Chinaman, when you are going through the kernel configuration, lets say " make xconfig " you will need to enable 8139 support in the " network device support > Ethernet 10 / 100Mbit " device list, I can't see that DVD yet, I am sure you have just omitted it's support, if you remember the configuration guide, the author leaves most multi media stuff out, his configuration is aimed for a bare bones server machine, I'll keep looking for it though.

Chinaman, I ran through the kernel config list again, I can't see any option for your DVD player, if it works with the standard kernel it's in there some where, load the standard kernel, see if you can figure out which module is loaded for the DVD " lsmod " find the option in your config file and enable it, multi media and sound are options I disable in my kernel configuration so I am useless in that area!!

Why did I post to this thread? Fair enough, cause mrhyde regardless of where the information that you posted came from what you advocated doing in regards to symlinking the kernel headers has been known to subtly break glibc. Breaking glibc is not a good thing . Yea I read that document back in '99 cover to cover when it was Securing and Optimizing Linux Red Hat Edition. I don't know why I would email the author about a document that is not maintained (it's two years old) and not specific to the distro in question (he is using RedHat, and yes I have a disagreement with how he advocates compiling a kernel). Why did I post a quote from Linus? Because a quote from Linus may make somebody think twice about doing something that could be harmfull to their system and well Linus has been pretty religious about this issue for quite some time (here is another example).

Ok mrhyde I do want to go back to your rc.local hack. I do question if hacking a BSD init feature into a SysV init system is a "standard" practice, but I don't really care one way or the another. The only problem with it is the commands 'init 1' and 'init S' are not going to stop anything in your rc.local script which is not a standard SysV init behavior. Of course the standard way would be to write a SysV init script and add the kill links for rcS.d and rc1.d.

And no I'm not the same person as Strike.

Chinaman what kernel sources are you using? If it's from debian and ends with bf, that is the sources that is used to build the boot floppies with and it has a few drivers removed so it will fit on the floppies. Yea I think the Realtek 8139 is one of them :/ . I would grab the sources from kernel.org. Yea you need to hack a few things to get another cdrom device to work. I would help here but the device naming conventions for scsi cdroms are pretty different than for ide cdroms (the only ide cdrom system I have here is using devfs). Basiclly you touch /dvdrw (or whatever you want to name your burner) edit your /etc/fstab, copy the /cdrom line, change the mount point to /dvdrw, and change the /dev entry to the approperate device. Just as long as you have ide cdrom support it should work (I think there is also a new filesystem for dvd's but I don't have a dvd player so whatever I may be mistaken there)

Hey Stiles,

I got 2.4.21 from kernel.org about a month or so ago when I first started with Linux and ran RH9. And the original kernel I selected was from the Debian Woody NONUS binary 1, and 2.4.18-bf2.4 is what it used. I think I selected that one because it said I could add ReiserFS with it, but I'm not sure at the moment.

When compiling 2.4.21 I added SCSI support to be able to use the DVD to burn CD's. Later I will try to add the functionality of burning DVD's. If I'm not mistaken, acid_kewpie (or another mod) wrote something for that - it's somewhere in my bookmarks to check out.

When I do cdrecord -scanbus I get
debian:~# cdrecord -scanbus
Cdrecord 1.10 (i686-pc-linux-gnu) Copyright (C) 1995-2001 Jörg Schilling
cdrecord: No such file or directory. Cannot open SCSI driver.
cdrecord: For possible targets try 'cdrecord -scanbus'. Make sure you are root.

I think if I add something like append="hdd=ide-scsi" to lilo, it will recognize the DVD as a CD-RW, which will be okay for now. After running the new kernel for a bit, and rebooting about 3 times in and out of Windoze, I added append="hda=ide-scsi" and that kept it from booting. Thought I lost it, but I got back in and edited and it's okay now.

Thanks for your help. I'll check out your advice in the last paragraph later. Our Chinese tutor will be here in a bit.

mrhyde,

debian:~# lsmod
Module Size Used by
lsmod: QM_MODULES: Function not implemented


This being my second experiment with compiling a kernel, and wanting nothing more than a lean, fast comp; I built a monolithic kernel. There are no modules installed.

I enabled 8139 support in the " network device support > Ethernet 10 / 100Mbit " device list - here's what dmesg gives:

tg3.c:v1.5 (March 21, 2003)
PCI: Found IRQ 9 for device 02:05.0
eth0: Tigon3 [partno(BCM95702A20) rev 1002 PHY(5703)] (PCI:33MHz:32-bit) 10/100/1000BaseT Ethernet 00:e0:18:a3:33:b3
8139too Fast Ethernet driver 0.9.26
PCI: Found IRQ 9 for device 02:0a.0
PCI: Sharing IRQ 9 with 02:03.0
eth1: RealTek RTL8139 Fast Ethernet at 0xf881e000, 00:50:bf:6c:0d:a3, IRQ 9
eth1: Identified 8139 chip type 'RTL-8139C'

When I compiled I used a combination of the original guide mentioned, "The Very Verbose Guide to Updating and Compiling Your Debian Kernel", and notes from reading man, docs, and forum posts.

