Ideally, the block should have 0xff written on the bad block position. Any value
other than 0xff implies a bad block. In practical situations, there can be
bit flips in the oob area as well which means that a block with 0x7f being read
at bad block position may imply a bad block but it is infact only a bit flip in
the bad block byte.
To resolve this problem, the block is marked as good if number of high bits is
greater than or equal to badblockbits (initialized to 7)
Signed-off-by: Vipin Kumar <vipin.kumar@st.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
ECC can correct up to 8 bits in 512 bytes data + 13 bytes ecc. This means that
the algorithm can correct a max of 8 bits in 4200 bits ie the error indices can
be from 0 to 4199. Of these 0 to 4095 are for data and 4096 to 4199 for ecc.
The driver flips the bit only if the index is <= 4096. This is a bug since the
data bits are only from 0 to 4095.
This patch modifies the check as < 4096
Signed-off-by: Vipin Kumar <vipin.kumar@st.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The ECC logic of FSMC works on 512 bytes data + 13 bytes ECC to generate error
indices of up to 8 incorrect bits. The FSMC driver reads 14 instead of 13 oob
bytes to accommodate for 16 bit device as well.
Unfortunately, the internal ecc state machine gets corrupted for 8 bit devices
reading 512 + 14 bytes of data resulting in error indices not getting reported.
Fix this by reading 14 bytes only for 16 bit devices
Signed-off-by: Vipin Kumar <vipin.kumar@st.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch reimplements the passing of partition information through platform
data. This was unintentionally deleted in commit
0d04eda143
"mtd: fsmc_nand.c: use mtd_device_parse_register"
Artem: fix gcc warning about passin 0 instead of NULL.
Signed-off-by: Vipin Kumar <vipin.kumar@st.com>
Acked-by: Stefan Roese <sr@denx.de>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Cc: stable@kernel.org [3.2+]
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch is based on Ville Herva's similar patch to block2mtd.
Trying to pass a parameter through the kernel command line when built-in would
crash the kernel, as phram_setup() was called so early that kmalloc() was not
functional yet.
This patch only saves the parameter string at the early boot stage, and parses
it later when init_phram() is called. The same happens in both module and
built-in cases.
With this patch, I can boot with a statically-compiled phram, and mount a
ext2 root fs from physical RAM, without the need for a initrd.
This has been tested in built-in and module cases, with and without a
parameter string.
Artem: amended comments a bit
Signed-off-by: Hervé Fache <h-fache@ti.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The last DMA command of ECC read page is used to disable the BCH module.
But the original code missed to set the pio[2] which is used to set the
GPMI_HW_GPMI_ECCCTRL register. fix it now.
Signed-off-by: Huang Shijie <b32955@freescale.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Flash device drivers initialize 'ecc_strength' in struct mtd_info, which is the
maximum number of bit errors that can be corrected in one writesize region.
Drivers using the nand interface intitialize 'strength' in struct nand_ecc_ctrl,
which is the maximum number of bit errors that can be corrected in one ecc step.
Nand infrastructure code translates this to 'ecc_strength'.
Also for nand drivers, the nand infrastructure code sets ecc.strength for ecc
modes NAND_ECC_SOFT, NAND_ECC_SOFT_BCH, and NAND_ECC_NONE. It is set in the
driver for all other modes.
Signed-off-by: Mike Dunn <mikedunn@newsguy.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The read function was so far requiring the reads to be aligned on page
boundaries, and be page length multiples in size. Relieve these
constraints to ease the userspace ubifs programs runs, which read ubifs
headers of 64 bytes.
Artem: squashed a later fix from Robert Jarzmik into this patch.
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Adds power management code with fine granularity. Every flash control
command is enclosed by runtime_put()/get()s. To make sure that no
overhead is generated by too frequent power state switches, a quality of
service request is issued.
