Linux-Misc Digest #585, Volume #27 Wed, 11 Apr 01 18:13:03 EDT
Contents:
Re: Using Sed : need of little help ("Jason C. Hill")
winking screen (Mark Fardal)
Resolution
Re: How to install RedHat-6.2 after RPM 4.0 upgrade?? ("Robert A. Knop Jr.")
Re: "sort" command in linux and unix (Dave Brown)
Re: (X == Moby_Dick)?...:... (David E. Fox)
Re: Clock skew on compiles.... (Ian Northeast)
Re: Won't boot: deleted libc.so.6 (mark)
Re: Using Sed : need of little help ("John W. Krahn")
Re: Compiler not working (David E. Fox)
Re: Keep fetchmail running as daemon for certain user (David E. Fox)
Re: [HELP] mount cdrom ("Kenny@BUI")
Re: maximal mount count? (Jean-Francois Landry)
Re: (X == Moby_Dick)?...:... ([EMAIL PROTECTED])
LD_PRELOAD and setuid programs (ELF) (Timothy J. Lee)
killing a process (Mladen Gavrilovic)
Re: A question on Newsgroup (Greg Howland)
Re: LD_PRELOAD and setuid programs (ELF) ("Peter T. Breuer")
Re: killing a process ("Peter T. Breuer")
Re: Xconfigurator help. (Craig Kelley)
Re: Mandrake 7.2 and Nvidia - A little bit closer (Craig Kelley)
----------------------------------------------------------------------------
From: "Jason C. Hill" <[EMAIL PROTECTED]>
Subject: Re: Using Sed : need of little help
Date: Wed, 11 Apr 2001 15:22:26 -0400
Except you so conveniently chopped off the rest of my message where I stated
if he wanted simple sed here is his answer.
So, please, stop trying to undercut me when you do such a horrible job at
it.
-J
------------------------------
Crossposted-To: comp.os.linux.hardware
Subject: winking screen
From: Mark Fardal <[EMAIL PROTECTED]>
Date: 11 Apr 2001 15:29:47 -0400
Hi,
My monitor screen has started winking at me. I like to think it's
because my sex appeal has increased... It's not a blank screen, or a
flickering screen, or a blinking screen. It just turns black for
about a second. At the same time, I hear a rapid whirring sound
that sounds like it's coming from the direction of my computer or my
UPS power supply; both are on the floor. (Hard to change the position
around to make sure, unfortunately.) The operation of the computer
itself is not affected in any way. Nothing in particular appears in
my console log window.
This winking maybe happens a few times a day. I've wiggled all the
cables with no result, so it's probably not a loose connection. This
started a few weeks ago, but I don't know of anything that changed on
my system then. The winks seem to be clustered--a few in an hour, then
none for several more.
some info that may or not be informative:
Nokia Multigraph 445Xpro monitor
Accelerated-X display server
uname -a gives
Linux weka 2.2.16 #1 SMP Fri Jul 14 10:44:25 EDT 2000 i686 unknown
Debian linux 2.2
Any ideas for how to fix this? It's not a problem now, but I wonder if
it indicates something that could get worse.
thanks,
Mark Fardal
UMass
------------------------------
From: <[EMAIL PROTECTED]>
Subject: Resolution
Date: Wed, 11 Apr 2001 19:30:17 -0000
How do you change the resolution on RedHat 6?
--
Posted via CNET Help.com
http://www.help.com/
------------------------------
From: "Robert A. Knop Jr." <[EMAIL PROTECTED]>
Subject: Re: How to install RedHat-6.2 after RPM 4.0 upgrade??
