Hi, On Wed, Jul 20, 2011 at 11:21:10AM +0000, Lin Tony-B19295 wrote: > > -----Original Message----- > > From: Felipe Balbi [mailto:balbi@xxxxxx] > > Sent: Wednesday, July 20, 2011 7:15 PM > > To: Lin Tony-B19295 > > Cc: linux-usb@xxxxxxxxxxxxxxx; linux-arm-kernel@xxxxxxxxxxxxxxxxxxx; > > balbi@xxxxxx; koen.beel.barco@xxxxxxxxx > > Subject: Re: [PATCH 5/7] ehci mxc: make it more flexible to be used for > > mx28 > > > > Hi, > > > > On Wed, Jul 20, 2011 at 07:08:24PM +0800, Tony Lin wrote: > > > @@ -165,14 +187,15 @@ static int ehci_mxc_drv_probe(struct > > platform_device *pdev) > > > } > > > > > > /* enable clocks */ > > > - priv->usbclk = clk_get(dev, "usb"); > > > - if (IS_ERR(priv->usbclk)) { > > > - ret = PTR_ERR(priv->usbclk); > > > - goto err_clk; > > > + if (!cpu_is_mx28()) { > > > > this should not be used in drivers, IMHO. > > > > > + priv->usbclk = clk_get(dev, "usb"); > > > > drivers should not have to care about clock names, are you sure your > > clkdev support is correct ? > > Yes, I agree with you about above two points. I feel the same as you. > But I had to following the existing framework. I didn't add any new special to the driver. > Just try some ways to workaround these ugly points. but in that case, since you're already there... why not cleaning those things up before making your changes ? -- balbi
Attachment:
signature.asc
Description: Digital signature
