Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   "Hyunki Koo" <hyunki00.koo@samsung.com>
To:     "'Krzysztof Kozlowski'" <krzk@kernel.org>,
        "'Greg Kroah-Hartman'" <gregkh@linuxfoundation.org>
Cc:     <hyunki00.koo@gmail.com>, "'Kukjin      Kim'" <kgene@kernel.org>,
        "'Jiri Slaby'" <jslaby@suse.com>,
        <linux-arm-kernel@lists.infradead.org>,
        <linux-samsung-soc@vger.kernel.org>,
        <linux-serial@vger.kernel.org>, <linux-kernel@vger.kernel.org>
In-Reply-To: <20200401091952.GA14425@kozik-lap>
Subject: RE: [PATCH] tty: samsung_tty: 32-bit access for TX/RX hold
 registers
Date:   Thu, 2 Apr 2020 18:44:58 +0900
Message-ID: <000001d608d3$5faa8260$1eff8720$@samsung.com>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Thread-Index: AQIlrsPG99gFy1AqjplcrZ0ldcCfeQKEs4Y1AmydwKICM8VWpaeNHlJA
Content-Language: ko
Content-Type: text/plain; charset="utf-8"
CMS-TYPE: 102P
DLP-Filter: Pass
References: <CGME20200401082749epcas2p2a774da515805bc3f761b6b5a8dc9e3d2@epcas2p2.samsung.com>
        <20200401082721.19431-1-hyunki00.koo@samsung.com>
        <20200401085548.GC2026666@kroah.com> <20200401091952.GA14425@kozik-lap>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Wed, Apr 01, 2020 at 6:20:20PM +0900, Krzysztof Kozlowski
wrote:
> On Wed, Apr 01, 2020 at 10:55:48AM +0200, Greg Kroah-Hartman
> wrote:
> > On Wed, Apr 01, 2020 at 05:27:20PM +0900, Hyunki Koo wrote:
> > > -	if (np)
> > > +	if (np) =7B
> > >  		of_property_read_u32(np,
> > >  			=22samsung,uart-fifosize=22, &ourport->port.fifosize);
> > >
> > > +		if (of_property_read_u32(np, =22reg-io-width=22, &prop) =3D=3D
> 0) =7B
> > > +			switch (prop) =7B
> > > +			case 1:
> > > +				ourport->port.iotype =3D UPIO_MEM;
> > > +				break;
> > > +			case 4:
> > > +				ourport->port.iotype =3D UPIO_MEM32;
> > > +				break;
> > > +			default:
> > > +				dev_warn(&pdev->dev, =22unsupported
> reg-io-width (%d)=5Cn=22,
> > > +						prop);
> > > +				ret =3D -EINVAL;
> > > +				break;
> > > +			=7D
> > > +		=7D
> > > +	=7D
> > > +
> >
> > Does this mean that reg-io-width is now a required property for all
> > samsung uarts?  Does this break older dts files?  Or should you
> > fall-back to the previous operation if the attribute is not there?
>=20
> Yes, it looks like silently breaking all boards.  Since
> of_property_read_u32() will return errno, the warning message won't be
> printed and all register reads will fail (return 0).
>=20
> This looks like not tested on real HW.
>=20
> Best regards,
> Krzysztof

=5BHyunki Koo=5D=20
reg-io-width =3D4 is required for Samsung uart
To do not break older dts files, I will set default value in else of of_pro=
perty_read_u32 like below.
+		if (of_property_read_u32(np, =22reg-io-width=22, &prop) =3D=3D 0) =7B
+ ...
+		=7D else =7B
+			ourport->port.iotype =3D UPIO_MEM;
+		=7D