Linux-Misc Digest #94, Volume #28 Wed, 13 Jun 01 00:13:03 EDT
Contents:
NYC LOCAL: Sunday 17 June 2001 Install Fest at Ruben's Place in Brooklyn
([EMAIL PROTECTED])
Re: Is someone trying to hack my system? (Martin Lichtin)
Re: rmmod output to syslog (Martin Lichtin)
Linux for old laptops (ross)
Re: 2GB File size limitation (Makkern)
Re: 2GB File size limitation (Makkern)
no sound from CD (Rick)
Switching disks (Leonard Evens)
Re: root password problem (Leonard Evens)
Re: Writing an OS from scratch ("bowman")
Re: Timers in threads ("bowman")
Re: Reiserfs problems? I Can't mount .. (Dances With Crows)
Re: backup/restore VFAT partition via Linux (Dances With Crows)
Re: Timers in threads ("bowman")
Re: Compiling kernel for Red Hat 7.1 (Jan Schaumann)
Re: Linux emus for Windows? (John Hasler)
two linux paritions require two separate swaps?! (uzon)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.advocacy,comp.unix.bsd.freebsd.misc
Subject: NYC LOCAL: Sunday 17 June 2001 Install Fest at Ruben's Place in Brooklyn
Date: 12 Jun 2001 23:08:05 -0400
The GNUbies and LXNY, with NYLUG in support, are throwing a big
Install Fest on Sunday 17 June 2001 in Brooklyn.
This Install Fest is free and open to all.
This Install Fest is made possible through the work and kind generosity of
the organizers and all the volunteers. We especially thank Ruben Safir,
Mrs. Safir, and HCE.
http://www.mrbrklyn.com
All free software welcome! We do not discriminate among free kernels based
upon the first letter of their names. We shall install as many free
systems on as many machines and on as many different kinds of machines as
possible. We shall install both Linux kerneled and free *BSD kerneled
systems. Any person who installs or helps install the Hurd or Squeak or
any Lisp OS or any ML OS will be treated to a family drink.
All experienced installers are invited to come and help.
All students of economics, of the history of engineering, and of the art
of propaganda are invited.
Date: Sunday June 17 2001.
Time: Starts at 1:00 pm.
Late comers are just as welcome as those who arrive at 1:00 pm.
Location: Ruben's Place
1600 East 17th Street
Brooklyn, NY 11230
By Train:
Brighton Line Q (D is local in Brooklyn) to Kings Highway.
Walk to E17th Street - 2 blocks. Turn Left.
One Block - cross Ave P - 3 houses on the Left.
By Car:
Note there is a Parking Lot in E15th and Kings Highway.
BQE to Prospect Expressway to the end, into Ocean Parkway.
Drive to Ave P and make a Left to E17th
one block past the Elevated Subway.
By Car from South Queens:
Belt to Knapp Street.
Knapp to the end to Gertitson Ave to the end to Nostrand Ave.
Left on Kings Highway (one block on Nostrand)
Kings Highway turns INTO Ave P. Left on E17th Street.
Hardware: Bring the boxes on which you wish to run a Free OS.
Though it is safest to bring your monitor, keyboard, and mouse,
you need not. We will have such available onsite.
Internet connections via ethernet will be provided.
Telephone lines will be provided, so we can help with PPP.
Software: Bring whatever distribution CDs, boot and rescue disks, boot
managers, tiny distributions, manuals, and anything else you
want. Again, you need not bring any of these things, since
they will be available onsite.
http://www.lxny.org
http://www.gnubies.org
http://www.nylug.org
http://www.sixgirls.org
http://www.fsf.org
http://www.debian.org
http://www.linux.org
http://www.debian.org/ports/hurd
http://www.squeak.org
http://www.freebsd.org
http://www.netbsd.org
http://www.openbsd.org
http://www2.ics.hawaii.edu/~esb/prof/proj/hello
http://www.daemonnews.org
http://slashdot.org
http://dmoz.org/Computers/Software/Operating_Systems
http://www2.tunes.org/Review/OSes.html
Jay Sulzberger <[EMAIL PROTECTED]>
Corresponding Secretary LXNY
LXNY is New York's Free Computing Organization.
http://www.lxny.org
------------------------------
From: Martin Lichtin <[EMAIL PROTECTED]>
Subject: Re: Is someone trying to hack my system?
