Category : UNIX Files
Archive : INTLX016.ZIP
Filename : INTLX016.TXT
Output of file : INTLX016.TXT contained in archive : INTLX016.ZIP
Subject: A couple of installation questions
Message-ID: <[email protected]>
Date: 20 Apr 92 16:12:31 GMT
Organization: Mankato State University
Lines: 12
I just downloaded the root and boot images and am having a few problems.
I decompressed the files and then rawrited them to 1.2 MB disks, this is where
the 1st problem is. When I rawrite the root disk it aborts with the message:
Adress Mark Not Found. It almost looks like the root image is bigger than
a 1.2 meg disk. (not by much though) I tried to run linux anyway ands it seems
to work fine. Also, Unix utilities like CAT and others are not included, do I
have to get those separatly, and if I do then will t I need to partition my
hardrive to use them. The root disk is already to full the way it is.
I know that these are novice questions, but hey I am a novice Unix user 🙂
Thanks for any advice
Kevin
[next article]
From: [email protected] (Haohui Wang)
Subject: Re: tape drive
Message-ID: <[email protected]>
Date: 20 Apr 92 15:17:34 GMT
References:
Sender: [email protected]
Distribution: comp
Organization: UB
Lines: 6
Nntp-Posting-Host: lictor.acsu.buffalo.edu
I have a Wangtek 60Mb tape drive, with Archive SC-499 QIC 24 controller.
Does anyone has a device driver for it? Or maybe source code I can look into?
thanks
howie
[next article]
From: [email protected] (Mun L Chan)
Subject: as86
Message-ID: <[email protected]>
Date: 20 Apr 92 15:22:28 GMT
Organization: Univ. of Pittsburgh, Computing & Information Services
Lines: 13
Hi,
After reading the FAQ about as86 (Bruce Evan's assembler) I decided to
get it. Is it a PC program?!! Can't seem to run it. Also, where can
I get the source and/or documentation for the assembler? Can someone
tell me why this assembler rather than other commercial assemblers
(i.e., Borland and MS) was used and if anyone has translated the
assembler source for these latter assemblers?
Many thanks,
M.L.Chan
[next article]
From: [email protected] (John B. Jr. Rogerson)
Subject: Kernel panic
Message-ID: <[email protected]>
Date: 20 Apr 92 16:53:44 GMT
Sender: [email protected] (USENET News System)
Organization: Princeton University
Lines: 21
Originator: [email protected]
Nntp-Posting-Host: phoenix.princeton.edu
could offer a suggestion on how to correct this. I copied by hand the
text of the panic message starting on the "login" line and I reproduce
it below in the hope that it will be meaningful to someone.
(none) login: general protection:0014
EIP: 0008:0000BF8E
EFLAGS: 00010046
fs: 0010
base: 00460000, limit: 00000018
Pid: 582, process nr: 0
8e e0 ff 05 b8 81 02 00 b0 20
Kernel panic: free_page_table called with wrong alignment
In swapper task - not syncing
One more question: If I wish to create a swapfile from my linux
partition (/dev/hda3) and I use the dd command with if=/dev/hda3 and
of=swapfile, can I be sure that swapfile will not contain any of my
linux files? I certainly wouldn't want them overwritten during
swapping.
Thank you. Jack Rogerson [email protected]
[next article]
From: [email protected] (Craig Setera)
Subject: Re: Yet another X question
Keywords: X Windows
Message-ID: <[email protected]>
Date: 20 Apr 92 17:00:53 GMT
References: <[email protected]> <1992Apr15.145806.16235@
ccu.umanitoba.ca> <[email protected]>
Organization: University of Minnesota, Minneapolis, CSci dept.
Lines: 17
>Yes, but this is the fault of stupid makefiles that leave .o's lying
>around. If you're clever and know what you're doing, you should be able to
>build it utility by utility, and I doubt that all the object and source for
>any one utility is more than a few megs.
>
>The reason X is such a hog is that its designers assumed people have a few
>hundred megs of disk lying around to waste.
>
>c
Ok, can someone make a reasonable estimation of what it would take to put
together JUST an X-Server? I'm interested in this for both Linux and possibly
for OS/2.
Craig
--
Craig Setera
([email protected])
[next article]
From: [email protected] (Joe Waters)
Subject: How do I PATCH stuff?
Message-ID:
Date: 15 Apr 92 16:47:33 GMT
Sender: [email protected]
Organization: NMSU Computer Science
Lines: 28
Okay, this is a semi-stupid question, BUT... I've looked all over for
docs for the patch util - can't find any, anywhere (I probably just
didn't look hard enough, but...) Can somebody tell me what the format
for patching diffs is? For example, I have the 2.1kernel.tar.Z
patches to apply to my 0.95c source code (in /usr/src/linux) - how do
I patch the untared/uncompressed file into whatever it's supposed to
be patched into? Same goes with the svga autoset patch...
E-mail would be nice since it appears as if everybody reading this
group except me already knows how to do this 🙂
OBTW Linus: Linux is way, way, way heavily great.
Thanks for any help anybody wants to give in advance,
Joe
/-------------------------+-------------------------------------------------\
| | NMSU doesn't care what I say as long as I don't |
| Joe Waters ("Falc") | use the /tmp drive. |
|([email protected]) | "Most any problem can be solved with duct tape |
| | and Froot Loops . . . or a good heavy club." |
\-------------------------+-------------------------------------------------/
--
/-------------------------+-------------------------------------------------\
| | NMSU doesn't care what I say as long as I don't |
| Joe Waters ("Falc") | use the /tmp drive. |
|([email protected]) | "Most any problem can be solved with duct tape |
| | and Froot Loops . . . or a good heavy club." |
\-------------------------+-------------------------------------------------/
[next article]
From: [email protected] (Joe Waters)
Subject: Question about svga autoset and selection mouse util
Message-ID:
Date: 17 Apr 92 06:22:33 GMT
Sender: [email protected]
Organization: NMSU Computer Science
Lines: 39
Okay, I have the new 0.95c+ kernel compiled and running, using
gcc2.1... Everything seems to work fine. Today I tried to install
the svga autoset patch that was posted here a few days ago. It
installed fine, and the kernel compiled and ran fine, with one slight
hitch: it didn't set the mode to the one I wanted. I have a paradise
512K card, and upon bootup I am usually given the choice of two modes:
0, which is 132x25, or 1, which is 132x43. I use mode 0 exclusively,
because my crummy monitor makes 132x43 come out hideously fuzzy. So I
followed the instructions on the patch and set -DSVGA_MODE=0 in the
top level make file, and it worked fine, the system didn't stop on
bootup to ask me what mode I wanted, but it defaulted to standard
80x24 text mode 🙁 I recompiled using the MODE=1 option, and it
worked - it came up in 132x43, as expected, only unfortunately hard to
read. So I think the problem has something to do with using a ZERO as
the mode number in the patch. Does anybody have any ideas/hints I
could follow towards working at a solution?
Oh, my other question. I also just tonight downloaded, patched, and
compiled the mouse cut/paste system SELECTION 1.0. Everything patched
and compiled fine once again, only this time I found myself looking
frantically through my /dev directory for a tty0 (the readme describes
it as "current tty") entry. I don't have one. Does anybody know the
correct mknod command to create one, or how I can go about getting a
/dev/tty0 into my system? Thanks very much for any help...
Joe
/-------------------------+-------------------------------------------------\
| | NMSU doesn't care what I say as long as I don't |
| Joe Waters ("Falc") | use the /tmp drive. |
|([email protected]) | "Most any problem can be solved with duct tape |
| | and Froot Loops . . . or a good heavy club." |
\-------------------------+-------------------------------------------------/
--
/-------------------------+-------------------------------------------------\
| | NMSU doesn't care what I say as long as I don't |
| Joe Waters ("Falc") | use the /tmp drive. |
|([email protected]) | "Most any problem can be solved with duct tape |
| | and Froot Loops . . . or a good heavy club." |
\-------------------------+-------------------------------------------------/
[next article]
From: [email protected] (Ted Dunning)
Subject: dynamic linking (was Re: GCC2.1 (Wonderful !))
Message-ID:
Date: 17 Apr 92 18:21:24 GMT
References: <[email protected]>
<[email protected]>
<[email protected]>
<[email protected]>
Sender: [email protected]
Reply-To: [email protected]
Organization: Computing Research Lab
Lines: 29
In-Reply-To: [email protected]'s message of 16 Apr 92 10:22:20 GMT
dynamic linking does not have to involve running bin/ld.
in the linux context, the only thing that has to be dynamicly linked
are the standard libraries. this can easily be done without either
the overhead of running ld, or the current kludge of staticly linking
programs with a stubs library pointing to a reserved part of virtual
memory.
the simplest approach is to just use a jump table and plug the stubs
into the jump table. this way, the actual version of the library can
change, and as long as the jump table is compatible, things will work
just fine. the startup routine in crt0 can even check to make sure
that the currently visible jump table is a compatible version.
on the other hand, true dynamic linking can also be done quite
handily. all that has to be done is make the stubs a bit more
intelligent. in particular, let each stub know its own name and write
a procedure which will resolve the jump table as needed. each stubs
library would have its own jump table in which all entries initially
point to a routine which resolves that entry and then calls the new
entry. all subsequent calls would be directed through the jump table.
but in fact, i can't see any compelling reasons why the first option
isn't just as acceptable as the second. it does feel a bit more
static and confining, but i can't see any reason that is a problem
here.
[next article]
From: [email protected] (John B. Jr. Rogerson)
Subject: Kernel panic
Message-ID: <[email protected]>
Date: 20 Apr 92 17:24:51 GMT
Sender: [email protected] (USENET News System)
Organization: Princeton University
Lines: 26
Originator: [email protected]
Nntp-Posting-Host: phoenix.princeton.edu
I'm very happy with the Linux operating system and send my thanks
to the developer(s). However, when booting up, about 10 percent of the
time I receive a kernel panic message. I would be most grateful if one
of the Linux system wizards could offer a suggestion on how to correct
this. I copied by hand the text of the panic message starting on the
"login" line and I reproduce it below in the hope that it will be
meaningful to someone.
(none) login: general protection:0014
EIP: 0008:0000BF8E
EFLAGS: 00010046
fs: 0010
base: 00460000, limit: 00000018
Pid: 582, process nr: 0
8e e0 ff 05 b8 81 02 00 b0 20
Kernel panic: free_page_table called with wrong alignment
In swapper task - not syncing
One more question: If I wish to create a swapfile from my linux
partition (/dev/hda3) and I use the dd command with if=/dev/hda3 and
of=swapfile, can I be sure that swapfile will not contain any of my
linux files? I certainly wouldn't want them overwritten during
swapping.
Thank you. Jack Rogerson [email protected]
[next article]
From: [email protected] (Joe Waters)
Subject: Question about svga preset patch
Message-ID:
Date: 20 Apr 92 07:01:08 GMT
Sender: [email protected]
Organization: NMSU Computer Science
Lines: 31
Okay, I got the 0.95c+ kernel patched and compiled with no problem,
and have lp and ps working, and I've nearly got a completely working
Unix system that's faster than just about anything I've seen for an
86-based machine... Kudos to all working on the project...
Now, a question of whoever (I lost his address) the guy is who wrote
the svga-preset patch. I patched my source with the patch, and it
compiled and ran fine, with one problem. My desired svga mode is mode
zero on my machine (paradise video card) - 132x25. Anything higher
than that (my only other choice is mode one, 132x43) is too fuzzy on
my cheap crummy monitor... The patch works fine with the video mode
set to one and compiled, but if I set it to zero (my preferred mode),
it compiles and runs, but defaults to standard 80x25 VGA mode. Is
there any fix available, even if it means patching other sources?
Thanks for any help...
Joe
/-------------------------+-------------------------------------------------\
| | NMSU doesn't care what I say as long as I don't |
| Joe Waters ("Falc") | use the /tmp drive. |
|([email protected]) | "Most any problem can be solved with duct tape |
| | and Froot Loops . . . or a good heavy club." |
\-------------------------+-------------------------------------------------/
--
/-------------------------+-------------------------------------------------\
| | NMSU doesn't care what I say as long as I don't |
| Joe Waters ("Falc") | use the /tmp drive. |
|([email protected]) | "Most any problem can be solved with duct tape |
| | and Froot Loops . . . or a good heavy club." |
\-------------------------+-------------------------------------------------/
[next article]
From: chojnows@osiris (Brian Chojnowski)
Subject: Re: Graphics and IPC questions...
Message-ID: <[email protected]>
Date: 20 Apr 92 17:59:09 GMT
References:
Sender: [email protected] (News)
Organization: University of Illinois at Urbana
Lines: 15
"Alex R.N. Wetmore"
>Why not implement these in vt100 mode, rather then with direct screen
>writes then? Linux seems to have very fast terminal emulation (at least
>on my 386/16 with a very old Paradise EGA), and you might as well take
>advantage of those using termcap. That way people on terminals (and
>telnet sessions, when those exist), can use the software too. It also
>doesn't restrict the software to linux, but allows it to work on any
>other unix platform (in theory at least).
I guess I havent made myself clear. I would like to do something like this,
where I would create a superset of the vt100. But the mode as it is, is
insufficient for the kind of graphics I am talking about. I want to also add
this new mode to the telnet program (when it gets written), so that the
graphics get displayed remotely.
[next article]
From: [email protected] (Darren Senn)
Subject: Re: TEX for Linux at tsx
Message-ID: <[email protected]>
Date: 20 Apr 92 19:57:04 GMT
References: <[email protected]>
Sender: [email protected]
Lines: 10
Speaking of TeX, is there a pascal compiler that is portable/has been ported to
linux?
I know where are the TeX source code is, but it's rather useless without the pas
cal
compiler. I've got the source for weave and tangle, so the fact that TeX is in
WEB isn't much of a problem.
Thanks much.
--
Darren Senn Phone: (408) 479-1521
[email protected] Snail: 1785 Halterman #1
Wasurenaide -- doko e itte mo soko ni anata wa iru yo. Santa Cruz, Ca 95062
[next article]
From: [email protected] (Curtis Yarvin)
Subject: Re: Yet another X question
Keywords: X Windows
Message-ID: <[email protected]>
Date: 20 Apr 92 20:07:56 GMT
References: <[email protected]> <1992Apr20.034455.24294@cs
.brown.edu> <[email protected]>
Sender: [email protected]
Organization: Brown University Department of Computer Science
Lines: 11
In article <[email protected]> [email protected] (Craig Setera)
writes:
>
>Ok, can someone make a reasonable estimation of what it would take to put
>together JUST an X-Server? I'm interested in this for both Linux and possibly
>for OS/2.
Five or six megs.
This estimate is on the high side.
c
[next article]
From: [email protected]
Subject: What next?
Summary: Suffering from FAQ-tigue...
Message-ID: <[email protected]>
Date: 20 Apr 92 20:18:58 GMT
Sender: [email protected] (USENET News System)
Reply-To: [email protected]
Organization: Jus' me.
Lines: 9
Saluton!
I've just gotten Linux 0.95a up and bootable (from floppy) on my HD.
So far, so good! ...but it isn't very handy just yet, y'know?
What next? Heck, how do I even get anything *into* Linux? Do I have
to rawrite .tar.Z'd files to a floppy, and read stuff in that way or what?
Rick Miller [email protected] [email protected] [email protected]
[next article]
From: [email protected] (Chris Justus)
Subject: Basic help with partitioning
Message-ID: <[email protected]>
Date: 20 Apr 92 17:48:07 GMT
Sender: [email protected]
Organization: Communications Research Centre, Ottawa, Canada
Lines: 32
Nntp-Posting-Host: debra.dgbt.doc.ca
I need some help with partitioning my harddrive - as I don't know
ANYTHING about partitions. I've made the boot and root disks for
Linux and have decided to run it from my hardrive. Currently, I have
a 120Meg drive (and just 1 partition). It was already formatted for
me with DOS 5.0 already installed. Now - from what I've read -
repartitioning my drive with erase all the data - that's fine, I've
backed everything up. I take it I'll have to reinstall DOS on my
hardrive (??). To make it DOS bootable - do I install DOS on the
primary partition?? Will I have to reformat my drive (I really don't
have a clue - the manuals I have are QUITE brief and don't mention
disk partitions at all. 🙁 ) The docs for linux say that I should put
the linux file system on the primary partition - why??? what will
happen if I don't??? From what I've read - I'll need about a 30 Meg
partition for Linux, is this right??? I'll need to run the compiler,
terminal programs (for the modem), and maybe rayshade (has anyone
compiled this for linux??). The remainder of the drive will go to DOS
as that's what all my software is right now...
Getting back to partitioning - should I partition from DOS
with fdisk or from linux with pfdisk(??) What's the difference?? It is
critical that I be able to boot my computer with DOS - any
suggestions, etc would be greatly appreciated as I don't have the
slightest CLUE - and it would be nice to not completely TRASH my
drive.
Email me as there's no sense in wasting bandwidth on something so
basic.
Thanks in Advance!!
--
Chris Justus [email protected]
Co-Op Student at Communications Research Centre, Communications Canada, Ottawa
[next article]
From: [email protected] (Dave Hirvela)
Subject: Gateway 2000 486/33/
Summary: help
Message-ID:
Date: 20 Apr 92 20:15:52 GMT
Sender: [email protected]
Lines: 38
As the proud new owner of a Gateway 2000 386/33 EISA system I would
like to get into the wonderful world of desktop unix. As BSD's version
seems to be a bit too flakey at the moment, I'd like to install Linux
and have been reading this group for a while getting a feel for some
of the problems I may run into.
My setup is the full up Gateway configuration, with a 8514/a Ultra graphics card
a 340 meg scsi drive with a 200 and 80 meg dos partion leaving 60 megs free
for a kinder and gentler os.
As far as I can tell, I should be able to run off floppies fine but I'm wonderin
g if anyone has got a machine just like this one going with one of the several s
csi patches out there.
My eisa scsi card is an Adaptec AHA-1742A currently in non bus mastering
mode and driving my two floppies and the 340 meg hd.
I have a v.32 modem going to an apollo network, so ftping will not be
too much a a problem, but I doubt I'll be able or want to recompile the
kernal unless it's absolutely necessary as I tend to use unix apps rather
than write them.
So, does anyone have this config running ? Or can provide any hints ?
I'll have to keep the rest of my system (windows 3.1 and Dos) so some
sort of boot block swapping setup would be nice so that I could boot up
from either at will without messing with floppies.
Thanks for the help, replies here or my mail address. If there is an
FAQ for scsi, please point me to it or mail it to me, I haven't seen it
and don't really want to get on any more mailing lists... 🙂
Dave
:wq
--
Any opinions expressed are obviously wrong. * [email protected]
Save Ferris ! Use only optical media ! * [email protected]
Try Telemate-PC and save feeping creatures ! * Booster Gold in mail on:
"I don't think that word means what you * grind.isca.uiowa.edu login:
think it does." _The Princess Bride_ * iscabbs - Try it today !!!
[next article]
From: [email protected] (Bob Rusk)
Newsgroups: comp.os.os2.misc,comp.os.linux
Subject: Re: Any comments on Quantum 240A ??
Message-ID:
Date: 20 Apr 92 21:07:43 GMT
References: <[email protected]>
Sender: [email protected]
Reply-To: [email protected]
Followup-To: comp.os.os2.misc
Organization: Harris Computer Systems Division
Lines: 7
In-reply-to: [email protected]'s message of 19 Apr 92 00:27:24 GMT
I have a Quantum 240A in my early Gateway 2000 386/20. It runs Linux
(and MS-DOS and MS-WINDOWS) with no problems. The drive is quite fast,
fairly quiet, reasonably priced, and well-gauranteed. I love it!
--
--
Bob Rusk
[email protected]
[next article]
From: [email protected] (Mika Pekka Liljeberg)
Subject: Drive performance in general
Message-ID:
Date: 20 Apr 92 20:50:39 GMT
References: <[email protected]>
Sender: [email protected] (Uutis Ankka)
Organization: Department of Computer Science, University of Helsinki, Finland
Lines: 78
In-Reply-To: [email protected]'s message of 18 Apr 92 16: 31:28 GMT
In article <[email protected]> [email protected]
(Guido Kueppers) wrote:
[stuff deleted]
> This is a test with the Seagate ST251 MFM drive and WD1006V controller:
> =======
> bash#df
>
> inodes inodes inodes blocks blocks blocks mount
> devices total used free total used free point
> -------------------------------------------------------------------
> /dev/hda1 6525 125 6400 199575 1373 18202 /
> dev/hda2 7395 3 7392 22185 239 21946 mnt/
> bash# sync
> bash# time cp /dev/kmem /kmem.dump <--- write 4MB from memory to disk
> 0.17u 55.08s 1:00.25 8pf
> ^^^^^^^
> bash# ls -l /kmem.dump
> -rw------- 1 root root 4194304 Apr 17 21:06 /kmem.dump
> bash# sync
> bash# time cp /kmem.dump /mnt/ <--- copy a 4MB file from hda1 to hda2
> 0.18u 10.12s 2:35.36 8pf
> ^^^^^^^
>
> bash#
>
> This is the same test performed on a Rodime 3259A IDE drive:
> =======
>
> / # df
>
> inodes inodes inodes blocks blocks blocks mount
> device total used free total used free point
> ------------------------------------------------------------------
> /dev/hda2 18504 163 18341 55512 14094 41418 /
> /dev/hda3 18504 706 17798 55512 19165 36347 /usr
> /dev/hda4 18576 1007 17569 55728 18105 37623 /usr/local
> / # sync
> / # time cp /dev/kmem /kmem.dump
> 0.19u 62.66s 1:28.08 8pf <--- About 50% more than with the MFM driv
e
>
> / # ls -l /kmem.dump
> -rw------ 1 root root 4194304 Apr 17 21:30 /kmem.dump
> / # sync
> / # time cp /kmem.dump /usr/local/bin/
> 0.23u 14.18s 2:58.79 8pf <--- about 30 secs more than with the MFM
> drive
>
> I don't know what this sort of testing is worth after all, but it seems to
> substantiate my doubts about IDE drive performance. Perhaps the problem
> lies with the particular drive I'm using. Perhaps it has to do with the IDE
> drive's internal buffering scheme. Someone willing to comment on this?
