David Dillow wrote:
Now that netroot is serialized and only one interface can be configured
at a time, netif is almost completely unneeded by the netroot handlers.
The remaining hold out is nfsroot, which uses it to get the command line
overrides and the DHCP root information.
Loosing arguments is always good.
The only thing I don't like about this is that netroot needs to
implement a few specialities just for NFS. I'd prefer netroot to be
(mount-)protocol independent.
On the other hand, I like the idea that network root handlers shouldn't
care a thing about network configuration and options.
Question: Is it possible that other netroot handlers we haven't
implemented or thought about might need to know about interfaces?
Discussion: Instead of implementing the dhcp options server failover
just for nfs, why not extend this to all protocols?
By adding that information to the calling convention, we can get rid
of netif, and prepare for the next steps of moving parsing and validation
into the handlers themselves.
---
modules.d/40network/netroot | 16 ++++++++++------
modules.d/95iscsi/iscsiroot | 7 +++----
modules.d/95nbd/nbdroot | 9 ++++-----
modules.d/95nfs/nfsroot | 20 ++++++++------------
4 files changed, 25 insertions(+), 27 deletions(-)
diff --git a/modules.d/40network/netroot b/modules.d/40network/netroot
index 645bbba..c9ebf4c 100755
--- a/modules.d/40network/netroot
+++ b/modules.d/40network/netroot
@@ -44,15 +44,20 @@ done
netif=$1
[ -e "/tmp/net.bootdev" ] && read netif < /tmp/net.bootdev
+# Load command line overrides to get the server override
+[ -e /tmp/net.$netif.override ] && . /tmp/net.$netif.override
+[ -n "$srv" ] && server_id=$srv
+
+# Load dhcp options and get the first fallback to the server
+[ -e /tmp/dhclient.$netif.dhcpopts ] && . /tmp/dhclient.$netif.dhcpopts
+[ -z "$server_id" ] && server_id=$new_dhcp_server_identifier
+
# Figure out the handler for root=dhcp by recalling all netroot cmdline
# handlers
if [ "$netroot" = "dhcp" ] ; then
# Unset root so we can check later
unset root
- # Load dhcp options
- [ -e /tmp/dhclient.$netif.dhcpopts ] && . /tmp/dhclient.$netif.dhcpopts
-
# If we have a specific bootdev with no dhcpoptions or empty root-path,
# we die. Otherwise we just warn
if [ -z "$new_root_path" ] ; then
@@ -92,9 +97,8 @@ done
[ -e /tmp/net.$netif.hostname ] && . /tmp/net.$netif.hostname
[ -e /tmp/net.$netif.resolv.conf ] && cp -f /tmp/net.$netif.resolv.conf /etc/resolv.conf
-# Run the handler; don't store the root, it may change from device to device
-# XXX other variables to export?
-if $handler $netif $netroot $NEWROOT; then
+# Run the handler to mount/login into the root device
+if $handler "$netroot" "$server_id" "$new_root_path" $NEWROOT; then
# Network rootfs mount successful
for iface in $IFACES ; do
[ -f /tmp/dhclient.$iface.lease ] && cp /tmp/dhclient.$iface.lease /tmp/net.$iface.lease
diff --git a/modules.d/95iscsi/iscsiroot b/modules.d/95iscsi/iscsiroot
index 2b041cc..923141c 100755
--- a/modules.d/95iscsi/iscsiroot
+++ b/modules.d/95iscsi/iscsiroot
@@ -11,14 +11,13 @@
PATH=$PATH:/sbin:/usr/sbin
if getarg rdnetdebug; then
- exec > /tmp/iscsiroot.$1.$$.out
- exec 2>> /tmp/iscsiroot.$1.$$.out
+ exec > /tmp/iscsiroot.$$.out
+ exec 2>> /tmp/iscsiroot.$$.out
set -x
fi
Maybe this is just me nagging around, but do the changes above have
anything to do with removing netif?
# root is in the form root=iscsi:[<servername>]:[<protocol>]:[<port>]:[<LUN>]:<targetname>
-netif="$1"
-root="$2"
+root="$1"
# read static conf settings
for conf in conf/conf.d/*; do
diff --git a/modules.d/95nbd/nbdroot b/modules.d/95nbd/nbdroot
index c7f24e8..f76211c 100755
--- a/modules.d/95nbd/nbdroot
+++ b/modules.d/95nbd/nbdroot
@@ -5,15 +5,14 @@
PATH=$PATH:/sbin:/usr/sbin
if getarg rdnetdebug; then
- exec > /tmp/nbdroot.$1.$$.out
- exec 2>> /tmp/nbdroot.$1.$$.out
+ exec > /tmp/nbdroot.$$.out
+ exec 2>> /tmp/nbdroot.$$.out
set -x
fi
Same here, is this netif related?
# root is in the form root=nbd:srv:port[:fstype[:rootflags[:nbdopts]]]
-netif="$1"
-root="$2"
-NEWROOT="$3"
+root="$1"
+NEWROOT="$4"
# If it's not nbd we don't continue
[ "${root%%:*}" = "nbd" ] || return
diff --git a/modules.d/95nfs/nfsroot b/modules.d/95nfs/nfsroot
index f005378..847ec60 100755
--- a/modules.d/95nfs/nfsroot
+++ b/modules.d/95nfs/nfsroot
@@ -43,16 +43,17 @@ root_to_var() {
PATH=$PATH:/sbin:/usr/sbin
if getarg rdnetdebug ; then
- exec > /tmp/nfsroot.$1.$$.out
- exec 2>> /tmp/nfsroot.$1.$$.out
+ exec > /tmp/nfsroot.$$.out
+ exec 2>> /tmp/nfsroot.$$.out
set -x
fi
And another one here. Is this Netif removal related?
# root is in the form root=nfs[4]:[server:]path[:options], either from
# cmdline or dhcp root-path
-netif="$1"
-root="$2"
-NEWROOT="$3"
+root="$1"
+server_id=$2
+root_path=$3
+NEWROOT="$4"
# Continue if nfs prefix
case "${root%%:*}" in
@@ -62,20 +63,15 @@ esac
root_to_var $root
-#Load other data that might provide info
-[ -f /tmp/net.$netif.override ] && . /tmp/net.$netif.override
-[ -f /tmp/dhclient.$netif.dhcpopts ] && . /tmp/dhclient.$netif.dhcpopts
-
#Empty path means try dhcp root-path, this is ok here since parse-nfsroot.sh
#already takes care of nfs:... formatted root-path
-[ -z "$path" ] && root_to_var $nfs:$new_root_path
+[ -z "$path" ] && root_to_var $nfs:$root_path
#Empty path defaults to "/tftpboot/%s" only in nfsroot.txt legacy mode
[ -z "$path" ] && [ "$(getarg root=)" = "/dev/nfs" ] && path="/tftpboot/%s"
if [ -z "$server" ] ; then
- # XXX new_dhcp_next_server is unconfirmed this is an assumption
- for var in $srv $new_dhcp_server_identifier $new_dhcp_next_server $new_root_path '' ; do
+ for var in $server_id $root_path '' ; do
[ -n "$var" ] && server=$var && break;
done
--
To unsubscribe from this list: send the line "unsubscribe initramfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
