Linux-Misc Digest #104, Volume #25 Tue, 11 Jul 00 07:13:05 EDT
Contents:
HELP with install... ("Jérôme Meyer")
Re: 1280x1024 Resolution (Valentin Guillen)
Re: ATI Rage Mobility and RH6.1 (Allen Ashley)
Re: PPP dialin with uugetty (M. Buchenrieder)
Re: Kernel too big (gLiTcH)
Re: Kernel too big (gLiTcH)
Interesting behavior of LILO ([EMAIL PROTECTED])
Re: Please Help W/My Various Problems (lost passwd, x, "LI", and just a couple more)
(Eric)
problems compiling 2.4.0-test3? (Tyler Durden)
Re: softraid problem with SuSE 6.4 and 2.2.x kernel (Johannes Niess)
Re: New HDD and Partion Magic or second PC. ("G-Man")
Re: Ethertap Devices ([EMAIL PROTECTED])
----------------------------------------------------------------------------
From: "Jérôme Meyer" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setu,alt.os.linux
Subject: HELP with install...
Date: Tue, 11 Jul 2000 11:35:18 +0200
Reply-To: [EMAIL PROTECTED]
I will install on my PC both Linux(Suse6.4) and Windows98 with
the size ~10GB for Windows and ~15GB for Linux.
I've just few question about the installation:
- What I should install first Windows or Linux (easy way)?
- I cannot understand how we can choose (on starting) Windows or Linux.
Because Linux
use the LILO boot system and not windows. How could I configure this
features?
Thanks a lot for your help,
Regards,
JM
------------------------------
From: Valentin Guillen <[EMAIL PROTECTED]>
Subject: Re: 1280x1024 Resolution
Date: Tue, 11 Jul 2000 01:20:35 -0600
johnvert,
I agree with another poster that you need to become familiar with the
man pages for XF86Config. This is because you need to understand how
the different parts of the file interact with each other and how the
file interacts with X.
Then, armed with this knowlege, you can then fire up your vi and edit
the file. First thing you need to do is remove all the blasted remarks
which many config utils like to plaster all over the file. How can one
possibly deal with the file in this condition of 10 remark lines for
every 1 valid line?
Anyway, the modelines which are not commented out, in order to be valid,
must concord with the frequency limits set earlier in the file. If they
don't agree, they are parsed out, or 'ignored' when X is firing up and
parsing it's config file. Of course, to set the frequency limits, one
needs that info from the manufacturer of the monitor. Lacking this
info, one can either extrapolate the data from similar monitors, or one
can use the standard VESA 2.x frequencies. Always better to use the
exact data, but we don't live in a perfect world.
So then, you need to have the data for your monitor or use standardized
data for defining the monitor's characteristics. Then, you need the
specs for your vid card, because you need to know the performance
abilities, in order to correctly set up the refresh rates, color depths,
and resolutions.
So, once a person has defined the frequency limits, and completed the
vid config process, you'll have an XF86Config file which will list all
manner of modelines for various valid (or invalid) resolutions and
refresh timings. You can then go in and comment out/delete nonsense
which you don't want cluttering up your file. I define nonsense as
things like double scan mode lines. No one in their correct mind would
ever invoke those modes, so why have them stuffed all over the config
file, slowing down the reading of the file, and doing nothing
constructive. Likewise with modelines which define refresh frequencies
which are so god awful un-ergonomic that you'll never conciously choose
to use them.
Hello.... anybody there? Modeline "512x384" 512x WHAT? Do you really
intend to use this resolution?
Excuse me, but 1800x1440 res? I don't think so! Or maybe this is the
one you really want: 320x200???
So then here is what that section you posted could look like, deleting
some of the extraneous stuff. This section below still has some
invalid entries, and we'll look at a couple of them to see why they're
invalid.