>
> Guido
I'm kind of suprised no-one has thought to run one of these tests
under good ole' MSDOGgie. Well, I did. I have a largish Prime RLL
drive with 1:1 interleave and a 26 ms average seek time. I copied
a 4.69 MB file from one partition to another using minimal buffering
(no cache, only 8 DOS buffers).
The whole operation took 27 seconds!!!
This bears some thinking about. Obviously the standard Linux HD driver
could use the attentions of a determined optimizing programmer. 😉
I understand that Linus has a lot of things with higher priorities to do,
but perhaps some of those guys, who liked to write turbo loaders for the
Commodore 64 are still around and would like another challenging project. 😉
Just thought I'd add my two, uhum, pennies worth.
Mika
--
Mika Liljeberg Email: [email protected]
Helsinki University [email protected]
Dept. of Computer Science
[next article]
From: [email protected] (Budi Rahardjo)
Subject: (0.95c+) How can I use my HD as the root directory ?
Keywords: boot 0.95c+
Message-ID: <[email protected]>
Date: 20 Apr 92 21:44:16 GMT
Sender: [email protected]
Followup-To: comp.is.linux
Organization: University of Manitoba
Lines: 18
Nntp-Posting-Host: ic11.ee.umanitoba.ca
I have just downloaded bootimage 0.95c+ and couldn't get it to use the
harddisk as root. INSTALL-0.11 says:
Once you have happily made a new root, you will want to boot up
with it. This is done by changing a word at offset 508 in the
boot-image. The word (in 386-order, ie low byte first) tells the
system which device to use as root:
I tried it and it didn't work. (It used to work with 0.95a though).
Did I miss something ?. Thanks.
-- budi
--
Budi Rahardjo
Electrical Engineering - University of Manitoba - Canada
[next article]
From: [email protected] (Adam Justin Thornton)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Uploaded _Dungeon_
Message-ID: <[email protected]>
Date: 20 Apr 92 21:23:05 GMT
Sender: [email protected] (News)
Organization: Milo's Meadow
Lines: 12
to tsx-11.mit.edu. Right now it's sitting in /incoming.
dungeon.Z is the compressed, statically linked binary without the Great
Debugging Tool. dungeon.tar.Z contains source, Makefile, and History,
as well as the binary.
Adam
--
"Adam Thornton plays the homosexual Horatio without succumbing to _too_ many
cliched stereotypes. His Horatio, Hamlet's bosom companion...has a penchant
for cream dresses and thigh-high black leather boots." | [email protected]
Rice and I don't share opinions, thankfully. | Retry Student A | 64,928 | 😉
[next article]
From: [email protected] (Michael Pereckas)
Subject: Re: shared versions of tex/mf/dvips available
Message-ID:
Date: 20 Apr 92 22:23:16 GMT
References: <[email protected]>
Sender: [email protected] (News)
Organization: University of Illinois at Urbana
Lines: 46
[email protected] (Peter Williams 8169821) writes:
>*.a files for shared versions of tex/mf/dvi programs are available in
>texmf.a.tar.Z and dvips.a.tar.Z in pub/linux at archsci.arch.su.oz.au
>(129.78.66.1)
>Makefiles are included (just change the target directory for the binaries)
>Remember tex, latex and slitex are created by linking to virtex and mf is
>created by linking to virmf.
I obtained this distribution last night and have been messing with it
all day. This hardly constitutes an exhaustive test, but it appears
to be complete, correct, and documented. Be advised that metafont
comes without any bases or .mf files, so you have to get them
separately. To generate plain.base, create the following simple file,
plain.ini:
input plain;
input local;
dump
Then run
inimf plain.ini
while in the bases directory, usually /usr/local/lib/mf/bases
The .mf files should go in /usr/local/lib/mf/macros
local.mf should contain the definitions for the devices you will use.
The MS-DOS \TeX\ and metafont system emTeX includes a local.mf with
modes for 300dpi normal laser printers, epson fx and lq printers
(360x180 and 360x360 dpi), and some others, and is what I used, since
I have emTeX installed.
I'm using dvips to produce postscript output which I print on my Epson
lq compatible printer with ghostscript. This works quite well. Dvips
can run metafont automagically to produce .pk bitmap files for any
fonts it needs if the fonts are not already available. Under DOS this
is not automatic....
This all installed easily and appears to work fine. There are man
pages for all the programs, which are most helpful. Format files are
included for plain TeX, LaTeX, and SliTeX. This is an excellent
distribution, and I encourage other Linux FTP sites to pick it up.
Michael Pereckas
[next article]
From: [email protected] (David A. Sinclair)
Newsgroups: comp.os.os2.misc,comp.os.linux
Subject: Re: Any comments on Quantum 240A ??
Summary: it's a good drive -- hard to find
Message-ID: <[email protected]>
Date: 20 Apr 92 22:40:58 GMT
References: <[email protected]>
Sender: [email protected]
Followup-To: comp.os.os2.misc,comp.os.linux
Organization: Brandeis University
Lines: 39
In a recent post, I asked for people's evaluation of the Quantum
LPS240AT drive, particularly its useability for OS/2 2.0 and for
Linux. All of the many replies I received were positive --
apparently, it's a really good drive. Unfortunately, it is also very
hard to find in stock!! I did, however, find what I believe is a good
substitute at a good price:
Seagate ST3283A
245 meg
IDE, 3.5" HH
12 (!!) ms access
128K cache
The cache is smaller than the Quantum's, but the access speed a bit
faster. I ordered mine today from Elco computer for a price of $605.
(That seemed like an okay price to me...(??))
If there is any interest, I will post a followup after having received
the drive and installed both OS/2 and Linux on it (and used it a bit!!
🙂
Cheers,
David Sinclair
[email protected]
disclaimer: I have absolutely no affiliation with either Elco
computer, or Seagate...I'm not even sure I'm a satisfied customer
yet!! 🙂
--
-----------------------------------------------------------------
David A. Sinclair - "Chester" | Bobby Fischer doesn't *want*
[email protected] | you to know where he is.
-----------------------------------------------------------------
[next article]
From: [email protected] (Peter Williams 8169821)
Subject: dvips.a.tar.Z broken. Sorry!
Message-ID: <[email protected]>
Date: 20 Apr 92 23:36:19 GMT
Sender: [email protected]
Organization: Design Computing Unit, The University of Sydney
Lines: 9
Nntp-Posting-Host: dean.arch.su.oz.au
The dvips.a.tar.Z that I posted yesterday seems to be broken so I have
withdrawn it temporarily. Sorry!
Should be mended and back up tomorrow.
--
Peter Williams |e-mail: [email protected]
Key Centre for Design Quality |phone: +61-2-692 2053 or +61-2-660 6156
University of Sydney |+++++++++++++++++++++++++++++++++++++++
[next article]
From: [email protected] (Mike Me)
Subject: Re: Gateway 2000 486/33/
Message-ID: <[email protected]>
Date: 20 Apr 92 21:48:31 GMT
References:
Sender: [email protected]
Organization: George Washington University
Lines: 53
In article
irvela) writes:
>My setup is the full up Gateway configuration 486 - 33 EISA, with a
8514/a Ultra graphics carda 340 meg scsi drive with a 200 and 80 meg
dos partion leaving 60 megs free for a kinder and gentler os.
>
>As far as I can tell, I should be able to run off floppies fine but I'm wonderi
ng if anyone has got a machine just like this one going with one of the several
scsi patches out there.
>
>My eisa scsi card is an Adaptec AHA-1742A currently in non bus mastering
>mode and driving my two floppies and the 340 meg hd.
>
Well, I'm in the same boat.
I have the same system, (only bigger drive).
Anyway this morning I was messing with it again, in the hope I could
get the SCSI patch to recognise my AHA-1742.
Here is the deal..... (at least partially)
The way the EISA system comes from gateway, they set the 1742 for
Enhanced operation.
when you run the scsi boot kernel It'll choke on it BIG-TIME!!
Use the EISA setup disk to switch it back down to standard mode.
Sellect the top entry in each one of the fields of the chooices.
(ie it'll give you a list of choices for int#, take the top one)
Ok, now switch the starting/boot address to 0C8000,
(also on the EISA config. disk)
Save the crap & restart.
The scsi boot image WILL (should) now recognize your drive as an
AHA-154x.
I still can not get fdisk (on linux) to work as it should, but I think
I'm real!! close.
fdisk still reports nothing (Not a damn thing!!!) I dont know why,
but at least the host/drive is recognised by the kernel/bootimage.
If anyone has a clue, please let us know.
Also people that have it running with the 154x adapter card, can you
send / post the numbers the kernel comes up with on bootup?
That will give me something to shoot for.
Who knows, maybee we can get this thing to work for EISA after-all.
Thanx for the help,
Dave if you have any more questions, e-amil me,
[next article]
From: [email protected] (A. V. Le Blanc)
Subject: Re: Partition tables suggestions...
Message-ID: <[email protected]>
Date: 20 Apr 92 10:24:19 GMT
References:
Followup-To: [email protected]
Organization: Computing Centre, University of Manchester
Lines: 18
In article
>I was wondering, would it be possible to implement logical partitions
>within a Linux physical partition? This would help us get around the
>4 partition limit in MS-DOS. The idea would be simmilar to what
>MS-DOS does with the Extended Partition can be partitioned into
>Logical drives.
This in fact already exists. You'll need a kernel of at least
0.95c+, and my fdisk 0.92 or higher. A complete installation version
of Linux 0.95c+ with fdisk 0.92, which together can create and use
up to 60 partitions on a hard disk, is available my anonymous ftp
from ftp.mcc.ac.uk [130.88.200.7] and from banjo.concert.net;
at MCC it is in /pub/linux/mcc-interim/0.95c+, and at banjo in
/pub/Linux/mirrors/uk.mirror/mcc-interim/0.95c+. Please read the
README files before installing.
-- Owen
[email protected]
[next article]
From: [email protected] (A. V. Le Blanc)
Subject: Re: expr, patches
Message-ID: <[email protected]>
Date: 20 Apr 92 11:13:14 GMT
References:
alter.bellcore.com> <[email protected]>
Followup-To: [email protected]
Organization: Computing Centre, University of Manchester
Lines: 20
In article <[email protected]> [email protected] (tim)
writes:
>
>are diffs for the shellutils, binutils, and fileutils available
>anywhere ? many of these (includeing expr since i changed to gcc 2.1)
>give alot of errors when i try to compile them.
As part of my project of making available all sources, patches, and
unlinked binaries used to create the mcc 'interim' version of Linux,
I have created directories /pub/linux/mcc-interim/0.95c+/patches and
/pub/linux/mcc-interim/0.95c+/files.a. Among other things, the
directories currently contain all patches for bash-1.12,
fileutils-3.2, groff-1.05, shellutils-1.6, and textutils-1.3.
The files.a directory contains unlinked binaries that can be used to
produce executables for all these commands, even if you do not wish
to get or install the mcc version; all you need is a linker and some
libraries. Mind you, I don't think these will work with gcc 1.40,
since the libraries are very dissimilar.
-- Owen
[email protected]
[next article]
From: [email protected] (H. Peter Anvin N9ITP)
Subject: Re: problems booting
Message-ID: <[email protected]>
Date: 20 Apr 92 20:50:00 GMT
References: <[email protected]>
Reply-To: [email protected] (Peter Anvin)
Organization: Northwestern University, Evanston, Illinois, USA
Lines: 25
In article <[email protected]> of comp.os.linux,
[email protected] (Richard Alan Brown) writes:
>
> A problem, not from rab@ariel... but it may interest him also.
>
> I used rawrite to make a bootable disk and a rootimage disk. When I
> boot my pc with the floppy in, I get the loading............
> and the SVGA mode question.
> when I press any other key, the system simply reboots, and cycles through
> this same process.
I experience exactly the same phenomenon on a 486/33 ISA with the OPTI
chipset and motherboard and an AMI 91/06/06 BIOS, an Orchid ProDesigner IIs
SuperVGA card and an IDE hard disk (Fujitsu).
I have heard it suggested that it *might* be due to a faulty IDE host
adapter card. I will have to find out on Wednesday whem the new card
comes.
/hpa
--
INTERNET: [email protected] TALK: [email protected]
BITNET: HPA@NUACC HAM RADIO: N9ITP, SM4TKN
IRC: Scalar NeXTMAIL: [email protected]
"Kissing a smoker is like licking an ashtray." -- Anon.
[next article]
From: [email protected] (Satish Chittamuru)
Newsgroups: comp.os.linux,comp.sys.ibm.pc
Subject: SCSI of ISA vs SCSI on EISA
Message-ID: <[email protected]>
Date: 16 Apr 92 01:29:36 GMT
Reply-To: [email protected] (Satish Chittamuru)
Followup-To: comp.os.linux
Organization: Microsoft Corp., Redmond WA
Lines: 36
Now that Linux can work with SCSI drives, I am thinking of getting a SCSI
controller instead of the IDE I have been considering so far. However I
am not sure if I will be utilizing the of SCSI setup efficiently on an ISA bus.
Basically the drive I am ordering is the Quantum LPS 240. It has a peak
transfer rate of 3.8 MBytes/sec (Quantum specs) and since it has a
sophisticated caching algorithm (with 256K cache), I am thinking it can
sustain a rate of 2.5MB+ per sec. The question is, can SCSI on the ISA bus
sustain this rate of transfer, especially once X is up and running on Linux.
I haven't delved deeply into the internals of a PC. But I keep reading of
machines with the ISA bus running at 8MHz, 10MHz or even 12.5MHz. Can the
bus be pushed to 12.5MHz? The machine will be a 486/33. Can I really make
use of the expensive SCSI controller and fast drive on an ISA bus or does
it make sense to shovel over more moola for an EISA bus? Could more
knowledgeable souls share their thoughts on the subject? Any comments,
suggestions, pointers would be greatly appreciated.
BTW, I am not getting the SCSI solely for speed. I believe even IDE can
theoretically transfer more than 5MB/sec. But I eventually want to add
a Floptical or the Quad Flextra from Brier Technologies. They are scsi
devices, I believe. No sense in paying for 2 controllers.
-csk.
p.s.: I am considering the Adaptec 1542b and the Ultrastor 14F controllersi
since they have bus mastering capability. Are there better/less expensive
controllers? If not, which of these two is better.
p.p.s: BTW, what book would you recommend for gaining a thorough knowledge
of the internals of a PC (tech details). A few that have been recommended are
the IBM technical reference manual and the Winn Rosch Hardware Book (I
don't remember the exact title).
--
Satish K. Chittamuru [email protected]
Microsoft Corporation Redmond, WA
=====
Veni! Vidi! Visa! (I came! I saw! I purchased!). -- Julius, Sieze Her!
[next article]
From: enewton@uucp (Eric C. Newton x2092)
Subject: Re: gdb(1) and ptrace(2)
Message-ID:
Date: 17 Apr 92 12:10:10 GMT
References: <[email protected]>
Sender: [email protected] (Usenet news user)
Reply-To: uunet.uu.net!srg!enewton
Organization: means "in a rut"
Lines: 39
In-Reply-To: [email protected]'s message of Thu, 16 Apr 1992 11:19:55
GMT
Path: srg!uunet!uunet!boulder!ophelia.cs.colorado.edu!drew
From: [email protected] (Drew Eckhardt)
Summary: gdb and ptrace
Organization: University of Colorado at Boulder
Date: Thu, 16 Apr 1992 11:19:55 GMT
1. gdb does not work with the new .95c+ kernel.
Breakpoints end up in wierd places, etc.
My guess is that internal structures changed
enough from .95a to .95c to break things
(ie task_info, etc)....
I don't still have my .95a sources to verify this,
but if it isn't just me someone should recompile
gdb with the new kernel header files.
(I have 8M free on my disk, and I don't think
it's enough.)
I have the whole beastie compiled on my machine at home. The problem
I have is that the ptrace() function is not in libc.a. I wrote a
dummy ptrace() function to link the program.
3. GDB 4.5 has been released. Is anyone building this?>
Yes. I will after 4.4 works.
Anyone care to volunteer the sources for making a ptrace system
call?
-Eric
Remember to mail to: uunet.uu.net!srg!enewton "The more you complain,
--
Remember to mail to: uunet.uu.net!srg!enewton "The more you complain,
[next article]
From: [email protected] (William Michael Lye)
Subject: Re: Drive performance in general
Message-ID:
Date: 21 Apr 92 02:42:24 GMT
References: <[email protected]>
Sender: [email protected]
Organization: Simon Fraser University, Burnaby, B.C., Canada
Lines: 29
[email protected] (Mika Pekka Liljeberg) writes:
>In article <[email protected]> [email protected]
e (Guido Kueppers) wrote:
>This bears some thinking about. Obviously the standard Linux HD driver
>could use the attentions of a determined optimizing programmer. 😉
>I understand that Linus has a lot of things with higher priorities to do,
>but perhaps some of those guys, who liked to write turbo loaders for the
>Commodore 64 are still around and would like another challenging project. 😉
>Just thought I'd add my two, uhum, pennies worth.
> Mika
>--
>Mika Liljeberg Email: [email protected]
>Helsinki University [email protected]
>Dept. of Computer Science
Oh, you mean, someone like lil' ol' me? I'm only *just* starting to get into
low-level programming on my clone. Its a real bugger to get hold of the info
I need. (I mean, what's this DMA stuff? 🙂 Seriously, if someone can
suggest a GOOD source of information on the low-level accessing of the drives
on a clone, I'm game to start bashing away, that is if someone hasn't already
done something similar...
--
Bill Lye, [email protected]
SFU claims these views....NOT!
[next article]
From: [email protected] (John B. Jr. Rogerson)
Subject: re: Kernel panic
Message-ID: <[email protected]>
Date: 21 Apr 92 02:06:07 GMT
Sender: [email protected] (USENET News System)
Distribution: usa
Organization: Princeton University
Lines: 10
Originator: [email protected]
Nntp-Posting-Host: phoenix.princeton.edu
With regard to a recent note of mine with the subject "Kernel
panic", I received a welcome instructional email note from [email protected]
that I should have included the level of the system that I am using and
the equipment that it is running on. Herewith, I offer the missing
information.
Linux system version 0.95c+, running on a 25Mhz 386 machine by
Gateway 2000 with 4Mb memory.
Regards, Jack Rogerson [email protected]
[next article]
From: [email protected] (Michael Pereckas)
Subject: Help, I can't backup my hard disk
Keywords: backup
Message-ID:
Date: 21 Apr 92 04:24:46 GMT
Sender: [email protected] (News)
Organization: University of Illinois at Urbana
Lines: 15
I can't back up my hard disk. Gnu tar 1.10 can't read files in
multivolume archives that are split across disks, cpio and pax can't
create multivolume archives at all, and diskbackup/diskrestore fails
randomly.
Does anyone have any ideas?
I'm running 0.95c+ on a 2Mbyte system with an MFM harddrive. I'm
using floppy drive /dev/PS1, which works fine with mounted filesystems
and single-volume tars. /dev/at0 causes the same problems with tar
(and db/dr modified for the smaller disks).
Michael Pereckas
[next article]
From: [email protected] (Ben Combee)
Subject: Oddities with MCC Release
Message-ID: <[email protected]>
Date: 21 Apr 92 04:45:29 GMT
Organization: Home of the Big G Burger
Lines: 16
I installed the MCC interim version of Linux, 0.95c+, tonight. It was
very well done (the easiest way I know of to get all the shared libraries
going), but when I installed the second compimag disk (with g++ and bash)
and rebooted, sh quit working. It would not let commands complete output
and gave the error message after any external like sh [8 1]: EINVAL.
I rebooted using the original root and copied the sh from there, and it
worked again.
Any ideas -- running a 386/33 (no cache) with 150M ESDI drive. Linux 0.95
worked fine.
--
Benjamin L. Combee | -===-
Internet: [email protected] | = ___ "Home of the Big G Burger"
1991-1992 joint enrollee | - = -- MST3K, Comedy Central
C, Modula-2, C++ resource | -===-
[next article]
From: [email protected] (Hongjiu Lu -- Graduate Student)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: How to change the path : Followup
Message-ID: <[email protected]>
Date: 21 Apr 92 03:13:03 GMT
References: <[email protected]>
Sender: [email protected] (USENET News System)
Organization: Washington State University
Lines: 26
In article <[email protected]> [email protected]
sw.edu.au (Matthew Jackson) writes:
>A short while ago I posted a query on how to change the path. Attached are the
>responses I received. I have not tried them as my Linux machine is at home and
>my net access is at work.
>
>It seems those who have had the same problem replied after they worked it out
>and also some more familiar with Unix than myself replied too, for which I am
>grateful.
>
>Anyway, I thought it would be of some use to others so I post a summary.
>
>Summary: For historical reasons, the root user does not have the
> current directory in the path.
WRONG, WRONG, WRONG. That is not for historical reasons. Do your
homework about Unix. That is for security reason.
>
> Under ash and bash, PATH=dir1:dir2:...
> Then export PATH
>
> Under csh or tcsh, PATH=(dir1 dir2 ...)
Please DO NOT DO THAT. KEEP THE WAY IT WAS.
