From:     Digestifier <Linux-Activists-Request@senator-bedfellow.mit.edu>
To:       Linux-Activists@senator-bedfellow.mit.edu
Reply-To: Linux-Activists@senator-bedfellow.mit.edu
Date:     Tue, 30 Mar 93 21:30:10 EST
Subject:  Linux-Activists Digest #693

Linux-Activists Digest #693, Volume #4           Tue, 30 Mar 93 21:30:10 EST

Contents:
  Re: Problems and Solutions/SLS Installation (Phil Wherry)
  Re: question about setuid (Fergus James HENDERSON)
  [BIG TROUBLE]  Buggy newsfeed and trouble with emacs (Ah)
  Re: Installation: SLS vs SCO (Walter D. Montalvo)
  Re: Cirrus Logic CL-GD5462 Driver (Mike R. Lyons)
  Re: swapping Ctrl/Caps ?? (Paul Nakada)
  Maxtor 7245AT HD timeout and reset problems (Ming-Huei Shieh)
  Re: 0.99p7A: doexpire calling df says cannot read mounted fs ? (Tim Themann)
  Has somebody ported Viola To Linux ?? (rmag@simula.efis.ucr.cr)
  Re: 386bsd, linux: which runs more out of the box? (Lawrence Fan)
  re: Taylor UUCP ignoring one site (news user)
  Mail servers...& linux (Mikemoe)
  tape recovery (Michael Jones/Signal Process Systems)
  HELP, can't boot A1 (Jyri L. Palm)

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

From: psw@carillon.mitre.org (Phil Wherry)
Subject: Re: Problems and Solutions/SLS Installation
Reply-To: psw@carillon.mitre.org (Phil Wherry)
Date: Tue, 30 Mar 1993 20:14:34 GMT

In article <1pa23fINNnu@snoopy.cis.ufl.edu>, fishwick@fish.cis.ufl.edu (Paul Fishwick) writes:

> SYMPTOM: Mouse failed to operate and the mouse test program yielded nothing.
> I had a Microsoft Bus (Inport) Mouse which, at this time, is not known
> to have a driver.
> FIX: Bought a Logitech Dexxa serial mouse for $29. Note that /dev/ttyS0 is
> COM1 and /dev/ttyS1 is COM2.

This is the second time today I've seen this. I'm a new Linux user, but
managed to get a Microsoft bus (InPort) mouse to work. I simply put the
line:

BusMouse        "/dev/mouse"

in the Xconfig file -- despite what the documentation says, this DOES work.

Phil

-- 
Phillip Wherry
Member of the Technical Staff
The MITRE Corporation, McLean, VA
psw@mitre.org

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

From: fjh@munta.cs.mu.OZ.AU (Fergus James HENDERSON)
Subject: Re: question about setuid
Date: Tue, 30 Mar 1993 20:16:01 GMT

urlichs@smurf.sub.org (Matthias Urlichs) writes:

>Wouldn't it be possible to open the script from within the kernel, on
>behalf of the shell, pass /proc/self/fd/XX as the script's name to it,
>and thus make shell scripts secure?

Yes. The following patch to /usr/src/linux/fs/exec.c seems to work.

>Obviously, for this to work you'd have to check the file's setuid-ness
>_after_ opening it. ;-)

Not necessarily. You don't have to prevent the race condition from
occuring, it's enough to detect when it does occur and take some
reasonable action like printk'ing the cracker's user id, heh heh :-)
and returning EPERM. Have a look at the code.

--- exec.c.orig Thu Apr  1 01:34:52 1993
+++ exec.c      Thu Apr  1 05:57:41 1993
@@ -17,6 +17,7 @@
  * was less than 2 hours work to get demand-loading completely implemented.
  */
 
+#include <linux/config.h>
 #include <linux/fs.h>
 #include <linux/sched.h>
 #include <linux/kernel.h>
@@ -32,7 +33,7 @@
 
 #include <asm/segment.h>
 
-extern int sys_exit(int exit_code);
+extern int sys_open(const char * filename, int flag, int mode);
 extern int sys_close(int fd);
 
 /*
@@ -493,7 +494,11 @@
                 */
 
                char *cp, *interp, *i_name, *i_arg;
+               dev_t script_dev;
+               unsigned long script_ino;
 
