From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752340AbaE0Nh3 (ORCPT ); Tue, 27 May 2014 09:37:29 -0400 Received: from mout.kundenserver.de ([212.227.126.130]:54049 "EHLO mout.kundenserver.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752019AbaE0Nh2 (ORCPT ); Tue, 27 May 2014 09:37:28 -0400 From: Arnd Bergmann To: Daniel Thompson Cc: linux-arm-kernel@lists.infradead.org, Russell King , Paul Bolle , linaro-kernel@lists.linaro.org, patches@linaro.org, spear-devel@list.st.com, linux-kernel@vger.kernel.org Subject: Re: [RESEND PATCH v5 3/5] ARM: Conceal DEBUG_LL_UART_NONE from unsupported platforms Date: Tue, 27 May 2014 15:37:12 +0200 Message-ID: <8143025.UYbkItXldM@wuerfel> User-Agent: KMail/4.11.5 (Linux/3.11.0-18-generic; KDE/4.11.5; x86_64; ; ) In-Reply-To: <53848F7C.6020508@linaro.org> References: <1398355246-29571-1-git-send-email-daniel.thompson@linaro.org> <6624276.NBS809VbNB@wuerfel> <53848F7C.6020508@linaro.org> MIME-Version: 1.0 Content-Transfer-Encoding: 7Bit Content-Type: text/plain; charset="us-ascii" X-Provags-ID: V02:K0:j9MteLcMT0D0IoZra3mVhP1Zx4QWlonqHAbL1sdO3H1 AtJe44tCNv8e4erChiyJMCGINCgvDXmuGUIyHWjeSTiu5AIGyl vEnyW7DdTkxw4nUdyUM04XzwbRHAdd/zoMsE2TtEzPe9yyim5x h5JNJiz+5kTlw77mLWsVYi5zWjAKu8eeExfNTq+LQor3yQDJ4u J1AMLRt7vWrVCszMlSGgaqjdcPJjbqfddNOjAmF3UxDtq59gsV lYWq5YiaJ83Zp7j5g8Oiak7kVAw7Tif2M/kr9XjnOd5qlMFCLR fNYDTTKOKXPzpWJd7Z8aJZcxlgbrfwlNh7OwtP7NtenvhndvRO Wl2eyzzkrVY91w04BUBY= Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tuesday 27 May 2014 14:13:32 Daniel Thompson wrote: > On 26/05/14 14:39, Arnd Bergmann wrote: > >>> diff --git a/arch/arm/Kconfig.debug b/arch/arm/Kconfig.debug > >>> index eab8ecb..4c8d7db 100644 > >>> --- a/arch/arm/Kconfig.debug > >>> +++ b/arch/arm/Kconfig.debug > >>> @@ -842,7 +842,18 @@ choice > >>> > >>> config DEBUG_LL_UART_NONE > >>> bool "No low-level debugging UART" > >>> - depends on !ARCH_MULTIPLATFORM > >>> + depends on ARCH_AT91 || \ > >>> + ARCH_CLPS711X || \ > >>> + ARCH_FOOTBRIDGE || \ > >>> + ARCH_KS8695 || \ > >>> + ARCH_NETX || \ > >>> + ARCH_OMAP1 || \ > >>> + ARCH_SA1100 || \ > >>> + ARCH_S3C24XX || \ > >>> + ARCH_S3C64XX || \ > >>> + ARCH_S5P64X0 || \ > >>> + ARCH_S5PC100 || \ > >>> + ARCH_S5PV210 > >>> help > >>> Say Y here if your platform doesn't provide a UART option > >>> above. This relies on your platform choosing the right UART > >> > >> ARCH_S3C24XX has been moved over in linux-next to use debug/s3c24xx.S. > >> A few others will likely move soon (AT91, CLPS711x, S5PV210), as the > >> platforms get moved to ARCH_MULTIPLATFORM. > > The list in my patch came from: > > sundance$ find arch/arm -name "debug-macro.S" > arch/arm/mach-at91/include/mach/debug-macro.S > arch/arm/mach-clps711x/include/mach/debug-macro.S > arch/arm/mach-footbridge/include/mach/debug-macro.S > arch/arm/mach-ks8695/include/mach/debug-macro.S > arch/arm/mach-netx/include/mach/debug-macro.S > arch/arm/mach-omap1/include/mach/debug-macro.S > arch/arm/mach-s3c64xx/include/mach/debug-macro.S > arch/arm/mach-s5p64x0/include/mach/debug-macro.S > arch/arm/mach-s5pc100/include/mach/debug-macro.S > arch/arm/mach-s5pv210/include/mach/debug-macro.S > arch/arm/mach-sa1100/include/mach/debug-macro.S > > In other words this patch lists of all platforms where > DEBUG_LL_UART_NONE should not result in failure to compile the kernel. I see the same list, this is only half the story though. > This list doesn't seem to have changed in linux-next. > > >> As an alternative to your patch, we could decide to move the remaining > >> files as well. > > > > Follow-up: I also noticed that AT91, CLPS711X, FOOTBRIDGE and all the SAMSUNG > > platforms already have an entry in the list, so DEBUG_LL_UART_NONE is redundant > > for them. The below change should be enough. > > > > For AT91 this is actually better because the old AT91X40 support doesn't > > have working DEBUG_LL support. This gets handled correctly by > > AT91_DEBUG_LL_DBGU0/AT91_DEBUG_LL_DBGU1, but not DEBUG_LL_UART_NONE. > > I'm very happy to update my patch to follow this one. However since it > will render the corresponding debug-macro.S unreachable by KBuild I like > to nuke the file as well. Good idea? No. See the description under config DEBUG_LL_UART_NONE: Say Y here if your platform doesn't provide a UART option above. This relies on your platform choosing the right UART definition internally in order for low-level debugging to work. The platforms I listed do have "a UART option above", e.g. config AT91_DEBUG_LL_DBGU0 bool "Kernel low-level debugging on rm9200, 9260/9g20, 9261/9g10 and 9rl" depends on HAVE_AT91_DBGU0 Most of the platforms that have an option like this put the file into include/debug, as is required for ARCH_MULTIPLATFORM. The ones that are not converted to ARCH_MULTIPLATFORM can have it in either place, see config DEBUG_LL_INCLUDE string default "debug/8250.S" if DEBUG_LL_UART_8250 || DEBUG_UART_8250 default "debug/pl01x.S" if DEBUG_LL_UART_PL01X || DEBUG_UART_PL01X ... default "debug/zynq.S" if DEBUG_ZYNQ_UART0 || DEBUG_ZYNQ_UART1 default "mach/debug-macro.S" Anything that is not explicitly listed under DEBUG_LL_INCLUDE will default to "mach/debug-macro.S", and the list does not (have to) match the choice statements in "Kernel low-level debugging port". Arnd