Fwd: Re: [PATCH v2 15/16] clk: Detect failure to set defaults

[Harm Berntsen]

Hi Stefano and Peng,

There is an issue that prevents the imx8mn to boot in 2021.10-rc2. See
the conversation below. Could you help with this?

-- Harm

-------- Forwarded Message --------
From: Simon Glass <sjg at chromium.org>
To: Harm Berntsen <harm.berntsen at nedap.com>
Cc: u-boot at lists.denx.de <u-boot at lists.denx.de>, trini at konsulko.com
<trini at konsulko.com>
Subject: Re: [PATCH v2 15/16] clk: Detect failure to set defaults
Date: Fri, 20 Aug 2021 12:18:07 -0600

> Hi Harm,
> 
> On Wed, 18 Aug 2021 at 08:09, Harm Berntsen <harm.berntsen at nedap.com>
> wrote:
> > 
> > On Thu, 2021-05-13 at 19:39 -0600, Simon Glass wrote:
> > > When the default clocks cannot be set, the clock is silently
> > > probed and
> > > the error is ignored. This is incorrect, since having the clocks
> > > at the
> > > correct speed may be important for operation of the system.
> > > 
> > > Fix it by checking the return code.
> > > 
> > > Signed-off-by: Simon Glass <sjg at chromium.org>
> > > ---
> > > 
> > > (no changes since v1)
> > > 
> > >  drivers/clk/clk-uclass.c | 6 +++++-
> > >  1 file changed, 5 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/drivers/clk/clk-uclass.c b/drivers/clk/clk-uclass.c
> > > index 4ab3c402ed8..2a2e1cfbd61 100644
> > > --- a/drivers/clk/clk-uclass.c
> > > +++ b/drivers/clk/clk-uclass.c
> > > @@ -796,13 +796,17 @@ void devm_clk_put(struct udevice *dev,
> > > struct clk
> > > *clk)
> > > 
> > >  int clk_uclass_post_probe(struct udevice *dev)
> > >  {
> > > +       int ret;
> > > +
> > >         /*
> > >          * when a clock provider is probed. Call
> > > clk_set_defaults()
> > >          * also after the device is probed. This takes care of
> > > cases
> > >          * where the DT is used to setup default parents and
> > > rates
> > >          * using assigned-clocks
> > >          */
> > > -       clk_set_defaults(dev, 1);
> > > +       ret = clk_set_defaults(dev, 1);
> > > +       if (ret)
> > > +               return log_ret(ret);
> > > 
> > >         return 0;
> > >  }
> > 
> > Note that this patch broke booting my imx8mn based board on U-Boot
> > v2021.10-rc2. The failure is due to the clock-controller at 30380000
> > configuration in the imx8mn.dtsi file. I had to remove the
> > following
> > clocks from the device tree to get my device to boot again (all
> > from
> > the assigned-clocks of clock-controller at 30380000):
> > 
> > <&clk IMX8MN_CLK_A53_CORE>,
> > <&clk IMX8MN_CLK_NOC>,
> > <&clk IMX8MN_CLK_AUDIO_AHB>,
> > <&clk IMX8MN_CLK_IPG_AUDIO_ROOT>,
> > <&clk IMX8MN_SYS_PLL3>,
> > <&clk IMX8MN_AUDIO_PLL1>,
> > <&clk IMX8MN_AUDIO_PLL2>;
> > 
> > I looked into the clk-imx8mn.c code and I see that we indeed miss
> > clocks there. Unfortunately I could not port code from the Linux
> > kernel: we are missing the imx_clk_hw_mux2 function for the
> > IMX8MN_CLK_A53_CORE clock. I did not look into the other clocks.
> 
> 
> Perhaps the iMX maintainer could help with this? It does sound like a
> bug.
> 
> Regards,
> SImon
> 
> > 
> > 
> > -- Harm