+               script_dev = inode->i_dev;
+               script_ino = inode->i_ino;
                iput(inode);
                buf[127] = '\0';
                if ((cp = strchr(buf, '\n')) == NULL)
@@ -537,7 +542,46 @@
                 * This is done in reverse order, because of how the
                 * user environment and arguments are stored.
                 */
+
+               /*
+                * If we have the /proc filesystem, then we can allow
+                * setuid scripts. We avoid the race condition by
+                * opening the script before executing the interpreter
+                * and passing /proc/self/fd/%d as the script filename.
+                * Bug: this assumes you have a 'proc' filesystem
+                * mounted as /proc.
+                */
+
+#ifdef CONFIG_PROC_FS
+               if (e_uid != current->euid || e_gid != current->egid) {
+                       char script_filename[20];
+                       char *script_filename_ptr = script_filename;
+                       int fd;
+
+                       fd = sys_open(filename, O_RDONLY, 0);
+                       if (fd < 0)  {
+                               retval = fd;
+                               goto exec_error2;
+                       }
+                       if (current->filp[fd]->f_inode->i_dev != script_dev ||
+                           current->filp[fd]->f_inode->i_ino != script_ino) {
+                               printk("user %d attempted to exploit setuid script race "
+                                               "condition\n", current->uid);
+                               printk("device=%d,%d inode=%d\n",
+                                               MAJOR(script_dev), MINOR(script_dev), script_ino);
+                                sys_close(fd);
+                               retval = -EPERM;
+                               goto exec_error2;
+                       }
+                       sprintf(script_filename, "/proc/self/fd/%d", fd);
+                       p = copy_strings(1, &script_filename_ptr, page, p, 2);
+               } else {
+                       p = copy_strings(1, &filename, page, p, 1);
+               }
+#else
                p = copy_strings(1, &filename, page, p, 1);
+#endif
+
                argc++;
                if (i_arg) {
                        p = copy_strings(1, &i_arg, page, p, 2);
@@ -558,6 +602,10 @@
                set_fs(old_fs);
                if (retval)
                        goto exec_error1;
+#ifdef CONFIG_PROC_FS
+               current->suid = current->euid = e_uid;
+               current->sgid = current->egid = e_gid;
+#endif
                goto restart_interp;
        }
        if ((N_MAGIC(ex) != ZMAGIC && N_MAGIC(ex) != OMAGIC) ||

-- 
Fergus Henderson             fjh@munta.cs.mu.OZ.AU      
This .signature virus is a self-referential statement that is true - but 
you will only be able to consistently believe it if you copy it to your own
.signature file!

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

From: mdgauthi@mothra.syr.EDU (Ah)
Subject: [BIG TROUBLE]  Buggy newsfeed and trouble with emacs
Date: Tue, 30 Mar 93 13:32:59 EST

OK, the newsfeed is the schools problem so noone can fix that :)  But I 
just want you all to know that if you followup to this I most likely will
not be able to read it :(

        Here's the problem that couldn't wait :)  I have emacs sitting here
wasting all kinds if space ala the SLS installation.  Now, since the speed isn't
really terrible I was going to keep it until the arrow keys didn't work.

        Nonetheless, I persevered.  I  read the FAQ and copied vt220.el to
console.el; I did a rain dance.  Nothing worked.  Then things got spooky, I
tried it as root while copying the files and they worked!  AHA, permission
problems I thought instantly (and there were, why do those directories install
with permisions of 777 anyway?)  but no such luck.  I ever tried loading it
by hand with load-library console.el & load-  vt220.el.  It loads but the
arrow keys still make a mess.  I am going to keep working on this.   (as
vi is too much trouuble for multiple files although it's faster and it works
at the moment)  but if anyone has any  brillint ideas (or stupid ideas too
I guess, I haven't got a clue) feel free to clue me in

-- 
-Matt, still looking for sigs in all the wrong places. :)
 
Linux --the best thing ever to hit a PC.