#
**********************************************************************
# Monitor section
#
**********************************************************************
Section "Monitor"
Identifier "ViewSonic E771"
VendorName "Unknown"
ModelName "Unknown"
HorizSync 30-70.25
VertRefresh 50-120
# 800x600 @ 56 Hz, 35.15 kHz hsync
# 640x480 @ 60 Hz, 31.5 kHz hsync
# 800x600 @ 60 Hz, 37.8 kHz hsync
# 1024x768 @ 60 Hz, 48.4 kHz hsync
# 1152x864 @ 60 Hz, 53.5 kHz hsync
# 1280x1024 @ 61 Hz, 64.2 kHz hsync
# 1024x768 @ 70 Hz, 56.5 kHz hsync
# 1152x864 @ 70 Hz, 62.4 kHz hsync
# 1280x1024 @ 70 Hz, 74.59 kHz hsync
# 640x480 @ 72 Hz, 36.5 kHz hsync
# 800x600 @ 72 Hz, 48.0 kHz hsync
# 640x480 @ 75 Hz, 37.50 kHz hsync
# 1024x768 @ 76 Hz, 62.5 kHz hsync
# 1152x864 @ 78 Hz, 70.8 kHz hsync
1280x1024 @ 74 Hz, 78.85 kHz hsync #This is the one you want!
# 1280x1024 @ 76 Hz, 81.13 kHz hsync
# 640x480 @ 85 Hz, 43.27 kHz hsync
# 800x600 @ 85 Hz, 55.84 kHz hsync
1024x768 @ 85 Hz, 70.24 kHz hsync #This is the one you want!
# 1152x864 @ 84 Hz, 76.0 kHz hsync
# 1280x1024 @ 85 Hz, 91.15 kHz hsync
# 1024x768 @ 87 Hz interlaced, 35.5 kHz hsync
# 1280x1024 @ 87 Hz interlaced, 51 kHz hsync
# 1152x864 @ 89 Hz interlaced, 44 kHz hsync
# 640x480 @ 100 Hz, 53.01 kHz hsync
800x600 @ 100 Hz, 64.02 kHz hsync #This is the one you want!
# 1024x768 @ 100 Hz, 80.21 kHz hsync
# 1152x864 @ 100 Hz, 89.62 kHz hsync
# 1280x1024 @ 100 Hz, 107.16 kHz hsync
EndSection
*****************************************************************
We'll assume that this is correct:
HorizSync 30-70
VertRefresh 50-120
So, there's a hash mark at the beginning of each these lines. This
means that it's valueless, because it's just a remark. Removing the
hash mark renders the line a file entry. You would select the
resolutions to enable by removing the hash mark for JUST ONE line per
resolution. Thus, if you only want to be able to switch between 800x,
1024x, and 1280x, you would never bother to uncomment out ANY line for
640x resolution, out of the many available at the various frequencies.
So, example of an invalid entry: # 1280x1024 @ 85 Hz, 91.15 kHz hsync
This is an invalid entry is would be parsed out of validity for this
reason:
The frequency of this modeline for horizontal sync is 91.15 kHz, which
is outside of the range we specified as being between 30 and 70 kHz.
Therefore, if you were to uncomment this line, because you wanted to use
this res (1280x at an ergonomic vertical refresh rate of 85hz), you
would invalidate this resolution, and it wouldn't be available to you.
Another thing, 16 bit color is considerably faster to display than 24
bit, and is indistinguishable for nearly any purpose. Stay away from
deeper color. Be quite certain of the frequencies your monitor can
handle before specifying them in this file or in the vid config util.
Either that or stay in the VESA compliant range of 70~75 hz when
stipulating the value. You CAN damage your monitor by using incorrect
data, usually when attempting to kick up the refresh rates to higher
values.
Interlaced? Usually the interlaced resolutions are more flickery, but
they allow you resolutions your hardware ordinarily couldn't do. I
wouldn't choose to use them. I would also delete all entries which are
invalid or out of range for your hardware. This simplifies your file,
and speeds up the rate at which X can parse it when firing up.
