The default NAND read functions are relying on the underlying controller
driver to correct bitflips, but some of those controllers cannot properly
fix bitflips in erased pages.
Check for bitflips in erased pages in default core functions if the driver
delegated the this check by setting the NAND_ECC_GENERIC_ERASED_CHECK flag.
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Tested-by: Franklin S Cooper Jr. <fcooper@ti.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Commits such as commit 853f1c58c4 ("mtd: nand: omap2: show parent
device structure in sysfs") attempt to rely on the core MTD code to set
the MTD name based on the parent device. However, nand_base tries to set
a different default name according to the flash name (e.g., extracted
from the ONFI parameter page), which means NAND drivers will never make
use of the MTD defaults. This is not the intention of commit
853f1c58c4.
This results in problems when trying to use the cmdline partition
parser, since the MTD name is different than expected. Let's fix this by
providing a default NAND name, where possible.
Note that this is not really a great default name in the long run, since
this means that if there are multiple MTDs attached to the same
controller device, they will have the same name. But that is an existing
issue and requires future work on a better controller vs. flash chip
abstraction to fix properly.
Fixes: 853f1c58c4 ("mtd: nand: omap2: show parent device structure in sysfs")
Reported-by: Heiko Schocher <hs@denx.de>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Tested-by: Heiko Schocher <hs@denx.de>
Cc: Heiko Schocher <hs@denx.de>
Cc: Frans Klaver <fransklaver@gmail.com>
Cc: <stable@vger.kernel.org>
Now that the nand_chip struct directly embeds an mtd_info struct we can
get rid of the ->flash_node field and forward set/get_flash_node requests
to the MTD layer.
As a side effect, we no longer need the mtd_set_of_node() call done in
nand_dt_init().
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
nand_dt_init() function requires 3 arguments where it actually needs one
(dn and mtd can both be retrieved from chip). Drop these parameters.
Testing for dn != NULL inside nand_dt_init() also helps simplifying the
caller code.
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
mtd_to_nand() was recently introduced to avoid direct access to the
mtd->priv field. Update core code to use mtd_to_nand().
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
If multiple NAND chips are registered to the same controller, then when
rebooting the system, the first one will grab the controller lock, while
the second will wait forever for the first one to release it. i.e., a
classic deadlock.
This problem was solved for a similar case (suspend/resume) back in
commit 6b0d9a8412 ("mtd: nand: fix multi-chip suspend problem"), and
the shutdown state really isn't much different for us, so rather than
adding a new special case to nand_get_device(), we can just overload the
FL_PM_SUSPENDED state.
Now, multiple chips can "get" the same controller lock (preventing
further I/O), while we still allow other chips to pass through
nand_shutdown().
Original report:
http://thread.gmane.org/gmane.linux.drivers.mtd/59726http://lists.infradead.org/pipermail/linux-mtd/2015-July/059992.html
Fixes: 72ea403669 ("mtd: nand: added nand_shutdown")
Reported-by: Andrew E. Mileski <andrewm@isoar.ca>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Cc: Scott Branden <sbranden@broadcom.com>
Cc: Andrew E. Mileski <andrewm@isoar.ca>
Acked-by: Scott Branden <sbranden@broadcom.com>
Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
We should pass along our flash DT node to the MTD layer, so it can set
up ofpart for us.
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
If nand_wait_ready() times out, this is silently ignored, and its
caller will then proceed to read from/write to the chip before it is
ready. This can potentially result in corruption with no indication as
to why.
While a 20ms timeout seems like it should be plenty enough, certain
behaviour can cause it to timeout much earlier than expected. The
situation which prompted this change was that CPU 0, which is
responsible for updating jiffies, was holding interrupts disabled
for a fairly long time while writing to the console during a printk,
causing several jiffies updates to be delayed. If CPU 1 happens to
enter the timeout loop in nand_wait_ready() just before CPU 0 re-
enables interrupts and updates jiffies, CPU 1 will immediately time
out when the delayed jiffies updates are made. The result of this is
that nand_wait_ready() actually waits less time than the NAND chip
would normally take to be ready, and then read_page() proceeds to
read out bad data from the chip.
