Re: [PATCH] platform/x86: bitland-mifs-wmi: Add Xiaomi Book Pro 14 platform profile values

[Ilpo Järvinen]

FYI,

There are a few more of these in the queue:

https://patchwork.kernel.org/project/platform-driver-x86/patch/20260826132924.179614-1-kento@xxxxxxxx/
https://patchwork.kernel.org/project/platform-driver-x86/patch/fb4e0678-3028-463e-8a8c-1b680e40c4ee@xxxxxxxxx/

The first one is a bit messy, as it adds some keys as well (which should 
have been in a separate patch).

And some more in Mingyou's redmi-wmi merge series's last patch:

https://patchwork.kernel.org/project/platform-driver-x86/patch/20260816100813.300450-7-qby140326@xxxxxxxxx/

The profile mode abstraction in that series seems not general enough 
though so the generalization work we've discussed is still necessary and 
should be used as a base for that series.

--
 i.

On Mon, 28 Sep 2026, Matias Civadda wrote:

> Hi Chris,
> 
> No problem, your proposal works for me. Please CC me when you send
> the series so I can test the shared changes and base my TM2107 v3
> on your generic patch.
> 
> Thanks,
> Matias
> 
> > Thanks, makes sense.
> > 
> > Matias, one thing I ran into when trying your generic part on the
> > Xiaomi Book Pro 14: its values are sparse (0x02, 0x03, 0x04, 0x09,
> > 0x0a) and 0x02 and 0x0a both mean low-power. With a profile array
> > indexed by the firmware value, the unused slots read back as
> > PLATFORM_PROFILE_LOW_POWER (0), and the reverse lookup for low-power
> > finds index 0, which this firmware ignores.
> > 
> > So I'd suggest {value, profile} pairs for the shared part. I have that
> > as a standalone patch on pdx86/for-next, with only the profile table
> > in the quirk struct, so your response layout and AC check handling
> > would go on top in your TM2107 patch. If that works for you I'll send
> > it as v2 together with the Xiaomi entry, and you can base your v3 on
> > it. If you'd rather own the generic patch, that's fine too, I'll
> > rebase the Xiaomi part on yours.
> > 
> 

-- 
 i.