H.J.
[next article]
From: [email protected] (Erik Fichtner)
Subject: 0.95c+ console and gcc 2.1 problem.
Keywords: Yuck. Bleah. Oof.
Message-ID:
Date: 21 Apr 92 04:36:35 GMT
Organization: University of New Mexico, Albuquerque
Lines: 27
Hey everyone... I installed 0.95c and compiled the 0.95c+ patches last weekend
and now there's a major problem with the terminal emulation routines. whenever
I use less or elvis (these are the worst offenders, other things do it too)
data falls thru to the 25th line and then all my data comes out on one
line on the bottom of the screen. stty rows 24 seems to fix less, but elvis
is still screwed up. this never was a problem under 0.95 (no letter) kernel,
so I've gotta assume that something got messed up along the way from 0.95 to
0.95c+. Can someone point out what's different here? I'd really like to
fix this back to the way 0.95 handled it since 0.95 worked! 😉 the systems
kinda useless if I can't read and edit files.
Also having a problem with gcc 2.1. One of the things I'm intrested in is ray
tracing... I used to run DKBtrace under MS-DOS, and then i ran it under AT&T
SVR4 (my last enviornment. yuuuck). So, i went and got the source again and
tried to compile it under Linux. *KABOOM!* gcc got very angry and spit this
error message at me:
gcc: Internal compiler error: program cc1 got fatal signal 6
make: *** [parse.o] Error 1
Correct me if i'm wrong, but that's not supposed to happen, is it? 🙂
--
Techs [email protected]
Disclaimer: any inaccuracies are caused by the phase of the moon
"...whaddya mean I ain't kind... Just not *your* kind!" - Megadeth
[next article]
From: [email protected] (Demian A. Johnston)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: How to change the path : Followup
Message-ID:
Date: 21 Apr 92 05:10:30 GMT
Organization: Sophomore, Physics, Carnegie Mellon, Pittsburgh, PA
Lines: 8
In-Reply-To: <[email protected]>
If you are root or even and "regular" joe user and you decide to put '.'
in your path at least put it at the end of the path list to prevent
strange problems.
Demian
---> [email protected]
---> [email protected]
---> [email protected]
[next article]
From: [email protected] (Ken Hughes)
Subject: Re: Yet another X question
Keywords: X Windows
Message-ID: <[email protected]>
Date: 20 Apr 92 16:42:07 GMT
References:
<[email protected]>
Sender: [email protected]
Reply-To: [email protected] (Ken Hughes)
Organization: Univ of South Florida, Dept of Computer Science
Lines: 48
In article <[email protected]>,
[email protected] (Adam Thompson) writes:
|> In <[email protected]> [email protected]
|> (Hongjiu Lu -- Graduate Student) writes:
|>
|> >From my own experience, a real X11R4 can take as little as 10MB to
|> run.
|> >on a 386. I only keep the minimum stufffs I need. Beside MIT
|> stuffs,
|> >I added some other programs, like xgif and xdvi. I can put off X
|> source
|> >from the net and compile it without much problem. I believe, for
|> X11R5,
|> >20MB should be a safe bet.
|> >H.J.
|>
|> Perhaps you're assuming he doesn't want to compile it 🙂
|>
|> According to the docs in the R5 distribution, the source tree untars
|> to
|> roughly 150-200 Megs. To compile, expect to need at least another
|> 100Meg.
|>
|> As far as compiled binaries go, using shared libraries, on a Sun
|> i386,
|> the full MIT core plus about 50% again in added-on goodies takes
|> somewhere
|> in the range of 60-70 Megs. Of course, to simply run R5
|> 'functional', then,
|> yes, expect to use ~20 megs. Perhaps even a bit less if you only
|> have one
|> WM on call...
I've got X11R5 running under Mach2.6 on my 386/40, and it's quite functional
and used less than 30M. The major problem here is that the average X client
using libX11.a is at least 400K (stripped) because everything must be
statically linked. The corresponsing clients under SunOS are around 50-60K.
If the X11R5 can be made to use dynamic libraries then the installed size
should be reasonable.
--
-----------------------------------------------------------------------
Ken Hughes | "Which button do I press to turn
([email protected]) | it off?" "Try the red one,
FT-Ph D candidate, PT-ex-sysadm | allright?" "Which red button?"
Dept of Comp Sci and Eng |
University of South Florida | .... _Heathers_
-----------------------------------------------------------------------
[next article]
From: [email protected] (Richard Burgess)
Subject: If not SCSI, why not floppy?
Message-ID: <[email protected]>
Date: 21 Apr 92 06:33:29 GMT
Sender: [email protected]
Organization: University of Pennsylvania
Lines: 38
Nntp-Posting-Host: eniac.seas.upenn.edu
As new user, I have dutifully sifted through the
mass of help files and assorted FAQs. The beginner's
guide (and other guides) were enormously helpful in
obtaining a floppy bootable system.
Alas my system is st02 SCSI based. Any attempts to
boot a floppy containing the proper SCSI kernal
causes a kernal panic. The kernal correctly
identifies my drive and finds SCSI ID 0 as my one
(now MSDOS) partition. It then dies. Without
floppy-booting a SCSI kernal, I am not sure how to
get a SCSI-aware-Fdisk to repartition. Do I
have to repartition with DOS Fdisk first? Am I missing
something in all that documentation?
Not to be defeated by a lack of hard-drive, I
figured I would use my second floppy to access
uemacs, mtools or anything - just as a test.
This also failed. Floppy resets are plentiful.
Given that my system is a clone with A:1.2M and
B:1.44M shouldn't
mount /dev/at1 /mnt
or
mount /dev/PS0 /mnt
work to mount my B: drive?
Does this have to do with the SCSI card-I doubt it.
The drive light goes on but the mounting
errors-out, leading me to believe that the os knows
where to look (ie /dev/** works) but cannot
figure out what it is it finds their.
Thanks for any help
Richard Burgess
---------------------------------------------------------
DOS 5, OS/2 2.0, WINDOWS 3.1, WIN32NT (soon!), and LINUX:
life was much easier when DOS 3.3 was the only OS!
[next article]
From: [email protected]
Subject: Re: Kernel panic
Message-ID: <[email protected]>
Date: 21 Apr 92 03:38:31 GMT
References: <[email protected]>
Organization: Penn State University
Lines: 14
Well, couldn't actually tell you what they mean. But, I would love to know
as I got one when I firsted tried to use my hard drive as root under 0.95c+.
Of course it was a different error and different numbers. It wasn't causing
a kernel panic just not giving me a shell.
If anyone is interested, I still have the EIP, EFLAGS, fs, etc... and the
info from scroll-lock and control-shift-scroll-lock.
--
Steven A. Bairstow * "There ain't no rules around here!
InternNet : [email protected] * We're trying to accomplish something!"
or : [email protected] * - Thomas Edison
WWIV-Net : 216 @8450 *
[next article]
From: [email protected] (Hongjiu Lu -- Graduate Student)
Subject: Re: 0.95c+ console and gcc 2.1 problem.
Keywords: Yuck. Bleah. Oof.
Message-ID: <[email protected]>
Date: 21 Apr 92 06:23:50 GMT
References:
Sender: [email protected] (USENET News System)
Organization: Washington State University
Lines: 26
In article
>Also having a problem with gcc 2.1. One of the things I'm intrested in is ray
>tracing... I used to run DKBtrace under MS-DOS, and then i ran it under AT&T
>SVR4 (my last enviornment. yuuuck). So, i went and got the source again and
>tried to compile it under Linux. *KABOOM!* gcc got very angry and spit this
>error message at me:
>
>gcc: Internal compiler error: program cc1 got fatal signal 6
>make: *** [parse.o] Error 1
>
>Correct me if i'm wrong, but that's not supposed to happen, is it? 🙂
>
>
That may be bug in gcc 2.1 or the source code is not very portable.
First thing I would do is add -Wall to CFLAGS in Makefile and try to fix
ALL the error/warning messages by including the appropriate header
files.
If parse.c is generated from parse.y, I would use bison 1.16 or 1.15
(I heard something about 1.16 bugs.)
Bison in 2.1shared-A.tar.Z is 1.16 and untested. You may want to
generate parse.c from the other machines to see if they are the same.
H.J.
[next article]
From: [email protected] (Tim Tsai)
Newsgroups: comp.os.os2.programmer,comp.os.linux
Subject: generic scsi support in OS/2
Message-ID:
Date: 21 Apr 92 07:44:53 GMT
Organization: Mississippi State University
Lines: 17
I asked this before in comp.os.os2.misc but got no answers, so
I'm going to give it another shot. On OS/2 2.0 GA, IBM has
built-in support for scsi controllers that intercepts INT13.
Does anybody have comments on how this works? My guess is that
OS/2 uses the virtual-8086 mode to call the controller's BIOS
routines and pass data back to protected mode through some sort
of share memory. While there probably is a degradation in
performance, a much broader range of devices are now available.
Thoughts?
BTW, this is crossposted to comp.os.linux because I think support like
this would get a lot more people to use Linux. We can still have
controller-specific drivers that emphasis on performance, but those that
have exotic controllers can at least get in the door.
Tim
[next article]
From: [email protected] (Andrew Moore)
Subject: cpio doesn't accept multi-volume archives
Message-ID:
Date: 21 Apr 92 04:16:05 GMT
Organization: Netcom - Online Communication Services (408 241-9760 guest)
Lines: 10
cpio will prompt for "next tape" but immediately dies upon pressing
ENTER.
I am using the command:
cpio -itv
Under Coherent, multi-volume archives require using the raw device.
If this is true of Linux as well, what are the raw device names/numbers?
Thanks.
-Andrew Moore
[next article]
From: [email protected] (BATES, ROBERT PATRICK)
Subject: fmtflop in 0.95x...???
Message-ID: <[email protected]>
Date: 21 Apr 92 08:39:00 GMT
Sender: [email protected]
Organization: Texas A&M University, Academic Computing Services
Lines: 18
News-Software: VAX/VMS VNEWS 1.41
I was just tooling through my S5R4 reference and was trying to figure out how
to format a Unix floppy. I found a few pages concerning fmtflop, and tried it
in Linux to format some disks to use on a Sun for FTP and the like. However,
to my dismay, Linux spits it back up and says that there's no such critter. If
it's there, where is it? If it's not, why not?
Also, is the gcc2.ps (I hope I got that right :->) a PostScript version of the
softdocs for gcc2? Just wondering so's I make sure I'm on a PS printer when I
dump it...
Thanks,
Rob.
==============================================================================
Robert Bates Disclaimer:
[email protected] "I make no claims..."
Texas A & M University
==============================================================================
[next article]
From: [email protected] (Drew Eckhardt)
Subject: Re: If not SCSI, why not floppy?
Message-ID: <[email protected]>
Date: 21 Apr 92 09:43:17 GMT
References: <[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 33
Nntp-Posting-Host: kinglear.cs.colorado.edu
In article <[email protected]> [email protected] (Richard Burge
ss) writes:
>As new user, I have dutifully sifted through the
>mass of help files and assorted FAQs. The beginner's
>guide (and other guides) were enormously helpful in
>obtaining a floppy bootable system.
BUT NOT THE SCSI FAQ.
The guy running the SCSI project (me) gets lazy, and when rushed just
takes a snapshot of his /vmunix file. I mount root from a NORMAL
hard disk, major 3 minor 1. Obviously, if you don't have a minix
filesystem at that location, you will get a kernel panic about
being unable to read root.
>Alas my system is st02 SCSI based. Any attempts to
>boot a floppy containing the proper SCSI kernal
>causes a kernal panic. The kernal correctly
>identifies my drive and finds SCSI ID 0 as my one
>(now MSDOS) partition. It then dies. Without
>floppy-booting a SCSI kernal, I am not sure how to
>get a SCSI-aware-Fdisk to repartition. Do I
>have to repartition with DOS Fdisk first? Am I missing
>something in all that documentation?
>
See above. The solution is to change the bytes at offset 0508 and 0509
in the boot image to 0's. This will rectify the problem.
>work to mount my B: drive?
>Does this have to do with the SCSI card-I doubt it.
No.
[next article]
From: [email protected]
Newsgroups: alt.os.linux,comp.os.linux
Subject: ALT key under GNU Emacs
Message-ID: <[email protected]>
Date: 21 Apr 92 10:28:25 GMT
References:
Sender: [email protected]
Organization: Oxford University Computing Laboratory, UK
Lines: 12
Dear all,
before I radically rearrange my hard-drive to fit the GNU Emacs sources 🙂
Has anyone out already got GNU Emacs to work properly with the ALT key
so that ALT-
Andrew
Andrew Stevens
Programmming Research Group JANET: [email protected]
11 Keble Road, Oxford, England UUCP: ...!uunet!mcvax!ukc!ox-prg!as
OX1 3QD +44 0865 272563
[next article]
From: [email protected] (Mika Pekka Liljeberg)
Subject: Re: expr (GNU shellutils), gcc 2,1 and libc.a
Message-ID:
Date: 21 Apr 92 09:20:48 GMT
References:
<[email protected]>
<[email protected]>
<[email protected]>
<[email protected]>
<[email protected]>
Sender: [email protected] (Uutis Ankka)
Organization: Department of Computer Science, University of Helsinki, Finland
Lines: 21
In-Reply-To: [email protected]'s message of 19 Apr 92 19: 58:43 GMT
In article <[email protected]> [email protected] (Ho
ngjiu Lu -- Graduate Student) wrote:
> BTW, I tried to fix malloc(0) in stdlib.h. It will work with ANSI code.
> Take a look at it. Be sure they work right. I strongly suggest to use
> -Wall during compilation.
>
> H.J.
Talking about malloc(), malloc.h uses the macros PTR and FEXPR
(or something like that) from ansidecl.h to declare functions.
Malloc.h doesn't include ansidecl.h, however, and this can cause
a lot of spurious "parse error before '('" errors, if you include
malloc.h straight.
I just added an #include
This fixed quite a few little problems.
Mika
--
Mika Liljeberg Email: [email protected]
Helsinki University [email protected]
Dept. of Computer Science
[next article]
From: [email protected] (Jaakko Hyvatti)
Subject: ash shell script parameter bug
Message-ID: <[email protected]>
Date: 21 Apr 92 09:27:09 GMT
Organization: University of Helsinki
Lines: 24
Seems to me that ash inserts a '\201' character before every equal
sign '=' in shell script parameters $1, $2, $3 etc. Run the following
shell script named 'foo':
cat > bar <
EOF
od -c bar
with:
ash foo a=b
and you'll get the following output:
0000000 a 201 = b \n
0000005
This caused me trouble while compiling groff with ash as /bin/sh.
--
Jaakko Hyv{tti [email protected]
Lapinrinne 1 B 608 [email protected]
00180 Helsinki, Finland +358-0-6958 5368
[next article]
From: [email protected] (Peter Orbaek)
Subject: Re: init, respawn
Message-ID: <[email protected]>
Date: 21 Apr 92 10:34:45 GMT
References: <[email protected]>
Sender: [email protected] (Peter Orbaek)
Organization: DAIMI: Computer Science Department, Aarhus University, Denmark
Lines: 37
[email protected] (Daniel Hagan) writes:
>1. I would like to know if there is a version of the init program which
>supports the standard system V inittab format (i.e. states, labels,
>run levels, actions ) which exists for linux or which can be ported to linux.
Actually, I have one that is pretty SysV compatible. Someone (I forgot the
name, sorry) from Holland wrote it for Minix and sent it to me. I have made
it work with Linux, but I don't use it because I don't really need it and
I also have real work to do.
I really intended to include it with a new release of the poeigl stuff, but
with the current C compiler mess that may be some months from now.
I will go home and get what I have and release it with no guarantees what so
ever.
>2. Is it possible to defragment a linux partition
>3. Relating to (1), is there a simple way of permanently allowing a logi
n through a modem attached to a serial port.
If you don't want to be able to dial out with that modem, it is just a matter
of putting the right line in /etc/inittab and initialising the modem
properly.
Otherwise, it is rather tedious...
> Any help or pointers to these questions would be greatly appreciated sin
ce I am new to the unix world.
> Thanks,
> Daniel
- Peter. ([email protected])
--
Peter Orbaek ----------------- [email protected] | ///
Hasle Ringvej 122, DK-8200 Aarhus N, DENMARK | ///
| \\\///
"Strong typing is for people with weak memories" | \XX/
[next article]
From: [email protected] (Anthony Murdoch)
Subject: Information on linux ?
Message-ID:
Date: 21 Apr 92 12:14:24 GMT
Sender: [email protected] (USENET News System)
Organization: Prentice Centre, University of Queensland
Lines: 17
Howdy,
I was wondering if there was anywhere I could find some general information
on linux ? I've been reading this news group and I'm getting interested in
it. But I haven't seen any form of faq.
Questions I want answers to are;
1) Where do I get it ?
2) What format is it in ?
3) How do I install it ?
4) What hardware does it support ?
5) What standard utilities are available ?
Thanx in advance
ant
[next article]
From: [email protected] (Dennis Flaherty)
Subject: Re: Question about svga preset patch
Message-ID: <[email protected]>
Date: 21 Apr 92 13:08:53 GMT
References:
Organization: Miles, Inc. Diagnostics Divis.
Lines: 38
In article
writes:
> Okay, I got the 0.95c+ kernel patched and compiled with no problem,
> and have lp and ps working, and I've nearly got a completely working
> Unix system that's faster than just about anything I've seen for an
> 86-based machine... Kudos to all working on the project...
>
> Now, a question of whoever (I lost his address) the guy is who wrote
> the svga-preset patch.
That's me, they guy with a paper bag over his head...
> I patched my source with the patch, and it
> compiled and ran fine, with one problem. My desired svga mode is mode
> zero on my machine (paradise video card) - 132x25. Anything higher
> than that (my only other choice is mode one, 132x43) is too fuzzy on
> my cheap crummy monitor... The patch works fine with the video mode
> set to one and compiled, but if I set it to zero (my preferred mode),
> it compiles and runs, but defaults to standard 80x25 VGA mode. Is
> there any fix available, even if it means patching other sources?
> Thanks for any help...
No, the problem is that when I updated my patch, I took the # sign
out of the makefile and put it in the setup.S file so that I could
get 80x25 VGA mode preset to work. Otherwise the assembler chokes.
Now setting SVGA_MODE=0 might as well #undef it. I'll fix it and
send yet-another version to tsx-11.
Hey, I said I wanted feedback, didn't I? Nice to see people are
using my patch.
BTW, I won't be making a patch to let you go back and forth between
modes after bootstrap-- "I already have a job."
--
Dennis T. Flaherty [email protected]
Miles, Inc. [email protected]
Diagnostics Division
My doctor says Mylanta, but my boss says Alka-Seltzer!
[next article]
From: [email protected] (0134)
Subject: Re: Oddities with MCC Release
Keywords: mcc , mg
Message-ID: <[email protected]>
Date: 21 Apr 92 13:18:03 GMT
References: <[email protected]>
Organization: Universitaet Bielefeld
Lines: 18
after installing the mcc interim version (0.95c+) our mg bombed !!!
even recompiling it led to no results. is there a version that runs under
mcc 0.95c+ ???
we had the same shell problem.
folker
p.s.: What is the state of tcpip + a driver for wd8003e cards ???
i am trying to set up a linux 386 as a printer server and also
to offer mail service.
Plus we have a 386 with 8-12 serial ports
plus a bunch of old vt100-vt230 terminals !!!!!
Has anybody tried this out before ???
[next article]
From: [email protected] (Timothy F. Sipples)
Newsgroups: comp.os.os2.programmer,comp.os.linux
Subject: Re: generic scsi support in OS/2
Message-ID: <[email protected]>
Date: 21 Apr 92 14:19:50 GMT
References:
Sender: [email protected] (News System)
Reply-To: [email protected]
Followup-To: comp.os.os2.misc
Organization: Dept. of Econ., Univ. of Chicago
Lines: 17
In article
:
> I asked this before in comp.os.os2.misc but got no answers, so
>I'm going to give it another shot. On OS/2 2.0 GA, IBM has
>built-in support for scsi controllers that intercepts INT13.
>Does anybody have comments on how this works? My guess is that
>OS/2 uses the virtual-8086 mode to call the controller's BIOS
>routines and pass data back to protected mode through some sort
>of share memory. While there probably is a degradation in
>performance, a much broader range of devices are now available.
>Thoughts?
Bingo -- correct.
--
Timothy F. Sipples Keeper of the OS/2 Frequently Asked Questions
[email protected] List, available via anonymous ftp from
Dept. of Economics 128.123.35.151, directory pub/os2/faq, or via
Univ. of Chicago 60637 netmail from [email protected].
[next article]
From: [email protected] (Timothy L. Nali)
Subject: Re: A new game port, and mtools problems...
Message-ID:
Date: 21 Apr 92 16:02:55 GMT
References: <[email protected]>
Organization: Sophomore, Electrical and Computer Engineering, Carnegie Mellon, P
ittsburgh, PA
Lines: 21
In-Reply-To: <[email protected]>
Excerpts from netnews.comp.os.linux: 19-Apr-92 A new game port, and
mtools.. Adam Justin Thornton@owl (1460)
> I'm using the mtools.a supplied with DLE.tar.Z in the gcc2.1 distribution,
> and I have C defined as /dev/hda1 and D as /dev/hdb1, which is where they
> are. However, when I try mtools
> found error message. What am I missing? Is there some syntax I should be
> aware of?
In the readme file, there is a long ld line (ld mtools mdir mread ....)
You have to type that in. It links mtools to commands like mdir, dir
command, and mread, copy a dos file to linux. then you type in mdir c:,
mread foo.ext, etc.
