- 11 Mar, 2011 40 commits
-
-
Ivan Djelic authored
This is a new software BCH encoding/decoding library, similar to the shared Reed-Solomon library. Binary BCH (Bose-Chaudhuri-Hocquenghem) codes are widely used to correct errors in NAND flash devices requiring more than 1-bit ecc correction; they are generally better suited for NAND flash than RS codes because NAND bit errors do not occur in bursts. Latest SLC NAND devices typically require at least 4-bit ecc protection per 512 bytes block. This library provides software encoding/decoding, but may also be used with ASIC/SoC hardware BCH engines to perform error correction. It is being currently used for this purpose on an OMAP3630 board (4bit/8bit HW BCH). It has also been used to decode raw dumps of NAND devices with on-die BCH ecc engines (e.g. Micron 4bit ecc SLC devices). Latest NAND devices (including SLC) can exhibit high error rates (typically a dozen or more bitflips per hour during stress tests); in order to minimize the performance impact of error correction, this library implements recently developed algorithms for fast polynomial root finding (see bch.c header for details) instead of the traditional exhaustive Chien root search; a few performance figures are provided below: Platform: arm926ejs @ 468 MHz, 32 KiB icache, 16 KiB dcache BCH ecc : 4-bit per 512 bytes Encoding average throughput: 250 Mbits/s Error correction time (compared with Chien search): average worst average (Chien) worst (Chien) ---------------------------------------------------------- 1 bit 8.5 µs 11 µs 200 µs 383 µs 2 bit 9.7 µs 12.5 µs 477 µs 728 µs 3 bit 18.1 µs 20.6 µs 758 µs 1010 µs 4 bit 19.5 µs 23 µs 1028 µs 1280 µs In the above figures, "worst" is meant in terms of error pattern, not in terms of cache miss / page faults effects (not taken into account here). The library has been extensively tested on the following platforms: x86, x86_64, arm926ejs, omap3630, qemu-ppc64, qemu-mips. Signed-off-by:Ivan Djelic <ivan.djelic@parrot.com> Signed-off-by:
David Woodhouse <David.Woodhouse@intel.com>
-
Baruch Siach authored
When page size is 4KiB, ecc.total is set to 8*9, and this causes nand_write_page_hwecc() to read past the initialized part of the eccpos array, which corrupts chip->oob_poi with bogus values from ecc_calc. Fix this by creating a proper nand_ecclayout for 4KiB flashes. Signed-off-by:
Baruch Siach <baruch@tkos.co.il> Acked-by:
Sascha Hauer <s.hauer@pengutronix.de> Signed-off-by:
-
Sandeep Paulraj authored
The DaVinci NAND driver was including the mach-types.h header file. This prevented the DaVinci NAND driver from getting used in a DSP only device. The linux port on c6x devices can make use of the same driver and does not define a corresponding header file. This header file was required in the driver because earlier there was code dependent on a machine_* check. This piece of code has now been factored out and is in arch/arm/mach-davinci/aemif.c Thus removing the header file is harmless Signed-off-by:
Sandeep Paulraj <s-paulraj@ti.com> Signed-off-by:
Artem Bityutskiy <Artem.Bityutskiy@nokia.com> Signed-off-by:
-
Lei Wen authored
Use nand_scan_ident to unify the need of mtd member initilization for both normal detection and keep configuration method. Signed-off-by:
Lei Wen <leiwen@marvell.com> Signed-off-by:
Haojian Zhuang <haojian.zhuang@marvell.com> Signed-off-by:
-
Lei Wen authored
Different NAND driver may require its unique detection. For pxa3xx_nand, it use its self id database to get the necessary info. Signed-off-by:
-
Lei Wen authored
Make the interface simpler which could make both debug and enhancement easier. Signed-off-by:
-
Lei Wen authored
Since we have rework the irq process, we don't need additional delay in wait_for_event. Also write_cmd and __readid is also discarded for the same reason. Signed-off-by:
Eric Miao <eric.y.miao@gmail.com> Signed-off-by:
-
Lei Wen authored
