Subject: Linux-Activists Digest #50
From: Digestifier <Linux-Activists-Request@BLOOM-PICAYUNE.MIT.EDU>
To: Linux-Activists@BLOOM-PICAYUNE.MIT.EDU
Reply-To: Linux-Activists@BLOOM-PICAYUNE.MIT.EDU
Date:     Fri, 10 Apr 92 17:45:24 EDT

Linux-Activists Digest #50, Volume #2            Fri, 10 Apr 92 17:45:24 EDT

Contents:
  kernel building problems (System test)
  Compress/Uncompress (Were ?) (P Thomson)
  Re: kernel building problems (David Engel)
  Linux SCSI support (Drew Eckhardt)
  new stuff at tsx-11 (Michael K. Johnson)
  Meta-Question (Chet Murthy)
  Re: Problems with serial port drivers?? (Farhan H Garib)
  Re: Problems with serial port drivers?? (Farhan H Garib)
  Re: Meta-Question (Chet Murthy)
  Re: Meta-Question (Jim Winstead Jr.)
  Re: kernel building problems (Karl R. Buck)
  Re: Problem with tar (?) (Joe Smith)
  Re: Another bug? (Linus Benedict Torvalds)

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

From: news@solitary.Stanford.EDU (System test)
Subject: kernel building problems
Date: Fri, 10 Apr 1992 15:38:57 GMT

I've been going crazy trying to get either 0.95c or 0.95c+ to
compile with gcc 2.1.  Which of these is the "second patch" to
which H.J.'s patch is to be applied?  

In any event, the makefiles don't seem to work correctly for me and
I don't know how to fix them.  The include files aren't found
when the various directories are stepped into.

I screwed around until I got most everything to compile but can't
get build to link. ld complains about an invalid offset for _write.
Does anyone know what might cause this.

Thanks

Don Taber
news@solitary.stanford.edu


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

From: pthoms@castle.ed.ac.uk (P Thomson)
Subject: Compress/Uncompress (Were ?)
Date: 10 Apr 92 12:26:37 GMT


Hi Quick Q.    Can any body tell me were I can find Compress /Uncompress for
linux.

                Thanks in Advance Peter.

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

From: david@ods.com (David Engel)
Subject: Re: kernel building problems
Date: 10 Apr 92 16:29:06 GMT

news@solitary.Stanford.EDU (System test) writes:
: I've been going crazy trying to get either 0.95c or 0.95c+ to
: compile with gcc 2.1.  Which of these is the "second patch" to
: which H.J.'s patch is to be applied?  

It refers to 0.95c.  Unfortunately, Linus changed the Makefile around 
some in 0.95c+ causing H.J.'s patch and fix for combining the kernel
and gcc 2.1 headers to fail.

: In any event, the makefiles don't seem to work correctly for me and
: I don't know how to fix them.  The include files aren't found
: when the various directories are stepped into.
: 
: I screwed around until I got most everything to compile but can't
: get build to link. ld complains about an invalid offset for _write.
: Does anyone know what might cause this.

I got 0.95c+ to compile with gcc 2.1 by mucking with the Makefiles and
adding some symlinks in /usr/include to /usr/src/linux/include.  The
main thing to make sure of when compiling the kernel is that tools/build
should search /usr/include first and /usr/src/linux/include second, and
be linked with the gcc 2.1 libraries.  Everything else in the kernel
distribution (I'm not including ps) should be self-contained.

Aside to Linus:  Would you mind describing how you have your system set
up?  It might help aleviate problems like this if we knew which compiler
and system libraries you are using.

David
-- 
David Engel                        Optical Data Systems, Inc.
david@ods.com                      1101 E. Arapaho Road
(214) 234-6400                     Richardson, TX  75081

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

From: drew@hamlet.cs.colorado.edu (Drew Eckhardt)
Subject: Linux SCSI support
Date: Fri, 10 Apr 1992 16:48:33 GMT

The new version (.95c+ plus SCSI) is available.  It supports 
Adaptec, Seagate, and Ultrastor host adapters.

