Add a SPL test for the NAND load method. We use some different functions to
do the writing from the main test since things like nand_write_skip_bad
aren't available in SPL.
We disable BBT scanning, since scan_bbt is only populated when not in SPL.
We use nand_spl_loaders.c as it seems to be common to at least a few boards
already. However, we do not use nand_spl_simple.c because it would require
us to implement cmd_ctrl. The various nand load functions are adapted from
omap_gpmc. However, they have been modified for simplicity/correctness.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Add a sandbox NAND flash driver to facilitate testing. This driver supports
any number of devices, each using a single chip-select. The OOB data is
stored in-band, with the separation enforced through the API.
For now, create two devices to test with. The first is a very small device
with basic ECC. The second is an 8G device (chosen to be larger than 32
bits). It uses ONFI, with the values copied from the datasheet. It also
doesn't need too strong ECC, which speeds things up.
Although the nand subsystem determines the parameters of a chip based on
the ID, the driver itself requires devicetree properties for each
parameter. We do not derive parameters from the ID because parsing the ID
is non-trivial. We do not just use the parameters that the nand subsystem
has calculated since that is something we should be testing. An exception
is made for the ECC layout, since that is difficult to encode in the device
tree and is not a property of the device itself.
Despite using file I/O to access the backing data, we do not support using
external files. In my experience, these are unnecessary for testing since
tests can generally be written to write their expected data beforehand.
Additionally, we would need to store the "programmed" information somewhere
(complicating the format and the programming process) or try to detect
whether block are erased at runtime (degrading probe speeds).
Information about whether each page has been programmed is stored in an
in-memory buffer. To simplify the implementation, we only support a single
program per erase. While this is accurate for many larger flashes, some
smaller flashes (512 byte) support multiple programs and/or subpage
programs. Support for this could be added later as I believe some
filesystems expect this.
To test ECC, we support error-injection. Surprisingly, only ECC bytes in
the OOB area are protected, even though all bytes are equally susceptible
to error. Because of this, we take care to only corrupt ECC bytes.
Similarly, because ECC covers "steps" and not the whole page, we must take
care to corrupt data in the same way.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
NAND devices are destroyed in between unit tests. Provide a function to
reinitialize the subsystem at the beginning of each test.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Reviewed-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
This performs the opposite of nand_register, allowing drivers to unregister
nand devices. This is probably unnecessary for most regular drivers, but we
expect sandbox drivers to get repeatedly bound/unbound, so this will help
avoid dangling pointers.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Reviewed-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
This allows using these functions without ifdefs. OneNAND depends on MTD,
so this ifdef was redundant in the first place.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Reviewed-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Rename SPL_MTD_SUPPORT to SPL_MTD in order to match MTD. This allows using
CONFIG_IS_ENABLED to test for MTD support.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Since commit 34793598c8 ("mtd: nand: mxs_nand_spl: Remove the page aligned
access") there are no longer any users of nand_get_mtd. However, it is
still important to know what the page size is so we can allocate a
large-enough buffer. If the image size is not page-aligned, we will go off
the end of the buffer and clobber some memory.
Introduce a new function nand_page_size which returns the page size. For
most drivers it is easy to determine the page size. However, a few need to
be modified since they only keep the page size around temporarily.
It's possible that this patch could cause a regression on some platforms if
the offset is non-aligned and there is invalid address space immediately
before the load address. spl_load_legacy_img does not (except when
compressing) respect bl_len, so only boards with SPL_LOAD_FIT (8 boards) or
SPL_LOAD_IMX_CONTAINER (none in tree) would be affected.
defconfig CONFIG_TEXT_BASE
======================= ================
am335x_evm 0x80800000
am43xx_evm 0x80800000
am43xx_evm_rtconly 0x80800000
am43xx_evm_usbhost_boot 0x80800000
am43xx_hs_evm 0x80800000
dra7xx_evm 0x80800000
gwventana_nand 0x17800000
imx8mn_bsh_smm_s2 0x40200000
All the sitara boards have DDR mapped at 0x80000000. gwventana is an i.MX6Q
which has DDR at 0x10000000. I don't have the IMX8MNRM handy, but on the
i.MX8M DDR starts at 0x40000000. Therefore all of these boards can handle a
little underflow.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
All other implementations of nand_spl_load_image only read as many pages as
are necessary to load the image. However, nand_spl_loaders.c loads the full
block. Align it with other load functions so that it is easier to
determine how large of a load buffer we need.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
Contrary to what the help message says, this is the number of pages per
block. Calculate it automatically based on SYS_NAND_BLOCK_SIZE and
SYS_NAND_PAGE_SIZE. To better reflect its semantics, rename it to
SYS_NAND_BLOCK_PAGES.