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

From: walterm@iat.holonet.net (Walter D. Montalvo)
Subject: Re: Installation: SLS vs SCO
Date: Tue, 30 Mar 1993 19:18:07 GMT

sheldon@iastate.edu (Steve Sheldon) writes:
: 
:  Having had experience installing SLS, and SCO ODT 2.0...
: 
:  SCO's install isn't that bad really.  The first versions of ODT 2.0 had
: some big bugs and the install would frequently crash many times during the
: installation, with no chance of recovery(you started over)
: 
:  SCO ODT 2.0 personal system comes on some 40+ disks.  This includes the
: Unix, the networking, DOS emulation, and Xwindows X11R4 with the .Xdesktop
: tool, Motif, etc.
: 
[snip]
: 
:  Really I don't mind the ODT 2.0 installation.  I suppose the main
: difference would be:
: 
:  SCO doesn't come with quite as many nifty things as SLS.  i.e. you don't
: get any compilers with SCO
      ^^^^^^^^^^^^^
That's surprising!  Not even GNU compilers?

:  Not that I would suggest using SCO.  Next we can discuss all the braindead
: things which are broken in SCO, or at least seem to mysteriously don't work.
: :-)
: 

I am tempted to discuss those brain-dead things.
A part of me would refrain from that to avoid creating a holy war of
Linux vs SCO (we have enough of those already), but another part of me
would like to have a civilized discussion of the features in SCO
and SLS.  For people wanting to decide between SLS and SCO it would be
helpful (although most of the people reading this posting are probably
Linux fans already).

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

From: lyons@ennews.eas.asu.edu (Mike R. Lyons)
Subject: Re: Cirrus Logic CL-GD5462 Driver
Date: Tue, 30 Mar 1993 18:35:19 GMT

In article <1993Mar11.005754.3555@microsoft.com> t-alanj@microsoft.com (Alan Jenn) writes:

>   I too have this board and would like to know if a driver
>   is being written to handle this chip.  Or, is there an
>   alternative to buying a new board?
>
>                                  Alan

 I am having the same difficulty.  I am using a Genoa 8500 VL (VESA
Local Bus) board with the Cirrus CL-GD5462 chip.  

 If anyone is working on a driver for this chip and/or would like a
beta tester, I am very interested.  Please e-mail direct.


                                -- Linux Newbie


                                -- Mike Lyons


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

From: pnakada@oracle.com (Paul Nakada)
Subject: Re: swapping Ctrl/Caps ??
Date: 30 Mar 93 21:16:27 GMT

In article <1p4ifhINNh2k@mailgzrz.TU-Berlin.DE> lok@cs.tu-berlin.de (Andreas Stang) writes:
   I spend the afternoon trying to get `Ctrl' on the `Caps Lock' key, but it
   didn't work. Is there anybody who has an Idea how to modify keyboard.c?
   (0.99pl6 and US-Keyboard)

I just swapped the 

caps,ctrl,
and 
uncaps,unctrl

entries in the table at the end of keyboard.c.

Under X I use the xmodmap file as explained in the xmodmap man page.

-Paul
--
Paul Nakada | Oracle Corporation | pnakada@oracle.com | DoD #7773 | '91 R100C

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

From: smh@eng.umd.edu (Ming-Huei Shieh)
Subject: Maxtor 7245AT HD timeout and reset problems
Date: 30 Mar 1993 20:04:33 GMT

I installed Linux SLS 98p5 on my system.
        486DX2-66 (256K cache, 16M RAM) VESA-Local Bus
        NEC 3FGx 15" Monitor
        Logictec Mouseman
        Trident 8900C video card
                (I order Orchid Prodesigners IIs, but the vendor dosen't 
                have this card available now, so he lend me Trident card.)
        Maxtor HD 7245AT (245M)
        DTC 2270VL Local-Bus to IDE and Super IO Controller

Everything seems work fine, I can run X-window, Latex, gv, xfig, ...etc.
But sometimes my HD responsed very slowly, even stucked when I was working on 
X-window, especially when I opened a new window.
And I saw the following message on the screen when I quited X-window. 

HD time out
HD controller reset
HD time out
HD controller reset
:
:

What is the problem?
I own Maxtor 7245AT HD (245Mb).
I heard old Maxtor 7213A (PCBA number less than 63) has the same problem 
as mentioned here a while ago in this newsgroup. 
Does any one who own Maxtor 7245AT have the same experience? 
Is the problem from HD, or IDE IO controller, or video card?