Date: Wed, 11 Apr 2001 12:30:04 -0700
In article
<[EMAIL PROTECTED]>,
"tech2kjason" <[EMAIL PROTECTED]> wrote:
> When I created new updated CD's for my RH7, it included an update for
> the binary rpm that Anaconda executes during the install process ( from
> inside the "redhat/instimage/usr/bin/" directory... Once I replaced that
> file the new CD's were able to install all of the updated packages
> (including the experimental ones that I kept from Rawhide) without any
> problems... Anyway, it's just a possibilty. If you can locate the binary
> "rpm" program that Anaconda executes, you may be able to use this
> apporach... especially since all Anaconda wants is have a binary to run.
I think I managed to get it working. It seems with RedHat 6.2 that
anaconda doesn't actually use an rpm binary. It seems to use the
rpmmodule.so python module. (Is that proper terminology? I haven't yet
spent the time to really learn python. Someday.)
I was able to get it to work by sort of stealing the python module from
the rpm-python-4.0.2 RPM. However, I had to recompile it. I used the
rpm SRPM, and hacked the Makefile in the directory where rpmmodule.so was
built so that it linked statically with the rpm-4 and db-3.1 libraries.
(Static, so that I didn't have to mess with getting the right dynamic
libraries on the redhat-6.2 CD.) I then made a new version of the
anaconda updates disk, replacing the rpmmodule.so on that disk with my
version. Things now seem to work.
I should probably write this up on a web page somewhere.
-Rob
------------------------------
From: [EMAIL PROTECTED] (Dave Brown)
Subject: Re: "sort" command in linux and unix
Reply-To: [EMAIL PROTECTED]
Date: 11 Apr 2001 15:04:11 -0500
In article <2G1B6.2334$[EMAIL PROTECTED]>, John Wingate wrote:
>Dave Brown <[EMAIL PROTECTED]> wrote:
>> I've seen this also, but in trying to reproduce it, now it seems that
>> sort ignores the LC_ALL or LC_COLLATE values unless invoking it with
>> the "-l" option. (per the info page).
>
>> I don't know why I've previously seen the "sort" work like "sort -l",
>> but I'm sure I have.
>
>What version of sort do you have? Mine reports: sort (GNU textutils) 2.0e,
>and it doesn't have a -l option.
Alas, I have v 2.0a. It's what RH 6.2 put on my machine.
Perhaps I was on a RH 7.0 machine when I saw the default behavior using
LC_ALL or whatever.
--
Dave Brown Austin, TX
------------------------------
From: David E. Fox <[EMAIL PROTECTED]>
Subject: Re: (X == Moby_Dick)?...:...
Date: Wed, 11 Apr 2001 13:02:41 -0700
Reply-To: [EMAIL PROTECTED]
Billy Bob Jameson wrote:
> Hi.
>
> I did a top + M (sort by memory usage) and the first thing on top of the
> list ix X with 80 MB. Is there anyone out there that can tell me more
What fields were you looking at (Virtual vs. RSS)? The second is the
resident set size, or the actual amount of RAM the process is using at that
point in time.
The first (VSZ) is the process's requested memory. Sometimes it seems this
value is overinflated. X (4.0.2) is currently using 42 Megs of VSZ but only
13 megs of RSS, according to 'ps'. A VSZ of 42 megs (for example) really
doesn't mean the process is actually using that much RAM -- only that it
might want to use that much, or that is what it thinks it's using. That's
probably not quite the best or most technical explanation.
80 megs does seem to be a lot, but it depends on what version of X you are
running, what video hardware you are using (I think that a portion of X's
memory footprint represents the video card memory, so if you have a card
with a lot of video ram, your X may look like it's using a lot of RAM) and
how long your X session has been up, as well as how many apps/windows you
are displaying.
--
========================================================================
David E. Fox Thanks for letting me
[EMAIL PROTECTED] change magnetic patterns
[EMAIL PROTECTED] on your hard disk.
=======================================================================
------------------------------
From: Ian Northeast <[EMAIL PROTECTED]>
Subject: Re: Clock skew on compiles....
Date: Wed, 11 Apr 2001 21:03:57 +0100
Bill Unruh wrote:
>
> In <[EMAIL PROTECTED]> dave <[EMAIL PROTECTED]> writes:
>
> >How do I fix this?