Signed-off-by: Sean Anderson <seanga2@gmail.com>
When no manufacturer is matched, manufacturer_desc is NULL. Avoid
dereferencing it in that case.
Fixes: 4e67c57125 ("mtd,ubi,ubifs: sync with linux v3.15")
Signed-off-by: Sean Anderson <seanga2@gmail.com>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
Compiling sandbox_defconfig with CONFIG_BOOTSTD=n fails:
/usr/bin/ld: drivers/mtd/spi/sf_bootdev.o:
in function `sf_get_bootflow':
/drivers/mtd/spi/sf_bootdev.c:43:(.text+0x96):
undefined reference to `bootmeth_set_bootflow'
Add the missing Kconfig dependency.
Fixes: Fixes: 0c1f4a9fb1 ("bootstd: Add a SPI flash bootdev")
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Similar change was done by commit b4c2c151b1 ("Kconfig: Remove all
default n/no options") and again sync is required.
default n/no doesn't need to be specified. It is default option anyway.
Signed-off-by: Michal Simek <michal.simek@amd.com>
Reviewed-by: Svyatoslav Ryhel <clamor95@gmail.com> # tegra
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Angelo Dureghello <angelo@kernel-space.org>
On Rockchip SoCs the first boot stages are written on NAND
with help of manufacturer software that uses a different format
then the MTD framework. Skip the automatic BBT scan with the
NAND_SKIP_BBTSCAN option to be able to pass the driver probe
function and to let the original data unchanged.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Add support for XMC XM25QU128C (128M-bit) Serial Flash memory. Used on
the Xunlong Orange Pi 3B, 5 and 5 Plus boards.
Datasheet:
https://www.xmcwh.com/uploads/806/XM25QU128C_Ver2.0.pdf
Signed-off-by: Ricardo Pardini <ricardo@pardini.net>
[jonas@kwiboo.se: update commit message]
Signed-off-by: Jonas Karlman <jonas@kwiboo.se>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
This adds more supported spinand devices from the Linux kernel
implementation.
This does not include the latest kernel implementation as this would
require a substantial amount of extra work due to the missing
ECC engine abstraction layer in U-Boot.
Signed-off-by: Mikhail Kshevetskiy <mikhail.kshevetskiy@iopsys.eu>
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de> (commit message)
Link: https://lore.kernel.org/all/20230110115843.391630-3-frieder@fris.de
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
This brings us closer to the current Linux kernel implementation of
the spinand core and makes backporting features and fixes easier.
This does not include the latest kernel implementation as this would
require a substantial amount of extra work due to the missing
ECC engine abstraction layer in U-Boot.
Signed-off-by: Mikhail Kshevetskiy <mikhail.kshevetskiy@iopsys.eu>
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de> (add commit message)
Link: https://lore.kernel.org/all/20230110115843.391630-2-frieder@fris.de
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Currently there are 3 different variants of read_id implementation:
1. opcode only. Found in GD5FxGQ4xF.
2. opcode + 1 addr byte. Found in GD5GxGQ4xA/E
3. opcode + 1 dummy byte. Found in other currently supported chips.
Original implementation was for variant 1 and let detect function
of chips with variant 2 and 3 to ignore the first byte. This isn't
robust:
1. For chips of variant 2, if SPI master doesn't keep MOSI low
during read, chip will get a random id offset, and the entire id
buffer will shift by that offset, causing detect failure.
2. For chips of variant 1, if it happens to get a devid that equals
to manufacture id of variant 2 or 3 chips, it'll get incorrectly
detected.
This patch reworks detect procedure to address problems above. New
logic do detection for all variants separatedly, in 1-2-3 order.
Since all current detect methods do exactly the same id matching
procedure, unify them into core.c and remove detect method from
manufacture_ops.
