On 8/17/24 19:58, Simon Glass wrote:
Hi Mikhail,
On Wed, 14 Aug 2024 at 04:20, Mikhail Kshevetskiy mikhail.kshevetskiy@iopsys.eu wrote:
only spinand on_die ecc is supported for a moment
Signed-off-by: Mikhail Kshevetskiy mikhail.kshevetskiy@iopsys.eu
drivers/mtd/nand/Makefile | 2 +- drivers/mtd/nand/core.c | 130 +++++++++++++++- drivers/mtd/nand/ecc.c | 249 ++++++++++++++++++++++++++++++ drivers/mtd/nand/spi/core.c | 207 ++++++++++++++++++++----- drivers/mtd/nand/spi/foresee.c | 2 +- drivers/mtd/nand/spi/macronix.c | 7 +- drivers/mtd/nand/spi/micron.c | 2 +- drivers/mtd/nand/spi/toshiba.c | 10 +- drivers/mtd/nand/spi/winbond.c | 10 +- include/linux/mtd/nand.h | 261 ++++++++++++++++++++++++++++++-- include/linux/mtd/spinand.h | 13 +- include/spi-mem.h | 2 + 12 files changed, 830 insertions(+), 65 deletions(-) create mode 100644 drivers/mtd/nand/ecc.c
Somehow this should use driver model (along with a suitable API) for the ECC acceleration.
It's just a necessary part of ECC engine. Other engines (not on_die) should be a separate modules.
Maybe it could be shrink a bit more (by removing traces of other possible engines types). All what I want is: 1) get a working driver with on_die ECC support (it's actually a necessary part of spinand driver) 2) make the code as close to linux driver as possible 3) add the possibility to add other engines in the future.
Regards, Simon