Signed-off-by: Bastian Hecht <hechtb@gmail.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The first 3 arguments of 'mtd_device_parse_register()' are pointers,
but many callers pass '0' instead of 'NULL'. Fix this globally. Thanks
to coccinelle for making it easy to do with the following semantic patch:
@@
expression mtd, types, parser_data, parts, nr_parts;
@@
(
-mtd_device_parse_register(mtd, 0, parser_data, parts, nr_parts)
+mtd_device_parse_register(mtd, NULL, parser_data, parts, nr_parts)
|
-mtd_device_parse_register(mtd, types, 0, parts, nr_parts)
+mtd_device_parse_register(mtd, types, NULL, parts, nr_parts)
|
-mtd_device_parse_register(mtd, types, parser_data, 0, nr_parts)
+mtd_device_parse_register(mtd, types, parser_data, NULL, nr_parts)
)
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch is part of a set which fixes unnecessary flash erase and write errors
resulting from the MTD CFI driver turning off vpp while an erase is in progress.
This patch allows sa1100_set_vpp() calls to be nested by adding a reference
counter.
Signed-off-by: Paul Parsons <lost.distance@yahoo.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch is part of a set which fixes unnecessary flash erase and write errors
resulting from the MTD CFI driver turning off vpp while an erase is in progress.
This patch allows pcmciamtd_set_vpp() calls to be nested by adding a reference
counter.
Signed-off-by: Paul Parsons <lost.distance@yahoo.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch is part of a set which fixes unnecessary flash erase and write errors
resulting from the MTD CFI driver turning off vpp while an erase is in progress.
This patch allows l440gx_set_vpp() calls to be nested by adding a reference
counter.
Signed-off-by: Paul Parsons <lost.distance@yahoo.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch is part of a set which fixes unnecessary flash erase and write errors
resulting from the MTD CFI driver turning off vpp while an erase is in progress.
This patch allows physmap_set_vpp() calls to be nested by adding a reference
counter.
omap1_set_vpp() already used a reference counter. Since it is called from
physmap_set_vpp(), omap1_set_vpp() can now be simplified.
simtec_nor_vpp() already disabled hard interrupts. Since it is called from
physmap_set_vpp(), simtec_nor_vpp() can now be simplified.
Signed-off-by: Paul Parsons <lost.distance@yahoo.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch is part of a set which fixes unnecessary flash erase and write errors
resulting from the MTD CFI driver turning off vpp while an erase is in progress.
This patch ensures that only those flash operations which call ENABLE_VPP() can
then call DISABLE_VPP(). Other operations should never call DISABLE_VPP().
Signed-off-by: Paul Parsons <lost.distance@yahoo.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch is part of a set which fixes unnecessary flash erase and write errors
resulting from the MTD CFI driver turning off vpp while an erase is in progress.
This patch ensures that only those flash operations which call ENABLE_VPP() can
then call DISABLE_VPP(). Other operations should never call DISABLE_VPP().
Signed-off-by: Paul Parsons <lost.distance@yahoo.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The current patch is required to support EVALSPEAR1340CPU
Revision 2 where a new (ONFI compliant) MT29F16G08 NAND
flash from Micron is present.
This NAND flash device defines a OOB area which is
224 bytes long (oobsize).
Signed-off-by: Armando Visconti <armando.visconti@st.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch improves the error correction routine for bch8
- Loop only up to number of errors detected
- Improve the error index calculation procedure
Additionally, it also renames the "correct" routine to indicate that it is bch8
specific
Signed-off-by: Armando Visconti <armando.visconti@st.com>
Signed-off-by: Vipin Kumar <vipin.kumar@st.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Since change_bit() requires a (unsigned int *) as second arg,
the correct definition of err_idx[] array declared as
local variable of fsmc_correct_data() is the following:
u32 err_idx[8];
Signed-off-by: Armando Visconti <armando.visconti@st.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
ALE and CLE offsets can be different on different devices. Let devices
pass these offsets to the fsmc driver through platform data.
Signed-off-by: Shiraz Hashim <shiraz.hashim@st.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
ECC1 & ECC4 layout for NAND of different pages sizes for e.g. 512bytes,
2KiB, 4KiB and 8KiB are separated. Previously there existed one ECC4
layout for 2KiB & 4KiB page size due to which oob test module available
in drivers/mtd/nand/test was failing.