_____________________________________________________________________________
Tim Nali \ "We are the music makers, and we are the dreamers of
[email protected] \ the dreams" -Willy Wonka and the Chocolate Factory
[next article]
From: [email protected] (Michel Eyckmans (MCE))
Subject: Re: problems booting
Message-ID: <[email protected]>
Date: 21 Apr 92 10:32:44 GMT
References: <[email protected]> <[email protected]
u.edu>
Sender: [email protected]
Reply-To: [email protected]
Organization: IMEC vzw Leuven
Lines: 31
In article <[email protected]>, [email protected] (
H. Peter Anvin N9ITP) writes:
|> In article <[email protected]> of comp.os.linux,
|> [email protected] (Richard Alan Brown) writes:
|> >
|> > I used rawrite to make a bootable disk and a rootimage disk. When I
|> > boot my pc with the floppy in, I get the loading............
|> > and the SVGA mode question.
|> > when I press any other key, the system simply reboots, and cycles through
|> > this same process.
|>
|> I experience exactly the same phenomenon on a 486/33 ISA with the OPTI
|>
I've had exactly the same problem. After uncompressing the 0.95c+
bootimage under 0.12, I used cp to put it onto a floppy. This
resulted in the behaviour described above. It turned out that my
copy of uncompress was corrupt. After switching to the 0.95a
uncompress, all worked well.
MCE
PS : Anybody know where to get "mvdir"?
MCE
--
=============================================================================
#include
M. Eyckmans (MCE) [email protected]
=============================================================================
[next article]
From: [email protected] (Kees J. Bot)
Subject: Re: IDE drive performance with linux
Message-ID: <[email protected]>
Date: 21 Apr 92 15:02:22 GMT
References: <[email protected]>
Sender: [email protected]
Lines: 58
[email protected] (Guido Kueppers) writes:
>Greetings,
>thanks to all who have replied to my IDE drive problem. From the responses I
>got so far I gather that the drive's poor performance is specific to my drive
>and possibly due to inappropriate configuration (translation vs. native mode).
>If there is any interest I will post my newly accquired wisdom on IDE drive
>setup when the problem is solved (Or should I rather spare you the details?)
I don't think the problem is with your drive, I think it's better to
send the device driver writer back to the drawing board.
I am currently rewriting the Minix hard disk driver to increase its
performance, because it also has problems with IDE drives. I've got
access to three IDE drives, of 100Mb, 120Mb and 200Mb.
The current Minix HD driver reaches reading speeds of 250, 60, and
500 kbps on them. Note the slow speed for the 120Mb. At the usual
rotational speed of 3600rpm, this is exactly 1k per rotation! (The
native Minix block size).
I have got another HD driver that precomputes a lot of data in advance,
programs the controller to read/write a lot of sectors and then goes to
work at the interrupt level. It can make 450, 500, and 600 kbps on each.
My guess is that the first two disks do not have a track cache and the
third one does. The current Minix HD driver can't keep up with the
120Mb disk.
I am trying to make a device driver that works like the second one but
does not use the ugly interrupt routine (it seems to hang sometimes),
but uses message passing instead at the expense of a few context
switches. Early results (the paint is still wet) show speeds within 5%
of the ugly driver.
Further comments:
- I don't know what the Linux AT disk driver looks like, a Linux
enthousiast should check it out to see if I'm right. (I don't have
any Linux code, I would probably label everything as "ugly". 🙂
- Using a raw device and reads of 127kb it is possible to reach
800 kpbs on the 200Mb disk using the "ugly" driver. Getting the
controller to transfer many sectors at once seems to be the trick.
- The drive doesn't have to translate the parameters as often for large
transfers. (Not that I think that the "translate" vs. "native" mode
thing is very important.)
- The file system must be able to send large read-ahead requests to the
driver, so it has something to work with. Writing speed is not as
important, but must be looked at too. (Minix writes all dirty blocks
for the same device at once if it must write one.)
[Back to making to floppy driver go 30% faster reliably, slow devices
seem to give more trouble than fast ones!]
--
Kees J. Bot ([email protected])
Systems Programmer, Vrije Universiteit Amsterdam
[next article]
From: [email protected] (n.h.chandler)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Where can I get pfdisk-0.92?
Keywords: fdisk, extended-partitions
Message-ID: <[email protected]>
Date: 21 Apr 92 17:22:03 GMT
Followup-To: poster
Organization: AT&T
Lines: 6
Where can I locate pfdisk-0.92 and fdisk-0.92? I'd
prefer that they were not wrapped up with 2 megabytes
of other stuff that I'd have to discard.
Neville H. Chandler
nhc\@mtdcc.att.com
[next article]
From: [email protected] (Philippe Backouche)
Subject: Thread
Message-ID: <[email protected]>
Date: 21 Apr 92 18:09:49 GMT
Sender: [email protected]
Reply-To: [email protected] (Philippe Backouche)
Organization: SCRI, Florida State University
Lines: 10
Could someone explain what threads are ? What is the difference between 'fork
ing'
a new process and making a thread ?
Do threads exist in Linux ???
Thanks for any help,
Philippe Backouche
[email protected]
[next article]
From: [email protected] (M. Saggaf)
Subject: SCSI Adaptec 1522
Message-ID: <[email protected]>
Date: 21 Apr 92 19:32:29 GMT
Sender: [email protected] (News system)
Distribution: na
Organization: Massachusetts Institute of Technology
Lines: 7
Nntp-Posting-Host: e40-008-7.mit.edu
Does anybody know if the current SCSI version of Linux supports the
Adaptec 1522 host adapter? I'm looking into buying that controller and
I'd like to know if Linux can use it. It's about $100 less expensive
than the Adaptec 1542 since it is not bus mastering. Please let me
know. Thanks
-- M. Saggaf ([email protected])
[next article]
From: [email protected] (Kevin Cummings)
Subject: Re: GCC2.1 (Wonderful !)
Keywords: gcc, 2.1, 1.40
Message-ID: <[email protected]>
Date: 21 Apr 92 19:00:07 GMT
References: <[email protected]> <1992Apr15.015254.25547@
ucc.su.OZ.AU> <[email protected]> <1992Apr16.102220.3952@colorad
o.edu> <[email protected]>
Sender: cummings@hammer (Kevin Cummings)
Organization: Prime Computer R&D
Lines: 137
Sorry folks, but I can't pass this one up. Those who don't wish a lesson in
PRIMOS style dynamic linking should go on to your next article now.
In article <[email protected]>, [email protected] (Jo
hn Plevyak) writes:
>
> Concerning the various kinds of dynamic linking.
>
> As I remember (from long ago) the Primos (the Prime operating system)
> replaced all dynamic calls with calls to illegal addresses. A table in
Let's see if I can re-interpret this correctly. A link from one routine
to another in the Prime Segmented architecture is called an IP (indirect pointer
).
It is a pointer to the target object. A Dynamic Link is a SPECIAL pointer
to a name. The FAULT bit in the pointer is set. This causes the hardware
to raise a POINTER_FAULT$ condition (PL/I style condition, similar to a
C signal). The POINTER_FAULT is caught by any available condition handler
(ON UNIT in PL/I terms) currently on the stack, including those buried in
the operating system. If the condition gets through to the OS, the OS
checks to see that the pointer less its fault bit is a valid pointer
to a character string in memory. If so, it passes the pointer off to
the dynamic linking routines which methodically search (via a search list)
the dynamically linked libraries and OS entries. Due to software compatibility,
three types of libraries are searched: OS entries, EPF libraries, and
old static mode shared libraries that might be installed in our LIBTBL.
The entrypoint names are just a name. Currently they are not qualified by
a library name and therefore the ENTRY$ search list specifies the order of
libraries to search looking for a name match. If the name cannot be resolved
in ANY of the libraries in the search list, a new error condition is raised
(signalled): LINKAGE_FAULT$, and the name is a part of the information
associated with the error.
> the image mapped the index of the illegal address to a filename/symbol
> pair. The filename was found or loaded and the address was patched. This
If the name is found in a library, the library routine passes the address of
the routine back to the dynamic link handler which then REPLACES the faulted
pointer (which we call a DYNT) with the runtime address of the routine. This
IP is now considered SNAPPED (resolved), and further references through this
IP are direct to the resolved routine. In some respect this is doing the
final linking stage at runtime.
> has the advantage that you could dynamically replace the library without
> relinking (the linker was in the OS). This could be used on a 386 class
> machine.
Yes, it probably could, but I don't know enough about 386 pointers to know
how to do something similar. But, given that a 386 32-bit pointer consists
of a 16 bit segment number and a 16-bit byte offset, you would probably have
to make use of some kind of "missing memory" error or page fault to a
special non-existent page in order to do something similar, and tie the
dynamic linking support into the page fault handler in the kernel, which is
not as clean as having its own special fault handler.
> The disadvantages were:
>
> 1) your program could suddenly fail if the new library (installed by
> someone else) had a bug
Isn't this a problem with ANY new library that has to be re-linked to?
> 2) you could not page directly off the executable
Yes you can. You just have to seperate the code part from the data part. IPs
are a part of the data segment. You can page the machine code part just fine.
This is what a segmented architecture is all about.
> 3) your operating system was forced to intervene on the first of
> every dynamic call... thus slowing startup time.
Yes again. But see some of the advantages below.
Another disadvantage is that if you wish to remain upward compatible, you can ne
ver
remove an entrypoint from a library, lest some program which currently uses it
starts to fail. At Prime, this is the "Compatibility Shall Reign Forever" syndr
ome.
> The advantages were:
> 1) no need to relink shared library programs
Unless the library entrypoint name changes. But if that happens, you have
bigger problems anyways.
A side point here is that you don't need to keep around old compiled binaries
or pre-linked files either. AND you will only ever have one copy of a shared
library installed on your system at any one time. Saves on file system disk spa
ce.
Further more, you can still staticly link some programs if you really need it
if you have both shared and unshared copies of the libraries available on your
development system.
> 2) no need to link to libraries that this execution of the program
> did not require... finer grain linking could be a win here.
Possibly cuts down on working set. You don't have to drag along unused baggage!
> In general I don't think that you will find an obvious clean and simple winner
> in a war of dynamic linking strategies if you ask more that a few people.
Probably true, however, if you carry around multiple versions of a library just
because it is no longer possible to re-link some old programs which were linked
with a previous version of the library, I don't see what you are saving! Your
savings are being diminished by the "extra" libraries you have on disk, and code
that is not being shared at runtime!
> The simplest solution is usually the best though, and (not intending to start
> a war here) while Primos did dynamic linking which was nice, the lack of
> child processes, memory protection of 'shelled' programs, the segmented
> architecture, system calls with a mix PL1/G and FORTRAN interfaces and assorte
d
> other goodies made me wish for Unix with or without dynamic linking
> (perhaps it has changed, this was 1986).
Probably not dramatically enough for your tastes. The Prime 200 and 300 was a F
ORTRAN
machine. And we used FTN-66 as our systems programming language at that time.
When we released the 400 (and every subsequent machine), it was really a PL/I ma
chine.
And we changed our system programming language (and our operating system) to ref
lect that.
Compatibility left us with the result you saw. (Go ahead, and tell me UNIX does
n't have
compatibility problems from rev to rev. It is its WORST problem.)
> john ([email protected])
Sorry to drag this out, but every point made here needs to be considered when
we do dynamic linking for LINUX. I am interested in hearing other's
points of view, and listening to the way others have implemented
dynamic linking. But I beleive that Prime made "all the right decisions"
in their impe
=================================================================
Kevin J. Cummings Prime Computer Inc.
20 Briarwood Road 500 Old Connecticut Path
Framingham, Mass. Framingham, Mass.
Work: [email protected]
Home: [email protected]
Std. Disclaimer: "Mr. McKittrick, after careful consideration,
I've come to the conclusion that your new
defense system SUCKS..." -- War Games
=================================================================
[next article]
From: [email protected] (Demian A. Johnston)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: ALT key under GNU Emacs
Message-ID:
Date: 21 Apr 92 19:15:58 GMT
Organization: Sophomore, Physics, Carnegie Mellon, Pittsburgh, PA
Lines: 17
In-Reply-To: <[email protected]>
Excerpts from netnews.alt.os.linux: 21-Apr-92 ALT key under GNU Emacs by
[email protected]
>
> before I radically rearrange my hard-drive to fit the GNU Emacs sources 🙂
> Has anyone out already got GNU Emacs to work properly with the ALT key
> so that ALT-
>
All your have to do is add a "km:" in the middle of the console entries
in the /etc/termcap file....
km -- Tell the program that the terminal has a meta key.
Demian
---> [email protected]
---> [email protected]
---> [email protected]
[next article]
From: [email protected] (Niels Skov Olsen)
Newsgroups: alt.os.linux,comp.os.linux
Subject: Re: Keyboard lockups SOLVED (i.e. worked around)
Message-ID: <[email protected]>
Date: 21 Apr 92 18:27:01 GMT
References: <[email protected]>
Sender: [email protected]
Organization: Department of Computer Science, U of Copenhagen
Lines: 78
Here is the real fix to the trouble I and others have had
with keyboard lockups and LED malfunction. I think this
should be included in the next release, but I don't know
the proper procedure for doing so. I post it here because
others have reported the same or similar problem.
The problem was that the keyboard would lock up when pressing
either of CAPS, NUMLOCK, SCROLL LOCK. The 'bug' shows up only
on some hardware. The fix as suggested by Bruce Evans was to
add a routine called kb_ack that waits until 0xfa shows up on
port 0x64, and call after outputting data to port 0x60. I am not
sure what the exact meaning of these ports are, but the patch
that follows fixes the lockups and makes the LED's work!
Thank you Bruce Evans.
Niels
Apply this patch in kernel/chr_drv/ :
*** /tmp/linux/kernel/chr_drv/keyboard.S Thu Apr 9 22:15:22 1992
--- keyboard.S Tue Apr 21 19:49:33 1992
***************
*** 182,194 ****
--- 182,196 ----
je 1f
movb %al,old_leds
call kb_wait
movb $0xed,%al /* set leds command */
outb %al,$0x60
+ call kb_ack /* [NSO]: Suggested by Bruce Evans */
call kb_wait
movb _kleds,%al
outb %al,$0x60
+ call kb_ack /* [NSO]: Suggested by Bruce Evans */
1: ret
uncaps: andb $0x7f,_kmode
ret
scroll:
testb $0x03,_kmode
***************
*** 782,791 ****
--- 784,817 ----
pushl %eax
pushl %ebx
movl $10000,%ebx
1: inb $0x64,%al
testb $0x02,%al
+ je 2f
+ decl %ebx
+ jne 1b
+ 2: popl %ebx
+ popl %eax
+ ret
+
+ /*
+ * kb_ack waits for 0xfa to appear in port 0x60
+ *
+ * Suggested by Bruce Evans
+ * Added by Niels Skou Olsen [NSO]
+ * April 21, 1992
+ *
+ * Heavily inspired by kb_wait 🙂
+ * I dont know how much waiting actually is required,
+ * but this seems to work.
+ */
+ kb_ack:
+ pushl %eax
+ pushl %ebx
+ movl $10000,%ebx
+ 1: inb $0x60,%al
+ testb $0xfa,%al
je 2f
decl %ebx
jne 1b
2: popl %ebx
popl %eax
[next article]
From: [email protected] (Kevin Cummings)
Newsgroups: comp.os.linux,comp.os.minix
Subject: Re: HELP! SHOELACE IS STRANGLING ME!!!
Message-ID: <[email protected]>
Date: 21 Apr 92 20:23:07 GMT
References: <[email protected]>
Sender: cummings@hammer (Kevin Cummings)
Organization: Prime Computer R&D
Lines: 23
In article <[email protected]>, [email protected] (Gr
eenup) writes:
>
> I recently used shoelace to try & make linux boot. after no success,
> i want to go back. HOW DO I GET RID OF THIS BOOT THING THAT KEEPS
> ASKING WHICH PARTITIAN TO BOOT?!?!?!?!?
Just a wild guess, but boot from your ORIGINAL BOOT floppy using your
ORIGINAL ROOT floppy (you do still have those, don't you?). And mount
your current hard disk on /mnt. Now you should be able to "cd /mnt/etc"
and undo whatever you did with shoelace.
=================================================================
Kevin J. Cummings Prime Computer Inc.
20 Briarwood Road 500 Old Connecticut Path
Framingham, Mass. Framingham, Mass.
Work: [email protected]
Home: [email protected]
Std. Disclaimer: "Mr. McKittrick, after careful consideration,
I've come to the conclusion that your new
defense system SUCKS..." -- War Games
=================================================================
[next article]
From: [email protected] (Kevin Cummings)
Subject: Re: Software index for linux
Message-ID: <[email protected]>
Date: 21 Apr 92 20:17:31 GMT
References: <[email protected]>
Sender: cummings@hammer (Kevin Cummings)
Organization: Prime Computer R&D
Lines: 61
In article <[email protected]>, [email protected] (Bill Jones) writes:
> One thread that seems to be remaining at a constant level in both
> alt.os.linux and here in comp.os.linux is what software is in what
> package. I know that I have had that trouble often (what, for instance
> is in tubes.tar.Z??)
>
> There has been some mention of keeping track of these packages in a
> "master index" somewhere and I beleive that Marc Corsini has mentioned
> trying to do that along with the FAQ.
>
> Now that I have reliable and stable access to news and e-mail, I would
> be willing to undertake such a project if enough people felt that it
> would be useful. What I am envisioning is
>
> 1. An index of what is available, what package it is in,
> and what it does (so far, it seems as though most
> everything is available at banjo.concert.net but the
> descriptions might still help people determine what they
> need).
>
> 2. A list of software that is currently being ported to linux
> and the status of such a port (if this can be obtained).
>
> 3. A list of all of the patches that have been made available
> along with the expected source code that each patch expects
> to find.
>
> 4. Any other software-related items of interest.
>
>
> If something like this were available, I would hope that others, especially
> newcomers to linux, would be able to use it. In addition, it would help
> cut down on the duplication of porting efforts and let people who are
> working on the same project team up if they desire.
>
> This would require the assistance of the authors of various programs and
> ports so that it would remain up to date and, of course, would not be
> mandatory that anyone participate. But, it strikes me that this is the
> time to try and get something like this set up -- before the volume of
> linux software, familiar and unfamilar, becomes so great that no one
> would desire to keep track of it.
>
> As I said, if enough people express an interest in this, I would be more
> than willing to undertake it. Since I am not a software wizard (not even
> close) this would be my contribution to the linux community. So, how
> about it? Let me know.
Sounds like a GREAT idea. Go for it!
=================================================================
Kevin J. Cummings Prime Computer Inc.
20 Briarwood Road 500 Old Connecticut Path
Framingham, Mass. Framingham, Mass.
Work: [email protected]
Home: [email protected]
Std. Disclaimer: "Mr. McKittrick, after careful consideration,
I've come to the conclusion that your new
defense system SUCKS..." -- War Games
=================================================================
[next article]
From: [email protected] (Kevin Cummings)
Subject: Re: timezone support, date, CMOS clock support: archive sites please note
Message-ID: <[email protected]>
Date: 21 Apr 92 20:33:43 GMT
References:
Sender: cummings@hammer (Kevin Cummings)
Organization: Prime Computer R&D
Lines: 24
I happend to find timebin.tar.Z on tsx-11 yesterday, and downloaded
to my LINUX machine. I untarred it, configured it, and moved the
binaries to /usr/bin and /usr/lib. I then ran date, and it told me
it was Mon 20 Apr, 20:xx:xx 1992. No time zone info, even though
/usr/lib/localtime was present. I then ran clock to see what was in
my CMOS clock, and it was Sat 20 Apr, 14:xx:xx EDT 1992.
I think I can understand the 14:00 bit (If I chalk it up to
the GMT-EDT difference), but Saturday?????
Something's not right here.
=================================================================
Kevin J. Cummings Prime Computer Inc.
20 Briarwood Road 500 Old Connecticut Path
Framingham, Mass. Framingham, Mass.
Work: [email protected]
Home: [email protected]
Std. Disclaimer: "Mr. McKittrick, after careful consideration,
I've come to the conclusion that your new
defense system SUCKS..." -- War Games
=================================================================
[next article]
From: [email protected] (Kevin Cummings)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: all kinds of questions (autoparking)
Message-ID: <[email protected]>
Date: 21 Apr 92 20:39:28 GMT
References: <[email protected]> <1992Apr13.223428.29499@colo
rado.edu> <[email protected]> <[email protected]
sinki.FI>
Sender: cummings@hammer (Kevin Cummings)
Organization: Prime Computer R&D
Lines: 34
In article <[email protected]>, [email protected]
.FI (Risto Kankkunen) writes:
> In [email protected] (Kevin Cummings) writes:
> > So unless you are shutting down your system to reboot DOS (or something
> > else), your disk heads will not be parked by LINUX. Maybe this is the
> > cause of some of the trashed file systems some people are experiencing?
>
> But isn't head parking suggested only before moving the computer or the
> disk itself? Although there is a greater possibility for disk damage
> when the heads stay above the disk on power-down, it isn't a problem in
> practise on everyday desk-top use.
Head parking is suggested (by me anyways) as necessary before powering
off your computer. Who knows what the older hard disk controllers
will do when you turn the power off, and the write line happens to
accidently spit something out. Good way to trash whole tracks.
Used to happen all the time on my CP/M system, until I learned to
park the heads before powering down. If your drive doesn't recognize
the power fail, your disk controller probably doesn't either, and you could
lose something. Probably happens less today than a few years ago,
but if you're using MFM drives, the chance is significantly increased.
=================================================================
Kevin J. Cummings Prime Computer Inc.
20 Briarwood Road 500 Old Connecticut Path
Framingham, Mass. Framingham, Mass.