You can obtain it in the files 

headrest.woz.colorado.edu : 
/pub/linux/scsi.95c+.tar. 
/pub/linux/vmunix.Z (binary)

The Quick FAQ : 

Q : What SCSI hosts are supported?

A : The Adaptec 1542 (The EISA bus master 17xx may work),
        Seagate ST01/ST02, (These are 
        repackaged  Future Domain 88x series boards. 
        If you have an 8 bit future domain controller,
        you may be able to modify the seagate driver
        to detect your card to.  If any one does this,
        please send diffs.), and Ultrastor 14F are supported.


Q : What about SCSI tapes ? 

A : The stubs are sort of there.  Basically, the SCSI tape 
        driver needs to be changed from a block to character 
        device, a read/write function made that makes 
        appropriate calls to the lowlevel 
        scsi_do_cmd function, and enough 
        ioctl's calling the same function to
        implement a normal mt.

        My guess is these will be about as complex as the sd.c
        driver, or ~500 lines.

Q : what about extended partitions?

A : not yet.  

Q : I get a message saying READ CAPACITY FAILED.
        What does this mean?

A : In some cases, very old, brain dead drives (esp Seagate)
        don't support read capacity.  This is a warning - 
        the only real implication is that if you try to 
        read past the end of the drive, on the 
        raw device, it will have to timeout, instead of
        simply checking bounds and failing outright.

        In other cases, something just gets flakey.  
        I haven't had a chance to track this down,
        as until I patched .95c+, I didn't 
        have this problem.

Q : What are the major / minor numbers for SCSI drives?

A : Right now, it is theoretically possible to have 
        21 SCSI devices in the system.  With SCSI- 
        MFM / ESDI / SMD  adapters this can be even
        more drives, as each id may have several drives on it
        each set up as a LUN of that id.   Needless to say,
        if we have allowed for 16 partitions per SCSI drive,
        we're pushing the limits on minor numbers.  Therefore,
        a yucky dynamic numbering scheme is used.

        Block device major 8 is used for SCSI drives. I 
        personally like /dev/sd[n][p], where n is the drive
        number, and p partition letter.  
        
        Minors are assigned in increments of 16 to SCSI disks
        as they are found, scaning from host 0, ID 0 to host
        n, ID `7, excluding the host ID.  

        A minor where minor mod 16 = 0 is the whole drive,
        where minor mod 16 is between 1 and 4, that partition,
        and extended partitions are not yet implemented.

        Example : I have four SCSI disks, set up as follows 
        Seagate ST02, ID=0
        Seagate ST02, ID = 5
        Ultrastor 14, ID = 0
        Adaptec 1542, ID = 0

        The first disk on the seagate at ID 0 will become 
        minors 0-15 inclusive, the second at ID5 16-31
        inclusive, the disk on the Ultrastor 32-47, on the
        Adaptec 48-63.  

         
Q : The distribution kernel panics, saying that it can't mount
        root.

A : This is because my system mounts root from /dev/hd1 (old 
        numbering) or /dev/hda1 (new numbering), dev b 03 01.
        If you have no minix filesystem there, it will die.

        If you are booting from floppy, change the word at 
        offset 508 in the bootsector to  0000, or the
        appropriate value 

Q : How do I reduce kernel bloat and eliminate the drivers I
        don't want?

A : Simply #undef CONFIG_DISTRIBUTION in
        include/linux/config.h, and define the 
        macros for the SCSI hosts you want 
        enabled.

Q : What about distributions, information, etc

A : Subscribe to the SCSI mailing list, 

        linux-scsi-request@headrest.woz.colorado.edu

   and the latest release is maintained on
   headrest.woz.colorado.edu under /pub/scsi

Q : What about bugs?

A : It works on MY hardware.  It works on Tommy Thorn's 
        Adaptec system, Dave Gentzel's Ultrastor, and 
        quite a few other systems.  

        Bugs that are there will mostly be very hardware
        specific, and nasty to track down.  SCSI should 
        be basically error free - consequently, the 
        error code has not been heavily tested, and 
        I'm sure there are bugs left in there.  

        If you have found a bug, please mail it 
        to the mailing list with specifics of 
        your hardware.   Other people may have 
        the same problem, a solution, etc.  