The situation described above may seem unlikely, but in fact it can be
reproduced almost every boot on the MIPS Creator Ci20.
Therefore, this patch increases the timeout to 400ms. This should be
enough to cover cases where jiffies updates get delayed. In nand_wait()
the timeout was previously chosen based on whether erasing or
programming. This is changed to be 400ms unconditionally as well to
avoid similar problems there. nand_wait() is also slightly refactored
to be consistent with nand_wait{,_status}_ready(). These changes should
have no effect during normal operation.
Debugging this was made more difficult by the misleading comment above
nand_wait_ready() stating "The timeout is caught later" - no timeout was
ever reported, leading me away from the real source of the problem.
Therefore, a pr_warn() is added when a timeout does occur so that it is
easier to pinpoint similar problems in future.
Signed-off-by: Alex Smith <alex.smith@imgtec.com>
Signed-off-by: Harvey Hunt <harvey.hunt@imgtec.com>
Reviewed-by: Niklas Cassel <niklas.cassel@axis.com>
Cc: Alex Smith <alex@alex-smith.me.uk>
Cc: Zubair Lutfullah Kakakhel <Zubair.Kakakhel@imgtec.com>
Cc: David Woodhouse <dwmw2@infradead.org>
Cc: Niklas Cassel <niklas.cassel@axis.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
The ->read_xxx() methods are all passed the page number the NAND controller
is supposed to read, but ->write_xxx() do not have such a parameter.
This is a problem if we want to properly implement data
scrambling/randomization in order to mitigate MLC sensibility to repeated
pattern: to prevent bitflips in adjacent pages in the same block we need
to avoid repeating the same pattern at the same offset in those pages,
hence the randomizer/scrambler engine need to be passed the page value
in order to adapt its seed accordingly.
Moreover, adding the page parameter to the ->write_xxx() methods add some
consistency to the current API.
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
CC: Josh Wu <josh.wu@atmel.com>
CC: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
CC: Maxime Ripard <maxime.ripard@free-electrons.com>
CC: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
CC: Huang Shijie <shijie.huang@arm.com>
CC: Stefan Agner <stefan@agner.ch>
CC: devel@driverdev.osuosl.org
CC: linux-arm-kernel@lists.infradead.org
CC: linux-kernel@vger.kernel.org
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
The ->init_size() hook was introduced to let NAND controller drivers
support NAND devices that could not be described in the nand_ids table.
Since then, the core has added support for extended-id parsing and
full-id description, thus allowing to describe pretty much all existing
NANDs.
Moreover, this hook is not used by any mainline driver, and should not be
used by new drivers, because detecting the NAND chip is not something
controller specific.
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Add two helper functions to help NAND controller drivers test whether a
specific NAND region is erased or not.
Signed-off-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Use a more descriptive name for the device_node element in struct nand_chip .
This name matches the element name used for device_node property of a flash
in the spi-nor framework.
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
This is an example of why it doesn't make much sense to put this
information here in the first place. I don't really know what purpose it
serves.
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
These are already-documented common bindings for NAND chips. Let's
handle them in nand_base.
If NAND controller drivers need to act on this data before bringing up
the NAND chip (e.g., fill out ECC callback functions, change HW modes,
etc.), then they can do so between calling nand_scan_ident() and
nand_scan_tail().
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
As all four bytes are written in any case the memset() is in vain.
Signed-off-by: Richard Weinberger <richard@nod.at>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
If a NAND device is not really present or pin muxes are not correctly
configured we can lock up the kernel waiting infinitely for NAND_STATUS
to be ready.
This can be easily reproduced on TI's DRA7-evm board by booting it
without NAND support in u-boot and disabling NAND pin muxes in the kernel.
Add timeout when waiting for NAND_CMD_RESET completion. As per ONFi v4.0
tRST can be upto 250ms for EZ-NAND and 5ms for raw NAND.
Signed-off-by: Roger Quadros <rogerq@ti.com>
Tested-by: Nishanth Menon <nm@ti.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
We're not initializing the ooblen field. Our users don't care, since
they check that oobbuf == NULL first, but it's good practice to zero
unused fields out.