Signed-off-by: Bhavna Yadav <bhavna.yadav@st.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
A newly erased page contains ff in data as well as spare area. While reading an
erased page, the read out ecc from spare area does not match the ecc generated
by fsmc ecc hardware accelerator. This is because ecc of data ff ff is not ff
ff. This leads to errors when file system erases and reads back the pages to
ensure consistency.
This patch adds a software workaround to ensure that the ecc check is not
performed for erased pages. This problem is solved by checking the number of
bits (in 512 byte data + 13 byte ecc) which are 0. If these number of bits are
less than 8, the page is considered erased and correction algorithm is not tried
on that page
Signed-off-by: Vipin Kumar <vipin.kumar@st.com>
Acked-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch reverts a change that may have been mistakenly included with the set
of patches that introduced the new mtd api entry functions. Or perhaps I am
mistaken :)
The problem is in the partition wrapper functions, where the calls to the driver
methods were replaced with calls to the new mtd api functions. This causes the
api function to be called a second time, further down the call stack. This is
not only unnecessary and redundant - because the sanity checking code and (more
restrictive) bounds checks for the partition were done in the first call - but
is potentially problematic and confusing.
For example, the call stack for a call to mtd_read() on a partitioned device
currently looks like this:
mtd_read() gets struct mtd_info for the partition
|
+-> part_read() via the pointer assigned when the partition was created
|
+->mtd_read() this time gets struct mtd_info for the master
|
+->xyz_driver_read() via the pointer asigned by the driver
It seems that this can cause a variety of problems. For example, if you want to
add code to the api function that tests a value in mtd_info that is relevant
only to the partition. Or (in my case) you want the driver to return a value
that may be different from that returned by the mtd api function.
This patch eliminates the second call to the mtd api function. It was tested on
the docg4 nand driver with a subset of the api functions, but I inspected the
rest and don't see any problems.
Signed-off-by: Mike Dunn <mikedunn@newsguy.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Change the name of the mtd so that it is simpler, and is easier to
cope with by mtdparts.
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Add a register used in new FLCTL hardware and a feature flag for it.
Signed-off-by: Bastian Hecht <hechtb@gmail.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Instead of reading out the register, use a cached value. This will
make way for a proper runtime power management implementation.
Signed-off-by: Bastian Hecht <hechtb@gmail.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Implements the command to seek and read in pages.
Signed-off-by: Bastian Hecht <hechtb@gmail.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The nand base code wants to read out 8 bytes in the READID command.
Reflect this in the driver code.
Signed-off-by: Bastian Hecht <hechtb@gmail.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Reorders the calls to make it a bit shorter and match the calling
procedure displayed in the datasheet.
Signed-off-by: Bastian Hecht <hechtb@gmail.com>
Acked-by: Laurent Pinchart <laurent.pinchart@ideasonboard.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Since commit ca97dec2ab the
command line parsing of MTD partitions does not work anymore.
Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@googlemail.com>
Signed-off-by: John Crispin <blogic@openwrt.org>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: stable@kernel.org [3.2+]
Acked-by: John Crispin <blogic@openwrt.org>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Some strange nand chip(such as Hynix H27UBG8T2A) can pass the `ONFI` signature
check. So the log can be printed out even it is not an ONFI nand indeed.
Change this log to the end of the function. Print out the log only when we
really detect an ONFI nand.
Signed-off-by: Huang Shijie <b32955@freescale.com>
Acked-by: Florian Fainelli <ffainelli@freebox.fr>
Acked-by: Brian Norris <computersforpeace@gmail.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
[1] Background :
The GPMI does ECC read page operation with a DMA chain consist of three DMA
Command Structures. The middle one of the chain is used to enable the BCH,
and read out the NAND page.
The WAIT4END(wait for command end) is a comunication signal between
the GPMI and MXS-DMA.