Work: [email protected]
Home: [email protected]
Std. Disclaimer: "Mr. McKittrick, after careful consideration,
I've come to the conclusion that your new
defense system SUCKS..." -- War Games
=================================================================
[next article]
From: [email protected] (Spencer Smith)
Subject: Re: SCSI Adaptec 1522
Message-ID:
Date: 21 Apr 92 22:01:56 GMT
References: <[email protected]>
Distribution: na
Organization: Kansas State University
Lines: 15
NNTP-Posting-Host: aleph.cis.ksu.edu
[email protected] (M. Saggaf) writes:
>Does anybody know if the current SCSI version of Linux supports the
>Adaptec 1522 host adapter? I'm looking into buying that controller and
>I'd like to know if Linux can use it. It's about $100 less expensive
>than the Adaptec 1542 since it is not bus mastering. Please let me
>know. Thanks
It doesnt work on my 1522, however if some kind soal out there
nows the register values for it, ill will be glad to try it out again.
I need the values for things like SCSI bus reset, READ, WRITE .....
Thanks
Spencer Smith
[email protected]
[next article]
From: [email protected] (Andy I. Liaw)
Newsgroups: comp.os.linux,alt.os.linux
Subject: where is bootlin.zip?
Message-ID: <[email protected]>
Date: 22 Apr 92 00:04:54 GMT
References: <[email protected]> <1992Apr21.031303.208
[email protected]>
Followup-To: comp.os.linux
Organization: Texas A&M University, College Station
Lines: 5
Can someone tell me where to find bootlin.zip? The one I got
from the posted article unzipped with errors after uudecode.
I can't find it in tsx-11.mit.edu.
Andy [email protected]
[next article]
From: [email protected] (A. V. Le Blanc)
Subject: Re: Kernel panic
Message-ID: <[email protected]>
Date: 21 Apr 92 06:27:45 GMT
References: <[email protected]>
Followup-To: [email protected]
Organization: Computing Centre, University of Manchester
Lines: 16
In article <[email protected]> [email protected]
(John B. Jr. Rogerson) writes:
> One more question: If I wish to create a swapfile from my linux
>partition (/dev/hda3) and I use the dd command with if=/dev/hda3 and
>of=swapfile, can I be sure that swapfile will not contain any of my
>linux files? I certainly wouldn't want them overwritten during
>swapping.
The dd command is used to create a file of a specific size. It needs
to be able to copy data from somewhere, so we allow it to copy data from
the hard disk; it doesn't matter what input it has, as long as it is
bigger than the swapfile to be created. There is no danger that
swapping will overwrite your files, at least not just from using dd
with if=/dev/hda to create them.
-- Owen
[email protected]
[next article]
From: [email protected] (A. V. Le Blanc)
Subject: Re: Graphics and IPC questions...
Message-ID: <[email protected]>
Date: 21 Apr 92 07:04:58 GMT
References:
uiuc.edu>
Followup-To: [email protected]
Organization: Computing Centre, University of Manchester
Lines: 28
In article <[email protected]> chojnows@osiris (Brian Choj
nowski) writes:
"Alex R.N. Wetmore"
>Why not implement these in vt100 mode, rather then with direct screen
>writes then? Linux seems to have very fast terminal emulation (at least
>on my 386/16 with a very old Paradise EGA), and you might as well take
>advantage of those using termcap. That way people on terminals (and
>telnet sessions, when those exist), can use the software too. It also
>doesn't restrict the software to linux, but allows it to work on any
>other unix platform (in theory at least).
I guess I havent made myself clear. I would like to do something like this,
where I would create a superset of the vt100. But the mode as it is, is
insufficient for the kind of graphics I am talking about. I want to also add
this new mode to the telnet program (when it gets written), so that the
graphics get displayed remotely.
I think Alex misled us a bit by saying 'vt100'. Perhaps he was thinking
of the vt100 emulation in MS-Kermit or NCSA telnet, where the emulator
switches to Tektronix 401x mode for graphics when it recognises certain
escape sequences. This might be a good place to start, since the source
code for both of these is available, though the Kermit version is a good
deal less buggy. Moreover, as far as conformity with existing interfaces,
there is a TeX previewer that works rather well with the vt100/t4012
emulator in Kermit.
-- Owen
[email protected]
[next article]
From: [email protected] (A. V. Le Blanc)
Subject: Re: Oddities with MCC Release
Message-ID: <[email protected]>
Date: 21 Apr 92 18:18:52 GMT
References: <[email protected]>
Followup-To: [email protected]
Organization: Computing Centre, University of Manchester
Lines: 20
In article <[email protected]> [email protected] (Ben Combee) writes
:
>I installed the MCC interim version of Linux, 0.95c+, tonight!ries
>goigggggggggggggggggggggggggggggggggggggggggggggg, but when I instaBl#ed the se
cond compimgggggggggggggggggggggggggggggggggdisk (with gB+ and bash)
>and rbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbbooted, sh quit working. It would not
let commands complete output
>and gave the error message after any external like sh [8 1]: EINVAL.
This was caused by an illegal (!) /dev/tty; it has now been fixed in the
release disks, and the larger (job control) version of bash moved to the
boot disk. To fix it on your system, type
rm /dev/tty
mknod /dev/tty c 5 0
chmod 222 /dev/tty
I have been trying to keep up a BUGS+WARNINGS file in the mcc-interim/0.95c+
directory with information like this in it.
[next article]
From: [email protected] (A. V. Le Blanc)
Subject: Re: 0.95c+ console and gcc 2.1 problem.
Message-ID: <[email protected]>
Date: 21 Apr 92 18:24:10 GMT
References:
Followup-To: [email protected]
Organization: Computing Centre, University of Manchester
Lines: 22
In article
>Hey everyone... I installed 0.95c and compiled the 0.95c+ patches last weekend
>and now there's a major problem with the terminal emulation routines. whenever
>I use less or elvis (these are the worst offenders, other things do it too)
>data falls thru to the 25th line and then all my data comes out on one
>line on the bottom of the screen. stty rows 24 seems to fix less, but elvis
>is still screwed up.
This may be caused by your having the wrong terminal type. Type
'echo $TERM' and see what it says. If it says 'vt100', vi will work
the way you describe. If it says 'console' or some such name, then this
note is a waste of time!
You might create a file /etc/profile, making sure it has universal read
permission, and put in it a line
EXPORT TERM=console
that is, unless you have login from serial ports.
-- Owen
[email protected]
[next article]
From: [email protected] (Darren Senn)
Subject: New Linux installation/problems
Message-ID: <[email protected]>
Date: 22 Apr 92 01:45:46 GMT
Sender: [email protected]
Lines: 25
Ok. I installed Linux yesterday (0.95c+), and so far I love it. I do, however,
have a few problems. My machine:
486/25 ISA bus with 4Mb RAM, OPTI chipset, AMI BIOS
ET4000A based 16-bit VGA card (1Mb memory) as primary display
Hercules monographics card as secondary display
ESDI 120Mb Micropolis hard drive.
5.25"/3.5" disk drives (3.5" is a:)
a) Shoelace doesn't come close to working. When I followed the instructions in
the
readme that came with 0.95c+, I created a test floppy, and tried booting
off
that. When I did, the BIOS boot sequence would work normally, check the
existance of drives, and then it would read the first sector of my 3.5"
drive and hang. 🙁
b) No matter what super-vga mode I try, it doesn't work. Typically it just show
s me
an 80x25 window into the specified screen size. In other words, things
scroll
correctly, but it only starts after n extra lines below the bottom of th
e screen.
c) Where can I ftp a bash manual?
Thanks for helping the neophyte.
--
Darren Senn Phone: (408) 479-1521
[email protected] Snail: 1785 Halterman #1
Wasurenaide -- doko e itte mo soko ni anata wa iru yo. Santa Cruz, Ca 95062
[next article]
From: [email protected] (Linus Benedict Torvalds)
Newsgroups: comp.os.linux,alt.os.linux
Subject: pre-0.96 (was Re: gdb still isn't working)
Message-ID: <[email protected]>
Date: 21 Apr 92 23:15:10 GMT
References:
[email protected]>
Organization: University of Helsinki
Lines: 128
In article <[email protected]> [email protected].
FI (Linus Benedict Torvalds) writes:
> [ trace not working in gdb ]
>
>My personal version handles this correctly (as well as doing some other
>things in a cleaner manner), but I'm not quite ready for a new release
>yet. I could make YAAR (yet another alpha-release) or just mail
>interested parties the fixes needed - mail me if you're interested, and
>depending on the number of messages I get I'll make it a new release.
Ok, the response seems to make a new pre-release appropriate: I have
uploaded "pre-0.96.tar.Z" to tsx-11 and nic.
Here is what the pre-release contains:
- truncate/ftruncate/fchmod/fchown system calls
note that there aren't any library functions for these, so they
aren't very useful yet...
[f]truncate needed a change in the logic of the internal
truncate VFS call - anybody that has any nonstandard filesystem
probably needs to look it up.
- io-bitmap syscalls giving root-processes access to selected io ports
from user space. There is a "ioperm()" system call that lets the
process select which ports it wants to enable/disable (all ports
disabled as default) as well as a (standard sysv?) ioctl interface
that X uses.
again, no library stubs, but it allows things like reading and
setting the cmos clock without using /dev/port, as well as
control over the VGA registers...
- mmap for /dev/mem
more things needed for X...
- the signal-handling fixes needed for gdb
These aren't yet complete: serial lines still send signals under
interrupts that can result in problems (ie ptrace doesn't
correctly get them), but that's pretty unlikely (and will be
fixed in the final 0.96). Breakpoints should work etc..
- multiple shared libraries
Up to 6 simultaneous shared libraries/process: the patches were
originally by pmacdona, but they were heavily changed by me, and
I think they work in a more natural manner now. One user-level
change is that the libraries are now checked for read and
execute permissions for safety-reasons.
- cleaned up special files.
read/write/ioctl no longer has special-case code: it is all
handled with tables to functions. This will mean that the SCSI
patches won't patch in quite cleanly into 0.96: you'll need to
add the code that sets up the functions.
Again: device drivers and vfs-filesystem hackers need to look
into the changes, although they are pretty logical (earlier
versions just didn't implement all the vfs-routines)
Note that the vfs-code for select is still not used: select is
hardcoded for the devices it supports right now.
- ptrace() has a new interface
as gdb for versions < 0.95c don't work on the new version, and
gdb won't work very well at all on 0.95c[+], there was no reason
not to break ptrace. Thus 0.96 has a new calling convention for
ptrace, and the old ptrace library function no longer works.
I'm including the new ptrace library function at the end of this
post.
- mount() takes 4 arguments, and checks that only the super-user can
mount/umount things.
Happily this shouldn't break any old binaries.
- some general cleanups
I've made the pre-release available only as pure source code: no diffs,
no binary. The reason is that most people that needed this release want
it for the gdb-fixes: and they should have no problem recompiling the
kernel. Others just have to wait for the real 0.96.
Changes that are NOT in this pre-release, but which I hope to have in
the real 0.96:
- more include-file cleanups - I'm still working on these
- the wd8003 driver and hopefully some other parts of biro's
config.
- select() using the vfs-tables.
And possibly bugfixes that people find in this pre-release...
Linus
---------- library ptrace.c (wants gcc-2.1) ----------
#define __LIBRARY__
#include
#include
int ptrace(int request, int pid, int addr, int data)
{
long ret;
long res;
if (request > 0 && request < 4)
(long *)data = &ret;
__asm__ volatile ("int $0x80"
:"=a" (res)
:"0" (__NR_ptrace),"b" (request), "c" (pid),
"d" (addr), "S" (data)
: "si","bx","cx","dx");
if (res >= 0) {
if (request > 0 && request < 4) {
errno = 0;
return (ret);
}
return (int) res;
}
errno = -res;
return -1;
}
[next article]
From: [email protected] (Douglas E. Quale)
Subject: Re: cpio doesn't accept multi-volume archives
Message-ID: <[email protected]>
Date: 22 Apr 92 04:16:08 GMT
References:
Sender: [email protected] (The News)
Organization: Undergraduate Projects Lab, UW Madison
Lines: 15
In article
>
>cpio will prompt for "next tape" but immediately dies upon pressing
>ENTER.
>I am using the command:
>cpio -itv >
>Under Coherent, multi-volume archives require using the raw device.
>If this is true of Linux as well, what are the raw device names/numbers?
Unfortunately Linux does not yet have raw disk devices.
--
Doug Quale
[email protected]
[next article]
From: [email protected] (Elvis Leung)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Question - Boot Linux from HD
Summary: Boot, HD
Message-ID: <[email protected]>
Date: 22 Apr 92 04:11:16 GMT
Sender: [email protected]
Organization: Computer Power Software
Lines: 8
Is it possible to install and boot linux from a HD partition?
--
--------------------------------------------------------------------------------
Sek-Kit E. Leung Fax: +61-6-2836860
Computer Power Software Voice: +61-6-2836777
Multimedia Research Group ACSnet: [email protected]
[next article]
From: [email protected] (Charles Hedrick)
Subject: Re: timezone support, date, CMOS clock support: archive sites please note
Message-ID:
Date: 22 Apr 92 03:35:44 GMT
References:
mer.Prime.COM>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 19
[email protected] (Kevin Cummings) writes:
>I happend to find timebin.tar.Z on tsx-11 yesterday, and downloaded
>to my LINUX machine. I untarred it, configured it, and moved the
>binaries to /usr/bin and /usr/lib. I then ran date, and it told me
>it was Mon 20 Apr, 20:xx:xx 1992. No time zone info, even though
>/usr/lib/localtime was present. I then ran clock to see what was in
>my CMOS clock, and it was Sat 20 Apr, 14:xx:xx EDT 1992.
I hope you mean /usr/lib/zoneinfo/localtime was present. Otherwise
you don't have time zones.
I'm not sure whether the day of the week is worth worrying about. It
wasn't set properly on my machine either. I didn't know enough about
DOS to feel safe in ignoring it, so I display it and set it
appropriately. But I have a suspicion that DOS may not be using it.
Certainly Unix doesn't care. clock -s will set the Unix date based on
the date and time from the clock. It will ignore the day of the week,
since that is computed from the date.
[next article]
From: [email protected] (Peter Williams 8169821)
Subject: dvips.a.tar.Z fixed + new directory organization at archsci.arch.su.oz.au
Message-ID: <[email protected]>
Date: 22 Apr 92 01:35:34 GMT
References: <[email protected]>
Sender: [email protected]
Organization: Design Computing Unit, The University of Sydney
Lines: 24
Nntp-Posting-Host: dean.arch.su.oz.au
In article <[email protected]>, [email protected] (P
eter Williams 8169821) writes:
|> The dvips.a.tar.Z that I posted yesterday seems to be broken so I have
|> withdrawn it temporarily. Sorry!
|>
|> Should be mended and back up tomorrow.
|>
|> --
|> Peter Williams |e-mail: [email protected]
|> Key Centre for Design Quality |phone: +61-2-692 2053 or +61-2-660 6156
|> University of Sydney |+++++++++++++++++++++++++++++++++++++++
The mended version is now in directory pub/linux/tex-etc along with all of the
other tex related stuff at archsci.arch.su.oz.au (129.78.66.1).
This version of dvips does not work with font libraries whereas the statically
linked version in dvips.tar.Z does.
N.B. this is a different directory than previously posted (neatness) and now
includes a README file which gives a very brief description of the various files
'
contents.
--
Peter Williams |e-mail: [email protected]
Key Centre for Design Quality |phone: +61-2-692 2053 or +61-2-660 6156
University of Sydney |+++++++++++++++++++++++++++++++++++++++
[next article]
From: root (System PRIVILEGED Account)
Subject: A test post to comp.os.linux
Message-ID: <[email protected]>
Date: 22 Apr 92 05:15:25 GMT
Sender: [email protected] (Theodore Y. Ts'o)
Reply-To: root
Organization: The Internet
Lines: 2
I am testing the mail feed to comp.os.linux. Please ignore this message
(assuming it actually gets out!)
[next article]
From: V5068U%[email protected] (jim wiegand)
Subject: Mail access to list.
Message-ID: <[email protected]>
Date: 22 Apr 92 06:00:37 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 3
hi all;
It seems as if I am having some trouble posting messages.
Sorry to waste bandwidth.
[next article]
From: [email protected] (david nugent)
Subject: Re: gdb(1) and ptrace(2)
Message-ID:
Date: 22 Apr 92 03:18:06 GMT
References:
Reply-To: [email protected]
Organization: Unique Computing Pty Ltd, Melbourne, Australia
Lines: 27
enewton@uucp (Eric C. Newton x2092) writes:
> 1. gdb does not work with the new .95c+ kernel.
[...]
>
> I have the whole beastie compiled on my machine at home. The problem
> I have is that the ptrace() function is not in libc.a. I wrote a
> dummy ptrace() function to link the program.
>
> 3. GDB 4.5 has been released. Is anyone building this?>
>
> Yes. I will after 4.4 works.
>
> Anyone care to volunteer the sources for making a ptrace system
> call?
Linus has detailed some further fixes that were necessary, and I now have
a working gdb 4.5 binary here. It does require those kernel fixes to run.
When the new kernel is posted for download, I'll be uploading the gdb 4.5
binary, cdiffs to the 4.5 distribution and documentation (to tsx-11.mit.edu).
..............................................................................
david nugent Public Access Usenet "Only Nixon can go to China"
[email protected] +61-3-792-3507 - ancient Vulcan proverb
3:632/400@fidonet, 58:4100/1@intlnet, 199:4242/5@rainbownet, 33:300/6@admin
PO Box 260, Endeavour Hills, Victoria, Australia, 3802.
[next article]
From: [email protected] (Steve Sykes)
Subject: Re: Yet another X question
Keywords: X Windows
Message-ID: <[email protected]>
Date: 21 Apr 92 18:20:26 GMT
References: <[email protected]> <[email protected]
.edu> <[email protected]>
Sender: [email protected] (Network News)
Organization: Department of Computer Science, Warwick University, England
Lines: 12
Nntp-Posting-Host: stone
I'm very keen to get something like X up. At the moment my biggest
problem is a lack of knowledge on how to implement i/o mapping.
Rather than thinking up some obscure and probably bad way to allow
user processes access to io-space, I was hoping someone could point me
in the right direction. I think it's normally done as an ioctl call
but have now idea how to implement it.
Any help gratefully received
Please reply by e-mail
Laurence Culhane
[email protected]
[next article]
From: [email protected] (Rob Hooft)
Subject: Result of porting 'f2c' and 'ispell3.00.07'
Summary: Good OS! but why is ispell from emacs so slooooww?
Keywords: f2c linux ispell gcc2.1 interprocess communications.
Message-ID: <[email protected]>
Date: 22 Apr 92 07:07:14 GMT
Organization: Physics Department, University of Utrecht, The Netherlands
Lines: 33
Hello all,
As my first post to this group (listening for a few months already)
I would like to share my experience on installing 2 packets.
1) f2c, ftp'ed from research.att.com, compiles ok, but if it is run
with the shared libraries, it crashes on a segmentation violation.
static libraries work ok. My main problem is that I do not understand
the full error message.
Executables from fortran, produced with the f2c/gcc2.1 pair do not
terminate. Only when I press the ENTER key after the program stops,
the prompt is returned.
2) ispell3.00.07 compiles great! I've never seen a workstation
compiling this software in just under a minute real time!
Running ispell from the command-line works really very well.
Running ispell from EMACS, however, takes AGES! May have something
to do with interprocess communication?
I'll have to look into the problems I mentioned. I just thought: If I
post it, may be somebody can assist me in locating them (I'm lazy).
I'd like to compliment Linus: It is unbelievable how LINUX feels on my
ETEQ486/33 with 16Mb of memory. During compilations of medium-sized
packages with GCC2.1, my 340Mb IDE drive led only flashes every 30
seconds for update.... Compilations run at 98+% of the CPU.....
Greetings, Rob Hooft, Utrecht University, The Netherlands.
--
Rob Hooft, Bijvoet Center for Biomolecular Research,
Chemistry department University of Utrecht, the Netherlands
[email protected] [email protected] [email protected] [email protected]
[next article]
From: [email protected] (Elvis Leung)
Subject: Is there an archive server for the linux files ?
Message-ID: <[email protected]>
Date: 22 Apr 92 07:08:42 GMT
Sender: [email protected]
Organization: Computer Power Software
Lines: 6
Subject says it all.
--
--------------------------------------------------------------------------------
Sek-Kit E. Leung Fax: +61-6-2836860
Computer Power Software Voice: +61-6-2836777
Multimedia Research Group ACSnet: [email protected]
[next article]
From: [email protected] (Lars Lindner)
Subject: HD boot using shoelace. HELP!
Message-ID: <[email protected]>
Date: 22 Apr 92 07:44:36 GMT
Sender: [email protected]
Organization: Ericsson Telecom AB
Lines: 31
Nntp-Posting-Host: eos10c02.ericsson.se
The question have probably already been answered but i could not find it
so here comes the question again.
I have installed shoelace on my computer, changed /etc/config to contain
" boot /vmlinux", made a file /vmlinux with "dd if=/dev/at0 of=/vmlinux"
with my boot floppy in drive A.
After the commands "laceup /dev/hda2 wini" and "laceup -w 2 /dev/hda"
i tryed to reboot and here is what a got:
/dev/fd0 not found - guessing major and minor device numbers.
processor = -1
ramimagedev = 512 /dev/fd0
ramsize = 0
rootdev = 770 bootdet
scancode = 13
Component base text data bss stack symbols model
shoelace : 80000 5688 C0C B80 91EC 0 ComI&D
Load Parts
kernel : kernel not found
Load Image
And after that nothing happend. What have i done wrong. Is there something
wrong in the /etc/config or shall i use another way to create /vmlinux?
All help is appreciated.
[email protected]
[next article]
From: [email protected] (Leslie Griffiths (Griff))
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: How to change the path : Followup
Message-ID: <[email protected]>
Date: 22 Apr 92 08:27:20 GMT
References: <[email protected]> <1992Apr21.031303.208
[email protected]>
Sender: [email protected] (USENET News Owner)
Organization: Coventry Polytechnic, Coventry, UK
Lines: 54
Nntp-Posting-Host: cc_sysh
In article <[email protected]> [email protected] (Ho
ngjiu Lu -- Graduate Student) writes:
>In article <[email protected]> [email protected]
nsw.edu.au (Matthew Jackson) writes:
>>A short while ago I posted a query on how to change the path. Attached are the
>>responses I received. I have not tried them as my Linux machine is at home and
>>my net access is at work.
>>
>>It seems those who have had the same problem replied after they worked it out
>>and also some more familiar with Unix than myself replied too, for which I am
>>grateful.
>>
>>Anyway, I thought it would be of some use to others so I post a summary.
>>
>>Summary: For historical reasons, the root user does not have the
>> current directory in the path.
>
>WRONG, WRONG, WRONG. That is not for historical reasons. Do your
>homework about Unix. That is for security reason.
>
>>
>> Under ash and bash, PATH=dir1:dir2:...
>> Then export PATH
>>
>> Under csh or tcsh, PATH=(dir1 dir2 ...)
>
>Please DO NOT DO THAT. KEEP THE WAY IT WAS.
>
>H.J.
Hmm. Actually, under bash/sh and other Bourne-like shells
$ PATH=dir1:dir1:...
$ export PATH
Under Csh, Tcsh and other C-shells
% set path=( dir1 dir2 dir3 )
(In which case the environment variable PATH is automatically updated)
or
% setenv PATH dir1:dir2:dir3
Which does it directly. Preferred method (I believe) is the former (i.e.
"set path=").
Above applies to sh, bash, csh, tcsh. Don't know about ash or others.
Cheers
--
Griff \\ [email protected]
^^^^^ :-= ^^^^^^^^^^^^^^^^
//
[next article]
From: [email protected] (Joerg Pommnitz)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: Miscellaneous questions.
Message-ID:
Date: 22 Apr 92 10:38:59 GMT
References: <[email protected]>
Sender: [email protected] (Owner of all binaries)
Organization: tu-chemnitz
Lines: 18
[email protected] (Scott Dunn) writes:
I'd like to ask the same question:
>2. What is a working version of tar and un/compress?
> A lot of problems seem to be solved by getting new versions of the
> aforementioned. However, nobody has indicated how to identify which
> are the dodgy copies.
> Sometimes when uncompressing it seems to get stuck in an infinite
> loop which takes all system resources. ie I can't login to another
> terminal and kill the process. Is this what is known as a 'race
> condition' ? Is this indicate a dodgy copy of un/compress ?
How can I fix these problems ?
What's the reason ?
I must say, I'm really glad somebody else ran into the same problem.
I was very angry that this would be a problem of my specific hardware.
[next article]
From: [email protected] (Peter Orbaek)
Subject: Announcement: SYSV init, shutdown, last, and more...
Keywords: init last setgrp shutdown reboot halt
Message-ID: <[email protected]>
Date: 22 Apr 92 06:49:40 GMT
Sender: [email protected] (Peter Orbaek)
Organization: DAIMI: Computer Science Department, Aarhus University, Denmark
Lines: 36
I just put
/pub/Linux-source/admutil-1.0.tar.Z
on
ftp.daimi.aau.dk (130.225.16.27)
It includes source for a better last(1) and setgrp(1) both by Michael Haardt,
plus a SYSV compatible init(8) program with runlevels and all.
This init was written by Miquel van Smoorenburg for Minix, and I ported it to
Linux with little effort.
There's also a shutdown(8) program as well as passwd(1) and chsh(1).
You should be able to build the binaries with both gcc 1.40 and gcc 2.1.
Follow the instructions in the Makefile and README.
Be warned that the init program is fairly untested, and I don't use it myself.
It's very easy to lock your self out of the machine if you do something wrong
with init and the new inittab format, so be sure you know what you are doing
before you install the new init.
There's also a problem with putpwent(3) in the older versions of libc, and if
you don't fix this or use the newest gcc 2.1 libc (later than 10-Apr-92) then
chsh and passwd will trash the /etc/passwd file and make it hard to log into
the machine afterwards.
- Peter ([email protected])
--
Peter Orbaek ----------------- [email protected] | ///
Hasle Ringvej 122, DK-8200 Aarhus N, DENMARK | ///
| \\\///
"Strong typing is for people with weak memories" | \XX/
[next article]
From: [email protected] (Greg Lee)
Subject: Re: timezone support, date, CMOS clock support: archive sites please note
Message-ID: <[email protected]>
Date: 22 Apr 92 13:52:29 GMT
References:
mer.Prime.COM>
Sender: [email protected] (News Service)
Organization: University of Hawaii
Lines: 23
Nntp-Posting-Host: uhunix.uhcc.hawaii.edu
In article <[email protected]> [email protected] (Kevin
Cummings) writes:
}I happend to find timebin.tar.Z on tsx-11 yesterday, and downloaded
}...
}Something's not right here.
Yes, something is not. I set the clock and the date with the new
programs 'clock' and 'date', and it seemed to work ok. 'date' did
mention the time zone (HST, -10 from GMT). But at the next reboot,
the 'clock -s' I had put in /etc/rc didn't set the date and gave
an error message that I probably wasn't root. I was root, but 'clock'
was suid bin and /dev/port was owned by bin. I had to change 'clock'
to suid root to get it to work. Time passed. After the next reboot
and 'clock -s' executed from /etc/rc, 'clock' said it was
Sun 4/21 15:21, but 'date' claimed it was
Tue 4/21 05:21. It was actually Tue 4/21 15:21. So I did
another 'clock -s', and then I got
Sun 4/21 15:22 from 'clock', and
Tue 4/21 15:22 from 'date'. That's correct, except 'clock'
still displays the wrong day-of-week. After 'clock -w', 'clock' and
'date' finally both displayed the correct date and time. Whew.
--
Greg Lee
[next article]
From: [email protected] (Bennett E. Todd)
Subject: floptools.sh -- handle multiple-floppy archives
Message-ID: <[email protected]>
Date: 22 Apr 92 17:30:57 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Bennett E. Todd)
Organization: The Internet
Lines: 189
This is some vile, yucko throwaway code. I suspect it should be
completely rewritten in perl. Then the same may well be true of most of
Unix:-).
I wrote this pair of programs, rdflops and wrflops. Each one prompts for
multiple floppies in sequence. I can back up my hard disk with something
like
tar cvf - / | compress -v | wrflops /dev/fd0
and extract files back from the archive with
rdflops /dev/fd0 | compress -d | tar xvf - filenames
I'm running the 0.95c+ kernel on the 0.95a rootimage, with utils out of
collections/095binar.tar.Z. I pulled all this from tsx-11. Ash as
/bin/sh was bombing off periodically, so I blew it away and stuffed
bash in there. You will probably want to change the #! line on these
scripts (especially wrflops!) to point to wherever you keep your bash.
The rdflops script seems to me to be simple, correct, and robust.
Catbytes could be more elegant, but it seems to do the job. With an
extra cat(1) squatting behind it to smooth the buffering (as I did in
wrflops), it seems to do a really clean job of driving the floppy. I
couldn't get dd(1) to do this job, since it pays attention to the
incoming blocksize. I wrote catbytes the way I did (avoiding printf and
so on) to keep the binary from bloating up too badly. Eventually I will
come to trust and appreciate shared libraries:-).
The wrflops script has some *really* fragile code in it for trying to
follow symlinks and deduce floppy sizes from major and minor device
numbers; you may well want to rip that all out.
The dirname script is probably not correct, but it seems to work well
enough for wrflops.
The reason wrflops is so careful to attempt to follow symlinks, is I use
/dev/fd0 and /dev/fd1. fd0 is a symlink to PS0 and fd1 is a symlink to
at1.
The only problem I have had, and it hasn't resulted in any loss of data,
has been the lack of EOF handling. When you run the restore, if you are
using tar(1) straight then it works perfectly. If you have a compress in
there then it works fine until you get to the end; then, instead of
everything exiting, compress proceeds to blow up. Happily, Linux seems
to be nicely robust in the face of programs trying to allocate up all of
the VM.
-Bennett (curmudgeon)
[email protected]
#!/bin/sh
echo 'x - catbytes.c'
sed 's/^ //' <<'EOF' >'catbytes.c'
#include
#include
/*
* catbytes n
*
* Copies ``n'' bytes from stdin to stdout. Unlike ``dd'' it doesn't
* care how stdin comes in buffered; unlike everything else I can find
* it doesn't gobble any excess bytes from stdin. Why does this have
* to be so hard?
*/
#define BS 65536
static char msg1[] = "syntax: catbytes count\n";
static char msg2[] = "short input\n";
static char msg3[] = "short write\n";
int main(int argc, char **argv) {
static char buffer[BS];
int i, n, total;
extern int read(int, char *, int);
extern int write(int, char *, int);
extern int atoi(char *);
if (argc != 2 || !isdigit(argv[1][0])) {
(void) write(2, msg1, sizeof(msg1));
exit(1);
}
total = atoi(argv[1]);
while (total > 0) {
i = (total > BS) ? BS : total;
if ((n = read(0, buffer, i)) <= 0) {
(void) write(2, msg2, sizeof(msg2));
exit(1);
}
if (write(1, buffer, n) != n) {
(void) write(2, msg3, sizeof(msg3));
exit(1);
}
total -= n;
}
return(0);
}
EOF
echo 'x - wrflops'
sed 's/^ //' <<'EOF' >'wrflops'
#!/bin/sh
progname=`basename $0`
syntax="syntax: $progname device"
die(){
echo "$progname: $*" 1>&2
exit 1
}
case $# in
1) device="$1" ;;
*) die "$syntax" ;;
esac
follow(){
[ $# -eq 1 ] || die "fn follow: syntax: follow dirname"
cd `dirname $1`
f=`basename $1`
(
while [ -h "$f" ] ; do
cd `dirname "$f"`
f=`ls -l "$f" | sed 's/^.* -> //'`
done
echo `pwd`/`basename $f`
)
}
f=`follow $1`
[ -b $f ] || die "I only know about block devices"
id=`ls -l $f | awk '{print $5 " " $6}' | tr -d ' '`
case "$id" in
'2,28') secs=18;;
'2,29') secs=18;;
'2,8') secs=15;;
'2,9') secs=15;;
*) die "I don't know that device, sorry."
esac
count=$[ $secs * 80 * 1024 ]
while catbytes $count | cat >"$device" ; do
echo -n 'Insert another floppy: ' >/dev/tty
read ans /dev/null
done
EOF
echo 'x - dirname'
sed 's/^ //' <<'EOF' >'dirname'
#!/bin/sh
progname=`basename $1`
syntax="syntax: $progname name"
die(){
echo "$progname: $*" 1>&2
exit 1
}
[ $# -eq 1 ] || die "$syntax"
s=`echo "$1" | sed 's,/[^/]*$,,'`
if [ "$s" = "$1" -o -z "$s" ] ; then
echo .
else
echo "$s"
fi
EOF
echo 'x - rdflops'
sed 's/^ //' <<'EOF' >'rdflops'
#!/bin/sh
progname=`basename $0`
syntax="syntax: $progname device"
die(){
echo "$progname: $*" 1>&2
exit 1
}
case $# in
1) device="$1" ;;
*) die "$syntax" ;;
esac
while cat "$device"; do
echo -n 'Insert another floppy: ' >/dev/tty
read ans /dev/null
done
EOF
exit 0
[next article]
From: [email protected] (Kevin Cummings)
Subject: Re: timezone support, date, CMOS clock support: archive sites please note
Message-ID: <[email protected]>
Date: 22 Apr 92 17:49:57 GMT
References:
mer.Prime.COM>
Sender: cummings@hammer (Kevin Cummings)
Organization: Prime Computer R&D
Lines: 61
In article
.edu (Charles Hedrick) writes:
> [email protected] (Kevin Cummings) writes:
>
> >I happend to find timebin.tar.Z on tsx-11 yesterday, and downloaded
> >to my LINUX machine. I untarred it, configured it, and moved the
> >binaries to /usr/bin and /usr/lib. I then ran date, and it told me
> >it was Mon 20 Apr, 20:xx:xx 1992. No time zone info, even though
> >/usr/lib/localtime was present. I then ran clock to see what was in
> >my CMOS clock, and it was Sat 20 Apr, 14:xx:xx EDT 1992.
>
> I hope you mean /usr/lib/zoneinfo/localtime was present. Otherwise
> you don't have time zones.
Yes, you are right, I meant to say /usr/lib/zoneinfo/localtime.
Also I got the circumstances backward. When I ran date, it came back:
Mon Apr 20, 16:xx:xx EDT 1992
When I ran clock it came back
Sat Apr 20, 20:xx:xx 1992
Furthermore I ran them again last night! New results:
date:
Tue Apr 21, 17:xx:xx EDT 1992
clock:
Jan Apr 21,, 21:xx:xx 1992
> I'm not sure whether the day of the week is worth worrying about. It
> wasn't set properly on my machine either. I didn't know enough about
> DOS to feel safe in ignoring it, so I display it and set it
> appropriately. But I have a suspicion that DOS may not be using it.
> Certainly Unix doesn't care. clock -s will set the Unix date based on
> the date and time from the clock. It will ignore the day of the week,
> since that is computed from the date.
Is the author aware that the DOW is stored as a number from 3-9 instead of
a number from 0-6 or 1-7? I found this out while playing with a DOS
program to read the CMOS memory. Looks like the clock program accessed
beyond Sat into Jan! Can you say Oops? Sure ya can.
If clock won't display the correct date, I certainly don't trust it to
SET the correct date! Last thing I want in LINUX is bad dates.
=================================================================
Kevin J. Cummings Prime Computer Inc.
20 Briarwood Road 500 Old Connecticut Path
Framingham, Mass. Framingham, Mass.
Work: [email protected]
Home: [email protected]
Std. Disclaimer: "Mr. McKittrick, after careful consideration,
I've come to the conclusion that your new
defense system SUCKS..." -- War Games
=================================================================
[next article]
From: [email protected] (Darren Senn)
Subject: Re: timezone support, date, CMOS clock support: archive sites please note
Keywords: Solution
Message-ID: <[email protected]>
Date: 22 Apr 92 18:18:43 GMT
References:
mer.Prime.COM>
hammer.Prime.COM>
Sender: [email protected]
Lines: 11
I installed the timebin package on my system, and found the same problems that t
he
rest of you had. I witnessed both of the problems that the rest of you have. M
y
solution to the first problem was to change the ownership of clock and date. To
fix the latter problem, I changed the mode of the zoneinfo directory, the localt
ime
file, and the posixrules file to be publicly readable. That fixed everything up
.
Now my timezone is correct, my day and date are correct.
--
Darren Senn Phone: (408) 479-1521
[email protected] Snail: 1785 Halterman #1
Wasurenaide -- doko e itte mo soko ni anata wa iru yo. Santa Cruz, Ca 95062
[next article]
From: [email protected] (Jim Winstead Jr.)
Subject: Re: two questions
Message-ID: <[email protected]>
Date: 22 Apr 92 19:50:30 GMT
References:
Sender: [email protected] (The News System)
Organization: Harvey Mudd College, WIBSTR
Lines: 23
In article
etmore) writes:
>1) Is there a man written yet? I was going to write my own and port
>over man pages for all of the stuff that I don't have them for yet, but
>I figured I would ask first.
I found a package a while back labelled man-1.0.tar.Z, which
identified itself as the GNU 'man' package. It's not on
prep.ai.mit.edu, so I think the GNU Folks may have dropped it in favor
of TeXinfo, but I've been using it on my system for some time with
good luck. I believe it can be found on tsx-11.mit.edu, possibly
compiled. Either that, or try looking it up with archie.
>2) What is the MCC release of Linux?
It's the Manchester Computing Center (?) release, which serves as a
sort of extended-release, as I see it. It includes many more
utilities than the basic release from Linus and myself, and as a
result comes on three (four?) floppies.
--
Jim Winstead Jr. (CSci '95) | "Catch a fish!"
Harvey Mudd College | -Geddy Lee,
[email protected] | San Diego Sports Arena
Disclaimer: Mine, not theirs! | January 20, 1992
[next article]
From: [email protected]
Subject: How do I get BOOTANY to work with Linux?
Message-ID: <[email protected]>
Date: 22 Apr 92 20:06:13 GMT
Sender: [email protected] (USENET News System)
Organization: Jus' me.
Lines: 14
Thanks to those few who answered my question about tar-ing from a
rawrit' diskette... I've got mtools up and running now, so that
won't be necessary for file transfer any more. 🙂
I'm *REALLY* thrilled with Linux so far. Now I think it's about
time to put it on an 'even standing' with DOS (on this machine,
at least). The obstacle is this annoying boot-from-floppy business.
BOOTANY was *mentioned* in the various FAQs and notes, but I don't
recall seeing a detailed description of how to make it work with
Linux. Could some kind soul give me some hints?
Thanks in advance,
Rick
[next article]
From: [email protected] (Darren Senn)
Subject: What is Linux's programming environemnt?
Message-ID: <[email protected]>
Date: 22 Apr 92 20:22:12 GMT
Sender: [email protected]
Lines: 13
The subject pretty much says it all. If I want to go out and get programming ma
nuals,
what should I get? SVR4? BSD? POSIX? SCO (god forbid)? I realize that Linux is
_not_
complete yet, so perhaps my question should be "To which programming environment
does
Linux aspire to be most similar?"
P.S.: By programming environment, I refer to the API, not the tools.
Thanks.
--
Darren Senn Phone: (408) 479-1521
[email protected] Snail: 1785 Halterman #1
Wasurenaide -- doko e itte mo soko ni anata wa iru yo. Santa Cruz, Ca 95062
[next article]
From: [email protected] (Paul Allen)
Subject: Re: all kinds of questions (autoparking)
Message-ID: <[email protected]>
Date: 22 Apr 92 20:40:35 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 35
[email protected] (Kevin Cummings) wrote:
|Head parking is suggested (by me anyways) as necessary before powering
|off your computer. Who knows what the older hard disk controllers
|will do when you turn the power off, and the write line happens to
|accidently spit something out. Good way to trash whole tracks.
|
|Used to happen all the time on my CP/M system, until I learned to
|park the heads before powering down. If your drive doesn't recognize
|the power fail, your disk controller probably doesn't either, and you could
|lose something. Probably happens less today than a few years ago,
|but if you're using MFM drives, the chance is significantly increased.
My first hard disk was a 5.5Mb Shugart that I ran on my old Z-100 with
an Adaptec SASI controller, a homebrew SASI S-100 host adapter, and my
own driver. It was one of the original 5.25" winchester drives, with
an access time in excess of 100ms. (But man, was it cheap! :-)) It
would sometimes lose data if shut down without being parked, so I wrote
a head parking utility for it. I replaced it with a 10Mb drive out of
an IBM XT, and have never parked the heads since.
Basically, if you're running on 10-year-old hardware, you might need to worry
about this. You better believe that the drive manufacturers figured out
a long time ago that customers will not tolerate disks that lose data.
Head parking just isn't a problem anymore. Old disks that don't autopark
should be parked before moving, but that's about it. Anything newer than
an IBM XT will do the right thing when you turn it off. (Unless it's
broken, of course, but that's a special case. :-))
If someone can provide a real-life story about a post-XT drive that
lost data consistently if not parked before power-down, please do. I
always enjoy being proven wrong! 🙂
Paul Allen
[email protected]
[next article]
From: [email protected] (Giles D Malet)
Subject: Bug in GCC 2.1 mktemp()
Message-ID: <[email protected]>
Date: 21 Apr 92 21:57:33 GMT
Reply-To: [email protected] (Giles D Malet)
Organization: You gotta be kidding !
Lines: 81
Attempts to run a newly installed mailer kept producing messages along the
lines of "Can't create temporary file - already exists !". Much digging
later I discovered the source of the problem - mktemp().
mktemp should produce a unique filename, based on a template, for each call.
Checking the source for libc.a in the GCC 2.1 distribution revealed the
fault. For example, running the following little program should produce
output along the lines of :
Got
Got
Got
Got
Got
but instead it produces something like :
Got
Got
Got
Got
Got
which is not quite unique.
This huge patch to ..../compiler/lib/posix/mktemp.c (I think) fixes
the problem, once you manage to reproduce the library :
----- cut -----
52c52
< count = 0;
---
> count = 0, oldpid = pid;
----- cut -----
/*----------------------------------------------------------------------
* Test mktemp() function.
