[PATCH] phy: renesas: phy-rzg3e-usb3: Fix runtime PM underflow during suspend

[Biju Das]

Hi Ovidiu,

> -----Original Message-----
> From: Ovidiu Panait <ovidiu.panait.rb at renesas.com>
> Sent: 28 April 2026 10:21
> Subject: RE: [PATCH] phy: renesas: phy-rzg3e-usb3: Fix runtime PM underflow during suspend
> 
> Hi Biju,
> 
> >
> > Hi Ovidiu,
> >
> > Thanks for the patch
> >
> > > -----Original Message-----
> > > From: Ovidiu Panait <ovidiu.panait.rb at renesas.com>
> > > Sent: 27 April 2026 20:48
> > > Subject: [PATCH] phy: renesas: phy-rzg3e-usb3: Fix runtime PM
> > > underflow
> > during suspend
> > >
> > > On the Renesas RZ/V2H platform, if the xhcd driver is unbound and
> > > the
> > system is suspended afterwards, a
> > > PM underflow error will occur:
> > >
> > >  # echo 15850000.usb >
> > > /sys/bus/platform/drivers/xhci-renesas-hcd/unbind
> > >  # systemctl suspend
> > >  15870000.usb-phy: PM: dpm_run_callback(): genpd_resume_noirq
> > > returns -
> > 13
> > >  15870000.usb-phy: PM: failed to resume noirq: error -13
> > >  15870000.usb-phy: Runtime PM usage count underflow!
> > >
> > > Since the PHY framework is managing the runtime PM of the PHY via
> > phy_power_on()/phy_power_off(), there
> > > is no need for the PHY driver to manipulate the runtime PM state
> > > during
> > suspend.
> > >
> > > To fix this, remove the runtime PM calls from the suspend/resume
> > > paths
> > and add a get/put pair inside
> > > rzg3e_phy_usb3_init_helper() to make sure the clock is enabled
> > > during
> > init, even when there is no
> > > consumer for the PHY.
> > >
> > > Also, change the suspend ops from NOIRQ_SYSTEM_SLEEP_PM_OPS to
> > SYSTEM_SLEEP_PM_OPS because runtime PM
> > > is disabled during the noirq phase and pm_runtime_resume_and_get()
> > > would
> > not actually enable the device
> > > clock.
> > >
> > > Fixes: ee5f1a3f90a4 ("phy: renesas: Add Renesas RZ/G3E USB3.0 PHY
> > driver")
> > > Signed-off-by: Ovidiu Panait <ovidiu.panait.rb at renesas.com>
> > > ---
> > >  drivers/phy/renesas/phy-rzg3e-usb3.c | 31
> > > ++++++++++++++++------------
> > >  1 file changed, 18 insertions(+), 13 deletions(-)
> > >
> > > diff --git a/drivers/phy/renesas/phy-rzg3e-usb3.c
> > b/drivers/phy/renesas/phy-rzg3e-usb3.c
> > > index 6b3453ea0004..055775e1a0f7 100644
> > > --- a/drivers/phy/renesas/phy-rzg3e-usb3.c
> > > +++ b/drivers/phy/renesas/phy-rzg3e-usb3.c
> > > @@ -64,6 +64,7 @@
> > >  #define USB3_TEST_LANECONFIG0_DEFAULT		(0xd)
> > >
> > >  struct rz_usb3 {
> > > +	struct device *dev;
> >
> > Do you need dev as struct phy has dev member [1]?
> >
> > [1]
> > https://elixir.bootlin.com/linux/v6.0-
> > rc4/source/include/linux/phy/phy.h#L153
> >
> > >  	void __iomem *base;
> > >  	struct reset_control *rstc;
> > >  	bool skip_reinit;
> > > @@ -130,11 +131,21 @@ static int rzg3e_phy_usb3test_phy_init(void
> > __iomem *base)
> > >  	return 0;
> > >  }
> > >
> > > -static int rzg3e_phy_usb3_init_helper(void __iomem *base)
> >
> > you can add one more function parameter dev here.
> >
> > static int rzg3e_phy_usb3_init_helper(struct device *dev, void __iomem
> > *base)
> >
> > > +static int rzg3e_phy_usb3_init_helper(struct rz_usb3 *r)
> > >  {
> > > -	rzg3e_phy_usb2test_phy_init(base);
> > > +	int ret;
> > > +
> > > +	ret = pm_runtime_resume_and_get(r->dev);
> > > +	if (ret)
> > > +		return ret;
> > > +
> > > +	rzg3e_phy_usb2test_phy_init(r->base);
> > >
> > > -	return rzg3e_phy_usb3test_phy_init(base);
> > > +	ret = rzg3e_phy_usb3test_phy_init(r->base);
> > > +
> > > +	pm_runtime_put_sync(r->dev);
> > > +
> > > +	return ret;
> > >  }
> > >
> > >  static int rzg3e_phy_usb3_init(struct phy *p) @@ -143,7 +154,7 @@
> > static int
> > > rzg3e_phy_usb3_init(struct phy *p)
> > >  	int ret = 0;
> > >
> > >  	if (!r->skip_reinit)
> > > -		ret = rzg3e_phy_usb3_init_helper(r->base);
> > > +		ret = rzg3e_phy_usb3_init_helper(r);
> >
> > ret = rzg3e_phy_usb3_init_helper(&p->dev, r->base);
> >
> > >
> > >  	return ret;
> > >  }
> > > @@ -187,6 +198,7 @@ static int rzg3e_phy_usb3_probe(struct
> > platform_device *pdev)
> > >
> > >  	platform_set_drvdata(pdev, r);
> >
> > >  	phy_set_drvdata(phy, r);
> > > +	r->dev = dev;
> >
> > Drop this.
> >
> > >
> > >  	provider = devm_of_phy_provider_register(dev, of_phy_simple_xlate);
> > >  	if (IS_ERR(provider))
> > > @@ -199,7 +211,6 @@ static int rzg3e_phy_usb3_suspend(struct device
> > *dev)  {
> > >  	struct rz_usb3 *r = dev_get_drvdata(dev);
> > >
> > > -	pm_runtime_put(dev);
> > >  	reset_control_assert(r->rstc);
> > >  	r->skip_reinit = false;
> > >
> > > @@ -215,27 +226,21 @@ static int rzg3e_phy_usb3_resume(struct device
> > *dev)
> > >  	if (ret)
> > >  		return ret;
> > >
> > > -	ret = pm_runtime_resume_and_get(dev);
> > > +	ret = rzg3e_phy_usb3_init_helper(r);
> >
> > 	ret = rzg3e_phy_usb3_init_helper(dev, r->base);
> 
> "dev" is the platform device, which is different from phy->dev. Phy->dev is a child of the platform
> device.

You are correct. I missed that.

> 
> We could be referencing the platform device from phy->dev with &phy->dev.parent, but I don't think that
> will make the code cleaner.

Thought of avoiding another global variable that saves 4 bytes.

Changing "&phy->dev"->""&phy->dev.parent" is not making the code cleaner
then leave as it is.

Cheers,
Biju