Date: Tue, 12 Jun 2001 21:13:45 -0600
Anthony F wrote:
>
> When I came to work I noticed a bunch of messages that I've never seen
> before. at the login I saw
>
> svc: unknown program 0 (me 100021)
> these were followed by about 20 lines that looked similar
> svc: unknown program 100004 (me 100021)
>
> the numbers would be different on each line though.
>
> the last line read
> svc: unknown version (0)
>
> Is someone trying to hack in?
That's a client requesting ACL information from nfsd on fright. SunOS does
this on every mount request. Since Linux doesn't have ACLs, nfsd is simply
saying "hey, i don't know what this is".
It's nothing to worry about...
------------------------------
From: Martin Lichtin <[EMAIL PROTECTED]>
Subject: Re: rmmod output to syslog
Date: Tue, 12 Jun 2001 21:16:35 -0600
> Cron job "rmmod -as" fills up /var/log/messages. How can I stop it
> from writing to the syslog?
Try changing the cron command to
rmmod -a >/dev/null 2>&1
------------------------------
From: ross <[EMAIL PROTECTED]>
Subject: Linux for old laptops
Date: Wed, 13 Jun 2001 03:30:09 -0000
I have two old laptops and I was wondering which(if any) versions of
Linux or even BSD will I be able to install. One of them is a AST
pentium75 with 650meg hard drive and 8meg ram, and the other one is a
Magitronic 486 25mhz with 250meg hard drive and 4meg ram? Can anyone
help?
--
Posted via CNET Help.com
http://www.help.com/
------------------------------
From: Makkern <[EMAIL PROTECTED]>
Subject: Re: 2GB File size limitation
Date: Wed, 13 Jun 2001 04:22:35 +0100
>>An audio enginer that uses a lot of computerised equipment.
>>2Gb = ~7:17 min of 32 channel audio
>>So when recording say a 2 hour session I expect to eat the greater part of
>>a 40Gb hard drive.
[chunk-talk]
> Everytime this issue is brought up, the core filesystem is blamed and
> chunking is never discussed. Certainly a mountain made out of a molehill.
A file should be an abstraction for a place to keep data. It is very
unhelpful to have arbitrary limits on the size of this data. Do you really
think it is sound engineering for every application to have its own
hacked-on way of dealing with large files? It should be a service provided
by the file system.
One could take your argument further and say "well, I don't need file
systems at all. Give me a raw disk partition and I'll store all the data
you want".
------------------------------
From: Makkern <[EMAIL PROTECTED]>
Subject: Re: 2GB File size limitation
Date: Wed, 13 Jun 2001 04:27:18 +0100
> No - I am trying to download the databses from NCBI, which take over
> 2GB in space for some of the files just to download.
On a related note, do you know how to rm a file once it's got too big? I
just got:
# rm file
rm: cannot remove `file': Value too large for defined data type
how can I rm the damn thing? This is my root filesystem, I don't want to
blast it to free up the space.
cheers,
Makkern
------------------------------
From: Rick <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.powerpc
Subject: no sound from CD
Date: Tue, 12 Jun 2001 23:38:00 -0400
I have not been able to get sound from my CD. I can get xmms to play
streaming mp3s. I can get event sounds from X-Chat, but not a peep
frommy CD. I do not believe I g=had sound from GNOME either.
Can anyone point me to some solution?
Any and all help appreciated,
Rick
------------------------------
From: Leonard Evens <[EMAIL PROTECTED]>
Crossposted-To: redhat.general
Subject: Switching disks
Date: Tue, 12 Jun 2001 22:00:45 -0500
I just installed a new disk as /dev/sdc and put RH7.1 on it. I
still have RH6.2 on /dev/sdb. At some point, I will want to switch
these disks or possibly remove the present sdb entirely. I can
think of several ways to go about doing it, but this was the subject
of extensive discussion not too long ago. Can someone remind me of
what the most efficient method would be? Also, with RH7.1, they
use disk labels in /etc/fstab, so the file systems with labels
would presumably not require any attention. Is that correct?
In that case the main problem would be getting lilo right and
being able to boot quickly with a minimal of fussing. I don't
want to have to use a rescue disk if I can avoid it.
P.S. With my present setup, when I run lilo I get lots of
complainst that the BIOS may not be able to access 0x82, which
I presume is the third disk. (In fact it can, and lilo works
fine.) Can someone remind me of where I look to relearn those
disk codes?
