Hi Alan, On Sat, May 10, 2014 at 10:35:49AM -0400, Alan Stern wrote: > > @@ -206,6 +208,19 @@ static int ehci_platform_probe(struct platform_device > > *dev) > > break; > > } > > } > > + > > + priv->rst = devm_reset_control_get_optional(&dev->dev, > > + NULL); > > I hate the style that matches arguments on a continuation line with the > opening paren of the function call, for a couple of reasons. Instead I > simply indent continuation lines by two or more tab stops. But some > people seem to be incurably attached to it.
Ok, I'll change it.
> > + if (IS_ERR(priv->rst)) {
> > + err = PTR_ERR(priv->rst);
> > + if (err == -EPROBE_DEFER)
> > + goto err_put_clks;
> > + priv->rst = NULL;
> > + } else {
> > + err = reset_control_deassert(priv->rst);
> > + if (err)
> > + goto err_put_clks;
> > + }
> > }
> >
> > if (pdata->big_endian_desc)
>
> The new code was added inside an "if" statement, which will cause it to
> apply only to OF devices. Is there any reason not to put the new code
> outside the "if" statement, so it applies to all devices?
Hmmm, I did this because so far, the reset framework only handles the
DT case. But that's right that it can be extended in the future, I'll
update it.
Thanks!
Maxime
--
Maxime Ripard, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com
signature.asc
Description: Digital signature
