Re: [PATCH v2] hwmon: Add driver for MSI Mystic Light controller voltage sensors

[Guenter Roeck]

On 11/7/26 12:33, Tomasz Bojanowski wrote:
> Hi Guenter,
>
> On 11/7/26 11:40, Guenter Roeck wrote:
> > Please let me know if the code still works for you.
>
> It does. With commit 64e4888, the MS-7E61 reads the same as with v2
> (CPU 1P8 1.79 V, CHIP Core 1.04 V, VDD MISC 1.10 V).
>
> > Offset 11 does not seem to show a valid voltage. On the other side,
> > offset 15 would reflect a value that at least makes sense.
>
> The vendor software reads CPU VDDP from offset 11 on the MS-7C94 and
> does not use offset 15. It would show the same 0.002 V, so I cannot
> tell what offset 15 is.

That makes me wonder if the vendor software is buggy. I have
two of those boards, and they both show similar values. If we can't
figure that out, I am inclined to drop support for that board:
It is not worth the trouble with people complaining that it does
not work.

> One small thing: if the read fails, msi_ml_read_report() returns
> without hid_hw_power(..., PM_HINT_NORMAL).

Good point. I wonder what the best way to handle this might be:
Push it back into the device specific read function, or always
call it from the common read function. Or maybe only keep it for
the 7e61 board and assume that it isn't needed for the older boards
since they are expected to send raw events. Any thoughts ?

Either case, do you have evidence that this is even needed in the
first place ? The call is rarely used elsewhere in the kernel,
and none of the other HID hwmon drivers uses it.

Thanks,
Guenter