--
Leonard Evens [EMAIL PROTECTED] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
------------------------------
From: Leonard Evens <[EMAIL PROTECTED]>
Subject: Re: root password problem
Date: Tue, 12 Jun 2001 22:06:18 -0500
Thomas Corriher wrote:
>
> On Thu, 07 Jun 2001 13:09:52 GMT, Peter T. Breuer <[EMAIL PROTECTED]>
> wrote:
> >Richard Kimber <[EMAIL PROTECTED]> wrote:
> >> bot403 wrote:
> >
> >>> Why would you do everythign as root anyways? Its a bad idea to always use
> >>> root. I keep a normal user around and only use root when i have to. Its
> >>> good security and
> >>> linux common sense. Trust me ive done some pretty stupid stuff as root.
> >
> >> I'm always a bit puzzled when people give the impression that one can avoid
> >
> >Then perhaps you should realize that you are missing something!
> >
> >> using root all the time. The way my machine is set up I always have to use
> >> root: to bring the eth0 interface up and down, install new software, run
> >
> >No you don't. Use sudo for all that.
> >
> >> various security checks, update software, do backups, and tweak the system
> >> in various ways.
> >
> >sudo.
> >
> >> Perhaps I'm doing things wrongly.
> >
> >Indeed you are.
> >
> >Peter
>
> Peter, there is no justification for you to act like an arrogant ass.
>
> --
> From the desk of Thomas Corriher
>
> The real email address is:
> tcorriher at earthlink.
> net
Peter acts as Peter thinks best, and we should accept him as he
is. Whether you like his tone or not,
he is a source of invaluable information which he freely shares.
We should be grateful for that, and calling him names is not
very useful or polite.
--
Leonard Evens [EMAIL PROTECTED] 847-491-5537
Dept. of Mathematics, Northwestern Univ., Evanston, IL 60208
------------------------------
From: "bowman" <[EMAIL PROTECTED]>
Subject: Re: Writing an OS from scratch
Date: Tue, 12 Jun 2001 21:36:04 -0600
"Chen Wang" <[EMAIL PROTECTED]> wrote in message
news:9g5tr3$4kf$[EMAIL PROTECTED]...
>
> 1. When the machine boots up on say a floppy, it's told to automatically
> to read the first sector (boot sector) for instructions and load them
> into memory and then execute it. Is this correct?
There are a few preliminary steps. When you plug the box in, the cpu starts
processing instructions from some sort of non-volatile memory. To keep it
simple, the cpu starts at 00000H, though this depends on the part. The
stack(s) are iniitalized, interrupt vectors set up, and whatever other
housekeeping needed is done. Usually, there is a POST (power on self test)
to make sure everything is working. Timer/counters, ports, and so forth are
set up.
Then, the FDC is initialized. This might include sector translations, DMA
addresses, and so forth.
The desired data is read, placed in volatile memory, and executed.
> If so, what are the
> requirements for the instructions and what kind of tools are available
> to compile and put it on the floppy? (Small wonder how the FIRST
> OS gets started without any existing OS..)
Typically an assembler, though C is used in many cases. Bootstrapping an OS
is interesting. Way back when, you could toggle a few instructions in from
the front panel, fire it up enough to read the reast in from paper tape. Or
with a ROM (or EPROM), you could rig up a bunch of switches, set an address,
and burn in the data. The first thing you do is get some sort of little
monitor going so you can talk to the board, and go from there.
> 3. What must be done for example, just to print a Hello World on the
> screen? (I suppose that means some kind of manipulation with the video
> memory to form the text)
How far down do you want to go? At the bottom, you take an ascii byte and
use it as an index into a table of character dot matrix type data. Build up
a row of bits for the top line of Hello, and pump
the bitstream out after receiving a cue from the crt retrace.
> 4. I am guessing that the bootloader part is the only thing that must be
done
> in assembly, but I am thorougly confused on when OS developmenet
crosses over
> to C. Given it's a higher language, where does its compiler come from
in the
> first place? Does one simply get a compiled version for the targe
machine and
> load it on? What exactly happens?
Well, after you get the monitor going, you write an assembler. Then you
write a simple compiler. And use that to write a little more complex
compiler.
This may seem primitive, but it is the condensed history of the last forty
or fifty years of computers. The fairly luxurious developement environments
we have today represent a lot of grunt work over the years.