*/
#include
#include
extern char *mktemp(char *name);
char *do_mktemp(void);
main()
{
int i;
char *name;
for ( i=0 ; i<5 ; i++ ) {
if ( name = do_mktemp() ) {
printf("Got <%s>\n",name);
free(name);
} else {
exit(-1);
}
}
}
char *do_mktemp(void)
{
static char template[] = "/tmp/dummy.XXXXXX";
char *f;
if ( (f = malloc(32)) == NULL ) {
fprintf(stderr,"malloc failed\n");
exit(-1);
}
(void) strcpy(f, template);
if (mktemp(f) == NULL) {
fprintf(stderr,"can't create temporary file");
return NULL;
}
return f;
}
/*----------------------------------------------------------------------
*/
--
Giles D Malet [email protected] Waterloo, Ont, Canada +1 519 725 5726
[next article]
From: [email protected] (Rajat Datta)
Subject: Suspending emacs and bringing it back
Message-ID: <[email protected]>
Date: 22 Apr 92 21:53:23 GMT
Sender: [email protected] (NNTP News Poster)
Organization: IBM T.J. Watson Research Center
Lines: 14
Disclaimer: This posting represents the poster's views, not necessarily those of
IBM
Nntp-Posting-Host: shravani.watson.ibm.com
I picked up emacs from tsx-11 and seem to be having a problem with
bringing emacs back up to the foreground once I've suspended it with
C-z. The console seems to come back in cooked mode. The screen is
not updated with the emacs screen, and any control chars I type show
up as C-
Anyone else seen this problem? I'm on 95c+. I'll try it on the
pre-96 (95c++; the uninformed might think it was a rewrite in c++ 🙂
release that's out there today and see if the problem is still there.
--
rajat ([email protected])
[next article]
From: [email protected] (Rajat Datta)
Subject: gcc 2.1, ld and -g
Message-ID: <[email protected]>
Date: 22 Apr 92 22:09:13 GMT
References: <[email protected]>
Sender: [email protected] (NNTP News Poster)
Organization: IBM T.J. Watson Research Center
Lines: 10
Disclaimer: This posting represents the poster's views, not necessarily those of
IBM
Nntp-Posting-Host: shravani.watson.ibm.com
I'm able to do pretty much everything with gcc 2.1 except for loading
programs with the -g flag. gcc -g -c a.c goes fine for any file a.c,
but gcc -g -o a a.c does not. ld gives an error message about not
finding library libg.a. Doing a gcc -v seems to pass the -g flag to
ld which doesn't know what to do with it. Should there be a libg.a?
If so, where is it? Is it the same as libgcc.a? Most likely I'm
doing something wrong. What is it? Thanks for any help.
--
rajat ([email protected])
[next article]
From: [email protected] (Charles Hedrick)
Subject: Re: IDE drive performance with linux
Message-ID:
Date: 22 Apr 92 22:41:36 GMT
References: <[email protected]> <[email protected]> <199
[email protected]>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 11
If you're getting one block per rotation, that means that the system
can't keep up with the rotation speed of the disk. This is not at all
unusual. Classically what you do is arrange the files on disk using
every other block or every third block, or whatever you need in order
that the system can keep up. With system v, there's an argument to
mkfs to specify this interleave. With the Berkeley Fast File System,
it's specified to mkfs or tunefs. Since folks are busy implementing
an improved file system for Linux, I'm going to hope that this feature
is included. It would presumably be easy to put into the current one
as well, but I haven't looked at the code to see what would be
involved.
[next article]
From: [email protected] (gary a moyer)
Newsgroups: alt.os.linux,comp.os.linux
Subject: help, cannot mount file system...
Summary: what the hell did I do?
Keywords: mount umount /dev
Message-ID: <[email protected]>
Date: 22 Apr 92 23:00:48 GMT
Sender: [email protected]
Lines: 7
Hey netters! I _accidentally_ messed up my ability to mount
a file system! I did it inadvertantly while trying to incorporate
some _protection_. Anyhow, what other files does mount access besides
the actual device in /dev ? I am getting the message:
mount: unable to open lock-file
thx. [email protected]
[next article]
From: [email protected] (Ross Biro)
Subject: Re: gcc 2.1, ld and -g
Message-ID: <[email protected]>
Date: 22 Apr 92 23:03:21 GMT
References: <[email protected]> <1992Apr22.220913.27276@wats
on.ibm.com>
Sender: [email protected] (News Service)
Organization: DSG, Stanford University, CA 94305, USA
Lines: 24
>ld which doesn't know what to do with it. Should there be a libg.a?
There seems to be a lot of confusion about libg.a. When gcc
-g attempts to link a program, it links in libg.a before libc.a to
include any special debugging functions. So as a minimum libg.a you
could simply include nothing, or as a maximum you could compile libc.a
with the -g flag and extra debugging statments, and then use that for
libg.a. I chose a different route by creating a symbolic link from
libg.a to libc.a. Currently linux has no standard libg.a, and
probably only needs an entry in the FAQ to tell people about it.
If you intend to do serious debugging, you should compile the
libraries with the -g flag, call the result libg.a and then tell gdb
to look in the library source tree for source files. That way you
could step through both your program and the library without any
problems.
--
Ross Biro [email protected]
Member League for Programming Freedom (LPF)
mail [email protected] for information
[next article]
From: [email protected] (Charles Hedrick)
Subject: Re: timezone support, date, CMOS clock support: archive sites please note
Message-ID:
Date: 22 Apr 92 23:06:09 GMT
References:
mer.Prime.COM>
hammer.Prime.COM>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 68
[email protected] (Kevin Cummings) writes:
>Also I got the circumstances backward. When I ran date, it came back:
>Mon Apr 20, 16:xx:xx EDT 1992
>When I ran clock it came back
>Sat Apr 20, 20:xx:xx 1992
>Furthermore I ran them again last night! New results:
>date:
>Tue Apr 21, 17:xx:xx EDT 1992
>clock:
>Jan Apr 21,, 21:xx:xx 1992
I'm not sure what you think is wrong here. The date and time look OK.
They're what I would expect if the CMOS clock is set to GMT, which is
the recommended configuration. EDT is 4 hours to the west of GMT.
You can get clock to do the conversion for you as it displays the
time. Use 'clock -u'.
If you want the CMOS clock to be using local time, I think the
documentation explains how to do it. If your complaint is about the
day of week, you're right.
I'm sorry about the -u stuff. If everybody is willing to run the CMOS
clock on GMT, I can make that the only behavior. My guess was that
most people would be using both Linux and DOS, and would want to run
the CMOS clock on local time in order to avoid oddities in DOS. Thus
I set up clock to default to having the CMOS clock on local time. I
could also have a file in zoneinfo that specifies whether which way
the system is set up, so you don't have to remember to type -u.
>Is the author aware that the DOW is stored as a number from 3-9 instead of
>a number from 0-6 or 1-7? I found this out while playing with a DOS
>program to read the CMOS memory. Looks like the clock program accessed
>beyond Sat into Jan! Can you say Oops? Sure ya can.
OK, I'll fix it. I was unable to find any documentation on the day
of week. 3-9???
>If clock won't display the correct date, I certainly don't trust it to
>SET the correct date! Last thing I want in LINUX is bad dates.
Huh? As far as I can tell, all that's wrong is the day of week. That
is really irrelevant to Unix. The -s function (set Unix day-time
based on CMOS clock) completely ignores the day of week from the CMOS
clock. All Unix keeps is the number of seconds since the beginning of
time. It computes the date, day of week, etc., based on that. As
long as the date and time are right, everything should work OK on
Linux. I'll still fix it, but it doesn't affect setting the time for
Linux. (What it does affect is setting the CMOS clock, if you decide
to do that under Linux. The program will set the day of week based on
0-6 instead of 3-9.)
I'll take a look at ownerships and protections again when I make up
the new copy of timebin. Some postings make it look like I got them
wrong, though it could also be an error in setting things up. At
least I'll note in the documentation that /usr/lib/zoneinfo and all
files in it should be world readable, and that clock should be setuid
to the same user that owns /dev/port.
[next article]
From: [email protected] (H. Peter Anvin N9ITP)
Subject: Re: problems booting
Message-ID: <[email protected]>
Date: 22 Apr 92 15:47:27 GMT
References: <[email protected]> <[email protected]
u.edu> <[email protected]>
Reply-To: [email protected] (Peter Anvin)
Organization: Northwestern University, Evanston, Illinois, USA
Lines: 25
In article <[email protected]> of comp.os.linux,
[email protected] writes:
> In article <[email protected]>, [email protected]
(H. Peter Anvin N9ITP) writes:
> |> In article <[email protected]> of comp.os.linux,
> |> [email protected] (Richard Alan Brown) writes:
> |> >
> |> > I used rawrite to make a bootable disk and a rootimage disk. When I
> |> > boot my pc with the floppy in, I get the loading............
> |> > and the SVGA mode question.
> |> > when I press any other key, the system simply reboots, and cycles through
> |> > this same process.
>
> I've had exactly the same problem. After uncompressing the 0.95c+
> bootimage under 0.12, I used cp to put it onto a floppy. This
> resulted in the behaviour described above. It turned out that my
> copy of uncompress was corrupt. After switching to the 0.95a
> uncompress, all worked well.
I uncompressed it under SunOS, FTP'd it to a PC floppy and used rawrite to
put it on a disk. Is there a bug in the 0.95a compress?
--
INTERNET: [email protected] TALK: [email protected]
BITNET: HPA@NUACC HAM RADIO: N9ITP, SM4TKN
IRC: Scalar NeXTMAIL: [email protected]
"Kissing a smoker is like licking an ashtray." -- Anon.
[next article]
From: [email protected] (Sheldon Bainwohl)
Subject: Confused.
Message-ID: <[email protected]>
Date: 22 Apr 92 18:37:00 GMT
Organization: Defence and Civil Institute of Environmental Medicine
Lines: 51
I've noticed a few problems upon boot-up recently that have me
confused. I'm basically a beginner to Un*x (linux), however, up
to this point I've been relatively successful at getting linux
(V. 0.95c+) up and running on my machine.
Once in a while, while booting up, besides the nornal messages..
8 virtual consoles
4 ptty's
....etc....
... I get the message "memory error" followed by a bunch of other
stuff (which I wasn't able to record, sorry), however, the messages
continue to tell me that the partition table is ok and that all
systems are functioning. Linux comes up ok, and everything seems to
be fine. Is there a problem that I should be concerned about?
In addition, more recently, I've noticed that during boot up, instead
of getting the message:
8 virtual consoles
4 ptty's
etc...
I get:
5 virtual consoles
4 ptty's
etc...
What's going on, are these problems related. I haven't really noticed
if they tend to occur together.
My machine is a 386, 20 MHz, with 8 MEGS of RAM and a Quantum 245 MEG
IDE drive. I should also point out that the boot disk has been edited
to boot off the hard drive using zipzap (i.e. offsets 508 and 509 have
been changed to boot off hda1).
In addition, I do make use of extended partitions, however, the partition
table shown upon boot up is always accurate.
In an other unrelated/related/? problem, if I try to boot off the hard
drive (i.e DOS v.5) which is hd4, the machine just hangs..... no memory
check or nothing. What has me confused is that if I use a bootable
DOS disk, there is no problem. DOS comes up just fine. My DOS drive
(C:\) seems perfectly happy. I have designated the DOS partition hda4 to
be active. Any suggestions would be greatly appreciated.
Thanks very much in advance. I apologize if these are FAQ's.
Sheldon
[next article]
From: [email protected] (M. Saggaf)
Subject: Bad Sectors
Message-ID: <[email protected]>
Date: 22 Apr 92 23:28:14 GMT
Sender: [email protected] (News system)
Organization: Massachusetts Institute of Technology
Lines: 15
Nntp-Posting-Host: e40-008-13.mit.edu
I have been getting many hard disk I/O errors and timeouts under
Linux. Those seem to occur at the same blocks. My disk was low-level
formatted recently with the proper bad sector information and neither Disk
Manager nor the AMI BIOS disk surfece scan program could find any more
bad sectors. Hence, I think either Linux tries to write to sectors
that have been marked bad by low-level formatting, or it is more
sensitive to flaky sectors than both Disk Manger and the BIOS program.
'mkfs' does report those bad blocks when invoked by the -c (verify)
switch, but apparently it does not try to do anything to prevent Linux
from using them. Is there any Linux utility that would mark those
blocks as bad and prevent Linux from using them? Please let me know.
(Maybe such a utility should be incorporated into mkfs). Thanks.
-- M. Saggaf ([email protected])
[next article]
From: [email protected] (Todd Radel)
Subject: Re: Graphics and IPC questions...
Message-ID: <[email protected]>
Date: 22 Apr 92 23:36:40 GMT
References:
uiuc.edu>
Organization: University of Delaware
Lines: 21
In article <[email protected]> chojnows@osiris (Brian Choj
nowski) writes:
>I guess I havent made myself clear. I would like to do something like this,
>where I would create a superset of the vt100. But the mode as it is, is
>insufficient for the kind of graphics I am talking about. I want to also add
>this new mode to the telnet program (when it gets written), so that the
>graphics get displayed remotely.
This has already been done. There's a package called UnixWindows (or UW)
that does this. From your remote machine, you use a terminal program
written for UW to dial in. After you've logged in, you type "UW" and the
server kicks in. You can have up to 9 (I think) movable windows on the
screen. Of course, you can only communicate in one at a time, and at
2400bps it's pretty slow, but it works...
--
* Todd Radel Sysop, Dickinson Domain BBS *
| CIS/English FidoNet: 1:150/160; VirtualNET @3023 |
| University of Delaware REGISTERED Alpha Colony VI! |
* "You don't need to be crazy to be a sysop, but it helps!" *
[next article]
From: [email protected] (Demian A. Johnston)
Subject: BootLin and side effects..
Message-ID:
Date: 22 Apr 92 23:42:31 GMT
Organization: Sophomore, Physics, Carnegie Mellon, Pittsburgh, PA
Lines: 16
Okay, maybe someone can explain this to me.
I have an ATI VgaWonder graphics card. I only and able to get
SuperVga modes of 132x40 and 132x25 (is that right.) I wanted to be
able to get 80x50 or
whatever but I couldn't. Now I installed bootlin and get the startup
stuff. But when I type space I get the normal 80x25 and instead of the
Menu for SuperVga modes I get 80x50 which is what I wanted but I can
nolonger get 132xXX modes..??
Any Ideas??
Demian
---> [email protected]
---> [email protected]
---> [email protected]
[next article]
From: [email protected] (Joshua E Gentry)
Subject: Can't get linux on my hard drive
Message-ID: <[email protected]>
Date: 22 Apr 92 23:14:28 GMT
Organization: Worcester Polytechnic Institute
Lines: 12
Well, I have a 386sx with an IDE hard drive with two partitions. One
partition has all my DOS stuff and on the other I want to put linux. Anyway
using edpart.exe I made both of them primary. Now here is the problem: If I
set my DOS partition to bootable and the linux partion to non bootable I can
mkfs the partition fine but when I try to mount it I get a mount error: 2.
Next I make the DOS partition non bootable and the linux partition
bootable. I can still mkfs it but this time when I try to mount it I get an
error about how mount can't open a locked file.
Any suggestions, any idea's, I really hope there are. Thanks for
continuing reading and hopefully responding.
Josh
[next article]
From: [email protected] (A. V. Le Blanc)
Subject: Re: Where can I get pfdisk-0.92?
Keywords: fdisk, extended-partitions
Message-ID: <[email protected]>
Date: 22 Apr 92 15:25:07 GMT
References: <[email protected]>
Followup-To: [email protected]
Organization: Computing Centre, University of Manchester
Lines: 49
In article <[email protected]> [email protected] (n.
h.chandler) writes:
>Where can I locate pfdisk-0.92 and fdisk-0.92? I'd
>prefer that they were not wrapped up with 2 megabytes
>of other stuff that I'd have to discard.
I have written to Neville Chandler privately, but as I have received
several queries about this, I believe the result will be of general
interest.
I will -- soon, I hope -- put together another packaged fdisk and
upload it to nic.funet.fi and to tsx-11.mit.edu, as well as sending
bits to a couple of people who requested them. In the interim, if
anyone wishes for a copy of fdisk, it is available in several forms:
(1) Obviously it is in the 'interim' version of Linux I put together.
This is a shared library executable. You do need to get about 1.5
mb of files to be able to extract both fdisk and the shared library.
(2) An UNLINKED binary, compiled by gcc 2.1, is in the file
misc-linux.tar.Z in the directory /pub/linux/mcc-interim/0.95c+/files.a
at ftp.mcc.ac.uk [130.88.200.7] and at its mirror at banjo.concert.net.
To link it after untarring (assuming you have gcc 2.1 installed),
type 'gcc -o fdisk fdisk.a'. The misc-linux.tar.Z file is 86835 bytes.
(3) The source, as well as a Makefile, are in the file misc.tar.Z
in the directory /pub/linux/mcc-interim/0.95c+/sources. This file
is 134820 bytes.
(4) The latest version of the README file is README.fdisk, in
/pub/linux/mcc-interim/0.95c+. It was 9295 bytes when I wrote
this note, but I suspect it was rather larger only minutes later.
I am doing my best to have sources orand patches for absolutely
everything in the 'interim' version, and I am also working on a
complete collection of man pages, as well as a man command which
someone else contributed, but which I am almost, but not quite,
satisfied with. This will make the fifth and final disk of the
'interim' version, and will include groff as well, since that is
useful together with man. I hope they all fit on a disk!
All of this should eventually be explained in a note somewhere,
but I haven't written it yet. Please be patient! Interim
versions are likely to be a little chaotic in their organisation....
After all, none of this is official.
-- Owen
[email protected]
(3)
[next article]
From: [email protected] (A. V. Le Blanc)
Newsgroups: comp.os.linux,comp.os.minix
Subject: Re: HELP! SHOELACE IS STRANGLING ME!!!
Message-ID: <[email protected]>
Date: 22 Apr 92 15:29:19 GMT
References: <[email protected]> <[email protected]
ime.COM>
Followup-To: [email protected]
Organization: Computing Centre, University of Manchester
Lines: 21
In article <[email protected]> [email protected] (Kevin
Cummings) writes:
>In article <[email protected]>, [email protected] (G
reenup) writes:
>>
>> I recently used shoelace to try & make linux boot. after no success,
>> i want to go back. HOW DO I GET RID OF THIS BOOT THING THAT KEEPS
>> ASKING WHICH PARTITIAN TO BOOT?!?!?!?!?
>
>Just a wild guess, but boot from your ORIGINAL BOOT floppy using your
>ORIGINAL ROOT floppy (you do still have those, don't you?). And mount
>your current hard disk on /mnt. Now you should be able to "cd /mnt/etc"
>and undo whatever you did with shoelace.
I suspect it isn't that simple. Installing the shoelace menu program
in the boot partition overwrites the boot sector. You cannot replace
it as it was unless you have saved a copy of it, or unless you do a
low-level format of your hard disk! Norton's Disk Doctor, as I recall,
can write a pretty standard boot sector to your hard disk and save
having to boot from a floppy.
-- Owen
[email protected]
[next article]
From: [email protected] (CHAN KAM HUNG)
Subject: Getting rid of Bootlace (Wini-boot)..
Message-ID: <[email protected]>
Date: 23 Apr 92 00:47:40 GMT
Sender: [email protected] (NNTP Poster)
Distribution: na
Organization: University of California, at Berkeley
Lines: 11
Nntp-Posting-Host: cory
While everyone is trying to get Linux to boot from their
hardisks, I'm trying to do the reverse. I once had Linux on
a partition with bootlace, etc., now, I've got rid of Linux
on that partition and reformatted it. HOWEVER, that
wini boot table still prompts me every time I boot.
Can anyone tell me how to get rid of it? I don't have the
knowlegde or skills or programs to actually edit the boot
partition, and I don't have Linux on the machine anymore.
What can I do? Please reply by email. thanks.
-Kam
[next article]
From: [email protected] (Rick Kelly)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: all kinds of questions (autoparking)
Message-ID: <[email protected]>
Date: 22 Apr 92 20:20:49 GMT
References: <[email protected]> <1992Apr13.223428.29499@colo
rado.edu> <[email protected]> <[email protected]
sinki.FI>
Reply-To: [email protected] (Rick Kelly)
Organization: The Man With Ten Cats
Lines: 17
In article <[email protected]> [email protected].
FI (Risto Kankkunen) writes:
>In [email protected] (Kevin Cummings) writes:
>> So unless you are shutting down your system to reboot DOS (or something
>> else), your disk heads will not be parked by LINUX. Maybe this is the
>> cause of some of the trashed file systems some people are experiencing?
>
>But isn't head parking suggested only before moving the computer or the
>disk itself? Although there is a greater possibility for disk damage
>when the heads stay above the disk on power-down, it isn't a problem in
>practise on everyday desk-top use.
These days, most hard disks greater than 40 megs will park themselves on
power down.
--
Rick Kelly [email protected] unixland!rmkhome!rmk [email protected]
[next article]
From: [email protected]
Subject: test message (please e-mail)
Message-ID: <[email protected]>
Date: 23 Apr 92 00:45:56 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 5
Did this get through? please e-mail to [email protected] if you see this.
I have sent many posts to this list over the last few days, but I haven't
seen them yet(or gotten responses!)
PLEASE, e-mail me. Thanks in advance.
Jim Gifford [email protected] ...!uunet!attmail.com!jgifford
[next article]
From: [email protected] (Greg Lee)
Subject: Re: ln,mv, i_nlinks incorrect
Message-ID: <[email protected]>
Date: 23 Apr 92 01:22:45 GMT
References: <[email protected]>
Sender: [email protected] (News Service)
Organization: University of Hawaii
Lines: 24
Nntp-Posting-Host: uhunix.uhcc.hawaii.edu
In article <[email protected]> [email protected]
k writes:
}
}Ok, do a fsck -r and clear any errors (I tend to get quite a few unused
}but marked zones after only a few hours of kernel recompilations, there
}are clearly other problems in the fs other than the one I will describe, the
}moral is use fsck regularly or run out of disk space !)
I just did an fsck -v on a file system that I use a lot, and
and there were no errors. I haven't fsck'd that partition in
at least a week and a half. I went through the steps of your
test involving afile, bfile, cfile, and at the end, "ls -l bfile"
did not show two links, but correctly showed only one link.
