I know you already commited, but, just a little comment...
On Mon, Feb 16, 2009 at 04:33:24PM +0100, Hans de Goede wrote:
> Hi all,
>
> Use dmraid instead of setting up dm tables ourselves, this fixes stacked
> dmraid's like raid10. As an added bonus this means dmraid will also stop
> breaking when the disk numbering changes. This patch uses the new
> --rm_partitions dmraid option to tell the kernel to remove the partitions
> from the used disks, stopping programs probing partitons causing IO-errors
> like those reported in bug 475384
>
> Note that this patch also removes part of the (broken) multipath code, which
> was in the same code loop as the dyi dm table creation code for dmraid. This
> is not a problem, as multipath has never worked in Fedora.
> ---
> mkinitrd | 117 +++++++++++--------------------------------------------------
> 1 files changed, 21 insertions(+), 96 deletions(-)
>
> diff --git a/mkinitrd b/mkinitrd
> index 211be1f..2ce03a7 100755
> --- a/mkinitrd
> +++ b/mkinitrd
> @@ -60,7 +60,7 @@ PROBE="yes"
> MODULES=""
> GRAPHICSMODS=""
> PREMODS=""
> -DMDEVS=""
> +DMRAIDS=""
> ncryptodevs=0
> ncryptoparts=0
> ncryptolvs=0
> @@ -130,10 +130,6 @@ qpopd() {
> popd >/dev/null 2>&1
> }
>
> -dm_get_uuid() {
> - echo nash-dm get_uuid "$1" | /sbin/nash --force --quiet
> -}
> -
> resolve_device_name() {
> echo nash-resolveDevice "$1" | /sbin/nash --forcequiet
> }
> @@ -560,6 +556,7 @@ handledm() {
> EOF
> local name=$(dmsetup info --noheadings -c -j $major -m $minor -o name)
> local vg=$(lvshow "/dev/mapper/$name")
> + local raids=$(/sbin/dmraid -s -craidname 2>/dev/null | grep -vi "no
Is it not better to have the dmraid command without the full path?
> raid disks")
> if [ -n "$vg" ]; then
> vg=`echo $vg` # strip whitespace
> case " $vg_list " in
> @@ -569,6 +566,17 @@ EOF
> ;;
> esac
> fi
> + for raid in $raids ; do
> + if [ "$raid" == "$name" ]; then
> + case " $DMRAIDS " in
> + *" $raid "*) ;;
> + *) DMRAIDS="$DMRAIDS $raid"
> + [ -z "$nodmraid" ] && find_base_dm_mods
> + ;;
> + esac
> + break
> + fi
I just say that little issue before I realised that it was commited. :)
regards
--
Joel Andres Granados
Brno, Czech Republic, Red Hat.
_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
