Re: [PATCH 4/5] PCI/pwrctrl: Add APIs to power on/off the pwrctrl devices

[Bartosz Golaszewski]

On Tue, Nov 25, 2025 at 2:50 PM Manivannan Sadhasivam <mani@xxxxxxxxxx> wrote:
>
> >
> > This is not an error though, is it? If there are multiple deferalls, we'll
> > spam the kernel log.
> >
>
> Good question. Initially, I made it as a debug log, but then realized that
> people may wonder why their controller driver encounters probe deferral without
> much clue, especially when the driver spits out other logs before calling this
> API. So decided to make it dev_err() to give a visual indication.
>
> If it is not preferred, I can demote it to debug log.

If we must log it, I'd say a dev_dbg() is enough. Probe deferral is not unusual.

It would be awesome to be able to synchronize the controller probe
with the pwrctl device binding but I don't have an idea on how to do
it yet. :(

Bart