小发现：Ubuntu使用kexec重启系统

kexec使用Unix系统调用exec()的覆盖哲理，它产生一个新的内核覆盖正在运行的内核。这个过程不用启动机器上的固件（firmware），所以可以节省一些重启时间。

因为启动固件需要走查总线和标识设备，节省了这些时间就会加速系统的启动。

知道了这些就解决一下我的系统不能重启的问题吧，如图： 

怎么会启动不了呢？

1.grep一下/etc目录

apple@Ubuntu:~$ sudo grep -r kexec /etc
grep: /etc/alternatives/smbstatus.1.gz: No such file or directory
/etc/rc6.d/K18kexec-load:# Provides:            kexec-load
/etc/rc6.d/K18kexec-load:# Required-Stop:       $remote_fs kexec
/etc/rc6.d/K18kexec-load:# Short-Description: Load kernel image with kexec
/etc/rc6.d/K18kexec-load:test -r /etc/default/kexec && . /etc/default/kexec
/etc/rc6.d/K18kexec-load:       test -x /sbin/kexec || exit 0
/etc/rc6.d/K18kexec-load:       test “x`cat /sys/kernel/kexec_loaded`y” = “x1y” && exit 0
/etc/rc6.d/K18kexec-load:               kexec -l “$KERNEL_IMAGE” –append=”$REAL_APPEND”
/etc/rc6.d/K18kexec-load:               kexec -l “$KERNEL_IMAGE” –initrd=”$INITRD” –append=”$REAL_APPEND”
/etc/rc6.d/S85kexec:# Provides:         kexec
/etc/rc6.d/S85kexec:# Short-Description: Execute the kexec -e command to reboot system
/etc/rc6.d/S85kexec:test -r /etc/default/kexec && . /etc/default/kexec
/etc/rc6.d/S85kexec:    test “x`cat /sys/kernel/kexec_loaded`y” = “x1y” || exit 0
/etc/rc6.d/S85kexec:    test -x /sbin/kexec || exit 0
/etc/rc6.d/S85kexec:    log_action_msg “Will now restart with kexec”
/etc/rc6.d/S85kexec:        kexec -e
/etc/rc6.d/S85kexec:        log_failure_msg “kexec failed”
/etc/rc6.d/S60umountroot:# Should-Stop:       halt reboot kexec
/etc/init.d/kexec-load:# Provides:              kexec-load
/etc/init.d/kexec-load:# Required-Stop: $remote_fs kexec
/etc/init.d/kexec-load:# Short-Description: Load kernel image with kexec
/etc/init.d/kexec-load:test -r /etc/default/kexec && . /etc/default/kexec
/etc/init.d/kexec-load: test -x /sbin/kexec || exit 0
/etc/init.d/kexec-load: test “x`cat /sys/kernel/kexec_loaded`y” = “x1y” && exit 0
/etc/init.d/kexec-load:         kexec -l “$KERNEL_IMAGE” –append=”$REAL_APPEND”
/etc/init.d/kexec-load:         kexec -l “$KERNEL_IMAGE” –initrd=”$INITRD” –append=”$REAL_APPEND”
/etc/init.d/kexec:# Provides:           kexec
/etc/init.d/kexec:# Short-Description: Execute the kexec -e command to reboot system
/etc/init.d/kexec:test -r /etc/default/kexec && . /etc/default/kexec
/etc/init.d/kexec:      test “x`cat /sys/kernel/kexec_loaded`y” = “x1y” || exit 0
/etc/init.d/kexec:      test -x /sbin/kexec || exit 0
/etc/init.d/kexec:      log_action_msg “Will now restart with kexec”
/etc/init.d/kexec:        kexec -e
/etc/init.d/kexec:        log_failure_msg “kexec failed”
/etc/init.d/umountroot:# Should-Stop:       halt reboot kexec
/etc/init.d/kdump:# Short-Description: Load crashkernel image with kexec
/etc/init.d/kdump:      test -x /sbin/kexec || exit 0
/etc/init.d/kdump:      kexec -p “$KERNEL_IMAGE” –initrd=”$INITRD” –append=”$APPEND”
/etc/rc2.d/S00kdump:# Short-Description: Load crashkernel image with kexec
/etc/rc2.d/S00kdump:    test -x /sbin/kexec || exit 0
/etc/rc2.d/S00kdump:    kexec -p “$KERNEL_IMAGE” –initrd=”$INITRD” –append=”$APPEND”
/etc/default/kexec:# Defaults for kexec initscript
/etc/default/kexec:# sourced by /etc/init.d/kexec and /etc/init.d/kexec-load
/etc/default/kexec:# Load a kexec kernel (true/false)
/etc/rc0.d/S60umountroot:# Should-Stop:       halt reboot kexec
apple@ubuntu:~$

