[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#280571: initrd-tools: Should initialize only LVM VG needed for root fs, not all



On Sun, Nov 21, 2004 at 02:23:31AM +0100, Christian Grigis wrote:
> On Sun, Nov 21, 2004 at 12:24:35AM +0000, Martin Michlmayr wrote:
> > * Christian Grigis <glove@grigri.org> [2004-11-20 16:24]:
> > > In other words, there are still a few issues...
> > 
> > But your last patch which Horms applied didn't make it worse or did
> > it?  i.e should we back that out?
> 
> My first patch fixed it for me for the ROOT=/dev/real_root case, and made
> it slightly worse for the ROOT=probe case. I have now tried the following
> patch (that includes the first one), and it works for me for both cases:

So it can run multiple times for LVM.
That makes sense if we are using vgchange for specific volumes
rather than globally. Does anyone have an objection (I haven't tested
this :)

-- 
Horms

> --- /usr/sbin/mkinitrd.orig	2004-11-10 10:47:02.000000000 +0100
> +++ /usr/sbin/mkinitrd	2004-11-21 02:08:48.000000000 +0100
> @@ -82,7 +82,6 @@
>  	IDE_CORE=
>  	IDE_MODULE=
>  	SHARE=/usr/share/initrd-tools
> -	LVM=
>  }
>  
>  usage() {
> @@ -261,11 +260,6 @@
>  	vg=${vg%/*}
>  	DEVLINKS="$DEVLINKS $vg"
>  
> -	if [ $LVM ]; then
> -		return
> -	fi
> -	LVM=yes
> -
>  	if module_exists drivers/md/lvm-mod && [ -d /lib/lvm-10 ]; then
>  		echo lvm-mod
>  		echo /dev/lvm >&4
> @@ -273,7 +267,7 @@
>  			echo '[ -c /dev/lvm ] || mknod /dev/lvm c 109 0'
>  			echo mount_tmpfs /etc
>  			echo vgscan
> -			echo vgchange -a y
> +			echo vgchange -a y ${vg}
>  			echo umount -n /etc
>  		} >&5
>  		{
> @@ -303,7 +297,7 @@
>  				cat /var/lvm.conf > /etc/lvm/lvm.conf
>  			fi
>  			mount -nt devfs devfs /dev
> -			vgchange -a y
> +			vgchange -a y ${vg}
>  			umount /dev
>  			umount -n /var
>  			umount -n /etc/lvm
> 
> i.e. it allows lvm() to be executed more than once.
> 
> Can you see cases where this would break where it would originally work?
> 
> Best regards,
> 
> -Christian



Reply to: