Hello all,
we use an IMX6ULL with a NAND MLC MT29F32G08CBADBWP as in the schematics in the evaluation kit for the IMX6ULL.
U-Boot 2016.03-nxp/imx_v2016.03_4.1.15_2.0.0_ga+ga57b13b (Dec 15 2020 - 15:09:09 +0000)
Our NAND organization is done as:
- mtd0boot 64MB u-boot
- mtd1kernel16MB kernel
- mtd2dtb 16MB device tree
- mtd3init 64MB contains the initramfs image
- mtd4system with UBIFS stores the rest of the system; uses the rest of the NAND
the problem is that in some device we detect bad block from the first mtd0 to mtd3, if this happen the system not boot. This because the software stop when detect a bad block and is not able to try to fix it and copy this in another block. These mtd partition are always only readed.
This problem was appeared in device that: come from the production after only one nand write, turned on for some days and the leaved in warehouse turned off. Turned on before the shipment they didn't boot.
We have seen that kobs-ng duplicate uboot in two copy but the others mtd partition are unique. We are think to duplicate these too but before i have some questions.
We have read this https://community.nxp.com/pwmxy87654/attachments/pwmxy87654/imx-processors/45804/1/NAND%20Flash%20Ba... how we can do to enable the BBI swap + BBT in uboot and in the system? On the uboot DTS we don't have these lines:
nand-on-flash-bbt;
fsl,no-blockmark-swap;i don't understand fine what they do ... can they help ?
Could this problem come from ESD discharges? We have an MLC NAND... What other factor can produce this problem on new NAND?