看来系统重启需要借用/etc/init.d/kexec这个脚本

2.查看/etc/default/kexec文件

apple@Ubuntu:~$ cat /etc/init.d/kexec
#! /bin/sh
### BEGIN INIT INFO
# Provides:             kexec
# Required-Start:
# Required-Stop:        reboot
# X-Stop-After:         umountroot
# Default-Start:
# Default-Stop:         6
# Short-Description: Execute the kexec -e command to reboot system
# Description:
### END INIT INFO

PATH=/usr/sbin:/usr/bin:/sbin:/bin

. /lib/lsb/init-functions

test -r /etc/default/kexec && . /etc/default/kexec

do_stop () {
        test “x`cat /sys/kernel/kexec_loaded`y” = “x1y” || exit 0
        test -x /sbin/kexec || exit 0

        log_action_msg “Will now restart with kexec”
        kexec -e
        log_failure_msg “kexec failed”
}

case “$1” in
  start)
        # No-op
        ;;
  restart|reload|force-reload)
        echo “Error: argument ‘$1’ not supported” >&2
        exit 3
        ;;
  stop)
        do_stop
        ;;
  *)
        echo “Usage: $0 start|stop” >&2
        exit 3
        ;;
esac
exit 0
apple@ubuntu:~$

/etc/init.d/kexec脚本使用了/etc/default/kexec中的配置，查看/etc/default/kexec

apple@ubuntu:/$ cat /etc/default/kexec
# Defaults for kexec initscript
# sourced by /etc/init.d/kexec and /etc/init.d/kexec-load

# Load a kexec kernel (true/false)
LOAD_KEXEC=true

# Kernel and initrd image
KERNEL_IMAGE=”/vmlinuz”
INITRD=”/initrd.img”

# If empty, use current /proc/cmdline
APPEND=””
apple@ubuntu:/$ cat /etc/default/kexec
# Defaults for kexec initscript
# sourced by /etc/init.d/kexec and /etc/init.d/kexec-load

# Load a kexec kernel (true/false)
LOAD_KEXEC=true

# Kernel and initrd image
KERNEL_IMAGE=”/vmlinuz”
INITRD=”/initrd.img”

# If empty, use current /proc/cmdline
APPEND=””

原来这个/etc/default/kexec是配置重启过程中kexec的参数的

查看这两个符号链接：

lrwxrwxrwx   1 root root    30 Oct 16 00:43 vmlinuz -> /boot/vmlinuz-2.6.28-17-server
lrwxrwxrwx   1 root root    33 Oct 16 00:44 initrd.img -> /boot/initrd.img-2.6.28-17-server

原来装的内核太多了，这两个符号链接链接到错误的内核版本。
修正这个问题内核就可以正常重启了

apple@Ubuntu:/$ sudo rm /vmlinuz
apple@ubuntu:/$ sudo rm /initrd.img
apple@ubuntu:/$ sudo ln -s /boot/vm
vmcoreinfo-2.6.28-11-server   vmcoreinfo-2.6.28-3-rt        vmlinuz-2.6.28-17-generic     vmlinuz-2.6.28-3-rt
vmcoreinfo-2.6.28-17-generic  vmlinuz-2.6.28-11-server      vmlinuz-2.6.28-17-server
apple@ubuntu:/$ sudo ln -s /boot/vmlinuz-2.6.28-11-server vmlinuz
apple@ubuntu:/$ sudo ln -s /boot/initrd.img-2.6.28-11-server initrd.img
apple@ubuntu:/$