rootfstype is handled by the kernel, so we should handle it, too. ref: https://bugs.archlinux.org/task/45117 --- init_functions | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/init_functions b/init_functions index 11ce3ca..05f4ab0 100644 --- a/init_functions +++ b/init_functions @@ -124,6 +124,7 @@ parse_cmdline() { \#*) break ;; # special cases rw|ro) rwopt=$_w ;; + fstype) rootfstype=$_w ;; fsck.mode=*) case ${_w#*=} in force) @@ -323,7 +324,7 @@ default_mount_handler() { fi msg ":: mounting '$root' on real root" - if ! mount ${fstype:+-t $fstype} -o ${rwopt:-ro}${rootflags:+,$rootflags} "$root" "$1"; then + if ! mount ${rootfstype:+-t $rootfstype} -o ${rwopt:-ro}${rootflags:+,$rootflags} "$root" "$1"; then echo "You are now being dropped into an emergency shell." launch_interactive_shell msg "Trying to continue (this will most likely fail) ..." -- 2.8.3