[PATCH 4/4] clk: scmi: Add workaround for set_rate/enable/disable

[Peng Fan]

> Subject: Re: [PATCH 4/4] clk: scmi: Add workaround for
> set_rate/enable/disable
> 
> On Sun, Oct 06, 2024 at 05:38:25PM +0800, alice.guo at oss.nxp.com
> wrote:
> > From: Ye Li <ye.li at nxp.com>
> >
> > Add workaround to set_rate/enable/disable to bus clock that SM will
> > reply DENIED error.
> >
> > Signed-off-by: Ye Li <ye.li at nxp.com>
> > Signed-off-by: Alice Guo <alice.guo at nxp.com>
> > Reviewed-by: Peng Fan <peng.fan at nxp.com>
> 
> In general, please include a cover letter so it will be clearer what the
> overall goal is, especially once merged.
> 
> > ---
> >  drivers/clk/clk_scmi.c | 21 +++++++++++++++++++--
> >  1 file changed, 19 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/clk/clk_scmi.c b/drivers/clk/clk_scmi.c index
> > a01292c479..a860a653ba 100644
> > --- a/drivers/clk/clk_scmi.c
> > +++ b/drivers/clk/clk_scmi.c
> > @@ -73,7 +73,13 @@ static int scmi_clk_gate(struct clk *clk, int
> enable)
> >  	if (ret)
> >  		return ret;
> >
> > -	return scmi_to_linux_errno(out.status);
> > +	ret = scmi_to_linux_errno(out.status);
> > +	if (ret == -EACCES) {
> > +		debug("Ignore %s enable failure\n",
> clk_hw_get_name(clk));
> > +		ret = 0;
> > +	}
> > +
> > +	return ret;
> >  }
> 
> This seems like a generic change being made globally and not a work-
> around for a specific problem on (some?) iMX families.

We have changed Linux upstream to this behavior, but better
in firmware/clock.c as linux,
https://elixir.bootlin.com/linux/v6.11.2/source/drivers/firmware/arm_scmi/clock.c#L761

Regards,
Peng.

 Has this been
> tested on other platforms?
> 
> --
> Tom