Re: [PATCH] ARM: imx25: provide a fixed regulator for usb phys

["Uwe Kleine-König" <u.kleine-koenig@pengutronix.de>]

[-- Attachment #1: Type: text/plain, Size: 3178 bytes --]

On Fri, Nov 15, 2019 at 10:07:57AM +0100, Uwe Kleine-König wrote:
> Hello,
> 
> On Mon, Sep 23, 2019 at 12:28:25PM +0200, Uwe Kleine-König wrote:
> > On Wed, Jul 24, 2019 at 03:09:39PM +0200, Uwe Kleine-König wrote:
> > > On Thu, Jun 27, 2019 at 03:15:10AM +0000, Peter Chen wrote:
> > > >  
> > > > > On 19-06-26 02:40, Peter Chen wrote:
> > > > > >
> > > > > > > Subject: [PATCH] ARM: imx25: provide a fixed regulator for usb phys
> > > > > > >
> > > > > > > The usb phys are internal to the SoC and so it their 5V supply. With
> > > > > > > this regulator added explicitly the following (harmless) boot messages go away:
> > > > > > >
> > > > > > > 	usb_phy_generic usbphy:usb-phy@0: usbphy:usb-phy@0 supply vcc not found, using dummy regulator
> > > > > > > 	usb_phy_generic usbphy:usb-phy@1: usbphy:usb-phy@1 supply vcc not found, using dummy regulator
> > > > > > >
> > > > > >
> > > > > > To eliminate the warning message, I suggest doing below changes, as
> > > > > > vcc supply is not mandatory.
> > > > > >
> > > > > > diff --git a/drivers/usb/phy/phy-generic.c
> > > > > > b/drivers/usb/phy/phy-generic.c index a53b89be5324..01a5ff1a0515
> > > > > > 100644
> > > > > > --- a/drivers/usb/phy/phy-generic.c
> > > > > > +++ b/drivers/usb/phy/phy-generic.c
> > > > > > @@ -275,7 +275,7 @@ int usb_phy_gen_create_phy(struct device *dev, struct usb_phy_generic *nop,
> > > > > >                 }
> > > > > >         }
> > > > > >
> > > > > > -       nop->vcc = devm_regulator_get(dev, "vcc");
> > > > > > +       nop->vcc = devm_regulator_get_optional(dev, "vcc");
> > > > > 
> > > > > Is the regulator optional? IMHO this shouldn't be the fix. I think the right fix is Uwe's
> > > > > approach.
> > > > > 
> > > > 
> > > > Add Felipe.
> > > > 
> > > > Some USB PHY's power are from the core system's power (eg, DDR), and some are
> > > > fixed at the board and no switch for it. So, it is transparent for software at some cases.
> > > 
> > > It's not clear to me how to proceed. There are two opposing opinions and
> > > I don't know enough about USB on mx25 to judge myself.
> > > 
> > > Felipe?
> > 
> > This thread is still open in my inbox. Felipe, how can I lure you into
> > giving your opinion?
> > 
> > My original suggestion can be seen at
> > https://lore.kernel.org/linux-usb/20190625100412.11815-1-u.kleine-koenig@pengutronix.de/,
> > Peter's alternative is still in the quotes above. Which is the
> > right/better one?
> 
> The stuff said above is still true. I'd like to put this issue on file
> with "fixed in mainline", but currently this seems to be a stalemate.
> because nobody cares enough :-|

This topic is on stall since two years now and I still have the patch in
the patch stack for my mx25 boards.

In my eyes my patch is the right one, because no matter how, the usb
phys need some powering, so devm_regulator_get_optional() doesn't seem
to be the right approach.

Would resending the patch help?

Best regards
Uwe

-- 
Pengutronix e.K.                           | Uwe Kleine-König            |
Industrial Linux Solutions                 | https://www.pengutronix.de/ |

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]