Re: [PATCH v7 0/5] Add driver for EC found on Qualcomm reference devices

[Dmitry Baryshkov]

On Mon, Mar 30, 2026 at 07:08:45AM -0700, Rob Clark wrote:
> On Sun, Mar 29, 2026 at 11:57 PM Anvesh Jain P
> <anvesh.p@xxxxxxxxxxxxxxxx> wrote:
> >
> >
> >
> > On 3/29/2026 7:52 PM, Akhil P Oommen wrote:
> > > On 3/27/2026 3:38 PM, Anvesh Jain P wrote:
> > >> From: Anvesh Jain P <anvesh.p@xxxxxxxxxxxxxxxx>
> > >>
> > >> Add Embedded controller driver support for Hamoa/Purwa/Glymur Qualcomm
> > >> reference boards. It handles fan control, temperature sensors, access
> > >> to EC state changes and supports reporting suspend entry/exit to the EC.
> > >
> > > Just a question, is it possible to force fan to run at full speed
> > > constantly? That would be helpful to keep the passive cooling minimal
> > > and get a more consistent results while doing performance profiling.
> > >
> > > I see that you are registering a cooling device in this series. So I
> > > suppose we just need to update the cooling sysfs nodes.
> > >
> > > -Akhil
> > >
> >
> > Hi Akhil,
> >
> > Yes, that is possible. The cooling device registered by this driver
> > exposes the standard thermal sysfs interface. You can force the fan
> > to full speed by writing the maximum cooling state directly:
> >
> >   # Find the max state
> >   cat /sys/class/thermal/cooling_deviceX/max_state
> >
> >   # Set to max state (force full speed)
> >   echo <max_state> > /sys/class/thermal/cooling_deviceX/cur_state
> >
> > Where X is the index of the cooling device corresponding to the EC fan.
> 
> Getting a bit off topic, but is there a good way to associated cooling
> devices and thermal sensors to the GPU device?  I could see that being
> useful for nvtop (which currently supports shows GPU load and memory
> metrics, but thermal metrics are missing for us)

I had a dirty patch, providing temp readings in NVMem. Probably I need
to clean it up and submit upstream.

> 
> BR,
> -R
> 
> > --
> > Best Regards,
> > Anvesh
> >
> >

-- 
With best wishes
Dmitry