------------------------------
From: "bowman" <[EMAIL PROTECTED]>
Subject: Re: Timers in threads
Date: Tue, 12 Jun 2001 21:41:28 -0600
"Steve Wolfe" <[EMAIL PROTECTED]> wrote in message news:d216g9.5lh.ln@helix...
>
> While I'm not familiar with setitimer, a new thread is *not* a new
> process, so I wouldn't think that it would work.
with some Linux implementations, a ps will show that each thread has its own
pid. For many other OS's, like NT or AIX4.3, the threads do not show as
individual processes
------------------------------
From: [EMAIL PROTECTED] (Dances With Crows)
Subject: Re: Reiserfs problems? I Can't mount ..
Date: 13 Jun 2001 03:42:28 GMT
Reply-To: [EMAIL PROTECTED]
On Wed, 13 Jun 2001 00:35:36 GMT, VP staggered into the Black Sun and
said:
>I can't mount my sdb7 reiserfs partition (I have there some importante
>files, my sources codes)
If they were important, where's your {daily,every-3-days,weekly}
backups? If you don't back up regularly, you *WILL* lose data to
hardware or software problems, and it will be your own fault.
>..Jun 12 23:12:32 zendo kernel: ReiserFS version 3.6.25
>..ernel: reiserfs: checking transaction log (device 08:17) ...
>..kernel: scsi0: ERROR on channel 0, id 1,\
> lun 0, CDB: 0x28 00 01 13 54 70 00 00 08 00
>.. kernel: Info fld=0x1135470, Current sd08:17: sns = f0 3
>.. do kernel: ASC=11 ASCQ= 0
>.. o kernel: I/O error: dev 08:17, sector 50256
This indicates that there is some sort of problem at a lower level than
the filesystem. Usually, though, you get some extra data out of the
SCSI drivers, saying "Sense Key: Medium Error" or something similar.
Your disk is most likely br0ken.
> Is a Hardware problem i try averything with the reiserfs utilities ,
I have seen this sort of thing in the past, and managed to get over it
with a utility called dd_rescue, a bunch of free space on another disk,
and e2fsck. You need to take /dev/sdb7 out of your fstab, and never
mount it read/write again. You should create a partition that's the
same size or larger than /dev/sdb7 on another disk. You should then use
dd_rescue to copy /dev/sdb7 over to this new partition. Finally, you
should run reiserfsck --rebuild-tree on the new partition, and hope to
Cthulhu that most of your data is still there.
RedHat, for some unknown reason, doesn't include dd_rescue in their
standard install. It's a small download and a quick compile--check
Google for the location of the tarball....
--
Matt G|There is no Darkness in Eternity/But only Light too dim for us to see
Brainbench MVP for Linux Admin / Outside of a dog, a book is a man's best
http://www.brainbench.com / friend. Inside of a dog, it's too dark
=============================/ to read. ==Groucho Marx
------------------------------
From: [EMAIL PROTECTED] (Dances With Crows)
Crossposted-To: hp.os.linux
Subject: Re: backup/restore VFAT partition via Linux
Date: 13 Jun 2001 03:42:31 GMT
Reply-To: [EMAIL PROTECTED]
On Tue, 12 Jun 2001 17:27:44 -0700, Daniel Gunyan staggered into the
Black Sun and said:
>"Gratien D'haese (Hewlett-Packard)" wrote:
>> I'm trying to backup a Windows98 FAT32 partition mounted via
>> (multi-boot) Linux system with the help of good old tar command.
>> To restore from scratch I do the following:
[snip]
>> ==> FAILS
>> I guess with "dd" command I would have more luck?
>
>I have done precisely this when Partition Magic hosed my hard drive. I
>don't think I used anything other than 'tar cf' to back it up. I did,
>however, use the Windows95 boot disk to set up the windows partition and
>FAT32 file system--and I also did a minimal windows installation AFTER
>partitioning and BEFORE doing the 'tar xf'. So this may be the key to
>preserving file permissions, since tar doesn't know about FAT
>permissions. The system has worked flawlessly (well, as flawless as
>win98 can possibly be).
>
>Windows is picky about booting, so I'd try creating the windows
>partition and file system with windows tools.