We can drop the NULL initializations since we're memset()ing the whole
thing.
Noticed by Coverity, CID #200821, #200822
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Coverity noticed that these 'ret' assignments weren't being used. Let's
use them.
Note that nand_lock() and nand_unlock() are still not officially used by
any drivers.
Coverity CIDs #1227054 and #1227037
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Previously, we requested that drivers pass ecc.size and ecc.bytes when
using NAND_ECC_SOFT_BCH. However, a driver is likely to only know the ECC
strength required for its NAND, so each driver would need to perform a
strength-to-bytes calculation.
Avoid duplicating this calculation in each driver by asking drivers to
pass ecc.size and ecc.strength so that the strength-to-bytes calculation
need only be implemented once.
This reverts/generalizes this commit:
mtd: nand: Base BCH ECC bytes on required strength
Signed-off-by: Aaron Sierra <asierra@xes-inc.com>
Reviewed-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
The MTD API reports -EUCLEAN only if the maximum number of bitflips
found in any ECC block exceeds a certain threshold. This is done to
avoid excessive -EUCLEAN reports to MTD users, which may induce
additional scrubbing of data, even when the ECC algorithm in use is
perfectly capable of handling the bitflips.
This threshold can be controlled by user-space (via sysfs), to allow
users to determine what they are willing to tolerate in their
application. But it still helps to have sane defaults.
In recent discussion [1], it was pointed out that our default threshold
is equal to the correction strength. That means that we won't actually
report any -EUCLEAN (i.e., "bitflips were corrected") errors until there
are almost too many to handle. It was determined that 3/4 of the
correction strength is probably a better default.
[1] http://lists.infradead.org/pipermail/linux-mtd/2015-January/057259.html
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Acked-by: Huang Shijie <shijie.huang@intel.com>
Commit 7854d3f749 ("mtd: spelling, capitalization, uniformity") added
a correctly spelled line, but failed to remove the wrongly spelled one.
Commit 064a7694b5 ("mtd: Fix typo mtd/tests") then fixed the spelling
again, but left the duplication.
Fixes: 7854d3f749 ("mtd: spelling, capitalization, uniformity")
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Add nand_shutdown to wait for current nand operations to finish and prevent
further operations by changing the nand flash state to FL_SHUTDOWN.
This is addressing a problem observed during reboot tests using UBIFS
root file system: NAND erase operations that are in progress during
system reboot/shutdown are causing partial erased blocks. Although UBI should
be able to detect and recover from this error, this change will avoid
the creation of partial erased blocks on reboot in the middle of a NAND erase
operation.
Signed-off-by: Scott Branden <sbranden@broadcom.com>
Tested-by: Scott Branden <sbranden@broadcom.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
It may be useful info, e.g. if someone wants to use ubinize.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
update a comment in nand_command_lp() about specific requirements of
individual commands, the DEPLETE1 command was removed in the past and
the comment no longer applied
Signed-off-by: Gerhard Sittig <gsi@denx.de>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
NAND devices with page sizes over 4 KiB require more than 4-bits of ECC
coverage. This patch calculates the value of ecc_bytes based on a still
assumed 512-byte step size (13-bits) and the ecc_strength.
Example:
Micron M73A devices (8 KiB page) require 8-bit ECC per 512-byte
Signed-off-by: Jordan Friendshuh <jfriendshuh@xes-inc.com>
Signed-off-by: Aaron Sierra <asierra@xes-inc.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Add an onfi_timing_mode_default field to nand_chip and nand_flash_dev in
order to support NAND timings definition for non-ONFI NAND.
NAND that support better timings mode than the default one have to define
a new entry in the nand_ids table.
The default timing mode should be deduced from timings description from
the datasheet and the ONFI specification
(www.onfi.org/~/media/ONFI/specs/onfi_3_1_spec.pdf, chapter 4.15
"Timing Parameters").
You should choose the closest mode that fit the timings requirements of
your NAND chip.
