The World of Computer Software/The World of Computer Software.iso/lnxd3-79.zip/DIGEST3.799

Internet Message Format | 1993-01-01 | 22KB

From [e-mail] Tue Dec 15 15:15:54 1992
Return-Path: <[e-mail]>
Received: from bloom-picayune (BLOOM-PICAYUNE.MIT.EDU) by montefiore.ulg.ac.be (4.1/SMI-4.1)
    id AA09283; Tue, 15 Dec 92 15:14:42 +0100
Received:  by bloom-picayune (5.57/25-eef)
    id AA16131; Tue, 15 Dec 92 07:30:22 -0500
Message-Id: <9212151230.AA16131@bloom-picayune>
From: Digestifier <[e-mail]>
To: [e-mail]
Reply-To: [e-mail]
Date:     Tue, 15 Dec 92 07:30:12 EST
Subject:  Linux-Activists Digest #799
Status: R

Linux-Activists Digest #799, Volume #3           Tue, 15 Dec 92 07:30:12 EST

Contents:
  Re: How do I provide the floppys to normal users???? (Peter Orbaek)
  Re: ext fs question (.badblocks) (Remy CARD)
  Insite 21MB Floptical and Linux ([e-mail])
  Security problem with "rm" (Martin Kosina)
  Re: Gcc help needed (Gerry George)
  Re: A proposal for organizing comp.os.linux. [NEW-CHANNELS] (Robert Chen)
  Re: Can't make 0.98pl6 ! ( Ian Justman)
  Crossposting of Announcements (Matt Welsh)
  Re: network over modem... (Robert Chen)
  HELP X gods! .99 broke my xdm! ([e-mail])
  More info on non-booting .99 (Michael K. Johnson)
  SLIP vs. PPP (John Auer)
  Re: Floppy DMA, question and patch (?) (Linus Torvalds)
  tcp/ip support and the Makefile of 0.99 (Ammo Goettsch)
  SLS update: .99 kernel (Peter MacDonald)
  Re: TCPIP-config (David Summers)
  Re: Security problem with "rm" (David Engel)
  Help! .99 crashes on boot

----------------------------------------------------------------------------

From: [e-mail] (Peter Orbaek)
Subject: Re: How do I provide the floppys to normal users????
Date: 15 Dec 92 09:06:56 GMT

[e-mail] (Klaus Espenlaub) writes:

>Is there any possibility to give normal users the right to mount
>floppy drives???? I've been trying it with different success (it works only
>when floppys already in at boot time........)

>May anybody help me???????

>Thanks for all

If you just want users to be able to use messy dos disks, then you
might want mtools.

If you also want ordinary users to be able to mount linux disks or, say,
CDROMS, then you might want to use a mount program that I wrote
some time ago. It supports exactly this, by way of an /etc/ufstab file
that determines what and where ordinary users can mount.

I can post the mount/umount pair if anybody is interested, but the are still
quite alpha.


        - Peter.
--
Peter Orbaek <[e-mail]>
Hasle Ringvej 122, DK-8200 Aarhus N, DENMARK

------------------------------

From: [e-mail] (Remy CARD)
Subject: Re: ext fs question (.badblocks)
Date: 15 Dec 92 08:43:58 GMT

In article <[e-mail]> [e-mail] (David Fenyes,Neurobiol/Anatomy,5705,7901935) writes:
>In article <[e-mail]> [e-mail] (Mark Evans) writes:
>>Remy CARD ([e-mail]) wrote:
>>:     mkefs now uses a reserved inode (inode #2) to store the bad blocks.
>>: This inode does not appear in the filesystem tree any more because it was
>>: leading to problems (root could delete the .badblocks file or save it during
>>: a backup).
>
>Is there a way to easily read the bad blocks list to verify that all the
>bad blocks have been identified? (or even to modify it after the fs has
>been created?)

        You can add bad blocks to the list with efsck.  'efsck -t' tests for
bad blocks in a fs and adds new bad blocks to the list.  'efsck -b file' reads
a bad blocks list from the file and adds them to the list.

        Unfortunately, there is no way to print the bad blocks list.  Seems
like I must add an option to efsck for the next release.

