I'm running NFS root, so I had to dump scripts that mount swap for
example and I reordered the init scripts.
I also had to stop the root remount and the order between the mount
scripts and udev has changed.
The /dev/pts directory is not getting created at boot, so naturally,
this directory can't mount.
Another thing, probably related to the /dev/pts issue, I can't start
X. When I manually create /dev/pts and mount /dev/pts I can start X,
but I have to ssh in from another computer to shut X down because
there is no keyboard and no mouse control. I use a Logitech USB MK700
wireless keyboard and mouse which work at the console okay, but they
don't work in X.
I am wondering if my kernel is configured wrong for X because Xorg
-configure crashes with an error about the number of devices not
matching the number of screens or something similar.
# Begin /etc/fstab
# file system mount-point type options
dump fsck order
192.168.5.3:/nfsroot/lfs/lfsx86_64/eagle64 / nfs
rw,_netdev,rsize=8192,wsize=8192 1 1
192.168.5.3:/nfsroot/lfs/lfsx86_64/usr /usr nfs
ro,_netdev,rsize=8192 0 0
192.168.5.3:/nfsroot/lfs/lfsx86_64/home /home nfs
rw,_netdev,rsize=8192,wsize=8192 0 0
proc /proc proc
nosuid,nodev,noexec 0 0
sysfs /sys sysfs
nosuid,nodev,noexec 0 0
devpts /dev/pts devpts
gid=4,mode=620 0 0
tmpfs /run tmpfs defaults
0 0
devtmpfs /dev devtmpfs mode=0755,nosuid
0 0
192.168.5.3:/backup/eagle64 /backup nfs
rw,_netdev,rsize=8192,wsize=8192 0 0
192.168.5.3:/md0/eagle64 /backup/more nfs
rw,_netdev,rsize=8192,wsize=8192 0 0
# End /etc/fstab
Here is a directory listing of my links in rc.S and rc3.d to the init scripts:
michael [ /etc/rc.d/rcS.d ]$ ls -l
total 0
lrwxrwxrwx 1 root root 21 Aug 30 08:43 S01mountvirtfs -> ../init.d/mountvirtfs
lrwxrwxrwx 1 root root 17 Aug 25 23:37 S03mountfs -> ../init.d/mountfs
lrwxrwxrwx 1 root root 14 Aug 30 08:44 S05udev -> ../init.d/udev
lrwxrwxrwx 1 root root 17 Aug 25 23:37 S07modules -> ../init.d/modules
lrwxrwxrwx 1 root root 17 Aug 25 23:37 S45cleanfs -> ../init.d/cleanfs
lrwxrwxrwx 1 root root 20 Aug 25 23:37 S50udev_retry -> ../init.d/udev_retry
lrwxrwxrwx 1 root root 17 Aug 25 23:37 S70console -> ../init.d/console
lrwxrwxrwx 1 root root 17 Aug 26 17:02 S85sysstat -> ../init.d/sysstat
lrwxrwxrwx 1 root root 16 Aug 25 23:37 S90sysctl -> ../init.d/sysctl
michael [ /etc/rc.d/rcS.d ]$
michael [ /etc/rc.d/rc3.d ]$ ls -l
total 0
lrwxrwxrwx 1 root root 21 Aug 25 23:37 S00mountvirtfs -> ../init.d/mountvirtfs
lrwxrwxrwx 1 root root 17 Aug 26 00:15 S02mountfs -> ../init.d/mountfs
lrwxrwxrwx 1 root root 18 Aug 25 23:37 S10sysklogd -> ../init.d/sysklogd
lrwxrwxrwx 1 root root 18 Aug 28 02:31 S19iptables -> ../init.d/iptables
lrwxrwxrwx 1 root root 17 Aug 25 23:37 S22rpcbind -> ../init.d/rpcbind
lrwxrwxrwx 1 root root 20 Aug 25 23:37 S24nfs-client -> ../init.d/nfs-client
lrwxrwxrwx 1 root root 16 Aug 28 01:02 S25random -> ../init.d/random
lrwxrwxrwx 1 root root 15 Aug 25 23:38 S28netfs -> ../init.d/netfs
lrwxrwxrwx 1 root root 14 Aug 29 01:29 S29dbus -> ../init.d/dbus
lrwxrwxrwx 1 root root 14 Aug 26 18:52 S30sshd -> ../init.d/sshd
lrwxrwxrwx 1 root root 15 Aug 26 17:52 S40fcron -> ../init.d/fcron
lrwxrwxrwx 1 root root 20 Aug 26 18:47 S50ez-ntpc.sh -> ../init.d/ez-ntpc.sh
lrwxrwxrwx 1 root root 13 Aug 28 22:47 S70gpm -> ../init.d/gpm