I also replied to Stiles just now. I gotta run, but will do some more work and check back later.

Thanks for your help!

I'm going out for awhile, but your ethernet card is there, it probablly just hasn't been initialized (take a look in /etc/network/interfaces and look for eth1, also 'man interfaces' will give you a reference on that file).

# /etc/network/interfaces -- configuration file for ifup(8), ifdown(8)

# The loopback interface
auto lo
iface lo inet loopback

# The first network card - this entry was created during the Debian installation
# (network, broadcast and gateway are optional)
auto eth0
iface eth0 inet static
address 192.168.0.3
netmask 255.255.255.0
network 192.168.0.0
broadcast 192.168.0.255
gateway 192.168.0.1


That's it. Looking at the man page now. I think I can fix it easy enough. Will try after lunch.

Thanks for the help.

Just add something like this (I'm assuming you want a static IP set to 192.168.0.4):

auto eth1
iface eth1 inet static
address 192.168.0.4
netmask 255.255.255.0
network 192.168.0.0
broadcast 192.168.0.255
gateway 192.168.0.1


then 'ifup eth1' and you should be good to go on your NIC.

I don't think it matters if the system sees your DVD burner as a cdrom, the only thing that is different I think is that the userspace burning programs need to be up to date and know how to burn a DVD and you will need the UDF filesystem compiled in your kernel. Yea you will need to do the append= thing and then your dvd burnner will look like a scsi device. I think it will be /dev/scd0, but I'm not sure. Actually the 2.6 kernel can burn cd's without scsi emuliation, but don't worie about that right now.

" BSD init feature " I didn't say that this was standard practice, I was talking about the administration style I was using. On BSD systems the boot scripts are refered to as rc scripts.

are you trying to say that there are not /etc/rc* directories on system V? Ok here is the output of my ultra2 (running solaris 9 for your information, OMG could that actually be a SVR4 OS????? ooooohhhhhhh, I just had too go there):

Keep on spinnning I dare ya :P

BTW what you said about linux uses in effect RedHat init is not correct as you should have figured out by now. Debian uses SysV semantics (no it is not sysV init per say my above post sheds some light on that, the /sbin specifics are quite different), gentoo introduces functions and dependancy resolution, slackware is BSD init. Ohh wow a lot of different Linux systems use different init systems.

Stiles why did you post in this thread? You are just as stupid as strike is, the two of you are really annoying me posting quotes and misinforming everyone! Just to test your knock that the processes started by rc.local will stay running when switching to level 1 I used the following on a Debian woody box;
Edit the " rc.local " file created by the steps suggested at the beginning of the thread, enter the command " /usr/sbin/cron start " Next I have removed the cron daemon " start/stop" links from the " init " runlevel directories eg rc0.d, rc1.d, rc2.d, rc3.d, rc4.d, rc5.d rc6.d at this stage I reboot the machine to be sure that it was rc.local that started cron. To list the processes I issued the command " ps -ef " cron was there running away.
Next lets switch to level 1...
# init 1
INIT: Switching to runlevel 1
INIT: Sending processes the TERM signal
INIT: Sending processes the KILL signal
I put these lines in so you can see what init is printing on the screen. After init has changed to level 1 type the root password and list processes again, no cron, init stopped it. There is no kill link to the rc.local file. I suggest stiles that you go and study your distributions documentation more closely and others too, open your mind up a bit. I have to call it a day here, my wife is unhappy I have spend time giving my wits to idiot, I am so very disappointed with all of this rubbish.

BSD doesn't use init. You are a fool.

To date I've never felt it necessary to use the "Warn User" facility on LinuxQuestions but this may change unless this thread becomes a little more civil.

I've got no problem with people disagreeing about things, and arguing their point of view but this thread appears to be going around in circles so is getting nowhere. You've already had comments from Jeremy and Fin which you seem to have ignored.

If you can't agree then just agree to disagree.

Jamie...

Of course BSD has an initialization system commonly refered to as BSD init. And yes there is a BSD init command, though it is not used to control runlevels, it is used for process control during booting. Yes slackware uses BSD init.

Do you know why cron wasn't running? Debian (and prob most versions of linux) has a script that is started when you enter init 1 called single which runs the commands 'killall5 -15' and 'killall5 -9'. Now on a solaris (solaris is bad about this) that would have left dangeling processes.

Why did make that post, ahhh it was cause I was piss drunk :P

I don't have debian and have not looked into it specifically, here is what I would suggest.


have a look at /etc/inittab

this will lead you to the boot scripts

example taken from a redhat system..

si::sysinit:/etc/rc.d/rc.sysinit