>
>David.
>--
>David Fenyes                                 [e-mail]
>University of Texas Medical School           Houston, Texas


        Remy
--

        Remy Card
        [e-mail]

------------------------------

From: [e-mail]
Subject: Insite 21MB Floptical and Linux
Date: Mon, 14 Dec 1992 23:54:10 GMT

Hello all,

        A couple of months ago I read that someone was attemting to get the
Floptical drive from insite to work with Linux.  What is the status of this?
And will it be incorporated into the kernel when done?

Thanks in advance
Edwin

[e-mail]

------------------------------

From: [e-mail] (Martin Kosina)
Subject: Security problem with "rm"
Date: Mon, 14 Dec 1992 22:56:39 GMT


I am having a serious security problem with the 'rm' command.
I have discovered that non-superusers can REMOVE files owned
by root. I doubt that such a serious bug could exist in Linux,
it must be something in my setup:

mkosina:NotSt0PgElMbo:100:6:Martin Kosina:/home/mkosina:/bin/sh

This is the /etc/passwd entry for 'mkosina', my user login.
This is a file in /etc that I will try to delete:

-rw-r--r--  1 root    root         62 Dec 11 21:02 rc.local

When I do 'rm rc.local' rm responds:

rm: remove `rc.local`, overriding mode 0644?

If I than say yes, it will be DELETED !

All other security works, read-only files deny permissions
to be over-written, processes refuse to be killed,etc.

I am running an out-of-the-box release of SLS p1-36.
This MUST be a permission problem somewhere, the only
thing I can think of is that SOMETHING runs setuid 0.
It should not be rm, though, here is what it looks like:

-rwxr-xr-x   1 root   root       5056 Sep  4 23:20 /bin/rm*

Any Linux gurus out there who can help me ?
Thanks in advance.

Martin

P.S. If this is indeed a bug in rm (or the unlink sys call ?)
which I doubt, could you tell me how to upgrade ? I can ftp.
--
[e-mail]  Public Access User --- Not affiliated with TECHbooks
Public Access UNIX and Internet at [phone] (1200/2400, N81)

------------------------------

From: [e-mail] (Gerry George)
Subject: Re: Gcc help needed
Date: 14 Dec 92 08:47:03 GMT



I have a similar problem (or two).  whenever I try to compile, I get the
message
                make *** [main.o] IOT trap/ABORT (core dumped)

or some variation, depending on the program I try to make.  Also, I get
permission errors when attempting a compile when other than root.  Need some
help in figuring out those problems.

Gerry George

------------------------------

From: [e-mail] (Robert Chen)
Subject: Re: A proposal for organizing comp.os.linux. [NEW-CHANNELS]
Date: Tue, 15 Dec 1992 01:23:06 GMT

In article <[e-mail]> [e-mail] (Nelson Minar) writes:
>For what it's worth, I think comp.os.linux is damned near unreadable.
>It isn't just that there are 150 messages a day. The messages in this
>newsgroup are completely disorganized. There are 150 messages and 100
>different threads! I'm disappointed that the group reorg didn't pass.
>
>However, if a few of the louder voices on comp.os.linux really wanted
>to, we could still split comp.os.linux this week. Instead of splitting
>it according to the normal Usenet voting procedures, split it into the
>alt groups. Alt groups don't require a vote - a week's discussion on
>alt.config seems to be sufficient.
>
>[e-mail] \/   You're free, so anything could happen, so happy birthday

Sigh.  Do you people have any idea how much traffic is left in c.o.l linux
after you kill these stupid threads?  Not a lot.  The splitting of
c.o.l vote failed for a reason.  Why do you think doing it anyway will
solve anything?  What's more (I don't know why people cannot see this;
it has been mentioned a zillion times) - splitting newsgroups does
not decrease traffic!  It increases it.  I read comp.os.linux.announce
because I don't want to miss the announcement of any new toys.  Then I
have to read the same damn articles again in c.o.l!  How can any sane
person say this is an improvement?!

