linux.git/drivers/mtd/mtdcore.c, branch v4.9

treewide: remove redundant #include

2016-10-11T22:06:33+00:00

Kernel source files need not include  explicitly
because the top Makefile forces to include it with:

  -include $(srctree)/include/linux/kconfig.h

This commit removes explicit includes except the following:

  * arch/s390/include/asm/facilities_src.h
  * tools/testing/radix-tree/linux/kernel.h

These two are used for host programs.

Link: http://lkml.kernel.org/r/1473656164-11929-1-git-send-email-yamada.masahiro@socionext.com
Signed-off-by: Masahiro Yamada 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

mtd: introduce the mtd_pairing_scheme concept

2016-09-15T14:47:56+00:00

MLC and TLC NAND devices are using NAND cells exposing more than one bit,
but instead of attaching all the bits in a given cell to a single NAND
page, each bit is usually attached to a different page. This concept is
called 'page pairing', and has significant impacts on the flash storage
usage.
The main problem showed by these devices is that interrupting a page
program operation may not only corrupt the page we are programming
but also the page it is paired with, hence the need to expose to MTD
users the pairing scheme information.

The pairing APIs allows one to query pairing information attached to a
given page (here called wunit), or the other way around (the wunit
pointed by pairing information).
It also provides several helpers to help the conversion between absolute
offsets and wunits, and query the number of pairing groups.

Signed-off-by: Boris Brezillon 
Reviewed-by: Brian Norris

mtd: kill the nand_ecclayout struct

2016-05-05T21:51:51+00:00

Now that all MTD drivers have moved to the mtd_ooblayout_ops model we can
safely remove the struct nand_ecclayout definition, and all the remaining
places where it was still used.

Signed-off-by: Boris Brezillon

mtd: create an mtd_ooblayout_ops struct to ease ECC layout definition

2016-04-19T20:05:55+00:00

ECC layout definitions are currently exposed using the nand_ecclayout
struct which embeds oobfree and eccpos arrays with predefined size.
This approach was acceptable when NAND chips were providing relatively
small OOB regions, but MLC and TLC now provide OOB regions of several
hundreds of bytes, which implies a non negligible overhead for everybody
even those who only need to support legacy NANDs.

Create an mtd_ooblayout_ops interface providing the same functionality
(expose the ECC and oobfree layout) without the need for this huge
structure.

The mtd->ecclayout is now deprecated and should be replaced by the
equivalent mtd_ooblayout_ops. In the meantime we provide a wrapper around
the ->ecclayout field to ease migration to this new model.

Signed-off-by: Boris Brezillon

mtd: add mtd_ooblayout_xxx() helper functions

2016-04-19T20:05:47+00:00

In order to make the ecclayout definition completely dynamic we need to
rework the way the OOB layout are defined and iterated.

Create a few mtd_ooblayout_xxx() helpers to ease OOB bytes manipulation
and hide ecclayout internals to their users.

Signed-off-by: Boris Brezillon

mtd: Hook I/O activity to the MTD LED trigger

2016-04-13T08:23:16+00:00

Now that we've added the MTD LED trigger, we need
to call each I/O path to ledtrig_mtd_activity.

Signed-off-by: Ezequiel Garcia 
Acked-by: Boris Brezillon 
Signed-off-by: Jacek Anaszewski

mtd: Uninline mtd_write_oob and move it to mtdcore.c

2016-04-13T08:23:14+00:00

There's no reason for having mtd_write_oob inlined in mtd.h header.
Move it to mtdcore.c where it belongs.

Signed-off-by: Ezequiel Garcia 
Acked-by: Boris Brezillon 
Signed-off-by: Jacek Anaszewski

mtd: merge MTD development from v4.4 into for-v4.5 development

2016-01-06T00:05:56+00:00

Small conflict between some bugfixes for 4.4 and some refactoring for
4.5.

Signed-off-by: Brian Norris

mtd: fix cmdlinepart parser, early naming for auto-filled MTD

2016-01-04T18:54:18+00:00

Commit 807f16d4db95 ("mtd: core: set some defaults when dev.parent is
set") attempted to provide some default settings for MTDs that
 (a) assign the parent device and
 (b) don't provide their own name or owner

However, this isn't a perfect drop-in replacement for the boilerplate
found in some drivers, because the MTD name is used by partition
parsers like cmdlinepart, but the name isn't set until add_mtd_device(),
after the parsing is completed. This means cmdlinepart sees a NULL name
and therefore will not work properly.

Fix this by moving the default name and owner assignment to be first in
the MTD registration process.

[Note: this does not fix all reported issues, particularly with NAND
drivers. Will require an additional fix for drivers/mtd/nand/]

Fixes: 807f16d4db95 ("mtd: core: set some defaults when dev.parent is set")
Reported-by: Heiko Schocher 
Signed-off-by: Brian Norris 
Cc: Heiko Schocher 
Cc: Frans Klaver

mtd: partitions: support a cleanup callback for parsers

2015-12-09T22:57:44+00:00

If partition parsers need to clean up their resources, we shouldn't
assume that all memory will fit in a single kmalloc() that the caller
can kfree(). We should allow the parser to provide a proper cleanup
routine.

Note that this means we need to keep a hold on the parser's module for a
bit longer, and release it later with mtd_part_parser_put().

Alongside this, define a default callback that we'll automatically use
if the parser doesn't provide one, so we can still retain the old
behavior.

Signed-off-by: Brian Norris 
Reviewed-by: Boris Brezillon