Hope this stuff is useful
Valentin Guillen
------------------------------
From: [EMAIL PROTECTED] (Allen Ashley)
Crossposted-To: linux.redhat.misc;
Subject: Re: ATI Rage Mobility and RH6.1
Date: 11 Jul 2000 10:11:50 GMT
If your chip is a Rage 128 then it is fully supported by
SuSE 6.4 and nearly so with Mandrake 7.1. It is supported
by release 4 of the xf86 server. Mandrake is missing frame
buffer console support for the Rage 128, but it can be
cobbed together.
------------------------------
Crossposted-To: comp.os.linux.networking
From: [EMAIL PROTECTED] (M. Buchenrieder)
Subject: Re: PPP dialin with uugetty
Date: Tue, 11 Jul 2000 07:06:02 GMT
"Anurodh" <[EMAIL PROTECTED]> writes:
[...]
>When i reboot or type init q, I keep on getting an error saying:
>Id S1 respawining too fast. Disabled for 5 minutes
The invocation of uugetty (if you're actually starting uugetty)
is failing; the kernel tries to restart it and it fails again, etc.
[...]
>in /etc , i added the following line to inittab:
>S1:2345:respawn:/sbin/uugetty -D /etc/conf.uugety.ttyS1 ttyS1 f38400
^^^ ^^^
Typos.
Besides that; do not use uugetty. It's obsolete, and you'd be better
off using mgetty for modems, anyways.
[...]
># restarted so that the modem is re-initialized
>ALTLOCK=cua3
>
[...]
Argh. Where did you find this? Recent versions don't even have
the /dev/cua* devices anymore.
Junk all that and install mgetty on ttyS1.
Michael
--
Michael Buchenrieder * [EMAIL PROTECTED] * http://www.muc.de/~mibu
Lumber Cartel Unit #456 (TINLC) & Official Netscum
Note: If you want me to send you email, don't munge your address.
------------------------------
Date: Tue, 11 Jul 2000 06:27:39 -0500
From: gLiTcH <[EMAIL PROTECTED]>
Subject: Re: Kernel too big
Well, if your kernel is small enough, and most peoples' are, zImage will work
just fine. Only if you overload the kernel when you could use modules does this
error occur. I've only had the problem once because I don't need much in my
kernels; mostly just bare essentials.
Brandon
Victor wrote:
> Ooh, I see. So the problem is with the compression routines that zImage
> uses? Ok. bzImage suits me just fine as well, I was just curious why it
> failed. They should update the output make menuconfig makes because the
> instruction it gave was make zImage.
>
> Anyway, thanks for the heads up.
>
> "Dances With Crows" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > On Sun, 09 Jul 2000 02:17:50 GMT, Victor
> > <<ihR95.29514$[EMAIL PROTECTED]>> shouted forth into the
> ether:
> > >System is 580 kB
> > >System is too big. Try using bzImage or modules.
> > >The kernel only includes networking and one sound module. Why would it be
> > >too big. make bzImage worked fine. I am just curious why this error
> > >occurs??? I've seen on IRIX a kernel can be 10 megs. My bsd kernel was 2
> > >megs. Why is this error happening? Is there a setting I can set?
> >
> > zImage and bzImage use two different methods of loading themselves into
> > memory when they are read directly from disk during the boot process. The
> > method used for zImage has a size limit of 512K for the compressed
> > image. BTW, zImage itself will be deprecated soon--check the latest
> > Kernel Traffic at http://kt.linuxcare.com/latest.epl to hear what the
> > developers have to say.
> >
> > Also, I've noticed that LILO version 21 can't seem to deal well with
> > kernels made with zImage... says "kernel too big". Kernels made with
> > bzImage work fine. Strange, but I can live with it....
