Re: [PATCH v4 3/3] mtd: spi-nor: Add support for IS25LX256 operating in 1S-8S-8S octal read mode

[Tudor Ambarus <tudor.ambarus@linaro.org>]

On 27.12.2022 14:37, Michael Walle wrote:
> Am 2022-12-26 09:17, schrieb Tudor Ambarus:
>> On 26.12.2022 10:04, Tudor Ambarus wrote:
>>> Hi, Nathan,
>>>
>>> The series is starting to look good, but I'll need another version,
>>> please.
>>>
>>> On 02.12.2022 15:55, Nathan Barrett-Morrison wrote:
>>>> This adds the IS25LX256 chip into the ISSI flash_info parts table
>>>
>>> Describe your changes in imperative mood, e.g. "Add support for
>>> S25LX256" instead of "This adds ..."
>>>
>>> It may worth to re-read
>>> https://www.kernel.org/doc/html/latest/process/submitting-patches.html
>>> once in a while.
>>>
>>>>
>>>> Signed-off-by: Nathan Barrett-Morrison <nathan.morrison@timesys.com>
>>>> ---
>>>>   drivers/mtd/spi-nor/issi.c | 32 ++++++++++++++++++++++++++++++++
>>>>   1 file changed, 32 insertions(+)
>>>>
>>>> diff --git a/drivers/mtd/spi-nor/issi.c b/drivers/mtd/spi-nor/issi.c
>>>> index 89a66a19d754..362bc3603d8f 100644
>>>> --- a/drivers/mtd/spi-nor/issi.c
>>>> +++ b/drivers/mtd/spi-nor/issi.c
>>>> @@ -29,6 +29,35 @@ static const struct spi_nor_fixups 
>>>> is25lp256_fixups = {
>>>>       .post_bfpt = is25lp256_post_bfpt_fixups,
>>>>   };
>>>> +static int
>>>> +is25lx256_post_bfpt_fixups(struct spi_nor *nor,
>>>> +               const struct sfdp_parameter_header *bfpt_header,
>>>> +               const struct sfdp_bfpt *bfpt)
>>>> +{
>>>> +    /*
>>>> +     * IS25LX256 supports both 1S-1S-8S and 1S-8S-8S.
>>>> +     * However, the BFPT does not contain any information denoting 
>>>> this
>>>> +     * functionality, so the proper fast read opcodes are never setup.
>>>> +     * We're correcting this issue via the fixup below.  Page program
>>>> +     * commands are detected and setup properly via the 4BAIT lookup.
>>>> +     */
>>
>> Why don't you set the READ support when parsing the 4bait table?
> 
> That would deviate from the read handling of all the other modes.

Oh, I see now, so you need to amend BFPT with the 3B opcodes so that the
4bait table convert the 3B read opcodes to 4B. That's fine, yes.

> 
>> We need to see the SFDP dump to determine how we treat this. I'm not
>> sure a post_bfpt hook is the right thing to do for this flash.
> 
> See:
> https://lore.kernel.org/linux-mtd/e07105d41ec62a6ee47ca0295ca347dc@walle.cc/

Okay, it looks fine. I still need the write, erase, verify erase, write,
read and compare test.

> 
> But yes, I missed that you should include the sfdp dump and some other
> commands in the comment section of the patch which adds support for that
> flash (see Tudors reply).
> 
Nathan, please resubmit addressing those small comments if you can.

> Tudor, what is the status of your documentation patch? I'd really like
> to refer to the kernel docs instead of having write the same over and
> over again ;)

I hope to have it ready for 6.3.

Cheers,
ta