>
> >make: *** Warning: Clock skew detected. Your build may be incomplete.
>
> You changed your clock. the compiler found files which were altered in
> the future ( Ie at a time later than the time at which the compiler was
> running). This usually happens because of clock problems or changes.
I have seen this when building in an NFS mounted filesystem, even though
the client and server clocks are pretty closely synchronised. I've never
seen it cause a problem though.
Regards, Ian
------------------------------
From: mark <[EMAIL PROTECTED]>
Subject: Re: Won't boot: deleted libc.so.6
Date: Wed, 11 Apr 2001 14:00:13 -0500
Reply-To: [EMAIL PROTECTED]
Thanks to Bob and Juergen.
Made the Symlink and every things back to normal and I learned a Bunch to
boot.
Mark
Bob Hauck wrote:
> On Fri, 06 Apr 2001 15:04:12 -0500, mark <[EMAIL PROTECTED]> wrote:
>
> >Here's the heart of the matter: I overwrote the file libc.so.6.
>
> That's bad.
>
> >I have created a boot disk to boot with and can now mount the hda2 and
> >get access to the /lib directory.
>
> Ok, so all you have to do is restore /lib/libc-2.x.y.so and symlink it
> to /lib/libc.so.6.
>
> >init: /lib/ld-linux.so.2 version glibc.2.1.1 not found (required by
> >/lib/libc.so.6)
>
> Hmmm...musta put in the wrong libc-6.
>
> >error after creating the symlink libc.so.6 pointing to
> >/lib/ld-linux.so.2
>
> That won't work.
>
> --
> -| Bob Hauck
> -| To Whom You Are Speaking
> -| http://www.haucks.org/
------------------------------
From: "John W. Krahn" <[EMAIL PROTECTED]>
Subject: Re: Using Sed : need of little help
Date: Wed, 11 Apr 2001 20:08:23 GMT
"Jason C. Hill" wrote:
>
> Except you so conveniently chopped off the rest of my message where I stated
> if he wanted simple sed here is his answer.
>
> So, please, stop trying to undercut me when you do such a horrible job at
> it.
Well _excuse me_ for trying to help the OP.
From: "Jason C. Hill" <[EMAIL PROTECTED]>
Newsgroups: comp.os.linux.misc
Subject: Re: Using Sed : need of little help
Date: Wed, 11 Apr 2001 15:22:26 -0400
Organization: America Online, Inc.
Lines: 9
NNTP-Posting-Host: 10.2.165.140
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 5.50.4522.1200
X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4522.1200
At least I'm using the tools that I'm talking about.
John
--
use Perl;
program
fulfillment
------------------------------
From: David E. Fox <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Compiler not working
Date: Wed, 11 Apr 2001 13:12:10 -0700
Reply-To: [EMAIL PROTECTED]
Andy Walker wrote:
> I'm trying to get plib to install so I can then get FlightGear going but
> when I do ./configure it fails. The message I get is GCC failure but in
If you can email me the configure output (or the last portion of it) I might
be able to figure out what is wrong.
I just built FlightGear yesterday. It runs, but I don't think the
accelerated 3D is working (I am using a Matrox G450, with X 4.0.2, on a
Athlon (1ghz) based system (Asus A7V133 mb). But that's a side issue,
albeit a somewhat important one.
The docs don't seem to be quite clear in that you have to compile and
install the SimGear library before you start on the plib library.
--
========================================================================
David E. Fox Thanks for letting me
[EMAIL PROTECTED] change magnetic patterns
[EMAIL PROTECTED] on your hard disk.
=======================================================================
------------------------------
From: David E. Fox <[EMAIL PROTECTED]>
Subject: Re: Keep fetchmail running as daemon for certain user
Date: Wed, 11 Apr 2001 13:28:21 -0700
Reply-To: [EMAIL PROTECTED]
Rod Smith wrote:
> I've had the same problem, as well. My solution is to run fetchmail in
> NON-DAEMON mode in a cron job. I've found this to be much more reliable.