This is a rework of Chuanhong Guo <gch981213@gmail.com> patch
submitted to linux kernel
Signed-off-by: Mikhail Kshevetskiy <mikhail.kshevetskiy@iopsys.eu>
Signed-off-by: Frieder Schrempf <frieder.schrempf@kontron.de>
Link: https://lore.kernel.org/all/20230110115843.391630-1-frieder@fris.de
Signed-off-by: Dario Binacchi <dario.binacchi@amarulasolutions.com>
Adapt behaviour to Linux kernel driver.
The return value of gpio_request_by_name_nodev() was not checked before,
and thus in case 'rb-gpios' was missing in DT, rb.type was set to
ATMEL_NAND_GPIO_RB nevertheless, leading to output like this for
example (on sam9x60-curiosity with the line removed from dts):
NAND: Could not find valid ONFI parameter page; aborting
device found, Manufacturer ID: 0xc2, Chip ID: 0xdc
Macronix NAND 512MiB 3,3V 8-bit
512 MiB, SLC, erase size: 256 KiB, page size: 4096, OOB size: 64
atmel-nand-controller nand-controller: NAND scan failed: -22
Failed to probe nand driver (err = -22)
Failed to initialize NAND controller. (error -22)
0 MiB
Note: not having that gpio assigned in dts is possible, the driver does
not override nand_chip->dev_ready() then and a generic solution is used.
Fixes: 6a8dfd5722 ("nand: atmel: Add DM based NAND driver")
Signed-off-by: Alexander Dahl <ada@thorsis.com>
Reviewed-by: Eugen Hristev <eugen.hristev@collabora.com>
Acked-by: Michael Trimarchi <michael@amarulasolutions.com>
This old patch was marked as deferred. Bring it back to life, to continue
towards the removal of common.h
Move this out of the common header and include it only where needed.
Signed-off-by: Simon Glass <sjg@chromium.org>
The last user of the NEEDS_MANUAL_RELOC has been removed in commit
26af162ac8 ("arch: m68k: Implement relocation")
Remove now unused NEEDS_MANUAL_RELOC code.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
The sandbox_set_enable_memio() should only ever be set during
sandbox testing, not within driver itself, move it back to test/ .
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Cast the address such that it can be printed without warnings
on both 32bit and 64bit systems. This really should use some
better print formatter, but for the lack of it, do it this way.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Get rid of nvmxip_mmio_rawread() and just implement the readl()/readq()
reader loop within nvmxip_blk_read(). Cast the destination buffer as
needed and increment the read by either 4 or 8 bytes depending on if
this is systemd with 32bit or 64bit physical address.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Perform all the block device creation only once, after the driver itself
successfully bound. Do not do this in uclass post bind, as this might be
triggered multiple times. For example the ut_dm_host test triggers this
and triggers a memory leak that way, since there are now multiple block
devices created using the blk_create_devicef() .
To retain the old probe-on-boot behavior, set DM_FLAG_PROBE_AFTER_BIND
flag in uclass post_bind callback, so the driver model would probe the
driver at the right time.
Rename the function as well, to match similar functions in
other block-related subsystems, like the mmc one.
Signed-off-by: Marek Vasut <marek.vasut+renesas@mailbox.org>
Rockchip boot blocks are written per 4 x 512 byte sectors per page.
Each page must have a page address (PA) pointer in OOB to the next page.
Pages are written in a pattern depending on the NAND chip ID.
This logic used to build a page pattern table is not fully disclosed and
is not easy to fit in the MTD framework.
The formula in rk_nfc_write_page_hwecc() function is not correct.
Make hwecc and raw behavior identical.
Generate boot block page address and pattern for hwecc in user space
and copy PA data to/from the already reserved last 4 bytes before EEC
in the chip->oob_poi data layout.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
status and ecc_status are of unsigned type where they are compared for
negative value. This is pointed by below sparse warning. Change datatype
to int to fix this.
warning: comparison of unsigned expression in '< 0' is always false
[-Wtype-limits]
Signed-off-by: Algapally Santosh Sagar <santoshsagar.algapally@amd.com>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Ashok Reddy Soma <ashok.reddy.soma@amd.com>
Link: https://lore.kernel.org/r/20230614090359.10809-5-ashok.reddy.soma@amd.com
Signed-off-by: Michal Simek <michal.simek@amd.com>
Add support for Winbond 256M-bit flash w25q256jwm.