> >
> > --
> > Matt G / Dances With Crows /\ "Man could not stare too long at the
> face
> > \----[this space for rent]-----/ \ of the Computer or her children and
> still
> > \There is no Darkness in Eternity \ remain as Man." --David Zindell "So
> did
> > But only Light too dim for us to see\ they become Gods, or
> Usenetters?" --/me
------------------------------
Date: Tue, 11 Jul 2000 06:28:01 -0500
From: gLiTcH <[EMAIL PROTECTED]>
Subject: Re: Kernel too big
Well, if your kernel is small enough, and most peoples' are, zImage will work
just fine. Only if you overload the kernel when you could use modules does this
error occur. I've only had the problem once because I don't need much in my
kernels; mostly just bare essentials.
Brandon
Victor wrote:
> Ooh, I see. So the problem is with the compression routines that zImage
> uses? Ok. bzImage suits me just fine as well, I was just curious why it
> failed. They should update the output make menuconfig makes because the
> instruction it gave was make zImage.
>
> Anyway, thanks for the heads up.
>
> "Dances With Crows" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > On Sun, 09 Jul 2000 02:17:50 GMT, Victor
> > <<ihR95.29514$[EMAIL PROTECTED]>> shouted forth into the
> ether:
> > >System is 580 kB
> > >System is too big. Try using bzImage or modules.
> > >The kernel only includes networking and one sound module. Why would it be
> > >too big. make bzImage worked fine. I am just curious why this error
> > >occurs??? I've seen on IRIX a kernel can be 10 megs. My bsd kernel was 2
> > >megs. Why is this error happening? Is there a setting I can set?
> >
> > zImage and bzImage use two different methods of loading themselves into
> > memory when they are read directly from disk during the boot process. The
> > method used for zImage has a size limit of 512K for the compressed
> > image. BTW, zImage itself will be deprecated soon--check the latest
> > Kernel Traffic at http://kt.linuxcare.com/latest.epl to hear what the
> > developers have to say.
> >
> > Also, I've noticed that LILO version 21 can't seem to deal well with
> > kernels made with zImage... says "kernel too big". Kernels made with
> > bzImage work fine. Strange, but I can live with it....
> >
> > --
> > Matt G / Dances With Crows /\ "Man could not stare too long at the
> face
> > \----[this space for rent]-----/ \ of the Computer or her children and
> still
> > \There is no Darkness in Eternity \ remain as Man." --David Zindell "So
> did
> > But only Light too dim for us to see\ they become Gods, or
> Usenetters?" --/me
------------------------------
From: [EMAIL PROTECTED]
Subject: Interesting behavior of LILO
Date: Tue, 11 Jul 2000 10:18:52 GMT
Hi,
I am using LILO v0.21. I've found an interesting behavior.
I have only one hard disk and the partitions are:
hda1 hda2 < hda5 hda6 >
Linux is installed in hda5. If I install the LILO boot
loader into hda2 and set hda2 as the active partition,
it hangs at "LI" when booting. But if I install the loader
into the MBR, it works fine.
Can anyone explain why the location of the LILO boot
loader can make such a difference? Note that the kernel
itself is in hda5.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
Crossposted-To:
comp.os.linux.hardware,comp.os.linux.help,comp.os.linux.questions,comp.os.linux.redhat,comp.os.linux.setup
From: Eric <[EMAIL PROTECTED]>
Subject: Re: Please Help W/My Various Problems (lost passwd, x, "LI", and just a
couple more)
Date: Tue, 11 Jul 2000 09:39:22 GMT
Reply-To: [EMAIL PROTECTED]
!!DON'T CROSSPOST TO SO MANY GROUPS!!!
I've put in two follow-ups for you.
Blades RipRock wrote:
>
> Greetings,
>
> First, the specs and the details: PIII 600, 128MB RAM, 20GB HD, TNT2 64
> (32MB), Sblaster Live! Value, Aztech v.90 modem....Trying to install Linux
> (redhat 6.2) on my buddy's system (on the same HD which has win98 - yes, I
> have fdisk'd so that Linux has it's own partitions), here are the problems,
> please help me.