That works too, but you have a mail server under your feet (or wherever your
box is) so you might as well use it.
I have DSL as well, and I've run fetchmail periodically, or in deamon mode,
and while I haven't had the problems that the original poster had, I
figured a bertter way was just to have mail sent directly to my DSL account
(mine is m206-157.dsl.tsoft.com). So I migrate all my email over to this new
'domain'.
--
========================================================================
David E. Fox Thanks for letting me
[EMAIL PROTECTED] change magnetic patterns
[EMAIL PROTECTED] on your hard disk.
=======================================================================
------------------------------
From: "Kenny@BUI" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: [HELP] mount cdrom
Date: Wed, 11 Apr 2001 16:29:07 -0400
hey thanks man.
is ln command the link command?
once again thanks for all your input.
kenny.
> It doesn't only appear so, it is so.
> All you have wrong is a bad symlink.
>
> rm /dev/cdrom
> ln -s /dev/hda /dev/cdrom
>
> And now you're ok.
>
> Eric
>
>
------------------------------
From: [EMAIL PROTECTED] (Jean-Francois Landry)
Subject: Re: maximal mount count?
Reply-To: [EMAIL PROTECTED]
Date: Wed, 11 Apr 2001 16:00:10 -0400
Once upon a time, David Miller <[EMAIL PROTECTED]> wrote:
>Several times in the past few days I've received the following message
>during bootup:
>
>'drive' has reached maximal mount count, check forced.
>
>Yesterday, drive was /dev/hdd1. Today it is /root and /dev/hdb1. The system
>checks the filesystems and corrects a few things then proceeds.
That's a feature of the ext2 filesystem. It forces checks after N mount
counts. You can adjust it with tune2fs, or just stop rebooting so often.
A quick man tune2fs will give you the details.
Jean-Francois Landry
--
"I'm not sure if this is a good or a bad thing.
Probably a bad thing; most things are bad things."
-- Nile Evil Bastard
--
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: (X == Moby_Dick)?...:...
Crossposted-To: comp.os.linux.x
Date: Wed, 11 Apr 2001 20:47:34 GMT
David E. Fox <[EMAIL PROTECTED]> writes:
> Billy Bob Jameson wrote:
> > I did a top + M (sort by memory usage) and the first thing on top of the
> > list ix X with 80 MB. Is there anyone out there that can tell me more
> What fields were you looking at (Virtual vs. RSS)? The second is the
> resident set size, or the actual amount of RAM the process is using
> at that point in time.
> The first (VSZ) is the process's requested memory. Sometimes it
> seems this value is overinflated. X (4.0.2) is currently using 42
> Megs of VSZ but only 13 megs of RSS, according to 'ps'. A VSZ of 42
> megs (for example) really doesn't mean the process is actually using
> that much RAM -- only that it might want to use that much, or that
> is what it thinks it's using. That's probably not quite the best or
> most technical explanation.
For instance, a chunk of that may represent X code that was invoked at
initialization time, which is thus permanently "allocated," but which
will _never_ run again (until you restart X), and which could happily
swap out without affecting performance the slightest bit.
> 80 megs does seem to be a lot, but it depends on what version of X
> you are running, what video hardware you are using (I think that a
> portion of X's memory footprint represents the video card memory, so
> if you have a card with a lot of video ram, your X may look like
> it's using a lot of RAM) and how long your X session has been up, as
> well as how many apps/windows you are displaying.
I've got a card with 16MB of video RAM; that 16MB gets picked up as
part of the apparent footprint, even though it's not part of the, um,
640MB :-) on the system. [Frankly, the 94MB of "X" doesn't much
matter to me even if it _is_ lardly, when I've got 640MB...]
