U-Boot OMAP GPMC ECC change

Roger Quadros rogerq at kernel.org
Fri May 12 13:53:07 CEST 2023



On 10/05/2023 18:38, Colin Foster wrote:
> Hi Roger,
> 
> On Wed, May 10, 2023 at 12:42:43PM +0300, Roger Quadros wrote:
>> Hi Colin,
>>
>> On 09/05/2023 18:31, Colin Foster wrote:
>>> Hi Roger,
>>>
>>> I was looking to test my system against U-Boot 2023.04. I'm running an
>>> OMAP 4460 SOM (I've been waiting to get kernel acceptance before U-Boot,
>>> but that has slipped) and it boots from NAND.
>>>
>>> When I jumped from 2023.01 to 2023.04, I noticed I get spammed in the
>>> SPL by "omap-elm: uncorrectable ECC errors". I bisected, and this seems
>>> to be the result of commit 04fcd25873 ("mtd: rawnand: omap_gpmc: Fix
>>> BCH6/16 HW based correction").
>>
>> oops. Sorry about that.
> 
> No worries!
> 
>>>
>>> Is the multi-sector ECC generation something that should work in a
>>> backward-compatible way? Do you know of anything that might be going
>>> wrong here?
>>
>> The patch wasn't supposed to break anything.
>> I do not yet know what could be wrong. Most likely a wrong ECC
>> configuration is being used.
>>
>> Could you please share what ECC configuration you are using on your board?
> 
> This is still out-of-U-Boot. I have an include/configs/our_product.h
> file with this:
> 
> """
> #define CFG_SYS_NAND_ECCPOS             {2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, \
>                                          13, 14, 15, 16, 17, 18, 19, 20, 21, 22, \
>                                          23, 24, 25, 26, 27, 28, 29, 30, 31, 32, \
>                                          33, 34, 35, 36, 37, 38, 39, 40, 41, \
>                                          42, 43, 44, 45, 46, 47, 48, 49, 50, 51, \
>                                          52, 53, 54, 55, 56, 57}
> 
> #define CFG_SYS_NAND_ECCBYTES   14
> #define CFG_SYS_NAND_MAX_ECCPOS 57

This should be 56 i.e. (57 - 2 + 1)
But it won't fix the issue you are facing. :P

> #define CFG_SYS_NAND_ECCSIZE            512
> #define CFG_SYS_NAND_MAX_OOBFREE        2
> """
> 
> 
>> Can you please point me to the Linux device tree file if it exists?
> 
> This is the latest submission. Still not accepted - I need to find time
> to button everything up and resubmit. My plan of attack was Kernel
> Acceptance, then U-Boot. Unfortunately my company lets the pesky
> "Shipping products" step get in the way :-)
> 
> https://lkml.org/lkml/2023/2/22/939
> 
> Or if you just want the ECC part:
> 
> +	nandflash: nand at 0,0 {
> +		compatible = "ti,omap2-nand";
> +		reg = <0 0 4>;
> +		interrupt-parent = <&gpmc>;
> +
> +		nand-bus-width = <16>;
> +		ti,nand-ecc-opt = "bch8";
> +		ti,elm-id=<&elm>;
> +		linux,mtd-name = "micron,nand";
> 
> 
> I think that's all the info you're looking for. Let me know if I missed
> something.

Yes this is all I was looking for.

Is CONFIG_NAND_OMAP_ECCSCHEME_BCH8_CODE_HW set in your u-boot config?

-- 
cheers,
-roger


More information about the U-Boot mailing list