- Ken

------------------------------

From: [e-mail] ( Ian Justman )
Subject: Re: Can't make 0.98pl6 !
Date: Mon, 14 Dec 1992 01:07:10 GMT

[e-mail] (ARE WE HAVING FUN YET) writes:
:
: I'm having trouble making the 0.98pl6 kernel - everything goes fine until
: it gets to linking things, then I get a load of Undefined symbols - mostly
: _put_fs_long, _put_fs_byte, and things like _set_bit and _clear_bit.
:
: How do I get this to compile  ??? (I have the SLS 0.98 distribution with
: the 2.2.2d gcc libs)

The compiler that came with your system will work just fine.  It does for
me.  Anyway, it sounds like you have a few files not being included (or
more appropriate, the appropriate files are not being included).  This
is mainly because you need symbolic links in the /usr/include directory
which point to /usr/src/linux/include/linux and /usr/src/linux/include/asm
named linux and asm respectively.  Then the kernel will compile properly.

One note, if you already have directories called /usr/include/linux and
/usr/include/asm, and you don't want to erase the stuff in those director-
ies, use mv to rename them to something like "linux_bak" and "asm_bak"
respectively before applying the links.

Query: I myself have made said symbolic links.  I figure since now I'm using
pl6, should I keep the linux and asm links pointed to the directories con-
taining the Linux include files to guarantee compatiblity with pl6.  Is it
wise to operate under that philosophy?

Thanks for any pearls of wisdom.

--
Born to void warranties!        [e-mail]

------------------------------

From: [e-mail] (Matt Welsh)
Subject: Crossposting of Announcements
Date: Tue, 15 Dec 1992 01:58:07 GMT

In article <[e-mail]> [e-mail] (Robert Chen) writes:
>I read comp.os.linux.announce
>because I don't want to miss the announcement of any new toys.  Then I
>have to read the same damn articles again in c.o.l!  How can any sane
>person say this is an improvement?!

At the moment, this is my fault. I intentionally crosspost articles
between the two groups for those sites that don't get c.o.l.a (and for
FidoNet). We plan to change this soon once c.o.l.a is more widely
distributed (and once Fidonet can pick up the new group). This should be
soon. If you have problems with this, please let me know.

Thanks.
mdw

--
Matt Welsh     [e-mail]     Cornell Theory Center
  "Go on, emote! I was raised on thought balloons!"

------------------------------

From: [e-mail] (Robert Chen)
Subject: Re: network over modem...
Date: Tue, 15 Dec 1992 01:45:38 GMT

In article <[e-mail]> [e-mail] (Matt Welsh) writes:
>In article <[e-mail]> [e-mail] (Jerry Jeremiah 290-8941) writes:

>>I have a modem.  I have a site that has a modem that connects to a unix
>>machine.   How do I get telnet and ftp to start my modem connection
>>and do everything that way instead of an ethernet adapter?
>
>It's called SLIP (Single Line IP), and it's currently in development for

Isn't that Serial Line Interface Protocal? (I could be wrong)

>Linux. Most other operating systems support SLIP connections, however to be
>of any use you need a fast modem (9600bps at the very least), because you're
>actually transmitting packets over the modem connection, and telnet, ftp, etc.
>aren't as simple as just "dialing up" (which works fine for slower modems).
>
>Once it's available for Linux, it should be very easy to get yourself SLIP
>access from another system, thereby putting yourself on the net via modem.


No need to wait.  Get ka9q from tsx-11 or wherever and away you go.
Remote X11, telnet, ftp, ping...  What more can you ask for?  All
without even compiling tcpip into the kernel.

>
>Cheers,
>Matt Welsh     [e-mail]     Cornell Theory Center

- Ken

------------------------------

From: [e-mail]
Subject: HELP X gods! .99 broke my xdm!
Date: 15 Dec 1992 05:42:46 GMT
Reply-To: [e-mail]

I just finished compiling 0.99 with tcp.  I had to do some fun patching
on my own, because my ethernet card is an ne2000.  I got the driver code
from Matt Welsh at Cornell.  Anyway, I set up the tcp code like I did for
0.98.5, applied to emergency patches to .99 posted by linus, did the
appropriate making and voila!  0.99 booted without a hitch. No inode errors,
no tcp errors, no nothing.  I even successfully recompiled ps and ps -U
worked fine.