Enable all irq when we start the nand controller, and put all the transaction logic in the pxa3xx_nand_irq. By doing this way, we could dramatically increase the performance by avoid unnecessary delay. Signed-off-by:
-
Lei Wen authored
The previous probe function is some kind of big part. This patch seperate the resource allocation to keep the probe process more clear than before. Signed-off-by:
-
David Lambert authored
32-bit integers used in 'calc_speed()' may overflow and lead to incorrect results. Use 64-bit integers instead. Signed-off-by:
David Lambert <dave@lambsys.com> Signed-off-by:
-
Baruch Siach authored
Signed-off-by:
-
John Ogness authored
The number of corrected ECC errors should be reported since other MTD systems make use of this information (such as UBI data scrubbing). Signed-off-by:
John Ogness <john.ogness@linutronix.de> Signed-off-by:
-
Tracey Dent authored
Signed-off-by:
Tracey Dent <tdent48227@gmail.com> Signed-off-by:
-
Jason Liu authored
Signed-off-by:
Jason Liu <r64343@freescale.com> Signed-off-by:
-
Roman Tereshonkov authored
New variable skip_initial_unlocking is added to the omap_onenand_platform_data. This is used to inform the onenand driver to skip onenand unlocking when it is initialized. Signed-off-by:
Roman Tereshonkov <roman.tereshonkov@nokia.com> Signed-off-by:
-
Roman Tereshonkov authored
A new option ONENAND_SKIP_INITIAL_UNLOCKING is added. This allows to disable initial onenand unlocking when the driver is initialized. Signed-off-by:
-
David Griego authored
OMAP-L137/AM17x has limited number of dedicated EMIFA address pins, enough to interface directly to an SDRAM. If a device such as an asynchronous flash needs to be attached to the EMIFA, then either GPIO pins or a chip select may be used to control the flash device's upper address lines. This patch adds support for the NOR flash on the OMAP-L137/ AM17x user interface daughter board using the latch-addr-flash MTD mapping driver which allows flashes to be partially physically addressed. The upper address lines are set by a board specific code which is a separate patch. Signed-off-by:
David Griego <dgriego@mvista.com> Signed-off-by:
Aleksey Makarov <amakarov@ru.mvista.com> Signed-off-by:
Sergei Shtylyov <sshtylyov@ru.mvista.com> Signed-off-by:
Savinay Dharmappa <savinay.dharmappa@ti.com> Signed-off-by:
-
Jarkko Lavinen authored
Add a driver for allowing an mtd device to be used as a block device for swapping. The block device is volatile, and the mapping of swapped pages is not stored on flash. Signed-off-by:
Jarkko Lavinen <jarkko.lavinen@nokia.com> Tested-by:
-
Jarkko Lavinen authored
Set max_discard_sectors to UINT_MAX. Signed-off-by:
-
Jarkko Lavinen authored
Add a new background method into mtd_blktrans_ops, add background support into mtd_blktrans_thread(), and add mtd_blktrans_cease_background(). If the mtd blktrans dev has the background support, the thread will call background function when the request queue becomes empty. The background operation may run as long as needs to until mtd_blktrans_cease_background() tells to stop. Signed-off-by:
-
Anatolij Gustschin authored
When initializing mtd->writebufsize, we must take into account possible flash chip interleaving. Wrong writebufsize initialization caused UBIFS recovery issues resulting in unmountable UBIFS file system on NOR flash partitions. Signed-off-by:
Anatolij Gustschin <agust@denx.de> Acked-by:
Guillaume LECERF <glecerf@gmail.com> Signed-off-by:
-
Yoshihiro Shimoda authored
Signed-off-by:
Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Signed-off-by:
-
Roman Tereshonkov authored
New multiblock erase speed test is added to mtd_speedtest. It consists of 2-, 4-, 8-, 16-, 32- and 64-blocks at once multiblock erase tests. Signed-off-by:
-
Adrian Hunter authored
By default mtd_speedtest uses all the eraseblocks of the MTD partition being tested. For large partitions a smaller number is sufficient and makes running the test quicker. For that reason, add a parameter 'count' to specify the maximum number of eraseblocks to use for testing. Signed-off-by:
Adrian Hunter <adrian.hunter@nokia.com> Signed-off-by:
-
Roman Tereshonkov authored
In onenand_verify function the address can be writepage non-aligned. When a page is read for comparing the right offset should be used for "this->verify_buf" to get the right matching with compared "buf" buffer. Signed-off-by:
Kyungmin Park <kyungmin.park@samsung.com> Signed-off-by:
-
Adrian Hunter authored
When reading using the 4KiB page read function, I/O errors could be ignored if more than 1 page was read at a time. Signed-off-by:
-
Stanislav Fomichev authored
In 'verify_bbt_descr()', first check the "bd" pointer, then dereference it. Comments amended by Artem. Signed-off-by:
Stanislav Fomichev <kernel@fomichev.me> Signed-off-by:
-
Roman Tereshonkov authored
The ebcnt and pgcnt variable initialization is moved before printk which uses them. Signed-off-by:
-
Julia Lawall authored
map_destroy dereferences its argument. The call is furthermore only reachable when this argument is NULL. Thus the call is dropped. The semantic match that finds this problem is as follows: (http://coccinelle.lip6.fr/) // <smpl> @@ expression *x; @@ *if (x == NULL) { ... * map_destroy(x); ... return ...; } // </smpl> Signed-off-by:
Julia Lawall <julia@diku.dk> Signed-off-by:
-
Mathias Krause authored
Commit 4f678a58 (mtd: fix memory leaks in phram_setup) missed two cases where the memory allocated for name would be leaked. This commit frees the memory when register_device() fails and on unregister_devices(). Signed-off-by:
Mathias Krause <minipli@googlemail.com> Signed-off-by:
-
Artem Bityutskiy authored
Set the 'mtd->writebufsize' field to 64 to mimic modern CFI flashes. Signed-off-by:
-
Gabor Juhos authored
Signed-off-by:
Gabor Juhos <juhosg@openwrt.org> Signed-off-by:
-
Holger Brunck authored
If different chips with different writebufsize are concatenated, the writebufsize from the concat device has to be taken from the device with the largest writebuffer. This writebufsize is used later on in the UBI layer for the min I/O size. Signed-off-by:
Holger Brunck <holger.brunck@keymile.com> Signed-off-by:
-
Hong Xu authored
Some SAM9 chips have the ability to perform DMA between CPU and SMC controller. This patch adds DMA support for SAM9RL, SAM9G45, SSAM9G46,AM9M10, SAM9M11. Signed-off-by:
Hong Xu <hong.xu@atmel.com> Tested-by:
Ryan Mallon <ryan@bluewatersys.com> Acked-by:
-
H Hartley Sweeten authored
The function blktrans_dev_release and blktrans_dev_put are only used locally in this file and should be static. Signed-off-by:
H Hartley Sweeten <hsweeten@visionengravers.com> Signed-off-by:
-
H Hartley Sweeten authored
The three backing_dev_info symbols are only used in this file and should be static. Signed-off-by:
-
Jamie Iles authored
clk_get() returns a struct clk cookie to the driver and some platforms may return NULL if they only support a single clock. clk_get() has only failed if it returns a ERR_PTR() encoded pointer. Signed-off-by:
Jamie Iles <jamie@jamieiles.com> Reviewed-by:
Wolfram Sang <w.sang@pengutronix.de> Signed-off-by:
-
Dmitry Eremin-Solenikov authored
As there are no more dependencies on MTD_CONCAT or CONFIG_MTD_CONCAT, drop this entry from Kconfig entirely. Signed-off-by:
Dmitry Eremin-Solenikov <dbaryshkov@gmail.com> Acked-by:
Stefan Roese <sr@denx.de> Signed-off-by:
-
Dmitry Eremin-Solenikov authored
As MTD_CONCAT support is becoming an integral part of MTD core, there is no need for it's special treatment. So stop checking for MTD_CONCAT availability. Acked by Jesper Nilsson <jesper.nilsson@axis.com> for merging this via the MTD tree. Signed-off-by:
Jesper Nilsson <jesper.nilsson@axis.com> Signed-off-by:
-
Dmitry Eremin-Solenikov authored
As MTD_CONCAT is becoming a part of mtd core, it's now meaningless to to check for it in ifdefs. Drop such references from MTD code. Signed-off-by:
-