[2] The current DMA code sets the WAIT4END bit at the last one, such as:
+-----+ +-----+ +-----+
| cmd | ------------> | cmd | ------------------> | cmd |
+-----+ +-----+ +-----+
^
|
|
set WAIT4END here
This chain works fine in the mx23/mx28.
[3] But in the new GPMI version (used in MX50/MX60), the WAIT4END bit should
be set not only at the last DMA Command Structure,
but also at the middle one, such as:
+-----+ +-----+ +-----+
| cmd | ------------> | cmd | ------------------> | cmd |
+-----+ +-----+ +-----+
^ ^
| |
| |
set WAIT4END here too set WAIT4END here
If we do not set WAIT4END, the BCH maybe stalls in "ECC reading page" state.
In the next ECC write page operation, a DMA-timeout occurs.
This has been catched in the MX6Q board.
[4] In order to fix the bug, rewrite the last parameter of mxs_dma_prep_slave_sg(),
and use the dma_ctrl_flags:
---------------------------------------------------------
DMA_PREP_INTERRUPT : append a new DMA Command Structrue.
DMA_CTRL_ACK : set the WAIT4END bit for this DMA Command Structure.
---------------------------------------------------------
[5] changes to the relative drivers:
<1> For mxs-mmc driver, just use the new flags, do not change any logic.
<2> For gpmi-nand driver, and use the new flags to set the DMA
chain, especially for ecc read page.
Acked-by: Shawn Guo <shawn.guo@linaro.org>
Signed-off-by: Huang Shijie <b32955@freescale.com>
Acked-by: Vinod Koul <vinod.koul@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Move the header to a more common place.
The mxs dma engine is not only used in mx23/mx28, but also used
in mx50/mx6q. It will also be used in the future chips.
Rename it to mxs-dma.h, and create a new folder include/linux/fsl/ to
store the Freescale's header files.
change mxs-dma driver, mxs-mmc driver, gpmi-nand driver, mxs-saif driver
to the new header file.
Acked-by: Shawn Guo <shawn.guo@linaro.org>
Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
Signed-off-by: Huang Shijie <b32955@freescale.com>
Acked-by: Vinod Koul <vinod.koul@linux.intel.com>
Acked-by: Chris Ball <cjb@laptop.org>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
While debugging on SA11x0, the following message was observed:
"Flash device refused suspend due to active operation (state 20)"
Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The gpmi driver selects the MTD_CMDLINE_PARTS directly.
But we should not select a visible symbol.
Just remove the select.
Signed-off-by: Huang Shijie <b32955@freescale.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
In commit "c797533 mtd: abstract last MTD partition parser argument" the
third argument of "mtd_device_parse_register()" changed from start address
of the MTD device to a pointer to a struct.
The "ixp4xx_flash_probe()" function was not converted properly, causing
an oops during boot.
This patch fixes the problem by filling the needed information into a
"struct mtd_part_parser_data" and passing it to
"mtd_device_parse_register()".
Signed-off-by: Marc Kleine-Budde <mkl@blackshift.org>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Cc: stable@kernel.org [3.2+]
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This patch changes all the OTP functions like 'mtd_get_fact_prot_info()' and
makes them return zero immediately if the input 'len' parameter is 0. This is
not really needed currently, but most of the other functions do this, and it is
just consistent to do the same in the OTP functions.
This patch also moves the OTP functions from the header file to mtdcore.c
because they become a bit too big for being inlined.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
In many places in drivers we verify for the zero length, but this is very
inconsistent across drivers. This is obviously the right thing to do, though.
This patch moves the check to the MTD API functions instead and removes a lot
of duplication.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Reviewed-by: Shmulik Ladkani <shmulik.ladkani@gmail.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Some MTD drivers return -EINVAL if the 'phys' parameter is not NULL, trying to
convey that they cannot return the physical address. However, this is not very
logical because they still can return the virtual address ('virt'). But some
drivers (lpddr) just ignore the 'phys' parameter instead, which is a more
logical thing to do.
Let's harmonize this and:
1. Always initialize 'virt' and 'phys' to 'NULL' in 'mtd_point()'.
2. Do not return an error if the physical address cannot be found.
So as a result, all drivers will set 'phys' to 'NULL' if it is not supported.
