This is a nand flash driver for the eXcite series of intelligent
cameras manufactured by Basler Vision Technologies AG.
Signed-off-by: NThomas Koeller <thomas.koeller@baslerweb.com>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

Also limit the amount we scan to one eraseblock.
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

Correct the location of the recalculation of the FIS directory size,
and also add the same recalculation for the byte-swapped case.
Signed-off-by: NRod Whitby <rod@whitby.id.au>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

Add support for correcting errors detected by the
hardware ECC.
Signed-off-by: NBen Dooks <ben-linux@fluff.org>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

Remove unused and broken mtd->ecctype and mtd->eccsize fields
from struct mtd_info. Do not remove them from userspace API
data structures (don't want to breake userspace) but mark them
as obsolete by a comment. Any userspace program which uses them
should be half-broken anyway, so this is more about saving
data structure size.
Signed-off-by: NArtem Bityutskiy <Artem.Bityutskiy@nokia.com>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

Remove ugly and weird MTD_PROGREGION_CTRLMODE_VALID() and
MTD_PROGREGION_CTRLMODE_INVALID() macros. There is only one
user of them and they are used locally just for printing.

Anyway, this patch is a preparation for removing mtd->ecctype
and mtd->eccsize, but these macros use them. Fix this.
Signed-off-by: NArtem Bityutskiy <Artem.Bityutskiy@nokia.com>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

The cfi_staa_write_buffers() uses mtd->eccsize but means mtd->writesize.
BTW, mtd-eccsize is broken and is not initialized, which means the code
fixed by this patch is broken/unused anyway.
Signed-off-by: NArtem Bityutskiy <Artem.Bityutskiy@nokia.com>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

OneNAND has internal bufferRAMs.  The driver keeps track of
what is in the bufferRAM to save having to load from the
NAND core.  After an erase operation, the driver must
mark bufferRAM invalid if it refers to the erased block.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

This fixes a regression with the RedBoot parsing code introduced by
commit 0b47d654Signed-off-by: NMartin Michlmayr <tbm@cyrius.com>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

err_pos_lut[4096] of an array with 4096 elements is a bug.

Spotted by the Coverity checker.

While I was at it, I also converted it to ARRAY_SIZE().
Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

Add checks to ensure that out-of-band reads and writes are
not attempted with an invalid offset or length.  Specifically,
the offset must be less than the size of oob for a page
and the length must not go beyond the size of the device.
Additionally the checks must adjust for auto-placement
(MTD_OOB_AUTO) of oob data.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

In commit c172471b Nico switched to using
common code for polling for command completion. Unfortunately he also used
a common default timeout for both write and erase commands, despite the
fact that erases can take a _whole_ lot longer. Use a more sensible
default for erase timeout.
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

CONFIG_MTD_CK804XROM=y, CONFIG_PCI=n results in the following compile
error:

  CC      drivers/mtd/maps/ck804xrom.o
ck804xrom.c: In function 'ck804xrom_init_one':
ck804xrom.c:114: error: implicit declaration of function 'pci_dev_get'
ck804xrom.c:114: warning: assignment makes pointer from integer without a cast
make[4]: *** [drivers/mtd/maps/ck804xrom.o] Error 1

Considering what hardware this driver is driving, a dependency on PCI
also seems logical.
Signed-off-by: NAdrian Bunk <bunk@stusta.de>
Cc: Ryan Jackson <rjackson@lnxi.com>
Signed-off-by: NAndrew Morton <akpm@osdl.org>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

OneNAND double-density package (DDP) has two chips, each with
their own bufferRAM.  The driver will skip loading data from
the NAND core if the data can be found in a bufferRAM, however
in that case, the correct chip's bufferRAM must be selected
before reading from bufferRAM.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

Provide the bad block scan with its own read function so that important error
messages that are not from the the bad block scan, can always be printed.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

When a write is done, the length written is returned.  When a
single subpage is written the length returned should be the
subpage size, however the page size was being returned.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

OneNAND can write oob to successive pages, but NAND
does not do that.  For compatibility, disallow OneNAND
from writing past the end of the page.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

Noticed by Ingo.
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

In previos patch, there's typo so fix it
Remove unnecessary goto statement
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

It use blockpage instead of a pair (block, page). It can also cover a small chunk access. 0x00, 0x20, 0x40 and so on.

And in JFFS2 behavior, sometimes it reads two pages alternatively.
e.g., It first reads A page, B page and A page.
So we check another bufferram to find requested page.
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

- Iterations of the patch to add oob auto-placement support to OneNAND left a line of code that was meant to have been deleted.
- read mtd->oobsize in onenand_transfer_auto_oob to optimized memcpy
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

Enable the use of oob operation mode MTD_OOB_AUTO with OneNAND.
Note that MTD_OOB_RAW is still not supported.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

Going over the bugs and warnings I found this one left over. The other
changes have already been correctly done for this driver but the actual
switch to pci_get_device that they assume has not.
Signed-off-by: NAlan Cox <alan@redhat.com>
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

- Remove unused fields
- Fix typo
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

When write-verify is enabled (CONFIG_MTD_ONENAND_VERIFY_WRITE),
the data written is read back and compared.  The comparison
was being made between dataRAM buffers, but this does not
verify that the data made it to the dataRAM correctly in
the first place.  This patch amends write-verify to
compare back to the original buffer.  It also now verifies
sub-page writes.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

The precise timings are board-specific (or NAND chip specific) and don't
belong here. If they're set already, then use what we find there.
Otherwise, revert to the most conservative default values (and whinge).
Signed-off-by: NDavid Woodhouse <dwmw2@infradead.org>

OneNAND records bad block information in the out-of-band area of either the first or second page of a block.  Due to a logic error, only the first page was being checked.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

OneNAND does 2 memory allocations for bad block information.
Only one of them was being freed.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

Update copyrights and code cleanup
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

- DDP code clean-up
- Reduce block & bufferram operations in DDP
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

If OneNAND is operating within specification, all operations should easily be
completed within the 20 millisecond timeout.
This patch faithlessly adds a check for the timeout and returns an error in
that case.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

We have to set ONENAND_REG_START_BLOCK_ADDRESS to avoid status error
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

The read-while-load method of reading from OneNAND needs to allow
for the change of bufferRAM address at the boundary between the
two chips in a double density (DDP) device.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

we don't need to return ecc error when 1-bit ecc.
We only return error code when 2-bit ecc error
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

Read-while-load enables higher performance read operations.
Signed-off-by: NAdrian Hunter <ext-adrian.hunter@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

Even though there is ECC error. OneNAND driver updates the buffram as valid
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

This patch teaches OneNAND to release processor in
read/write/erase cycles and let other processes proceed.
Also, remove buggi touch watchdog call which only hides
the problem instead of solving it.
Signed-off-by: NArtem Bityutskiy <Artem.Bityutskiy@nokia.com>
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>

OneNAND supports up to 4 writes at one NAND page. Add support of this feature.
Signed-off-by: NKyungmin Park <kyungmin.park@samsung.com>