IMHO, the best thing to do is to make the 'Doze C: partition as small as
you can during the initial install. 1G is about right; just install all
aftermarket programs to D:\whatever\ and *leave C: alone* as much as
possible. C: is equivalent to / , and many experienced Unix admins keep
/ tiny and try not to fsck with it. What's good for the Penguin can be
good for the Paperclip.
If you do that, there's a much simpler way of dealing with backups and
restores of C: under Unix:
mount /dev/hda1 /mnt/doze
dd if=/dev/zero of=/mnt/doze/zero bs=16k
rm /mnt/doze/zero
umount /mnt/doze
dd if=/dev/hda1 bs=16k | gzip -9 -c > /somewhere/hda1-raw-backup.gz
(You should not mess with the partition table on /dev/hda1, but if you
want, just "fdisk -l /dev/hda1 > /somewhere/partitiontable-backup" and
you have everything you need to recreate the table exactly the way Doze
expects it.)
Restore:
cat /somewhere/hda1-raw-backup.gz | gzip -dc | dd of=/dev/hda1 bs=16k
This may not work if you have a Doze partition > 3G and your
kernel/glibc/applications are too old to handle files > 2G.
--
Matt G|There is no Darkness in Eternity/But only Light too dim for us to see
Brainbench MVP for Linux Admin / Outside of a dog, a book is a man's best
http://www.brainbench.com / friend. Inside of a dog, it's too dark
=============================/ to read. ==Groucho Marx
------------------------------
From: "bowman" <[EMAIL PROTECTED]>
Subject: Re: Timers in threads
Date: Tue, 12 Jun 2001 21:48:42 -0600
"Robert L. Klungle" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
>
> Also, when a read() function hangs due to no data, killing
> the thread doesn't seem to clear the device handler. Only
> killing the main program seems to do it.
look at pthread_cleanup_push, pthread_cleanup_pop, and pthread_testcancel.
When you are blocked on a read, a pthread_cancel will end the block, with an
EINTR. do the pthread_testcancel if this is returned by read. The testcancel
is a cancellation point, so the code specified in the cleanup will be run by
the pthread_exit(). This will allow you to close the port and do whatever
other cleanup you need.
------------------------------
From: [EMAIL PROTECTED] (Jan Schaumann)
Crossposted-To: comp.os.linux.redhat
Subject: Re: Compiling kernel for Red Hat 7.1
Date: Wed, 13 Jun 2001 03:49:51 -0000
* Meeko wrote:
> I'm sure there is a very simple solution to this, but I just can't seem
> to figure it out. I am trying to compile kernel 2.4.5 for my RH7.1
> system. I have done all the steps I normally do to compile a kernel (I
> have done this repeatedly for several RH distributions without a problem,
> so its not like I don't know what I'm doing).
>
> I downloaded the kernel source, untarred it to /usr/src/linux, cd'd to
> linux, make mrproper, make xconfig, spent 30 minutes making sure every
> setting is how I want it, saved configuration and exited, make dep, make
> clean, make bzImage, make modules, make modules_install, moved the
> kernel from arch/i386/blah blah... to /boot/vmlinuz-2.4.5-source, moved
> System.map to /boot, removed the old symlinks and created new symlinks
> for vmlinuz -> vmlinuz-2.4.5-source; System.map -> System.map-2.4.5,
> edited /etc/lilo.conf, ran lilo, and rebooted.
What's the output of /sbin/lilo ?
Can you boot the new kernel if you specify it at the Lilo prompt?
-Jan
--
Jan Schaumann <http://www.netmeister.org>
Reasons Why a Beer is Better Than a Woman:
(7) A beer won't ask you to pick up some tampons when you go to the store.
------------------------------
From: John Hasler <[EMAIL PROTECTED]>
Subject: Re: Linux emus for Windows?
Date: Wed, 13 Jun 2001 02:53:00 GMT
ComWizz wrote:
> Linux emus...
Nono, not emus. Penguins.
--
John Hasler
[EMAIL PROTECTED] (John Hasler)
Dancing Horse Hill
Elmwood, WI
------------------------------
From: [EMAIL PROTECTED] (uzon)
Subject: two linux paritions require two separate swaps?!
Date: 12 Jun 2001 21:03:49 -0700
hi,
well i sort of put the question in the "subject"-
if i have two different linux installations on my system,
do i need 2 swap partition? if so, how do i specify which one is used for which?
(i'm guessing i only need one swap partition. but i'm just making sure)
thanks,
~Uz0n~
------------------------------
** 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
******************************
