From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail.windriver.com (mail.windriver.com [147.11.1.11]) by mail.openembedded.org (Postfix) with ESMTP id 5A54E71634 for ; Wed, 14 Jan 2015 03:38:00 +0000 (UTC) Received: from ALA-HCA.corp.ad.wrs.com (ala-hca.corp.ad.wrs.com [147.11.189.40]) by mail.windriver.com (8.14.9/8.14.5) with ESMTP id t0E3bwIF005014 (version=TLSv1/SSLv3 cipher=AES256-SHA bits=256 verify=FAIL); Tue, 13 Jan 2015 19:37:58 -0800 (PST) Received: from [128.224.162.226] (128.224.162.226) by ALA-HCA.corp.ad.wrs.com (147.11.189.50) with Microsoft SMTP Server (TLS) id 14.3.174.1; Tue, 13 Jan 2015 19:37:57 -0800 Message-ID: <54B5E495.7020104@windriver.com> Date: Wed, 14 Jan 2015 11:37:57 +0800 From: ChenQi User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:31.0) Gecko/20100101 Thunderbird/31.3.0 MIME-Version: 1.0 To: "Moseley, Drew" References: <1420818765-28571-1-git-send-email-drew_moseley@mentor.com> <54B32BD4.3080209@windriver.com> In-Reply-To: X-Originating-IP: [128.224.162.226] Cc: "openembedded-core@lists.openembedded.org" Subject: Re: [PATCH 1/2] init-install.sh: Switch to using static device paths. X-BeenThere: openembedded-core@lists.openembedded.org X-Mailman-Version: 2.1.12 Precedence: list List-Id: Patches and discussions about the oe-core layer List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 14 Jan 2015 03:38:12 -0000 Content-Type: text/plain; charset="windows-1252"; format=flowed Content-Transfer-Encoding: 7bit On 01/13/2015 10:46 PM, Moseley, Drew wrote: >> On Jan 11, 2015, at 9:05 PM, ChenQi wrote: >> >> On 01/09/2015 11:52 PM, Drew Moseley wrote: >>> Use static device paths for mounting from /dev/disk/ in >>> case the device ordering changes. >>> >>> Signed-off-by: Drew Moseley >>> --- >>> meta/recipes-core/initrdscripts/files/init-install.sh | 14 ++++++++++---- >>> 1 file changed, 10 insertions(+), 4 deletions(-) >>> >>> diff --git a/meta/recipes-core/initrdscripts/files/init-install.sh b/meta/recipes-core/initrdscripts/files/init-install.sh >>> index fb537ee..f924418 100644 >>> --- a/meta/recipes-core/initrdscripts/files/init-install.sh >>> +++ b/meta/recipes-core/initrdscripts/files/init-install.sh >>> @@ -156,6 +156,12 @@ mkfs.ext3 $rootfs >>> echo "Formatting swap partition...($swap)" >>> mkswap $swap >>> +# Determine the static device paths >>> +sync; udevadm settle >>> +bootfs_static="$(udevadm info --query=property --name=$bootfs | grep DEVLINKS= | cut -d= -f2 | cut -d\ -f1)" >>> +swap_static="$(udevadm info --query=property --name=$swap | grep DEVLINKS= | cut -d= -f2 | cut -d\ -f1)" >>> +rootfs_uuid="$(udevadm info --query=property --name=$rootfs | grep ID_PART_ENTRY_UUID | cut -d= -f2)" >>> + >> Would ID_PART_ENTRY_UUID be always there? > That's a good question. In my test systems (two VirtualBox configs and two AMD based PCs) it is always there. Taking a second look at my developer Ubuntu system it does not seem to have that field so maybe I'm just getting lucky on my test systems. > > Anybody have a suggestion for a replacement? We can use blkid command. > >>> mkdir /tgt_root >>> mkdir /src_root >>> mkdir -p /boot >>> @@ -166,8 +172,8 @@ mount -o rw,loop,noatime,nodiratime /run/media/$1/$2 /src_root >>> echo "Copying rootfs files..." >>> cp -a /src_root/* /tgt_root >>> if [ -d /tgt_root/etc/ ] ; then >>> - echo "$swap swap swap defaults 0 0" >> /tgt_root/etc/fstab >>> - echo "$bootfs /boot ext3 defaults 1 2" >> /tgt_root/etc/fstab >>> + echo "$swap_static swap swap defaults 0 0" >> /tgt_root/etc/fstab >>> + echo "$bootfs_static /boot ext3 defaults 1 2" >> /tgt_root/etc/fstab >>> # We dont want udev to mount our root device while we're booting... >>> if [ -d /tgt_root/etc/udev/ ] ; then >>> echo "/dev/${device}" >> /tgt_root/etc/udev/mount.blacklist >>> @@ -186,7 +192,7 @@ if [ -f /etc/grub.d/00_header ] ; then >>> cat >$GRUBCFG <<_EOF >>> menuentry "Linux" { >>> set root=(hd0,1) >>> - linux /vmlinuz root=$rootfs $rootwait rw $5 $3 $4 quiet >>> + linux /vmlinuz root=PARTUUID=$rootfs_uuid $rootwait rw $5 $3 $4 quiet >> I thought PARTUUID is for GPT only. >> > Perhaps so but it does seem to be there on my VirtualBox systems. My Ubuntu host uses "root=UUID=blah" but that was't working for me at some point. I'll dig back into it and see what I can find. > > Drew > > > > root=UUID=xxx has to be resolved by an initramfs because kernel itself cannot handle it. There's a patchset I sent half a year ago trying to solve this problem. http://cgit.openembedded.org/cgit.cgi/openembedded-core-contrib/log/?h=ChenQi/initramfs-uuid This offers one solution, maybe an ugly one. Best Regards, Chen Qi