Thanks for help,

Ming-Huei









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

From: tthemann@isys-hh.hanse.de (Tim Themann)
Subject: Re: 0.99p7A: doexpire calling df says cannot read mounted fs ?
Date: Tue, 30 Mar 1993 18:10:35 GMT

michaelw@desaster.hanse.de (Michael Will) writes:

>Now calling 
>       /usr/local/lib/news/bin/expire/doexpire & 
>as user news gave me:

>df: cannot read mounted filesystems: permission denied
>test: integer expression expected before -le
>df: cannot read mounted filesystems: permission denied
>test: integer expression expected before -gt

change /usr/local/lib/news/bin/spacefor, line 56 form
        df $arg | awk "BEGIN { nf = 4 ; nr = 2 }
to      /usr/bin/df $arg | awk "BEGIN { nf = 4 ; nr = 2 }

>to mail me something (which failed because a mail.lock could not be created,
>but this will be solved with a newer mail I think.)

Check your permissions

Tim (tthemann@isys-hh.hanse.de)

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

From: rmag@simula.efis.ucr.cr
Subject: Has somebody ported Viola To Linux ??
Date: 30 Mar 1993 20:29:05 -0500
Reply-To: rmag@simula.efis.ucr.cr

Recently I got the sources for viola (an X-client to access the WWW), but it
seems that the compiling is going to be a rough ride. Before I spend too
much time in the project, I will like to know if somebody has already done
it. Thanks,     
                        Roberto.


-- 
**********************************************************************
      <rmag@simula.efis.ucr.cr>   J.R. Maga\~na (TeX convention)
  also: <rmag@inforisc.cr>         Laboratorio de Simulaciones.
        <rmag@ucrvm2.bitnet>   Escuela de Fisica, Universidad de C.R.
**********************************************************************      

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

Crossposted-To: comp.os.386bsd.questions
From: lawf@base.bellcore.com (Lawrence Fan)
Subject: Re: 386bsd, linux: which runs more out of the box?
Date: Tue, 30 Mar 93 19:48:16 GMT

In article <1993Mar23.052609.11978@serval.net.wsu.edu>, hlu@eecs.wsu.edu (HJ Lu) writes:
|> 
|> We have tried very hard to make porting to Linux easy. The Linux C library
|> is ANSI/POSIX compliant. Unless the software is developed for 386BSD or
|> something like that, porting to Linux should be easier. Also the floating
|> point part of Linux C library should be less buggy. But there are a few
|> exceptions:
|> 
|> 1. signal in Linux is POSIX. You can do everything with POSIX except for
|>    sigvector. Sun also supports POSIX signal.
|> 2. tty in Linux is POSIX. Sun also supports POSIX tty.
|> 3. time functions are POSIX, plus a few BSD and SYSV extensions.
|> 4. setjmp/longjmp functions are POSIX. But you can use -D__FAVOR_BSD
|>    to make it BSD or use sigsigjmp/siglongjmp.
|> 
|> From my experiences, 95% of codes from public sites should be compiled
|> right out of box, at most with a few warning if those codes are not
|> ANSI/POSIX ready. You can try to compile one of your favorite packages. 
|> 

Talking about porting to Linux.  I have some questions:

1) Do we have SIGBUS?

2) I see there are include files under /usr/include that has the
same file names as /usr/src/linux/include/linux.  As we update
the linux kernel version, the files under /usr/src/linux/include/linux
gets updated but not the ones in /usr/include.  Why is that?

==========================================================================
Lawrence Fan                                                (908) 699-8051
Bellcore                     RRC-1R217 444 Hoes Lane, Piscataway, NJ 08854
lawf@base.bellcore.com

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

From: news@cck.coventry.ac.uk (news user)
Subject: re: Taylor UUCP ignoring one site
Date: 29 Mar 93 14:07:45 GMT


>Unless Charles can make it "worth his time" to try and help track down a
>bug in efs, his comments are utterly worthless. The only thing we know is
>that Charles doesn't think efs is stable. So what? That doesn't help! 
>

Are you saying that only those who can fix problems can comment? If something
doesn't work then anyone should say so. If that person or anyone else is
interested they might try to fix it but we shouldn't keep quiet just because
we don't know what is wrong. Perhaps you are saying he is a liar and it
is really working all the time. You are right that he should have said
what his configuration is but people are unlikely to learn that whe the first
response they receive is a "get a grip type message". If I remember correctly
his post was in response to a claim that efs is bug free. It was not a
request for help or even a criticism. It was a contradiction of a claim which
he believed was unfair.