        Chances are I will provide you with a debugging
        version of the kernel, which will dump out important 
        SCSI information and let me see what exactly is 
        going on.

Happy hacking...


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

From: johnsonm@stolaf.edu (Michael K. Johnson)
Subject: new stuff at tsx-11
Reply-To: johnsonm@stolaf.edu (Michael K. Johnson)
Date: Fri, 10 Apr 1992 17:18:47 GMT

tsx-11.mit.edu:
/pub/linux/binaries/usr.bin/uucp-1.03-bin.tar.Z
                            xcomm.tar.Z (sz/rz comm package)
/pub/linux/sources/usr.bin/uucp-1.03-src.tar.Z

michaelkjohnson
johnsonm@stolaf.edu

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

From: murthy@cs.cornell.edu (Chet Murthy)
Subject: Meta-Question
Date: 10 Apr 92 17:34:27 GMT

Please do not misunderstand this question.  I am very much 
in favor of what Linux is about.  What Ihave been wondering, though,
as I watch this group, and see the activity, and wish I had a laptop
to run Linux on, is this:

  What relation does Linux have to the GNU  Hurd Project?
  Essentially, what I'm trying to figure out, is why all
  this work on Linux is being done in what seems complete
  disconnection from GNU, when the aims of this group,
  and GNU in general, are almost identical.

(Of course, I mght be mistaken about that, but I think I saw
where Linus had written that he didn't want people to hoard
source code, or prevent its free dissemination, which, in a sense,
is the central spirit of GNU)

In any case, I think that what you guys are doing is great, and
first chance I get to get myself a 386/486 machine, I will be
ftp-ing Linux, and getting busy using it.

Thanks,
--chet--

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

From: fhg2@cunixb.cc.columbia.edu (Farhan H Garib)
Crossposted-To: alt.os.linux
Subject: Re: Problems with serial port drivers??
Date: 10 Apr 92 16:29:44 GMT
Reply-To: gs45@cunixb.cc.columbia.edu

In article <65178@apple.Apple.COM> erc@Apple.COM (Ed Carp) writes:
>Has anyone seen any problems with the serial port drivers under
>linux?  I'm running 0.95a and am having a hard time making uucp fly
>at 9600 baud and above, either straight through or with a modem.
>This same machine and comm board runs XENIX HDB UUCP flawlessly,
>and there have not been any reported complaints about taylor-uucp
>dropping packets, etc.
>
>This might not show up under light loads -- just under a heavy
>load, like uucp at 19200.
>-- 
>Ed Carp  N7EKG/6       erc@khijol.UUCP         erc@apple.com
>                       Cupertino, CA           408/252-5947
>
>-- Absolutely unabashed Gates McFadden groupie! --

Kermit seems to drop packets at 19200 also.

-Greg

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

From: fhg2@cunixb.cc.columbia.edu (Farhan H Garib)
Crossposted-To: alt.os.linux
Subject: Re: Problems with serial port drivers??
Date: 10 Apr 92 16:29:44 GMT
Reply-To: gs45@cunixb.cc.columbia.edu

In article <65178@apple.Apple.COM> erc@Apple.COM (Ed Carp) writes:
>Has anyone seen any problems with the serial port drivers under
>linux?  I'm running 0.95a and am having a hard time making uucp fly
>at 9600 baud and above, either straight through or with a modem.
>This same machine and comm board runs XENIX HDB UUCP flawlessly,
>and there have not been any reported complaints about taylor-uucp
>dropping packets, etc.
>
>This might not show up under light loads -- just under a heavy
>load, like uucp at 19200.
>-- 
>Ed Carp  N7EKG/6       erc@khijol.UUCP         erc@apple.com
>                       Cupertino, CA           408/252-5947
>
>-- Absolutely unabashed Gates McFadden groupie! --

Kermit seems to drop packets at 19200 also.

-Greg

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

From: murthy@cs.cornell.edu (Chet Murthy)
Subject: Re: Meta-Question
Date: Fri, 10 Apr 1992 17:55:06 GMT

Another question that occurred to me was:

Is there any relatoin between this project and the 386 BSD project?
It seems that their aims are quite similar; hence one might imagine
that quite a lot of work could be saved by collaboration.