Signed-off-by: Boris BREZILLON <boris.brezillon@free-electrons.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
This follows Chapter 2 of Linux's CodingStyle:
> However, never break user-visible strings such as printk messages,
> because that breaks the ability to grep for them.
Signed-off-by: Rafał Miłecki <zajec5@gmail.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
chip->pagebuf is a 32-bit type (int), so the shift will only be applied
as 32-bit. Fix this for 64-bit safety.
Caught by Coverity.
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Do nand reset before write protect check.
If we want to check the WP# low or high through STATUS READ and check bit 7,
we must reset the device, other operation (eg.erase/program a locked block) can
also clear the bit 7 of status register.
As we know the status register can be refreshed, if we do some operation to trigger it,
for example if we do erase/program operation to one block that is locked, then READ STATUS,
the bit 7 of READ STATUS will be 0 indicate the device in write protect, then if we do
erase/program operation to another block that is unlocked, the bit 7 of READ STATUS will
be 1 indicate the device is not write protect.
Suppose we checked the bit 7 of READ STATUS is 0 then judge the WP# is low (write protect),
but in this case the WP# maybe high if we do erase/program operation to a locked block,
so we must reset the device if we want to check the WP# low or high through STATUS READ and
check bit 7.
Signed-off-by: White Ding <bpqw@micron.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
In commit 67a9ad9b8a ("mtd: nand: Warn the user if the selected ECC
strength is too weak"), a check was added to inform the user when the
ECC used for a NAND device is weaker than the recommended ECC
advertised by the NAND chip. However, the warning uses WARN_ON(),
which has two undesirable side-effects:
- It just prints to the kernel log the fact that there is a warning
in this file, at this line, but it doesn't explain anything about
the warning itself.
- It dumps a stack trace which is very noisy, for something that the
user is most likely not able to fix. If a certain ECC used by the
kernel is weaker than the advertised one, it's most likely to make
sure the kernel uses an ECC that is compatible with the one used by
the bootloader, and changing the bootloader may not necessarily be
easy. Therefore, normal users would not be able to do anything to
fix this very noisy warning, and will have to suffer from it at
every kernel boot. At least every time I see this stack trace in my
kernel boot log, I wonder what new thing is broken, just to realize
that it's once again this NAND ECC warning.
Therefore, this commit turns:
------------[ cut here ]------------
WARNING: CPU: 0 PID: 1 at /home/thomas/projets/linux-2.6/drivers/mtd/nand/nand_base.c:4051 nand_scan_tail+0x538/0x780()
Modules linked in:
CPU: 0 PID: 1 Comm: swapper Not tainted 3.16.0-rc3-dirty #4
[<c000e3dc>] (unwind_backtrace) from [<c000bee4>] (show_stack+0x10/0x14)
[<c000bee4>] (show_stack) from [<c0018180>] (warn_slowpath_common+0x6c/0x8c)
[<c0018180>] (warn_slowpath_common) from [<c001823c>] (warn_slowpath_null+0x1c/0x24)
[<c001823c>] (warn_slowpath_null) from [<c02c50cc>] (nand_scan_tail+0x538/0x780)
[<c02c50cc>] (nand_scan_tail) from [<c0639f78>] (orion_nand_probe+0x224/0x2e4)
[<c0639f78>] (orion_nand_probe) from [<c026da00>] (platform_drv_probe+0x18/0x4c)
[<c026da00>] (platform_drv_probe) from [<c026c1f4>] (really_probe+0x80/0x218)
[<c026c1f4>] (really_probe) from [<c026c47c>] (__driver_attach+0x98/0x9c)
[<c026c47c>] (__driver_attach) from [<c026a8f0>] (bus_for_each_dev+0x64/0x94)
[<c026a8f0>] (bus_for_each_dev) from [<c026bae4>] (bus_add_driver+0x144/0x1ec)
[<c026bae4>] (bus_add_driver) from [<c026cb00>] (driver_register+0x78/0xf8)
[<c026cb00>] (driver_register) from [<c026da5c>] (platform_driver_probe+0x20/0xb8)
[<c026da5c>] (platform_driver_probe) from [<c00088b8>] (do_one_initcall+0x80/0x1d8)
[<c00088b8>] (do_one_initcall) from [<c0620c9c>] (kernel_init_freeable+0xf4/0x1b4)
[<c0620c9c>] (kernel_init_freeable) from [<c049a098>] (kernel_init+0x8/0xec)
[<c049a098>] (kernel_init) from [<c00095f0>] (ret_from_fork+0x14/0x24)
---[ end trace 62f87d875aceccb4 ]---
Into the much shorter, and much more useful:
nand: WARNING: MT29F2G08ABAEAWP: the ECC used on your system is too weak compared to the one required by the NAND chip
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
In addition to mtd_block_isbad(), which checks if a block is bad or
reserved, it's needed to check if a block is reserved only (but not
bad). This commit adds an MTD interface for it, in a similar fashion to
mtd_block_isbad().
