From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: [PATCH] ARM: dts: OMAP3-N900: Fix offset for smc91x ethernet Date: Tue, 24 Feb 2015 10:02:28 -0800 Message-ID: <20150224180228.GC11056@atomide.com> References: <1424364590-14603-1-git-send-email-pali.rohar@gmail.com> <7411731.0J5JRAOyMm@wuerfel> <201502201052.47515@pali> <20150220152856.GH32521@atomide.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Content-Disposition: inline In-Reply-To: <20150220152856.GH32521@atomide.com> Sender: linux-kernel-owner@vger.kernel.org To: Pali =?utf-8?B?Um9ow6Fy?= Cc: Arnd Bergmann , =?utf-8?Q?Beno=C3=AEt?= Cousson , Rob Herring , linux-omap@vger.kernel.org, devicetree@vger.kernel.org, linux-kernel@vger.kernel.org List-Id: devicetree@vger.kernel.org * Tony Lindgren [150220 07:47]: > * Pali Roh=C3=A1r [150220 01:56]: > > On Friday 20 February 2015 10:24:35 Arnd Bergmann wrote: > > > On Thursday 19 February 2015 17:49:50 Pali Roh=C3=A1r wrote: > > > > Offset for smc91x must be zero otherwise smc91x linux kernel > > > > driver does not detect smc91x ethernet hardware in qemu > > > > N900 machine. > > > >=20 > > > > Signed-off-by: Pali Roh=C3=A1r > > >=20 > > > Is that the same offset on real hardware, or could this be a > > > mistake in the qemu model? > > >=20 > > > Arnd > >=20 > > In original Nokia 2.6.28 kernel is offset also set to 0, see: > >=20 > > https://gitorious.org/linux-n900/linux-n900/source/629fc5ab00cafb31= 272c478efa2c2b35fabd4c70:arch/arm/mach-omap2/board-rx51-peripherals.c#L= 42 > >=20 > > https://gitorious.org/linux-n900/linux-n900/source/629fc5ab00cafb31= 272c478efa2c2b35fabd4c70:arch/arm/mach-omap2/board-rx51-peripherals.c#L= 274 > >=20 > > Tony already tested offset 0 on real HW and wrote that it is > > working too, just show some warning. But qemu with offset 0x300 > > does not work. >=20 > Yes it works with offset 0, except the smc91x warns about not > using the default offset 0x300 based on the EEPROM. As only > three address lines seem to be connected, we can use offset > 0 just fine, and update the EEPROM configuration as needed. Applying into omap-for-v4.0/fixes with updated comments. Regards, Tony