But perhaps I'm just not understanding.  I really don't want to start a flame
war or anything - I think that the freer Unix is, the better.
But on the other hand, it sure would be a waste to constrct two
free unixes which are "almost", rather than one, which is "there".

--chet--


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

From: jwinstea@jarthur.claremont.edu (Jim Winstead Jr.)
Subject: Re: Meta-Question
Date: Fri, 10 Apr 1992 20:04:04 GMT

In article <1992Apr10.173427.15796@cs.cornell.edu> murthy@cs.cornell.edu (Chet Murthy) writes:
>Please do not misunderstand this question.  I am very much 
>in favor of what Linux is about.  What Ihave been wondering, though,
>as I watch this group, and see the activity, and wish I had a laptop
>to run Linux on, is this:

>  What relation does Linux have to the GNU  Hurd Project?

None, really.  The GNU Hurd is a POSIXized BSDish set of servers built
on top of the Mach 3.0 kernel, as I understand it.  Linux is a
POSIXized System Vish kernel, built from scratch.  (With a bit of
Minix influence....)

>  Essentially, what I'm trying to figure out, is why all
>  this work on Linux is being done in what seems complete
>  disconnection from GNU, when the aims of this group,
>  and GNU in general, are almost identical.

They are almost identical in the ends (free unix!), but the means are
quite different, as I see it.  GNU has fairly limited development team
working on the Hurd.  Linux has been available publically since almost
the beginning, and is going through a much wider, but less formalized,
alpha/beta testing process.

>(Of course, I mght be mistaken about that, but I think I saw
>where Linus had written that he didn't want people to hoard
>source code, or prevent its free dissemination, which, in a sense,
>is the central spirit of GNU)

It is, but the FSF does it in more of a business-like way.  They put
out real applications/utilities for widespread use.  As Linux stands
now, it is primarily a hacker's OS, and that's how it started.

Linux is, however, headed in the direction of widespread use.  You can
watch the activity in comp.os.linux and alt.os.linux and see the
number of people looking for "real" applications (spreadsheets,
databases) that would be necessary outside of the hacker's domain.

I don't know if I answered your questions very directly (or
correctly), but this is what Linux means to me, and seems to mean to a
vast number of people.
-- 
Jim Winstead Jr. (CSci '95)    | "Catch a fish!"
Harvey Mudd College            |  -Geddy Lee,
jwinstea@jarthur.Claremont.EDU |   San Diego Sports Arena
Disclaimer: Mine, not theirs!  |   January 20, 1992

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

From: kxb@matt.ksu.ksu.edu (Karl R. Buck)
Subject: Re: kernel building problems
Date: 10 Apr 1992 14:48:02 -0500

news@solitary.Stanford.EDU (System test) writes:

>I've been going crazy trying to get either 0.95c or 0.95c+ to
>compile with gcc 2.1.  Which of these is the "second patch" to
>which H.J.'s patch is to be applied?  

Yes, I'm having problems too. I recently posted asking specifically
what the "second patch" was and if it was the 0.95c patches that was
the latest (at that time). I received one helpful email message from
some helpful person saying that the "second" patch was included in
this diff. I assumed (perhaps wrongly) that the latest 0.95c+ sources
also include this "second patch."

>In any event, the makefiles don't seem to work correctly for me and
>I don't know how to fix them.  The include files aren't found
>when the various directories are stepped into.

What exactly are the errors?

Now, I'm (trying) to compile a virgin 0.95c+ kernel with GCC 2.1
without much luck. I had to fix a couple of minor things in the
Makefile and completely (I hope) got rid of the last bits of gcc 1.40. 

>I screwed around until I got most everything to compile but can't
>get build to link. ld complains about an invalid offset for _write.
>Does anyone know what might cause this.

