Hello everyone,

I have a Gumstix Overo (OMAP3) COM here that I am evaluating BCH8 ECC scheme on boot-loader (2014.10) and the kernel (3.17.7). Traditionally the board has been configured with 1 bit HAM, as are other OMAP3 boards.

I have these changes in my board config:

+#define CONFIG_BCH +#define CONFIG_SYS_NAND_MAX_ECCPOS 56 +#define CONFIG_SYS_NAND_ECCPOS {2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, \ + 13, 14, 16, 17, 18, 19, 20, 21, 22, \ + 23, 24, 25, 26, 27, 28, 30, 31, 32, \ + 33, 34, 35, 36, 37, 38, 39, 40, 41, \ + 42, 44, 45, 46, 47, 48, 49, 50, 51, \ + 52, 53, 54, 55, 56} +#define CONFIG_SYS_NAND_ECCBYTES 13 +#define CONFIG_NAND_OMAP_ECCSCHEME OMAP_ECC_BCH8_CODE_HW_DETECTION_SW +#define CONFIG_SPL_MAX_SIZE (58 * 1024)

And other than MLO, I make sure 'nandecc sw' before I do any NAND write/read.

Now the first problem occurs after MLO executes and before U-Boot is loaded. I get a whole bunch of "ecc unrecoverable error" from `nand_spl_load_image` calls. Weird thing is that U-Boot image eventually gets loaded after a couple hundred lines of these error messages.

This problem however does not occur when I write the U-Boot image from the Kernel using flash_erase and nandwrite tools (my DTS has ti,nand-ecc-opt = "bch8"; ".

Any pointers would be greatly appreciated,

Adam