From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from nommos.sslcatacombnetworking.com (nommos.sslcatacombnetworking.com [67.18.224.114]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client did not present a certificate) by ozlabs.org (Postfix) with ESMTP id 9035E68743 for ; Thu, 17 Nov 2005 02:24:33 +1100 (EST) In-Reply-To: <1132136755.21643.40.camel@hades.cambridge.redhat.com> References: <1131357857.27347.28.camel@baythorne.infradead.org> <20051107160939.GD3839@smtp.west.cox.net> <1131380422.27347.72.camel@baythorne.infradead.org> <20051107165453.GA29111@lst.de> <1131382971.27347.79.camel@baythorne.infradead.org> <1132131116.28963.75.camel@baythorne.infradead.org> <17274.63508.501761.748278@cargo.ozlabs.ibm.com> <1132136755.21643.40.camel@hades.cambridge.redhat.com> Mime-Version: 1.0 (Apple Message framework v746.2) Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed Message-Id: <8CE48849-147D-4140-96B4-9727E1E509C7@kernel.crashing.org> From: Kumar Gala Date: Wed, 16 Nov 2005 09:24:51 -0600 To: David Woodhouse Cc: Tom Rini , linuxppc-dev@ozlabs.org Subject: Re: [PATCH] Fix 8250 probe on ppc32 List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Nov 16, 2005, at 4:25 AM, David Woodhouse wrote: > On Wed, 2005-11-16 at 20:12 +1100, Paul Mackerras wrote: >> Yes, it works on newworld (at least it does on my G4 powerbook). It >> should work on 32-bit CHRP with any luck, but for oldworld we'll have >> to generate a COFF file. > > I'll try it on Pegasos some time tonight or tomorrow. AFAICT I'm going > to need the patch which started this thread, if I want any serial > ports > on my Pegasos. The newly-created asm-powerpc/serial.h file is > devoid of > serial port definitions. > > Here it is again, with Tom's three criticisms addressed, deliberately > ignored, and addressed. In that order. > > It no longer touches asm-ppc/serial.h, so doesn't break the legacy > zImage. > > --- > [PATCH] Fix 8250 probe on ppc32 > > There are no serial ports defined in the new asm-powerpc/serial.h, and > rightly so. The 64-bit setup code has a routine to look in the device > tree for them and register a platform_device for any we find there; > this > patch steals that routine and makes it available on 32-bit machines > too... > > Signed-off-by: David Woodhouse > [snip] > --- linux-2.6.13/drivers/serial/Kconfig.sof 2005-10-20 > 12:54:48.000000000 +0100 > +++ linux-2.6.13/drivers/serial/Kconfig 2005-10-20 > 13:05:39.000000000 +0100 > @@ -77,6 +77,11 @@ config SERIAL_8250_CS > > If unsure, say N. > > +config SERIAL_8250_OF > + bool > + default y > + depends on PPC_OF && SERIAL_8250 && PPC_MERGE > + > config SERIAL_8250_ACPI > bool "8250/16550 device discovery via ACPI namespace" > default y if IA64 [snip] Should this really depend on PPC_OF? Does it require true OpenFirmware or just a flat dev tree to work? I ask because I'd like to extended this in the future to handle 8250 serial ports that aren't on a PCI or ISA bus, but on SoCs like 83xx, 85xx, 86xx, 4xx? These will most likely not be running with true OF but I've got a flat dev tree working today. The following is what I've been toying with for an description of the 8250 node (in .dts format): serial@4500 { device_type = "serial"; compatible = "ns16550"; reg = <4500 100>; clock-frequency = <0>; interrupts = <1a 3>; interrupt-parent = <40000>; }; - kumar