None of the 'mtd_point()' users use 'phys' anyway, so this should not break
anything. I guess we could also just delete this parameter later.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
This header is tiny and contains only pmc551-private stuff, so it should
not live in 'include/linux' - let's just merge it with pmc551.c.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The MTD API function now zero the 'retlen' parameter before calling
the driver's method — do not do this again in drivers. This removes
duplicated '*retlen = 0' assignent from the following methods:
'mtd_point()'
'mtd_read()'
'mtd_write()'
'mtd_writev()'
'mtd_panic_write()'
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Many drivers check whether the partition is R/O and return -EROFS if yes.
Let's stop having duplicated checks and move them to the API functions
instead.
And again a bit of noise - deleted few too sparse newlines, sorry.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
We already verify that offset and length are within the MTD device size
in the MTD API functions. Let's remove the duplicated checks in drivers.
This patch only affects the following API's:
'mtd_erase()'
'mtd_point()'
'mtd_unpoint()'
'mtd_get_unmapped_area()'
'mtd_read()'
'mtd_write()'
'mtd_panic_write()'
'mtd_lock()'
'mtd_unlock()'
'mtd_is_locked()'
'mtd_block_isbad()'
'mtd_block_markbad()'
This patch adds a bit of noise by removing too sparse empty lines, but this is
not too bad.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Add verification of the offset and length to MTD API functions and verify that
MTD device offset and length are within MTD device size.
The modified API functions are:
'mtd_erase()'
'mtd_point()'
'mtd_unpoint()'
'mtd_get_unmapped_area()'
'mtd_read()'
'mtd_write()'
'mtd_panic_write()'
'mtd_lock()'
'mtd_unlock()'
'mtd_is_locked()'
'mtd_block_isbad()'
'mtd_block_markbad()'
This patch also uninlines these functions and exports in mtdcore.c because they
are not performance-critical and do not have to be inlined.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
The 'mtd_unpoint()' API function should be able to return an error code because
it may fail if you specify incorrect offset. This patch changes this MTD API
function and amends all the drivers correspondingly.
Also return '-EOPNOTSUPP' from 'mtd_unpoint()' when the '->unpoint()' method is
undefined. We do not really need this currently, but this just makes
sense to be consistent with 'mtd_point()'.
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Currently, the flash-based BBT implementation writes bad block data only
to its flash-based table and not to the OOB marker area. Then, as new bad
blocks are marked over time, the OOB markers become incomplete and the
flash-based table becomes the only source of current bad block
information. This becomes an obvious problem when, for example:
* bootloader cannot read the flash-based BBT format
* BBT is corrupted and the flash must be rescanned for bad
blocks; we want to remember bad blocks that were marked from Linux
So to keep the bad block markers in sync with the flash-based BBT, this
patch changes the default so that we write bad block markers to the proper
OOB area on each block in addition to flash-based BBT. Comments are
updated, expanded, and/or relocated as necessary.
The new flash-based BBT procedure for marking bad blocks:
(1) erase the affected block, to allow OOB marker to be written cleanly
(2) update in-memory BBT
(3) write bad block marker to OOB area of affected block
(4) update flash-based BBT
Note that we retain the first error encountered in (3) or (4), finish the
procedures, and dump the error in the end.
This should handle power cuts gracefully enough. (1) and (2) are mostly
harmless (note that (1) will not erase an already-recognized bad block).
The OOB and BBT may be "out of sync" if we experience power loss bewteen
(3) and (4), but we can reasonably expect that on next boot, subsequent
I/O operations will discover that the block should be marked bad again,
thus re-syncing the OOB and BBT.
Note that this is a change from the previous default flash-based BBT
behavior. If your system cannot support writing bad block markers to OOB,
use the new NAND_BBT_NO_OOB_BBM option (in combination with
NAND_BBT_USE_FLASH and NAND_BBT_NO_OOB).
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Signed-off-by: Artem Bityutskiy <artem.bityutskiy@linux.intel.com>
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
