AM625: AM62x SPI NAND Flash Bad Block Management

stephen zhang said:

We tested two winbond SPI-Nand Flash, W25N02 (QSPI) and W35N02 (OSPI), on customer boards, and both encountered bad block problems.There was no problem with reading and writing at the beginning, but after a few days of testing, many bad blocks appeared.The read and write commands are as follows

Are you saying you executed those raw commands in an endless loop??? How often?

Usually raw flash like OSPI/QSPI NAND/NOR  is used with some kind of wear-leveling driver (like the UBI subsystem, used by UBIFS), rather than "beating on it" in a raw fashion, which may lead to premature wear.

Regards, Andreas

Hi Andreas,
The test is not in a loop. The purpose of our operation is to test the Flash boot mode, write uboot image and ubifs to the Flash and then test the startup function. If there is a problem with startup after writing, we will reconfigure uboot or ubifs and then write it again. The number of writes is not very large, not exceed 10 times/day.

Regards,
Stephen

stephen zhang said:

The number of writes is not very large, not exceed 10 times/day.

Ok good, you should be able to do such raw flash operations 100s of times without issues. You just can't do that continuously. The Flash datasheet would have the details on min. erase/write cycles for a given block, but this is likely not your issue.

Perhaps there's some other SW problem at play here, accidentally marking the blocks as bad. You can reset the bad block information in U-Boot by executing the `nand scrub` command to see if this fixes things. You will need to turn on CONFIG_CMD_NAND in U-Boot to get access to that command.

https://git.ti.com/cgit/ti-u-boot/ti-u-boot/tree/doc/README.nand?h=ti-u-boot-2023.04#n215

"scrub [offset length]"
like "erase" but don't skip bad block. Instead erase them.
DANGEROUS!!! Factory set bad blocks will be lost. Use only
to remove artificial bad blocks created with the "markbad" command.

If this fixes things we still need to figure out what is going wrong in the first place.

Regards, Andreas

Andreas Dannenberg said:

You will need to turn on CONFIG_CMD_NAND in U-Boot to get access to that command.

Does CMD_NAND only can be used for parallel NAND flash, but not SPI NAND Flash?

I tried enable nand cmd in uboot before but cannot operate the device.

enabled CONFIG_CMD_NAND and CONFIG_MTD_RAW_NAND, and the error encountered during linking is as follows.

  CC      drivers/mtd/nand/raw/nand.o
  CC      drivers/mtd/nand/raw/nand_bbt.o
  CC      drivers/mtd/nand/raw/nand_ids.o
  CC      drivers/mtd/nand/raw/nand_util.o
  CC      drivers/mtd/nand/raw/nand_ecc.o
  CC      drivers/mtd/nand/raw/nand_base.o
  CC      drivers/mtd/nand/raw/nand_amd.o
  CC      drivers/mtd/nand/raw/nand_hynix.o
  CC      drivers/mtd/nand/raw/nand_macronix.o
  CC      drivers/mtd/nand/raw/nand_micron.o
  CC      drivers/mtd/nand/raw/nand_samsung.o
  CC      drivers/mtd/nand/raw/nand_toshiba.o
  CC      drivers/mtd/nand/raw/nand_timings.o
  AR      drivers/mtd/nand/raw/built-in.o
  CC      drivers/mtd/nand/spi/core.o
  CC      drivers/mtd/nand/spi/gigadevice.o
  CC      drivers/mtd/nand/spi/macronix.o
  CC      drivers/mtd/nand/spi/micron.o
  CC      drivers/mtd/nand/spi/toshiba.o
  CC      drivers/mtd/nand/spi/winbond.o
  AR      drivers/mtd/nand/spi/spinand.o
  AR      drivers/mtd/nand/spi/built-in.o
  CC      drivers/mtd/nand/core.o
  CC      drivers/mtd/nand/bbt.o
  AR      drivers/mtd/nand/nandcore.o
  AR      drivers/mtd/nand/built-in.o
  AR      drivers/mtd/onenand/built-in.o
  CC      drivers/mtd/spi/sf-uclass.o
  CC      drivers/mtd/spi/sf_probe.o
  CC      drivers/mtd/spi/spi-nor-ids.o
  CC      drivers/mtd/spi/spi-nor-core.o
  AR      drivers/mtd/spi/spi-nor.o
  AR      drivers/mtd/spi/built-in.o
  LDS     u-boot.lds
  LD      u-boot
aarch64-none-linux-gnu-ld.bfd: drivers/mtd/nand/raw/nand.o: in function `nand_init_chip':
/home/z98999/am62x/ti-u-boot/drivers/mtd/nand/raw/nand.c:92: undefined reference to `board_nand_init'
/bin/sh: line 1: 23223 Segmentation fault      (core dumped) aarch64-none-linux-gnu-ld.bfd -z noexecstack -pie --gc-sections -Bstatic --no-dynamic-linker -z notext --build-id=none -Ttext 0x80800000 -o u-boot -T u-boot.lds arch/arm/cpu/armv8/start.o --whole-archive arch/arm/cpu/built-in.o arch/arm/cpu/armv8/built-in.o arch/arm/lib/built-in.o arch/arm/mach-k3/built-in.o board/ti/am62x/built-in.o board/ti/common/built-in.o boot/built-in.o cmd/built-in.o common/built-in.o disk/built-in.o drivers/built-in.o drivers/usb/cdns3/built-in.o drivers/usb/common/built-in.o drivers/usb/dwc3/built-in.o drivers/usb/emul/built-in.o drivers/usb/eth/built-in.o drivers/usb/gadget/built-in.o drivers/usb/gadget/udc/built-in.o drivers/usb/host/built-in.o drivers/usb/isp1760/built-in.o drivers/usb/mtu3/built-in.o drivers/usb/musb-new/built-in.o drivers/usb/musb/built-in.o drivers/usb/phy/built-in.o drivers/usb/ulpi/built-in.o env/built-in.o fs/built-in.o lib/built-in.o net/built-in.o --no-whole-archive -L /home/z98999/cross_compile/arm-gnu-toolchain-11.3.rel1-x86_64-aarch64-none-linux-gnu/bin/../lib/gcc/aarch64-none-linux-gnu/11.3.1 -lgcc -Map u-boot.map
/home/z98999/am62x/ti-u-boot/Makefile:1754: recipe for target 'u-boot' failed
make[1]: *** [u-boot] Error 139
make[1]: *** Deleting file 'u-boot'
make[1]: Leaving directory '/home/z98999/am62x/ti-u-boot/out/a53'
Makefile:177: recipe for target 'sub-make' failed
make: *** [sub-make] Error 2

It seems that the Raw Nand driver should be enabled. There is no Winbond driver in Raw NAND Device Support, so I enabled CONFIG_NAND_OMAP_GPMC and some related configurations, the compilation passed, but the device could not be found when using the nand command. 

=> nand info

=> nand bad

no devices available
=> nand scrub

no devices available

I am using SDK9.0, uboot version 2023.04.

Regards,
Stephen

Is there a way to clear bad blocks in Linux? I didn't find a command similar to scrub. mtd-utils can read and write flash, but it can't erase bad blocks.

Hi Stephen,

I would need to research this some more, please allow a couple of days for a response. Thanks for your patience.

Regards, Andreas

Hi, Andreas：

    Is there any progress so far?

Unfortunately I have not been able to look at this yet due to other priorities.