Re: [PATCH v5 04/28] mtd: spi-nor: swp: Improve locking user experience

[Miquel Raynal <miquel.raynal@bootlin.com>]

>> Technically speaking all four first patches are fixes, except I don't
>> ask the first one to be backported. The reason why we ask fixes to be
>> first in the series is because we want them to be as independent as
>> possible from previous cleanups/changes. Here each four first patch are
>> targeting completely different places and should not interact with each
>> other. Anyway, I will re-shuffle the patches.
>
> you don't need to resend just for that I think. Pratyush or Michael can
> re-shuffle when applying.

It's a bit painful to do while applying, I will send a v6.

>> As for Sashiko's feedback, the AI raises the same point as our previous
>> discussion: the QE bit handling is really bad, and I am working on
>
> I forgot what we talked about, sorry.

The fact that there is a helper in the core that blindly sets a
particular QE bit, but this helper might (theoretically) be used by
chips which have the QE bit somewhere else.

I plan on reworking this by storing the location of the QE bit during
SFDP parsing. Then I will use that location in a helper to set or clear
it, which will also be dependent on the QER bitfield. But that field is
a real mess, so I'm not sure it's gonna work :-)

Cheers,
Miquèl