Hi Adam,

Adam Ford [off-list ref] wrote on Tue, 19 Jan 2021 09:49:36 -0600:

On Tue, Jan 19, 2021 at 8:36 AM Miquel Raynal [off-list ref] wrote:

quoted

Hi Adam,
 

quoted

quoted

---8<---

Author: Miquel Raynal [off-list ref]
Date:   Tue Jan 19 12:27:07 2021 +0100

    wip: fix omap

    Signed-off-by: Miquel Raynal [off-list ref]
 

Thanks for fixing this.

I tested your patch, and I no longer get a Panic and the MTD device
appears to appear correctly:

mtdoops: mtd device (mtddev=name/number) must be supplied
omap2-nand 30000000.nand: GPIO lookup for consumer rb
omap2-nand 30000000.nand: using device tree for GPIO lookup
of_get_named_gpiod_flags: parsed 'rb-gpios' property of node
'/ocp@68000000/gpmc@6e000000/nand@0,0[0]' - status (0)
gpio gpiochip6: Persistence not supported for GPIO 0
nand: device found, Manufacturer ID: 0x2c, Chip ID: 0xbc
nand: Micron MT29F4G16ABBDA3W
nand: 512 MiB, SLC, erase size: 128 KiB, page size: 2048, OOB size: 64
nand: using OMAP_ECC_BCH8_CODE_HW_DETECTION_SW
6 cmdlinepart partitions found on MTD device omap2-nand.0
Creating 6 MTD partitions on "omap2-nand.0":
...  

Good to know. Can you just tell me if the values of
- chip->ecc.bytes vs. engine_conf->code_size
- chip->ecc.steps vs. engine_conf->nsteps
are the same in both cases (your patch and mine)? Otherwise your data
might appear corrupted somehow.  

I didn't fully vet my hack, beyond eliminating the Kernel panic, so I
felt more comfortable comparing the values to a stable release.
I compared the values in 5.10 to 5.11-rc4 + your patch, and the number
of steps and oobregion->length are identical between them.

5.10.5:
  chip->ecc.steps = 4
  oobregion->length = d

5.11-rc4 + patch:
  engine_conf->nsteps = 4
  oobregion->length = d

Great, thanks for checking.

I'll send the formal patch then.

Thanks,
Miquèl