X winds up generally growing in size, particularly with having to
allocate space for backing store, it may free() up memory, but that
_doesn't_ mean the memory gets taken off the size of the process.
There probably needs to be a big, bold comment when you run top that
indicates something to the effect of:
"Don't believe the numbers you see here unless you truly understand
what RSS, SIZE, and SHARE really mean."
--
(concatenate 'string "cbbrowne" "@acm.org")
http://vip.hex.net/~cbbrowne/resume.html
"Motto for a research laboratory: What we work on today, others will
first think of tomorrow." -- Alan Perlis
------------------------------
From: [EMAIL PROTECTED] (Timothy J. Lee)
Subject: LD_PRELOAD and setuid programs (ELF)
Reply-To: [EMAIL PROTECTED] (this is a valid address for a limited time)
Date: Wed, 11 Apr 2001 21:15:23 GMT
For obvious reasons, arbitrary use LD_PRELOAD is generally ignored
for setuid programs.
But is there a way in Linux to have setuid programs use LD_PRELOAD
in a controlled way? For example, in Solaris, shared libraries in
/usr/lib/secure (but not elsewhere) can be named by LD_PRELOAD and
setuid programs will use them. Is there a similar capability in Linux?
(It does not seem to work by putting the shared library in /usr/lib
or /usr/lib/secure, and there is no man page for the ELF ld-linux.)
--
========================================================================
Timothy J. Lee
Unsolicited bulk or commercial email is not welcome.
No warranty of any kind is provided with this message.
------------------------------
From: Mladen Gavrilovic <[EMAIL PROTECTED]>
Subject: killing a process
Date: Wed, 11 Apr 2001 21:24:07 GMT
Hi all,
Is there any way to kill a process that refuses to die? In particular
I'm talking about cdda2wav here.
I use xcdroast to burn cds, and if it dies it leaves some relics behind
it (it is a front end so it leaves the real programs behind). All of
them can be killed except cdda2wav. It refuses to die no matter what.
I've tried:
killall cdda2wav
killall -9 cdda2wav
kill -9 <pid>
I've also tried deleting /proc/<pid> but it doesn't let me do this
("Operation not permitted.").
The worst thing about it is that cdda2wav soon grows to take up 99.9%
CPU, as shown by top.
Anyone know how to kill this thing?
Regards,
Mladen
------------------------------
Subject: Re: A question on Newsgroup
From: Greg Howland <[EMAIL PROTECTED]>
Date: Wed, 11 Apr 2001 21:29:33 GMT
"ThanhVu Nguyen" <[EMAIL PROTECTED]> writes:
> I switched my ISP and the new one doesn't carry this NG with it. Is
> there a free NG server...
Most ISP's will add the group to their server if you ask.
Did they already say that they wouldn't?
Greg
------------------------------
From: "Peter T. Breuer" <[EMAIL PROTECTED]>
Subject: Re: LD_PRELOAD and setuid programs (ELF)
Date: Wed, 11 Apr 2001 23:38:24 +0200
Timothy J. Lee <[EMAIL PROTECTED]> wrote:
> For obvious reasons, arbitrary use LD_PRELOAD is generally ignored
> for setuid programs.
> But is there a way in Linux to have setuid programs use LD_PRELOAD
> in a controlled way? For example, in Solaris, shared libraries in
> /usr/lib/secure (but not elsewhere) can be named by LD_PRELOAD and
> setuid programs will use them. Is there a similar capability in Linux?
I believe they honour the contents of /etc/ld.so.preload.
/etc/ld.so.preload File containing a whitespace separated
list of ELF shared libraries to be
loaded before the program. libraries
and an ordered list of candidate
libraries.
> (It does not seem to work by putting the shared library in /usr/lib
> or /usr/lib/secure, and there is no man page for the ELF ld-linux.)
man ld.so is what you want.
