From mboxrd@z Thu Jan  1 00:00:00 1970
From: l.majewski@samsung.com (Lukasz Majewski)
Date: Wed, 31 Oct 2012 15:44:07 +0100
Subject: [PATCH v7 2/5] usb: s3c-hsotg: Adding phy driver support
In-Reply-To: <20121031140548.GE10998@arwen.pp.htv.fi>
References: <1351573057-22249-1-git-send-email-p.paneri@samsung.com>
 <1351573057-22249-3-git-send-email-p.paneri@samsung.com>
 <20121031140548.GE10998@arwen.pp.htv.fi>
Message-ID: <20121031154407.4eb54dba@amdc308.digital.local>
To: linux-arm-kernel@lists.infradead.org
List-Id: linux-arm-kernel.lists.infradead.org

Hi Felipe,

> Hi,
> 
> On Tue, Oct 30, 2012 at 10:27:34AM +0530, Praveen Paneri wrote:
> > @@ -3501,20 +3511,27 @@ static int __devinit s3c_hsotg_probe(struct
> > platform_device *pdev) int ret;
> >  	int i;
> >  
> > -	plat = pdev->dev.platform_data;
> > -	if (!plat) {
> > -		dev_err(&pdev->dev, "no platform data defined\n");
> > -		return -EINVAL;
> > -	}
> > -
> >  	hsotg = devm_kzalloc(&pdev->dev, sizeof(struct s3c_hsotg),
> > GFP_KERNEL); if (!hsotg) {
> >  		dev_err(dev, "cannot get memory\n");
> >  		return -ENOMEM;
> >  	}
> >  
> > +	plat = pdev->dev.platform_data;
> > +	if (!plat) {
> > +		/* Fallback for transceiver */
> > +		phy = devm_usb_get_phy(dev, USB_PHY_TYPE_USB2);
> > +		if (IS_ERR_OR_NULL(phy)) {
> > +			dev_err(&pdev->dev, "no platform data or
> > transceiver defined\n");
> > +			return -EPROBE_DEFER;
> > +		} else {
> > +			hsotg->phy = phy;
> > +		}
> > +	} else {
> > +		hsotg->plat = plat;
> > +	}
> 
> I think this should be the other way around, meaning you try to grab
> the phy, if you can't, then you fallback to pdata.
> 

I agree.

The new approach is to use new PHY driver. And only when failed we
shall use legacy approach.

-- 
Best regards,

Lukasz Majewski

Samsung Poland R&D Center | Linux Platform Group