From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-1.0 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,SPF_PASS,URIBL_BLOCKED autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id B4C3FC43441 for ; Mon, 26 Nov 2018 23:34:04 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 83A01208E4 for ; Mon, 26 Nov 2018 23:34:04 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 83A01208E4 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=linux-foundation.org Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=linux-kernel-owner@vger.kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727763AbeK0K3u (ORCPT ); Tue, 27 Nov 2018 05:29:50 -0500 Received: from mail.linuxfoundation.org ([140.211.169.12]:47570 "EHLO mail.linuxfoundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726516AbeK0K3u (ORCPT ); Tue, 27 Nov 2018 05:29:50 -0500 Received: from localhost.localdomain (c-24-6-170-16.hsd1.ca.comcast.net [24.6.170.16]) by mail.linuxfoundation.org (Postfix) with ESMTPSA id 86DEF6C; Mon, 26 Nov 2018 23:34:01 +0000 (UTC) Date: Mon, 26 Nov 2018 15:34:00 -0800 From: Andrew Morton To: Li Zhijian Cc: linux-kernel@vger.kernel.org, Dominik Brodowski , zhijianx.li@intel.com, Philip Li Subject: Re: [PATCH] initramfs: clean old path before creating a hardlink Message-Id: <20181126153400.4985207970edfb39cb47eb61@linux-foundation.org> In-Reply-To: <1542352368-13299-1-git-send-email-lizhijian@cn.fujitsu.com> References: <1542352368-13299-1-git-send-email-lizhijian@cn.fujitsu.com> X-Mailer: Sylpheed 3.5.1 (GTK+ 2.24.31; x86_64-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 16 Nov 2018 15:12:48 +0800 Li Zhijian wrote: > Previously, sys_link() will fail due to the new path is already existed. > this case ofen appears when we use a concated initrd, below is an > sample: > > 1) prepare a basic rootfs, it contains a regular files rc.local > lizhijian@:~/yocto-tiny-i386-2016-04-22$ cat etc/rc.local > #!/bin/sh > echo "Running /etc/rc.local..." > yocto-tiny-i386-2016-04-22$ find . | sed 's,^\./,,' | cpio -o -H newc | gzip -n -9 >../rootfs.cgz > > 2) create a extra initrd which also includes a etc/rc.local > lizhijian@:~/lkp-x86_64/etc$ echo "append initrd" >rc.local > lizhijian@:~/lkp/lkp-x86_64/etc$ cat rc.local > append initrd > lizhijian@:~/lkp/lkp-x86_64/etc$ ln rc.local rc.local.hardlink > append initrd > lizhijian@:~/lkp/lkp-x86_64/etc$ stat rc.local rc.local.hardlink > File: 'rc.local' > Size: 14 Blocks: 8 IO Block: 4096 regular file > Device: 801h/2049d Inode: 11296086 Links: 2 > Access: (0664/-rw-rw-r--) Uid: ( 1002/lizhijian) Gid: ( 1002/lizhijian) > Access: 2018-11-15 16:08:28.654464815 +0800 > Modify: 2018-11-15 16:07:57.514903210 +0800 > Change: 2018-11-15 16:08:24.180228872 +0800 > Birth: - > File: 'rc.local.hardlink' > Size: 14 Blocks: 8 IO Block: 4096 regular file > Device: 801h/2049d Inode: 11296086 Links: 2 > Access: (0664/-rw-rw-r--) Uid: ( 1002/lizhijian) Gid: ( 1002/lizhijian) > Access: 2018-11-15 16:08:28.654464815 +0800 > Modify: 2018-11-15 16:07:57.514903210 +0800 > Change: 2018-11-15 16:08:24.180228872 +0800 > Birth: - > > lizhijian@:~/lkp/lkp-x86_64$ find . | sed 's,^\./,,' | cpio -o -H newc | gzip -n -9 >../rc-local.cgz > lizhijian@:~/lkp/lkp-x86_64$ gzip -dc ../rc-local.cgz | cpio -t > . > etc > etc/rc.local.hardlink <<< it will be extracted first at this initrd > etc/rc.local > > 3) concate 2 initrds and boot > lizhijian@:~/lkp$ cat rootfs.cgz rc-local.cgz >concate-initrd.cgz > lizhijian@:~/lkp$ qemu-system-x86_64 -nographic -enable-kvm -cpu host -smp 1 -m 1024 -kernel ~/lkp/linux/arch/x86/boot/bzImage -append "console=ttyS0 earlyprint=ttyS0 ignore_loglevel" -initrd ./concate-initr.cgz -serial stdio -nodefaults > > In this case, sys_link(2) will fail and return -EEXIST, so we can only > get the rc.local at rootfs.cgz instead of rc-local.cgz I can't claim to understand this, but I believe you ;) How serious is this problem? Do you think a 4.20 merge is justified? Or a -stable backport? If so, why? > --- a/init/initramfs.c > +++ b/init/initramfs.c > @@ -291,12 +291,16 @@ static int __init do_reset(void) > return 1; > } > > +static void __init clean_path(char *path, umode_t fmode); > + > static int __init maybe_link(void) > { > if (nlink >= 2) { > char *old = find_link(major, minor, ino, mode, collected); > - if (old) > + if (old) { > + clean_path(collected, 0); > return (ksys_link(old, collected) < 0) ? -1 : 1; > + } > } > return 0; > } The forward declaration is unpleasing. Why not simply move the function? --- a/init/initramfs.c~initramfs-clean-old-path-before-creating-a-hardlink-fix +++ a/init/initramfs.c @@ -291,7 +291,17 @@ static int __init do_reset(void) return 1; } -static void __init clean_path(char *path, umode_t fmode); +static void __init clean_path(char *path, umode_t fmode) +{ + struct kstat st; + + if (!vfs_lstat(path, &st) && (st.mode ^ fmode) & S_IFMT) { + if (S_ISDIR(st.mode)) + ksys_rmdir(path); + else + ksys_unlink(path); + } +} static int __init maybe_link(void) { @@ -305,18 +315,6 @@ static int __init maybe_link(void) return 0; } -static void __init clean_path(char *path, umode_t fmode) -{ - struct kstat st; - - if (!vfs_lstat(path, &st) && (st.mode ^ fmode) & S_IFMT) { - if (S_ISDIR(st.mode)) - ksys_rmdir(path); - else - ksys_unlink(path); - } -} - static __initdata int wfd; static int __init do_name(void) _