While here, fix mtd_block_isbad() so the out-of-bounds checking is done
before the callback check.
Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
Tested-by: Pekon Gupta <pekon@ti.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
This commit makes use of the chip->ecc_strength_ds and chip->ecc_step_ds which
contain the datasheet minimum requested ECC strength to produce a noisy warning
if the configured ECC strength is weaker.
Signed-off-by: Ezequiel Garcia <ezequiel.garcia@free-electrons.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
nand_base can be passed a kmap()'d buffers from highmem by
filesystems like jffs2. This results in failure to map the
physical address of the DMA buffer on various contoller
driver on different platforms. This change adds a chip option
to use preallocated databuf as bounce buffers used in
nand_do_read_ops() and nand_do_write_ops().
This allows for specific nand controller driver to set this
option as needed.
Signed-off-by: Kamal Dasu <kdasu.kdev@gmail.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
The nand_chip::erase_cmd callback previously served a dual purpose; for
one, it allowed a per-flash-chip override, so that AG-AND devices could
use a different erase command than other NAND. These AND devices were
dropped in commit 14c6578683 (mtd: nand:
remove AG-AND support). On the other hand, some drivers (denali and
doc-g4) need to use this sort of callback to implement
controller-specific erase operations.
To make the latter operation easier for some drivers (e.g., ST's new BCH
NAND driver), it helps if the command dispatch and wait functions can be
lumped together, rather than called separately.
This patch does two things:
1. Pull the call to chip->waitfunc() into chip->erase_cmd(), and return
the status from this callback
2. Rename erase_cmd() to just erase(), since this callback does a
little more than just send a command
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Tested-by: Lee Jones <lee.jones@linaro.org>
Now that the index variable is correctly set earlier in this function
we can use it in other places that compute the same thing too.
Signed-off-by: Ron Lee <ron@debian.org>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Commit 7351d3a5db added an index variable
as part of fixing checkpatch warnings, presumably as a tool to make some
long lines shorter, however it only set that index in the case of there
being no gaps in eccpos for the fragment being read. Which means the
later step of filling ecccode from oob_poi will use the wrong indexing
into eccpos in that case.
This patch restores the behaviour that existed prior to that change.
Signed-off-by: Ron Lee <ron@debian.org>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Mention to CONFIG_MTD_ECC_BCH in the warning message can be confusing as this
doesn't match the exact name of the configuration option.
This warning showed up once to me when I was starting to set up BCH. After
checking my .config file, it took a moment before realizing it is
CONFIG_MTD_NAND_ECC_BCH instead of CONFIG_MTD_ECC_BCH.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Acked-by: Lee Jones <lee.jones@linaro.org>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Add the "page" argument for the read_subpage hook. With this argument,
the implementation of this hook could prints out more accurate information
for debugging.
Signed-off-by: Huang Shijie <b32955@freescale.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
The nand_get_flash_type parameter "busw" input value is not used by any
branch, and it is updated before use it in the function, so remove it,
define the "busw" as an internal variable.
Signed-off-by: Cai Zhiyong <caizhiyong@huawei.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
Check the chip->jedec_version, and print out the right information
for JEDEC compliant NAND.
Signed-off-by: Huang Shijie <b32955@freescale.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>