michael [ /etc/rc.d/rc3.d ]$
michael [ /etc/rc.d/init.d ]$ ls
checkfs functions mountfs rc swap udev_retry
cleanfs gpm mountvirtfs reboot sysctl
console halt netfs rpcbind sysklogd
dbus iptables network sendsignals sysstat
ez-ntpc.sh localnet nfs-client setclock template
fcron modules random sshd udev
michael [ /etc/rc.d/init.d ]$
michael [ ~ ]$ cat init_scripts.md5
d6684c5e107b5e60bedfcd2e8a6de6ad checkfs
24a49e1a0627697f5c6c251a3634b9b3 cleanfs
418690da6bd4a55f4ad5bc3cf16c48b8 console
ceb3291a49f28cdf491ba6ac581a9271 dbus
a27e6b5f04d82f507e1f7609932f4c67 ez-ntpc.sh
c0cb5db7c93f40fbf2a7415841dcc75c fcron
46b19555629c098951794bedd7c5938e functions
dd9f264f16cc003a3da4f7634769090c gpm
346f4d9d895bd97ecc571b71f9646de0 halt
811dda2c333bebdd2b06e64c3cdf7359 iptables
0fd911e9bd4aae767d1ff13fc2fb54a3 localnet
f59706a3a5bd3f35ad8677921865afaf modules
438122e84f2161a196cb0f29af430769 mountfs
053ceb6322e46e76799e338f53d9ce94 mountvirtfs
ba42be3b0fc91d2f12bba6fc621101a9 netfs
b5f9911941b2db2cddef068a89488fa9 network
e1ca7b897728444a3a2b70362917b7c7 nfs-client
91601547bcf1004ebe1789baa206ac85 random
0ed75f192960ceaa674703750b0e4f40 rc
d9d088591b5c9fd3bbe2a9162d417b9c reboot
16876d1f76a0eed202e4a0e67274dea5 rpcbind
06a134630979c337f77f6143042f6910 sendsignals
da80ecd227b70f5795dbad717e808edb setclock
77bd4daaa7de53ad7c8a5149b7e6a85f sshd
de90ac9772da32a9ff0497f2f9ced620 swap
416352a69f03ca117141b4af4e9e73dc sysctl
26b18a8829d7a1f0d8f35b2d0e38612c sysklogd
3882fb52b06782910ae4075b09c6e0c2 sysstat
98d845dcec861728c219df8d78891b5f template
6593e5097e6cbca1bcdd125bbb0a68c9 udev
0fd63a6ee25ca73aefb875b803d31e08 udev_retry
michael [ ~ ]$
Someone if they don't have an obvious answer should be able to tell me
which script I modified from the md5 sums
and from there I can look at the changes made. I started with the LFS
7.1 book.
I recompiled udev per instructions in BLFS for LFS 7.1, but that
hasn't fixed this problem.
1 I messed up the init scripts and they are the source of the problem.
2 My kernel is the source of the problem because it is configured wrong.
3 My fstab is wrong.
4 Some other directory created for udev or a file created for udev is wrong.
What I like is that my LFS build is tighter than a typical commercial
Linux distribution. That is becoming
less and less the case though as I try to satisfy all of the
dependencies for X. My /usr partition is mushrooming to 5+ gigs. My
goal is to use this Linux system as a network booted recovery
environment, a swiss army knife that can be used to: backup, restore,
and rebuild computers. I'm surprised there isn't a custom open source
OS that can be network booted already put together. Commercial Linux
distributions are designed for installation to computers with a hard
disk, totally inappropriate in my situation.
As an aside:
A major disadvantage of not having package management is that it is
harder to say, test this package and tell me if I need to rebuild and
reinstall it because something that didn't exist before exists now and
I should take advantage of that. I'm used to rpm, but slackware style
package management is simpler.
-- Michael C. Robinson
--
http://linuxfromscratch.org/mailman/listinfo/blfs-support
FAQ: http://www.linuxfromscratch.org/blfs/faq.html
Unsubscribe: See the above information page