However, xdm locks up my machine.  xdm was working fine on 0.98.5, and nothing
has changed with the xdm (or X, for that matter) stuff since then.  I can
start X with startx just fine.  But xdm locks it up.  Once I was lucky
enough to get some errors in the xdm-errors.  Basically, it said 'hung in
XOpenDisplay (:0); aborting' and then 'cannot open display.'  Does anyone
know what my problem is here?  This is important to me since I have several
users, and xdm is their way to login and X is the environment of choice.  Do
I need to recompile the xdm source?  Do I have an old version of xdm?  I am
running XFREE86-1.1.

Any help will be greatly appreciated.

Jeff Little
[e-mail]

------------------------------

From: [e-mail] (Michael K. Johnson)
Subject: More info on non-booting .99
Date: Tue, 15 Dec 1992 05:34:35 GMT


I recently posted that I couldn't get .99 to boot, nohow.

I have tried a few more things, and everything has failed...

I made a backup of /etc/init and then made /etc/init be a hard link to
/bin/bash, so that I could be sure that /etc/init was not at fault.
Still no dice -- the linux_banner is printed, and no more, no disk
activity, no nuttin.

I restored /etc/init and figured that it must be in the do_execve()
code.  I diffed .98.6's exec.c and .99's exec.c and found 3 lines
difference, checked to see that those lines were called by
do_execve(), and borrowed the .98.6 exec.c.  Still no dice.  I have
restored the new .99 exec.c

The only think left that I can think of is the new readahead code that
Linus wrote.  I am not familiar with this code at all, and so would
ask that Linus and/or someone familiar with this code would look over
it for any subtle bugs that could be causing me this trouble.

As I mentioned before, I have put Linus's two one-line patches in,
including the initialization one, to no effect.

Thanks for any help,

michaelkjohnson

------------------------------

From: [e-mail] (John Auer)
Subject: SLIP vs. PPP
Date: Tue, 15 Dec 1992 02:26:42 GMT

After reading some of the PPP propaganda^H^H^H^H^H info :-)
it seems that PPP is, perhaps, a better way to go than the
older SLIP protocol.  I got the ka9q distribution, and it
talks about SLIP only.  I am aware that some linux'ers are
working on an Alpha SLIP.

Is this a true SLIP product?  Is the newer SLIP competitive
with PPP?  Has someone hacked a version of PPP for Linux?

I'm getting ready to start bringing SLIP/PPP online, and
obviously want to do it right the first time.  I don't
want this to turn into a flame war between PPP and SLIP;
just want to know the rational for the SLIP choice.

And while I've got you here, thanks for the great OS,
Linus and co.!  3 months without a hitch...

John

------------------------------

From: [e-mail] (Linus Torvalds)
Subject: Re: Floppy DMA, question and patch (?)
Date: 15 Dec 92 11:26:58 GMT

In article <If=[e-mail]> [e-mail] (Frank T Lofaro) writes:
>
>       Is the 1MB DMA restriction in the floppy code necessary, or can
>386/486's handle 16MB DMA for the floppy? I have this patch (suggested
>to me in e-mail) that ups the limit to 16MB from 1MB, and it seems to
>work.

It should work fine.  The only reason for the 1MB (as opposed to 16MB)
limit was that I originally thought of making only the low 1MB
identity-mapped, and anything above it could be moved around with the
pager hardware, so DMA wouldn't work on those without some address
calculations.

As it turned out, I never did anything like this, and I currently
guarantee that all kernel memory is identity-mapped up to the limit of
physical RAM (it's easier on the memory manager, as well as on devices
like some SCSI cards which use DMA).  So changing the limit to 16MB
should work fine.  Not that it probably matters very much: 99% of the
floppy IO reads go through the track buffer, and writes aren't exactly
timing-critical in most circumstances..

                Linus

------------------------------

From: [e-mail] (Ammo Goettsch)
Subject: tcp/ip support and the Makefile of 0.99
Reply-To: [e-mail] (Ammo Goettsch)
Date: Tue, 15 Dec 1992 01:11:14 GMT

I disabled TCP/IP throught the "make config" option of the kernel Makefile
for version 0.99. In .../net/Makefile, it still wants to link in
tcp/tcpip.o, which of course had not been compiled, since the make file
in .../net/tcp never gets executed ifndef CONFIG_TCPIP.