Performed basic erase/write/readback operations on
ZynqMP zc1751+dc1 board.
Signed-off-by: Venkatesh Yadav Abbarapu <venkatesh.abbarapu@amd.com>
Reviewed-by: Jagan Teki <jagan@amarulasolutions.com>
XTX changed full company name from "XTX Technology (Shenzhen) Limited
to "XTX Technology Limited" since 2020,So remove "(Shenzhen)".
Signed-off-by: Bruce Suen <bruce_suen@163.com>
Reviewed-by: Jagan Teki <jagan@amarulasolutions.com>
The CN9130 SoC (an ARMADA 8K type) has both a NAND Flash Controller and
a generic local bus controller (Device Bus Controller) that share common
pins.
With a board design that incorporates both a NAND flash and uses
the Device Bus (in our case for an SRAM) accessing the Device Bus device
fails unless the NfArbiterEn bit is set. Setting the bit enables
arbitration between the Device Bus and the NAND flash.
Since there is no obvious downside in enabling this for designs that
don't require arbitration, we always enable it.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>
The NAND flash controller (NFC) on the AC5/AC5X SoC is the same as
the NFC used on other Marvell SoCs. It does have the additional
restriction of only supporting SDR timing modes up to 3.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>
Move header to include to allow external code
to get the internal bdev structures to access
block device operations.
as at it, just add the UCLASS_NVMXIP string
so we get the correct output in partitions
listing.
Signed-off-by: Rui Miguel Silva <rui.silva@linaro.org>
flash_get_size() will get the flash size from the device itself and go
through all erase regions to read protection status. However, the device
mappable region (eg: devicetree reg property) might be lower than the
device full size which means that the above cycle will result in a data
bus exception. This change fixes it by reading the 'addr_size' during
probe() and also use that as one possible upper limit.
Signed-off-by: Nuno Sá <nuno.sa@analog.com>
The 'tiny' SPI nor functions have the same name as their big brothers,
which can be confusing. Use different names so it is clear which
version is in the image.
Signed-off-by: Simon Glass <sjg@chromium.org>
Acked-by: Vignesh Raghavendra <vigneshr@ti.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
With tiny SPI flash the erasesize is 0 which can cause a divide-by-zero
error. Check for this and return a proper error instead.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The fdt_addr_t and phys_addr_t size have been decoupled. A 32bit CPU
can expect 64-bit data from the device tree parser, so use
dev_read_addr_ptr instead of the dev_read_addr function in the
various files in the drivers directory that cast to a pointer.
As we are there also streamline the error response to -EINVAL on return.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The fdt_addr_t and phys_addr_t size have been decoupled. A 32bit CPU
can expect 64-bit data from the device tree parser, so use
dev_read_addr_index_ptr instead of the dev_read_addr_index function
in the various files in the drivers directory that cast to a pointer.
As we are there also streamline the error response to -EINVAL on return.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Sandisk SDTNQGAMA is a 8GB size, 3.3V 8 bit chip with 16KB page size,
1KB write size and 40 bit ecc support
Signed-off-by: Paweł Jarosz <paweljarosz3691@gmail.com>
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Acked-by: Michael Trimarchi <michael@amarulasolutions.com>
The MTD framework reserves 1 or 2 bytes for the bad block marker
depending on the bus size. The rockchip_nfc driver currently only
supports a 8 bit bus, but reserves standard 2 bytes for the BBM.
The first free OOB byte is therefore OOB2 at offset 2.
Page address(PA) bytes are moved to the last 4 positions before
ECC. Update the description for U-boot.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Add flash_node to the rockchip_nfc driver chip structure in order
to find the partitions in the add_mtd_partitions_of() function.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
The MTD framework in U-boot is not identical for drivers ported
from Linux. The rockchip_nfc driver was ported with OOB ops functions
while the framework expects a layout structure per chip.
Fix by adding a structure with OOB data and remove unused functions.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
The compatible string for rk3308 has as fallback string
"rockchip,rv1108-nfc". As there is no logic in probe priority between
the SoC orientated string and the fall back, so remove the compatible
string "rockchip,rk3308-nfc" from the driver.
Signed-off-by: Johan Jonker <jbx6244@gmail.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Michael Trimarchi <michael@amarulasolutions.com>