>
> 1. Once the installation process is over and it reboots, instead of the
> reassuring LILO prompt, I only got "LI". How can I resolve this? Does the
> HD have to be set as "normal", "LBA" or something else in BIOS?
No, you need not to change anything in the bios, you need a /boot
partition the bios can reach, ie. either make one on HDD (thus below
cyl. 1024) or use a floppy to boot from. After you've linux booted, you
can upgrade lilo to a newer version that can boot from beyond cyl 1024.
(If your bios supports it!)
> 2. When I tried to configure X, when it came to testing out the X
> configuration, once the screen goes blank to test, I'd get this floating
> bubble saying, "sync out of range". I tried to change the v&h sync ranges
> to various numbers, but I'd still get the same result. The monitor's from
> Dell.
Just try some more, Dell has monitors in the default list, choose the
monitor you have, or if it's not there set u manually using the specs
from your monitor
> 3a. How can I get into my own box for which I lost the password to root and
> the users I created (yeah, they both were same)? Someone told me to boot
> into it in single mode, but I don't know how to do that and also, once in
> single mode, I understand I'd have to mount the partitions correctly so
> that I can write to the files which need to be changed. Can someone please
> kindly instruct me how?
It'll all be mounted just nicely. At the liloprompt enter `linux single`
and unless you have this protected with a passwd (default setup leaves
this unprotected) you'll be the root user.
>
> 4. For those Redhat 6.2 users, once the installation process was and it was
> rebooting, it dawned on me that it never asked me or I missed the part
> about where LILO should be installed. If I did indeed miss it, where in
> the installation process should I be expecting it?
Where it asks you if you want it to install in the MBR, if you want a
boot floppy (YES you want one!) IIRC it's right after you created the
partitions.
> 5. How can I get sound to work with SBLive! Value?
By getting the emu10k module.
Lots of fun, :-)
Eric
>
> TIA! Please help!
>
> Regards,
> RipRock
------------------------------
From: Tyler Durden <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux
Subject: problems compiling 2.4.0-test3?
Date: Tue, 11 Jul 2000 10:39:23 GMT
Is anyone else having problems trying to compile the 2.4.0-test3 (not 2)
kernel? It's failing on make bzImage when trying to compile timer.c
with the following error messages:
make[2]: Entering directory `/usr/src/linux-2.4.0-test3/kernel'
gcc -D__KERNEL__ -I/usr/src/linux-2.4.0-test3/include [...] -c -o
timer.o timer.c
timer.c: In function `update_process_times':
timer.c:580: structure has no member named `priority'
timer.c:580: `DEF_PRIORITY' undeclared (first use in this function)
[...]
make[2]: *** [timer.o] Error 1
make[2]: Leaving directory `/usr/src/linux-2.4.0-test3/kernel'
make[1]: *** [first_rule] Error 2
make[1]: Leaving directory `/usr/src/linux-2.4.0-test3/kernel'
make: *** [_dir_kernel] Error 2
I initially patched the 2.4.0-test2 source tree, so I also tried
downloading the entire source, but I get the same error message either
way. I don't speak C, so I can't go in and try to fix whatever it is
that's fucking up.
Anyone got any bright ideas?
--Tyler
------------------------------
From: Johannes Niess <[EMAIL PROTECTED]>
Crossposted-To:
tu-graz.betriebssysteme.linux,de.comp.os.unix.linux.misc,linux.dev.raid,de.comp.os.unix.linux.hardware
Subject: Re: softraid problem with SuSE 6.4 and 2.2.x kernel
Date: 11 Jul 2000 12:54:49 +0200
"Clemens Ender" <[EMAIL PROTECTED]> writes:
> hi,
>
> I'm using SuSE 6.4 Distribution and the 2.2.16-kernel. When I compiled the
> kernel I enabled RAID-option and built raid-support as modules.
> I also have raidtools 0.9 installed.
> I did NOT patch the kernel or something.
