[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20190408100947.GI9224@smile.fi.intel.com>
Date: Mon, 8 Apr 2019 13:09:47 +0300
From: Andy Shevchenko <andriy.shevchenko@...el.com>
To: Mika Westerberg <mika.westerberg@...ux.intel.com>
Cc: Chris Chiu <chiu@...lessm.com>, heikki.krogerus@...ux.intel.com,
drake@...lessm.com, linus.walleij@...aro.org,
linux-gpio@...r.kernel.org, linux-kernel@...r.kernel.org,
linux@...lessm.com
Subject: Re: [PATCH] pinctrl:intel: Retain HOSTSW_OWN for requested gpio pin
On Mon, Apr 08, 2019 at 10:41:06AM +0300, Mika Westerberg wrote:
> On Mon, Apr 08, 2019 at 12:21:03PM +0800, Chris Chiu wrote:
> > +static void
static u32
(see below why)
> > +intel_gpio_update_pad_mode(void __iomem *hostown, u32 mask, u32 value)
> > +{
> > + u32 curr = readl(hostown);
> > + u32 updated = (curr & ~mask) | (value & mask);
>
> I think here we should first complain if the expected ownership is not
> correct. Warning or info level probably enough.
It is easy to achieve, something like
if ((value ^ saved) & requeted)
dev_warn(...);
See also below.
(I had mentioned this earlier)
> > + return writel(updated, hostown);
writel() is a void function, this is wrong. We need to return curr instead to
make above working to issue a warning message.
> Also if the pin is not requested and not changed we should not touch the
> register.
I don't think it brings any value here, if mask is 0 we will write back the
same value we read.
> Otherwise this looks good to me.
--
With Best Regards,
Andy Shevchenko
Powered by blists - more mailing lists