Peter
------------------------------
From: "Peter T. Breuer" <[EMAIL PROTECTED]>
Subject: Re: killing a process
Date: Wed, 11 Apr 2001 23:38:56 +0200
Mladen Gavrilovic <[EMAIL PROTECTED]> wrote:
> Is there any way to kill a process that refuses to die? In particular
By definition, no.
Peter
------------------------------
From: Craig Kelley <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.setup
Subject: Re: Xconfigurator help.
Date: 11 Apr 2001 15:46:38 -0600
"Kenny@BUI" <[EMAIL PROTECTED]> writes:
> hello,
> we just upgraded to from 6.2 to 7.0.
> the thing would not display properly. we are trying to use Xconfigurator to
> fix the problem. it appears that the monitor is being detected but when we
> type startx the
> screen does not refresh properly. the icons are there but you have to put
> the mouse over them for them to appear. there are also lines all over the
> screen. looks like the tile effect in win98.
7.0 uses XFree86 4.0.2, while 6.2 uses XFree86-3.x -- you may want to
downgrade your X11 server to the older one so that it works again.
--
It won't be long before the CPU is a card in a slot on your ATX videoboard
Craig Kelley -- [EMAIL PROTECTED]
http://www.isu.edu/~kellcrai finger [EMAIL PROTECTED] for PGP block
------------------------------
From: Craig Kelley <[EMAIL PROTECTED]>
Crossposted-To: alt.linux,alt.os.linux,alt.os.linux.mandrake,comp.os.linux.setup
Subject: Re: Mandrake 7.2 and Nvidia - A little bit closer
Date: 11 Apr 2001 15:50:35 -0600
mrsmith <[EMAIL PROTECTED]> writes:
> Seeming to have nothing else to do, I again attempted an install of
> the Nvidia drivers on Mandrake 7.2. Here is the procdedure I used on
> this attempt:
> Installed Mandrake 7.2
> Upgraded using initial ManFreq for XFree86 4.0.2 and KDE 2.1
> Rebooted to run level 1
> Installed XFree86 4.0.3 update using .tar.gz files and Xinstall.sh
> Installed Nvidia kernel and GLX, including making symbolic link from
> new X to old X, per various instructions.
> Configured XF86Config-4 per instructions.
> Executed nv_check.sh and followed reccommendations.
> Set run level back to 5 and rebooted.
>
> System goes through normal start-up and comes to auto logon portion.
> At this point X acts as if it will start but then falls back to run
> level 3. I check the X log in /var/log/ and it appears everything went
> well except for two problems. It gives "error loading module bitmap"
> and says it was not found. It does the same for module "pcidata".
>
> On previous excursions into the "modules" directory, I have seen both
> of these modules. But, now they are not there. Here are my questions:
>
> 1. Where might they have went based on my installation?
> 2. Can they be replaced out of a particular package?
> 3. Perhaps, are these modules that are created after the fact by
> executing something like Xconfigurator, etc?
>
> Thank to all for there patience...
You don't really need to do all that. Just take a stock install, and
follow the directions listed here:
http://www.nvidia.com/Products/OpenLinuxDwn.nsf/09769readmefaq
In a short sequence:
1) Setup XFree86 4.0.x to run your nvidia card (no 3D accel yet)
2) rpm --rebuild NVIDIA-kernel-blahblahblah.src.rpm
3) rpm -i /usr/src/redhat/RPMS/i386/NVIDIA-kernel.blah.rpm
4) rpm --force --nodeps -i NVIDIA-GLX.blahblahblah.i386.rpm
5) Edit /etc/X11/XF86Config
a) change "nv" to "nvidia" everywhere you find it
b) uncomment the GLX module
That's it.
--
It won't be long before the CPU is a card in a slot on your ATX videoboard
Craig Kelley -- [EMAIL PROTECTED]
http://www.isu.edu/~kellcrai finger [EMAIL PROTECTED] for PGP block
------------------------------
** 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 by posting to comp.os.linux.misc.
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
******************************