Clemens,
you are using the kernel portion of raid version 0.4 with raidtools
0.9. It does not fit. I'd suggest patching your kernel to raid
0.9. You can find the patches somewhere on the redhat server.
Johannes Nie�
------------------------------
From: "G-Man" <[EMAIL PROTECTED]>
Subject: Re: New HDD and Partion Magic or second PC.
Date: Tue, 11 Jul 2000 11:01:57 GMT
Which way should I go, New hard drive and partition magic or buy a second
computer? To install Linux on.
=G=
Juergen Leeb <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> You need help, but what is your question?
>
> G-Man schrieb:
> >
> > I want to install Mandrake 7.0. I am trying to decide if I should get a
> > second hard drive and partition magic and run a Windows98 C: Drive and a
> > dedicated second HDD for Mandrake 7.0. Or, do I save the $$$ and buy a
> > 233MHZ Pentium or similar AMD system and install it there? I can
probably
> > get an inexpensive system for around $400.00 with monitor. I figure with
2
> > systems, I can use one to access linuxdoc.org or linux-mandrake.com and
get
> > help if needed without having to reboot all the time. FWIW I run a PII
266
> > with 6.4 GB HDD 192 MB SDRAM.
> > Thanks for the help,
> > =G=
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To:
comp.os.linux.networking,comp.os.linux.development.system,comp.os.linux.development.apps
Subject: Re: Ethertap Devices
Date: Tue, 11 Jul 2000 11:05:17 GMT
In comp.os.linux.development.system Tim Godfrey <[EMAIL PROTECTED]> wrote:
|> Do you have this option compiled into your kernel?
|>
|> Kernel/User netlink socket (CONFIG_NETLINK) [Y/n/?] ?
|>
|
| Yes indeed I do...
Are you loading anything by module? Or have you compiled everything into
the kernel directly like I do? If you compile everything in, could you
maybe send me a copy of your kernel .config file? I could see exactly
what is different between yours and mine and see if any differences
suggest to me something that could be a problem.
| I have also tried opening /dev/tap0 as O_RDWR.
|
| Really stumped....
Has me stumped, too ... except that when I get stumped on my own machines
I'm usually trying a whole lot of things quite rapidly to see if anything
even sheds clues.
Focusing on the fact that your message says a resource is busy, I would
wonder if something opened the /dev/tap0 device. I can see that on my
own machine:
root@izar:/root 483> lsof -n | fgrep tap
readpacke 21835 root 3u CHR 36,16 8213 /dev/tap0
root@izar:/root 484>
You probably already looked around for anything like that, even though most
likely you have nothing that could have opened it.
I might look around deeper in the netlink device handling code to see what
kinds of things it could be testing for busy conditions and checking to see
if those conditions might exist. In the mean time, you might want to give
a shot at seeing if my little readpacket program will by some chance make
a difference. You can grab it at: http://phil.ipal.org/readpacket.tar.gz
It was something I used to do some testing of ethertap to see what I get
out, and I got stuff like:
root@izar:/root 517> ifconfig tap0
tap0 Link encap:Ethernet HWaddr FE:FD:00:00:00:00
inet addr:10.1.1.1 Bcast:10.1.255.255 Mask:255.255.0.0
UP BROADCAST RUNNING NOARP MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:36 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
root@izar:/root 518> route add -host 10.2.3.4 gw 10.1.1.1
root@izar:/root 519> ( ( sleep 4 ; date | sock -u 10.2.3.4 4660 ) & )
root@izar:/root 520> /root/readpacket
======================================== 73 bytes
01000000 00000000 fefd0000 00000800 45000039 9e840000 4011c428 0a010101
0a020304 04011234 00252e4f 54756520 4a756c20 31312030 353a3539 3a333920
43445420 32303030 0a
root@izar:/root 521>
--
| Phil Howard - KA9WGN | My current websites: linuxhomepage.com, ham.org
| phil (at) ipal.net +----------------------------------------------------
| Dallas - Texas - USA | [EMAIL PROTECTED]
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.misc) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Misc Digest
******************************