Well, I'm having a similar problem. I get several warnings but finally
get the error:

ld -M boot/head.o init/main.o kernel/kernel.o mm/mm.o fs/fs.o
fs/minix/minix.o kernel/blk_drv/blk_drv.a kernel/chr_drv/chr_drv.a
kernel/math/math.a lib/lib.a -o /tmp/system > /tmp/System.map
ld: malformed input file (not rel or archive) kernel/blk_drv/blk_drv.a(`
.SYMDEF)
make: *** [/tmp/system] Error 1

I thought It would be that I was using the wrong ld or a.out.h, but I
have tried the recommended combination with no luck. I also got this
message when trying to compile the 0.95c sources.

I'd be interested in hearing from anyone who has successfully compiled
the 0.95c+ kernel with 2.1. Thanks!
-- 
Karl Buck, kxb@matt.ksu.edu     :)

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

From: jes@mbio.med.upenn.edu (Joe Smith)
Subject: Re: Problem with tar (?)
Date: 10 Apr 92 19:20:59 GMT

In article <4778@mccuts.uts.mcc.ac.uk> zlsiial@uts.mcc.ac.uk (A. V. Le Blanc) writes:


>   I believe this is by design.  Tar with z simply pipes its output
>   through compress, and compress doesn't support multi-volume output.
>
>       -- Owen
>       LeBlanc@mcc.ac.uk

Hmm.  In the tar.texinfo file (which is admitted to be out of date) we
find:

   Note that the -z option will not work with the -M option,
   or with the -u, -r, -A, or -D commands.

However, I tried it with tar-1.10 under SunOS and it worked for
reading and creating compressed archives.  Certainly if tar isn't
designed to do this, it ought to gag on the options, not at the end of
the first volume.

In the same thread, I caught the bugger last night trying to untar the
2.1lib.tar.Z file from tsx-11.  First, it died several times with
memory faults.  Uncompressing first before unpacking got me around
this.  It also created several files with totally bogus group id's:
the affected files listed (tar -t...) fine (owner/group 'hlu/other'
just like the others) but the resulting files had group number 37820,
or some such.  Is that a legal group id?  I don't have the souce to
check on the group id type (I thought it would be a 16-bit int, but
apparently not).

Then gcc on hello.c burped up "parse error before '('" again.  This is
most intriguing.  It's not an include file problem.  Any source with a
parenthesis, even 'int i = (2+3)*4;' will fail.  No includes
necessary.  Anyway, I ran sum on my PC, and on the unpacked binaries
on the Sun, and sure enough: cc1 had the right size, but a different
CRC (the other files from that tar were fine).  Hopefully,
transferring the raw, uncompressed cc1 will get gcc2.1 up.  I can tell
you the first thing I'm going to rebuild is tar & compress!

<Joe

--
 Joe Smith
 University of Pennsylvania                    jes@mbio.med.upenn.edu
 Dept. of Biochemistry and Biophysics          (215) 898-8348
 Philadelphia, PA 19104-6059

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

From: torvalds@klaava.Helsinki.FI (Linus Benedict Torvalds)
Subject: Re: Another bug?
Date: 10 Apr 92 20:39:16 GMT

In article <1992Apr9.200244.28901@wam.umd.edu> joel@wam.umd.edu (Joel M. Hoffman) writes:
>In article <1992Apr8.090321.7673@klaava.Helsinki.FI> torvalds@klaava.Helsinki.FI (Linus Benedict Torvalds) writes:
>>Well, only one known bug so far, but a couple of problems. I thought I'd
>>mention them before anyone else does, and we'll call them "features" :^)
>
>More than once, I've gotten a message about mis-linked processes.  I
>usually get it when I'm compiling from one VC and using Kermit
>interactively on another, but most of the time I'm compiling and using
>Kermit....

The mis-linked process problem seems to be linked with swapping: I've
never seen it, and most (all?) that have reported it have only 2M ram -
it seems to be a race-condition in the exit code, that breaks while
swapping heavily.  I think that running gcc and kermit on a 2M system
should be enough to swap quite a bit... 

I have fixed some of the exit-code in 0.95c+ - I'd be very interested to
hear if this problem is gone with the new kernel.  The exit-code was
"interesting" - I think tytso optimized the sibling linking a bit too
much, and it broke when paging meant that the kernel sleeps in some new
circumstances.  I might be totally off base - apologies to tytso if I
am. 

                        Linus

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


** 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 0.95a released on March 17, 1992

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