Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751749AbaLCF25 (ORCPT ); Wed, 3 Dec 2014 00:28:57 -0500 Received: from szxga02-in.huawei.com ([119.145.14.65]:24010 "EHLO szxga02-in.huawei.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750815AbaLCF24 (ORCPT ); Wed, 3 Dec 2014 00:28:56 -0500 Message-ID: <547E9F89.5000601@huawei.com> Date: Wed, 3 Dec 2014 13:28:41 +0800 From: Wang Nan User-Agent: Mozilla/5.0 (Windows NT 6.1; rv:24.0) Gecko/20100101 Thunderbird/24.0.1 MIME-Version: 1.0 To: Masami Hiramatsu CC: , , , , Subject: Re: [PATCH v11 1/7] ARM: probes: move all probe code to dedicate directory References: <1417423513-47161-1-git-send-email-wangnan0@huawei.com> <1417423695-47521-1-git-send-email-wangnan0@huawei.com> <547D471B.2070908@hitachi.com> <547D9308.8080403@huawei.com> <547E93E3.70209@hitachi.com> In-Reply-To: <547E93E3.70209@hitachi.com> Content-Type: text/plain; charset="ISO-2022-JP" Content-Transfer-Encoding: 7bit X-Originating-IP: [10.111.69.90] X-CFilter-Loop: Reflected Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On 2014/12/3 12:38, Masami Hiramatsu wrote: > (2014/12/02 19:23), Wang Nan wrote: >> On 2014/12/2 12:59, Masami Hiramatsu wrote: >>> (2014/12/01 17:48), Wang Nan wrote: >>>> In discussion on LKML (https://lkml.org/lkml/2014/11/28/158), Russell >>>> King suggest to move all probe related code to arch/arm/probes. This >>>> patch do the work. Due to dependency on 'arch/arm/kernel/patch.h', this >>>> patch also move patch.h to 'arch/arm/include/asm/patch.h', and related >>>> '#include' directive are also midified to '#include '. >>> >>> Moving is good to me, but renaming files are also required I think. >>> >>>> Signed-off-by: Wang Nan >>>> --- >>>> arch/arm/Makefile | 1 + >>>> arch/arm/{kernel => include/asm}/patch.h | 0 >>>> arch/arm/kernel/Makefile | 16 ++-------------- >>>> arch/arm/kernel/jump_label.c | 2 +- >>>> arch/arm/kernel/patch.c | 3 +-- >>>> arch/arm/probes/Makefile | 15 +++++++++++++++ >>>> arch/arm/{kernel => probes}/kprobes-arm.c | 0 >>>> arch/arm/{kernel => probes}/kprobes-common.c | 0 >>>> arch/arm/{kernel => probes}/kprobes-test-arm.c | 0 >>>> arch/arm/{kernel => probes}/kprobes-test-thumb.c | 0 >>>> arch/arm/{kernel => probes}/kprobes-test.c | 0 >>>> arch/arm/{kernel => probes}/kprobes-test.h | 0 >>>> arch/arm/{kernel => probes}/kprobes-thumb.c | 0 >>>> arch/arm/{kernel => probes}/kprobes.c | 2 +- >>>> arch/arm/{kernel => probes}/kprobes.h | 0 >>>> arch/arm/{kernel => probes}/probes-arm.c | 0 >>>> arch/arm/{kernel => probes}/probes-arm.h | 0 >>>> arch/arm/{kernel => probes}/probes-thumb.c | 0 >>>> arch/arm/{kernel => probes}/probes-thumb.h | 0 >>>> arch/arm/{kernel => probes}/probes.c | 0 >>>> arch/arm/{kernel => probes}/probes.h | 0 >>>> arch/arm/{kernel => probes}/uprobes-arm.c | 0 >>>> arch/arm/{kernel => probes}/uprobes.c | 0 >>>> arch/arm/{kernel => probes}/uprobes.h | 0 >>> >>> As I did on x86, these would be better renamed as expressing what they do. >>> I guess most of the files may have emulate-*.c or decode-*.c :) >>> >>> Thank you, >>> >> >> OK. I posted another patch in this thread. The directory tree is as follow: >> >> arch/arm/probes/ >> |-- Makefile >> |-- decode-arm.c >> |-- decode-arm.h >> |-- decode-thumb.c >> |-- decode-thumb.h >> |-- decode.c >> |-- decode.h >> |-- kprobes >> | |-- actions-arm.c >> | |-- actions-common.c >> | |-- actions-thumb.c >> | |-- kprobes.c >> | |-- kprobes.h >> | |-- test-arm.c >> | |-- test-core.c >> | |-- test-core.h >> | `-- test-thumb.c >> `-- uprobes >> |-- actions-arm.c >> |-- uprobes.c >> `-- uprobes.h >> >> 2 directories, 19 files >> >> What do you think? > > Yeah, that looks better :) > > Btw, if you introduce probes/{kprobes,uprobes}/, *probes.c should be core.c too, > since the directories already show its name. And also, both dirs should have its > own Makefile. > Seprated Makefile may introduce extra complexity. Think about someone try to compile kprobe as a module (currently it is impossible due to dependencies between kprobe and kernel core, but decoupling is possible), seprated Makefiles may force him to create at least 3 modules for kprobe, even if one module is enough. Anyway, there may some features in kernel build system can help him. I'll post v12 patch series based on your suggestion. Thank you. > Thank you, > > -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/