Re: [PATCH v1] mtd: rawnand: meson: fix bitmask for length in command word

[Arseniy Krasnov]

On 22.03.2023 23:10, Martin Blumenstingl wrote:
> Hello Arseniy,
> 
> thank you for submitting this fix!
Thanks!
> 
> On Wed, Mar 22, 2023 at 7:45 PM Arseniy Krasnov
> <avkrasnov@xxxxxxxxxxxxxx> wrote:
>>
>> Valid mask is 0x3FFF, without this patch the following problems were
>> found:
>>
>> 1) [    0.938914] Could not find a valid ONFI parameter page, trying
>>                   bit-wise majority to recover it
>>    [    0.947384] ONFI parameter recovery failed, aborting
>>
>> 2) Read with disabled ECC mode was broken.
>>
>> Fixes: 8fae856c5350 ("mtd: rawnand: meson: add support for Amlogic NAND flash controller")
>> Signed-off-by: Arseniy Krasnov <AVKrasnov@xxxxxxxxxxxxxx>
> This matches what I can see in the old vendor driver, so:
Moreover it was clear that mask of 0x3f is too small for length of data in
bytes, for example for 2048 + OOB size.
> Acked-by: Martin Blumenstingl <martin.blumenstingl@xxxxxxxxxxxxxx>
> 
> [...]
>> -               cmd = (len & GENMASK(5, 0)) | scrambler | DMA_DIR(dir);
>> +               cmd = (len & GENMASK(13, 0)) | scrambler | DMA_DIR(dir);
> My understanding of the vendor driver is that this "len" is only used
> for "raw" access (my own words: any access that doesn't use the HW ECC
> engine).
Exactly, 'len' is only for raw access.
> As a future improvement (no need to update re-send this patch) it
> would be great to have a #define with a meaningful name for
> "GENMASK(13, 0)" (maybe something like NFC_CMD_RAW_LENGTH) as it's
> used in multiple places now
Ack

Thanks, Arseniy
> 
> 
> Best regards,
> Martin