Is the link line in the make file wrong or should it have been compiled?

Also, what will happen if I compile TCP/IP but have no network card?

Ammo Goettsch

------------------------------

Crossposted-To: comp.os.linux.announce
From: sanjuan![e-mail] (Peter MacDonald)
Subject: SLS update: .99 kernel
Date: Tue, 15 Dec 1992 03:00:53 GMT

I was so impressed with the .99 kernel, that I have already upgraded
SLS to it.  Just get SLS/a2 and "sysinstall -install /user/image.taz" it.
It may take Ted a bit to get the .99 source and change the symlink.
Be patient (don't you get tired of hearing that: give it ALL to me
right NOW!).

I am not upgrading the boot disk as yet, but let me know if the old
one causes problems (SCSI?).   The kernel has grown another 30K or so
so until/if loadable devices show up, I may have to be a little conservative
on the boot disk upgrades.  Maybe I could try rebuilding the kernel
without FP-EMU.  Do bash/tar/compress et al need FP?

Peter
[e-mail]

--
Send submissions for comp.os.linux.announce to: [e-mail]

------------------------------

From: [e-mail] (David Summers)
Subject: Re: TCPIP-config
Date: Tue, 15 Dec 1992 02:31:42 GMT

[e-mail] (Michael Kraehe) writes:

>Hy folks,

>I think config is a bad name for that tool, after a find -name config
>i fount 3 executabels and 2 other files.

>Wy dont we name it ifconfig or sock_config, to avoid naming conflicts with
>BSD.

>By Michael.

On SunOS it is called 'ifconfig' and it is also able to report the current
settings of the Ethernet interfaces as well as set them.  Does anyone know
how to do this with the current (0.99) version of Linux?

   - David Summers
   ([e-mail])


------------------------------

From: [e-mail] (David Engel)
Subject: Re: Security problem with "rm"
Date: 15 Dec 92 02:55:34 GMT

Martin Kosina ([e-mail]) wrote:
: I am having a serious security problem with the 'rm' command.
: I have discovered that non-superusers can REMOVE files owned
: by root. I doubt that such a serious bug could exist in Linux,
: it must be something in my setup:

: This is a file in /etc that I will try to delete:

: -rw-r--r--  1 root    root         62 Dec 11 21:02 rc.local

: When I do 'rm rc.local' rm responds:

: rm: remove `rc.local`, overriding mode 0644?

: If I than say yes, it will be DELETED !

What are the permissions on /etc?  You don't need access to a file to
be able to delete it.  All you need is write access to the directory
the file is in.

David
--
David Engel                        Optical Data Systems, Inc.
[e-mail]                      1101 E. Arapaho Road
[phone]                     Richardson, TX  75081

------------------------------

From: [e-mail] ()
Subject: Help! .99 crashes on boot
Date: Tue, 15 Dec 1992 02:44:30 GMT

Hello,
I have just compiled version .99 taken off sunsite and it compiled
fine, but when loading (off a floppy, made with make disk, everything
is fine...finds my three partitions, partition table ok and then...
Unable to handle kernel paging request at address c8c1ff06
Ooops: 0000
EIP 0008:0000B6C0
EFLAGS: 00010206
fs: 0017
base: c0000000, limit = 000A0000
pid:1 process nr=1
0b 13 85 d2 74 27 B3 3a 02 74


and then it just hangs there...
I'm using a 486/50 4MB RAM, IDE disk (120 Meg) and booting from 1.44 floppy

Any ideas what's going on???
Thanks for any help,
Jerry

------------------------------


** 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: [e-mail]

You can send mail to the entire list (and comp.os.linux) via:

    Internet: [e-mail]

Linux may be obtained via one of these FTP sites:
    nic.funet.fi                pub/OS/Linux
    tsx-11.mit.edu                pub/linux
    tupac-amaru.informatik.rwth-aachen.de    pub/msdos/replace

The current version of Linux is 0.98pl6 released on December 2, 1992

End of Linux-Activists Digest
******************************