same thing as above on the ams-delta
driver (just converted to ->exec_op()).

For the ST, Meson, Arasan and PLxxx drivers, I'll ask the authors to
rebase and do the necessary changes on top of your work.

Thanks,
Miquèl