>There are a lot of things "wrong" with Linux, but there are lots of things
>right about it as well.

The only problems I have had with the file system were in setting it up and with
a lttle help that was soon done. No files have been lost or orther problems
and although I have only been running it for a short time the file system has
been subjected to severe abuse and not a problem has surfaced. There isn't much
wrong with Linux that a relatively naive user such as myself would spot. 

>I am not a kernel hacker, or a writer of new filesystems, or any kind of major
>contributor to Linux, but I at least try to give back to the Linux community 
>by helping people out with problems that I know how to solve. Charles isn't 
>helping anything, just dishing out criticism.
>

I shall look forward to receiving help from you then. Perhaps now is a good time.
There are files missing from 0.99P7 compared to 0.99P6 in the directory ..linux/net/tcp.
For example, wd.c, ne.c and so on. I don't know why they are missing or perhaps why
they were present in 0.99P6. Am I allowed to mention this given that I don't know
why?

Colin Bruce                            (ccx004@coventry.ac.uk)
Coventry University, Priory Street, Coventry, UK

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

From: millerm@oakhill.sps.mot.com (Mikemoe)
Subject: Mail servers...& linux
Reply-To: millerm@oakhill.sps.mot.com
Date: Tue, 30 Mar 1993 15:34:02 GMT

2 part question...

Any one port a mail-server to Linux? I'm in need of one...

Otherwise,
I tried compiling "lists" from wuarchive.wustl.edu last night on my
Linux box. I was unable to do this because the include files were
goofed up (or something) in linux...Examples...

[from memory]
/usr/include/sys/types.h included /usr/include/linux/types.h which didn't
    exist.
    
/usr/include/signal.h included another signal.h file which turned out to be
    itself...
    
I could never get past this stage...did my include directory entail a problem,
or is it me not understanding something?

THanks,
Mike


---
+--------------------------------------------------------------+
| Michael Miller   millerm@oakhill.sps.mot.com                 |
| (512) 891-8582   Motorola - Semiconductors  Austin, TX       |
+--------------------------------------------------------------+
| Contrary to what most people say, the most dangerous animal  |
| in the world is not the lion or the tiger or even the        |
| elephant. It's a shark riding on an elephant's back, just    |
| trampling and eating everything they see.                    |
|                           --  Deep Thoughts by Jack Handey   |
+--------------------------------------------------------------+


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

From: spt@cscns.com (Michael Jones/Signal Process Systems)
Subject: tape recovery
Date: 30 Mar 93 20:03:21 GMT

I can not figure out the way to make a boot and root with
a tape driver.  There is not enough room on a root 1.2M floopy
for shared libs etc.  I know that the a1 and a2 disks make a ram
disk and uncompress the libs. etc.  How do I make a boot disk
with my image that will do the same thing?  An how can I make
a root disk wiht the correct /dev/tape files?

Mike
spt!wizard@spunky.cscns.com

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

From: palmj@rebecca.its.rpi.edu (Jyri L. Palm)
Subject: HELP, can't boot A1
Date: 30 Mar 93 19:12:44 GMT

Basically, I accidentally removed my /lib/libc.so.4 .... ddddduuuuhhh.
But anyway, I figured that if I could boot from my SLS disks I could
remount /dev/hda3 and recreate the link.

However, when booting from a1 (after selecting screen mode) I got

Uncompress Linux: check - work - invalid compressed data--crc error
--System halted

So I thought my disk was corrupted so I ftped and raw wrote a fresh a1
from tsx-11.mit.edu ... when I booted from this new disk, I got the
same error message ... *sigh* ... what can I do next ?

Configuration:
486-33 8M RAM
200M Maxtor
Tseng SVGA
/dev/hda3 / ext2
/dev/hda2 /usr ext2
/dev/hda4 swap 

Please send any suggestions to palmj@rpi.edu .... thanx VERY MUCH in
advance

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


** 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: Linux-Activists-Request@NEWS-DIGESTS.MIT.EDU

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

    Internet: Linux-Activists@NEWS-DIGESTS.MIT.EDU

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 0.99pl6 released on February 23, 1993

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