I ran fsck again, and there were no errors.
I've been using Linux since sometime in January, but I have never
encountered this problem of accumulating inodes marked in the bitmap
but not in use (though I do accumulate inodes with mode not cleared
-- I think that is harmless). I'm running 95c+ now.
Consider the hypothesis that your system has somehow contracted an
unusual sickness.
--
Greg Lee
[next article]
From: [email protected] (Rhys Weatherley)
Subject: Unix Windows (was Re: Graphics and IPC questions...)
Message-ID: <[email protected]>
Date: 23 Apr 92 01:24:35 GMT
References:
uiuc.edu> <[email protected]>
Sender: [email protected]
Reply-To: [email protected]
Lines: 53
>This has already been done. There's a package called UnixWindows (or UW)
>that does this. From your remote machine, you use a terminal program
>written for UW to dial in. After you've logged in, you type "UW" and the
>server kicks in. You can have up to 9 (I think) movable windows on the
>screen. Of course, you can only communicate in one at a time, and at
>2400bps it's pretty slow, but it works...
Up to 7 windows actually. 🙂
I'm not sure if UW is really what he was after, but anyway ...
I'm contemplating porting my program, UW/PC, for DOS (and soon MS-Windows), to
Linux (and hopefully the port will be general enough for other Unixes also).
Don't hold me to it, since I'm still getting the MS-Windows port finished, but
there's a possibility. I've looked at the code of my DOS version carefully,
and I think I can get away with just replacing the screen, keyboard, and
serial comms libraries with Unix equivalents. With a couple of other tweaks
to the main code, I should be able to re-compile the bulk of it for Linux
using g++. (Boy am I glad I designed it to be portable! 🙂
For those who don't know, UW/PC is a communications package (ala Kermit,
Pcomm, etc) that understands the UW protocol for managing up to 7 telnet-like
sessions at once. So, e.g., with a Linux version you could call a remote
system and have effectively more than one login on the remote end (rather than
on the Linux end). I sometimes refer to it as the poor-man's X/Windows (at
least for remote windowing). C++ source code is distributed with it.
Unlike other Unix comms packages (e.g. Kermit) that just dump incoming
characters to the tty device and rely on the terminal to do the emulation,
UW/PC for Linux will do its own internal emulation and map it to termcap
requests. This way, you would get a terminal emulator capable of the VT52,
VT100, VT102, ANSI and ADM31 emulations, and, through the use of the terminal
compiler TERMCC with UW/PC, you can build your own emulations in an
assembly-like language.
If you want to see the DOS version in action (if anyone here still uses
DOS that is 🙂 , then it's available on Simtel20 as the file:
PD1:
A new DOS/MS-Windows version is in the pipeline. If the Linux port comes
off, then anyone who is familiar with the DOS version should be able to use
it without any trouble.
Cheers,
Rhys.
+=====================+==================================+
|| Rhys Weatherley | The University of Queensland, ||
|| [email protected] | Australia. G'day!! ||
|| "I'm a FAQ nut - what's your problem?" ||
+=====================+==================================+
[next article]
From: [email protected] (Michael K. Johnson)
Subject: two questions
Message-ID: <[email protected]>
Date: 23 Apr 92 01:16:21 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Michael K. Johnson)
Organization: The Internet
Lines: 27
From: [email protected] (Alex R.N. Wetmore)
1) Is there a man written yet? I was going to write my own and port
over man pages for all of the stuff that I don't have them for yet, but
I figured I would ask first.
There is a man written, and there are manual pages being
written/adapted by the faithful members of the linux-man list. If you
would like to join in this effort (as you seem to have voiced a
willingness to work on this 🙂 mail [email protected] and
ask to be added. To mail articles to everyone on the list, mail to
[email protected]
We have man2, and people are working on man1, man3, man5, and man8 at
least, as well as beginners' guides and users' guides. There has been
talk about a kernel hackers' guide, but I don't know if anyone has
done work on it yet.
The man is at tsx-11.mit.edu:/pub/linux/sources/usr.bin/man-1.0.tar.Z
There are other man programs available.
Man pages are available as tsx-11.mit.edu:/pub/linux/docs/man/*
They are not all yet available, but we are working on them.
michaelkjohnson
[email protected]
[next article]
From: [email protected]
Subject: help with 0.95c+ and mcc-interim-0.95c+ problems!!
Message-ID: <[email protected]>
Date: 23 Apr 92 00:06:12 GMT
References:
Sender: [email protected]
Followup-To: [email protected]
Distribution: na
Organization: Michigan State University
Lines: 28
the machine is Zenith 386sx with 4Mb Ram, 40Mb IDE hdisk,
1.4Mb and 360K floppies.
linux upto 0.12 has worked w/o problems on this machine.
1. 0.95c+(boot) and 0.95a(root); mmc-interim(0.95c+) work
fine from floppies.
2. an attempt to install linux 0.95c+ on hdisk:
mkfs -c /dev/hda2 5950
results in an Reset-Floppy Called message display and the
machine hangs (it is reproducible).
3. an attempt to install linux using mcc-interim(0.95c+);
mkfs and mkswap go ok; on invoking 'install_root' more than
a screen full of Reset-Floppy Called messages and the system
hangs (reproducible problem).
4. hitting ^c twice logs me out (login prompt is seen..).
*5. in the floppy version of 95c+(boot) and 95a(root), i am having
problem with '>' key (shift .), it has to be typed twice before
the echo is seen -- generally used in redirection!
any pointers / help will be greatly appreciated.
- ishwar([email protected])
[next article]
From: [email protected] (Drew Eckhardt)
Subject: Re: IDE drive performance with linux
Message-ID: <[email protected]>
Date: 23 Apr 92 01:26:58 GMT
References: <[email protected]> <[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 39
Nntp-Posting-Host: ladymacb.cs.colorado.edu
In article
rutgers.edu (Charles Hedrick) writes:
>If you're getting one block per rotation, that means that the system
>can't keep up with the rotation speed of the disk. This is not at all
>unusual. Classically what you do is arrange the files on disk using
>every other block or every third block, or whatever you need in order
>that the system can keep up. With system v, there's an argument to
>mkfs to specify this interleave. With the Berkeley Fast File System,
>it's specified to mkfs or tunefs. Since folks are busy implementing
>an improved file system for Linux, I'm going to hope that this feature
>is included. It would presumably be easy to put into the current one
>as well, but I haven't looked at the code to see what would be
>involved.
You don't want to do this - you'll take X (where X is the interleve)
revolutions to read anything. Instead, you want to speed up reads
(switch to DMA), and read an entire track at once. This will take the
same amount of time as reading one sector.
You can read into a secondary buffer, or increase the buffer cache size
to be the same size as a track.
The problem with the first method is that you still have to do a copy for
each transfer to buffercache, which will be a performance hit. The
second method will have problems if you are using lots of small files,
as you will use more buffer cache than you would with the current
system, and stand a better chance of knocking something out of cache
and having to actually read it. Both have a problem in that track size
varies for different disks, and you can have a zone recorded device -
but anything that goes through the trouble of being zone
recorded, or has a really high density probably also has at least a track
buffer (some of the newer SCSI disks have 256K buffers), and getting the
whole thing in one shot so you don't get hit by rotational latency
isn't as much of an issue.
which will have some penalty want to buffer at the track level -
if you want, you can even DMA into it
[next article]
From: [email protected] (Drew Eckhardt)
Newsgroups: alt.os.linux,comp.os.linux
Subject: Re: help, cannot mount file system...
Keywords: mount umount /dev
Message-ID: <[email protected]>
Date: 23 Apr 92 01:31:12 GMT
References: <[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 16
Nntp-Posting-Host: ladymacb.cs.colorado.edu
In article <[email protected]> [email protected] (gary a moyer) writes:
>Hey netters! I _accidentally_ messed up my ability to mount
>a file system! I did it inadvertantly while trying to incorporate
>some _protection_. Anyhow, what other files does mount access besides
>the actual device in /dev ? I am getting the message:
>mount: unable to open lock-file
>
> thx. [email protected]
This is a FAQ - the mount programs use a lock file, /etc/mtab~ - so no
one tries to access the /etc/mtab file at the same time. Delete this,
or better yet put
rm -f /etc/mtab~
in your /etc/rc.
[next article]
From: [email protected] (Michael K. Johnson)
Subject: IDE drive performance with linux
Message-ID: <[email protected]>
Date: 23 Apr 92 01:37:38 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Michael K. Johnson)
Organization: The Internet
Lines: 22
From: [email protected] (Charles Hedrick)
If you're getting one block per rotation, that means that the system
can't keep up with the rotation speed of the disk. This is not at all
unusual. Classically what you do is arrange the files on disk using
every other block or every third block, or whatever you need in order
that the system can keep up. With system v, there's an argument to
mkfs to specify this interleave. With the Berkeley Fast File System,
it's specified to mkfs or tunefs. Since folks are busy implementing
an improved file system for Linux, I'm going to hope that this feature
is included. It would presumably be easy to put into the current one
as well, but I haven't looked at the code to see what would be
involved.
It is also possible to reformat your IDE drive (yes, it can be done)
with a different interleave, in which case your fs doesn't have to
know about the interleave. If anyone wants more info on this, ask me.
I know that ontrack can do it, and I think that there are a few other
dos programs at wuarchive.wustl.edu that can do it.
michaelkjohnson
[email protected]
[next article]
From: [email protected] (Peter Williams 8169821)
Subject: alpha test ispell for "proper" english available for ftp
Message-ID: <[email protected]>
Date: 23 Apr 92 01:06:37 GMT
Sender: [email protected]
Organization: Design Computing Unit, The University of Sydney
Lines: 20
Nntp-Posting-Host: dean.arch.su.oz.au
There is an alpha test version of Geoff Kuenning's ispell V3.0 for "proper"
(i.e. Australian/British) english available for ftp in pub/linux/ispell
at archsci.arch.su.oz.au (129.78.66.1).
There are 3 files:
README.ispell - brief installation instructions
ispell.tar.Z - tar file that should drop the various files into the
appropriate directories if untarred in / directory.
- includes man, info, el and elc files
- has the medium sized directory recommended by Geoff
ispell.texinfo - texinfo source for using ispell in emacs
As soon as I get time I will build and make available American English
and German dictionaries.
--
Peter Williams |e-mail: [email protected]
Key Centre for Design Quality |phone: +61-2-692 2053 or +61-2-660 6156
University of Sydney |+++++++++++++++++++++++++++++++++++++++
[next article]
From: [email protected]
Subject: Thanks, no more e-mail please!! 😉
Message-ID: <[email protected]>
Date: 23 Apr 92 02:01:50 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 6
Thanks, it got thru obviously, please cease the e-mail though.
I really appreciate it, but this system slows down BIG TIME the more mail I get.
🙁
Thanks!!!!!!
ljg
[email protected]
[next article]
From: [email protected]
Subject: resource sharing?
Message-ID: <[email protected]>
Date: 23 Apr 92 02:33:25 GMT
Sender: [email protected] (Theodore Y. Ts'o)
Reply-To: [email protected]
Organization: The Internet
Lines: 17
Ok, I have 2 linux machines:
1: 386-33 4MB 100MB as 50Mb root, 30MB/pub 20MB swap
2: 386-16sx laptop 8MB 40MB 20MB root, 20MB DOS/Stacker(approx 40MB tot.)
I want to bring the laptop home, and use null modem cable(for now, maybe etherne
t
later) to connect them, and share resources(diskspace, printers, modems...)
I would specifically like to mount the laptop's hd as /laptop, and use it's
internal modem, as well as mount my root as /desktop on the laptop so
it can use my modem and printers. also, I have a dumb dos box, and I
would like to use the desktop as a file/print/comm server.
(if this is currently impossible, let me know, otherwise I need to know
EXACTLY what I need to do this, as I only have e-mail ftpmail access.
)
Thanks in advance,
Jim Gifford
[email protected]
...!uunet!attmail.com!jgifford
[next article]
From: [email protected]
Subject: resource sharing?
Message-ID: <[email protected]>
Date: 23 Apr 92 02:33:50 GMT
Sender: [email protected] (Theodore Y. Ts'o)
Reply-To: [email protected]
Organization: The Internet
Lines: 17
Ok, I have 2 linux machines:
1: 386-33 4MB 100MB as 50Mb root, 30MB/pub 20MB swap
2: 386-16sx laptop 8MB 40MB 20MB root, 20MB DOS/Stacker(approx 40MB tot.)
I want to bring the laptop home, and use null modem cable(for now, maybe etherne
t
later) to connect them, and share resources(diskspace, printers, modems...)
I would specifically like to mount the laptop's hd as /laptop, and use it's
internal modem, as well as mount my root as /desktop on the laptop so
it can use my modem and printers. also, I have a dumb dos box, and I
would like to use the desktop as a file/print/comm server.
(if this is currently impossible, let me know, otherwise I need to know
EXACTLY what I need to do this, as I only have e-mail ftpmail access.
)
Thanks in advance,
Jim Gifford
[email protected]
...!uunet!attmail.com!jgifford
[next article]
From: [email protected] (Doug Evans)
Subject: Re: Graphics and IPC questions...
Keywords: Graphic standard and how to IPC?
Message-ID:
Date: 22 Apr 92 15:46:40 GMT
References: <[email protected]>
Organization: Edmonton Remote Systems, Edmonton, AB, Canada
Lines: 26
chojnows@osiris (Brian Chojnowski) writes:
>
> I plan to add a direct to screen library. Basically I envision
> creating a terminal type that is a superset of vt100. Nothing as fancy or
> system hog like X. Just a characterset based graphix system. So while we are
> at it, if anyone has some info somewhere on how to write assembly stuff
> under linux, I can start porting some of my dos-mode screen routines.
What does a "character set based graphics system" mean?
If it means you want something like what a vt220 gives you,
or perhaps some other terminal, wouldn't the better route be to
take the vt100 stuff and enhance it?
Direct screen writing stuff may be interesting and even necessary
on dumb 4.77 Mhz machines, but for more powerful machines, I think
it attaches too much hardware dependance to the application.
Can you not achieve what you want doing it the Terminfo/Termcap way?
P.S. I presume Linux uses Terminfo, right?
Doug Evans |
..!{canada,uunet}!sspiff!dje | Vancouver 11, Calgary 0
[email protected] | There is no joy in Cowtown.
[email protected] | But how 'bout those Canucks, eh?
[next article]
From: [email protected] (Theodore Ts'o)
Subject: Re: IDE drive performance with linux
Message-ID: <[email protected]>
Date: 23 Apr 92 03:15:40 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 33
From: [email protected] (Charles Hedrick)
Date: 22 Apr 92 22:41:36 GMT
If you're getting one block per rotation, that means that the system
can't keep up with the rotation speed of the disk. This is not at all
unusual. Classically what you do is arrange the files on disk using
every other block or every third block, or whatever you need in order
that the system can keep up.
We could certainly do that, although I think it would be more profitable
to try speeding up the hard disk driver; and only if that doesn't work,
then start mucking with block allocation algorithm for the filesystem.
(BTW, a very simple change to the block allocation algorithm for the
Minix fs would be to pass it the location of the previous block in the
file, and for it to try to allocate a block in the same "zone". If
there is no previous block, it should allocate a block in the zone with
the highest amount of free space, to spread out the allocated blocks
around. Such a change would give you a large part of the speed
advantages of the Berkeley FFS, with relatively little work. Of course,
it still wouldn't give you long filenames...)
In any case, there are two things about the current hard disk driver
which hopefully should be fixed. (1) It doesn't use DMA. This may or
may not be fixable; I'd need to do some research about the low-level
interface before I could answer that. (2) It doesn't coallesce
multiple block read requests over a single track into a single request.
Fixing this should significantly speed up the speed of non-caching IDE
drives, especially combined with the look-ahead which the filesystem is
already doing.
- Ted
[next article]
From: [email protected]
Subject: test message----please ignore 😉
Message-ID: <[email protected]>
Date: 23 Apr 92 03:27:26 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 19
^L------------- Begin Forwarded Message -------------
Date: Wed Apr 22 23:22:23 -0400 1992
From: internet!ATHENA.MIT.EDU!tytso (Theodore Ts'o)
Phone: +(617) 253-8091
Subject: Re: linux list
To: !jgifford
Content-Length: 443
From: [email protected]
Hi, Michael, this is a follow-up to my last message.
Hi, right after my last message, I was flooded with about 4 or 5 digests. 😉
However, my posts STILL aren't showing up. 🙁
Grump.... Whenever I try sending a test post, it always shows up. Can
you try sending a post to linux-activists, cc'ed to me? I can't figure
out what's going wrong, and perhaps that will help me figure it out.
- Ted
[next article]
From: [email protected] (Daniel L'Hommedieu)
Subject: PS/2, SCSI, etc
Message-ID:
Date: 23 Apr 92 02:45:29 GMT
Sender: [email protected] (USENET News System)
Organization: North Carolina State University
Lines: 18
Greetings. I have an IBM PS/2 Model 65sx, currently with 2 megs RAM
(soon to be 10 megs), a 60 megabyte Western Digital SCSI drive, and a 84
meg Seagate SCSI drive.
I have been watching the newsgroup for a few days, but haven't seen any
FAQ, or anything to do with IBM MicroChannel.
Is there anyone working on an MCA version of linux? If so, I would like
to correspond with this person(s).
Oh, someone too please point me toward the FAQ. Thanks so much.
Daniel
--
Daniel L'Hommedieu / CSC Major, NC State U, on the 6-year plan 🙂
Internet: eagle@{catt.ncsu.edu,Salzo.Cary.NC.US,grunt.Berkeley.edu}
10215 Hanover Glen Road / Charlotte NC 28210 / 704-541-5116
"Windows 3.1 is Microsoft's $50 bug-fix to Windows 3.0."-JWL
[next article]
From: [email protected]
Subject: Control-alt-del
Message-ID: <[email protected]>
Date: 23 Apr 92 03:30:42 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 10
Amazing discovery with NCS version of AMI BIOS. C-A-D has never worked
with Linux-0.95 on. A-C-D works 100 per cent of the time (since
discovered it pre-0.96 version) BTW: using left Alt left Ctrl rightmost
Del
BTW: With NCS AMI BIOS can set repeat speed and init delay, using int16
in keyboard.S with same value get less than 250msec initial delay Does
Linux alter the timer chips such that the keyboard thinks 250msec has
past?
John
[next article]
From: [email protected] (Matt Mosley)
Subject: linux
Message-ID:
Date: 22 Apr 92 20:03:03 GMT
Organization: Starnet BBS
Lines: 18
A few questions. I just got Linux running under a partition on my 386-25.
My primary partition, C:, is running DOS, which is how this message is
coming to you 🙂 (I am using Waffle-DOS for uucp).
Anyway, D: and E: are Linux filesys and swap, respectively. (swap is 4m).
I would like to know how to do the following:
1) Does Linux have *any* working mail program? I can't seem to find any.
2) How do I transfer my news and mail feeds over to Linux (I'd first need
a mail program, of course, plus readnews and uucp).
3) Where can I get troff/groff/nroff for Linux? (or at least workable).
Thanks!
-----------------------------------------------------------------------------
Matt Mosley (Shadow) | Systems manager of Starnet BBS and
[email protected] | programmer for Starcom Software
[next article]
From: [email protected] (Matt Mosley)
Subject: also
Message-ID: <[email protected]>
Date: 22 Apr 92 20:17:06 GMT
Organization: Starnet BBS
Lines: 5
I'm also interested in a finger command for Linux, if it exists...
-----------------------------------------------------------------------------
Matt Mosley (Shadow) | Systems manager of Starnet BBS and
[email protected] | programmer for Starcom Software
[next article]
From: [email protected] (Ng Pheng Siong)
Subject: How to pronounce "Linux"?
Message-ID: <[email protected]>
Date: 23 Apr 92 04:43:17 GMT
Sender: [email protected]
Organization: National University of Singapore
Lines: 10
I apologize if this is considered trivial, but what is the
proper pronunciation for "Linux"?
Li lee or lie?
Nux nuks or nerks? Nooks?
Anyone?
PhengSiong
Singapore (Which schools teach English not American.)
[next article]
From: [email protected] (Martin Peter Howell)
Newsgroups: alt.os.linux,comp.os.linux
Subject: Re: trouble with curses
Keywords: curses
Message-ID: <[email protected]>
Date: 19 Apr 92 04:50:13 GMT
References: <[email protected]>
Organization: RMIT Computer Centre, Melbourne Australia.
Lines: 15
[email protected] (gary a moyer) writes:
>Has anyone had trouble using curses? I am _trying_ to compile the
>visual version of gnuchess (that uses cureses), however, the farthest
>I get is "curses.h: parse error before line 100" and then it slowly
>chokes and dies. I can't remeber where exactly I got the curses
>package from (some anon ftp site). By chance is there a newer
>version floating around anywhere?
This is caused by sys/param.h #defining HZ whilst curses.h tries to
use it as a variable. A temporary hack is to add "#define HZ HZ" before
you include curses.h.
-Martin.
[email protected]
[next article]
From: [email protected] (Kellom{ki Pertti)
Newsgroups: comp.os.linux,comp.os.minix
Subject: Re: HELP! SHOELACE IS STRANGLING ME!!!
Message-ID:
Date: 23 Apr 92 07:51:51 GMT
References: <[email protected]>
Sender: [email protected] (#Kotilo NEWS system )
Organization: Tampere Univ. of Technology, Finland.
Lines: 16
In-Reply-To: [email protected]'s message of 18 Apr 92 01: 25:17 GMT
Nntp-Posting-Host: talitiainen.cs.tut.fi
In article <[email protected]> [email protected] (Gre
enup) writes:
I recently used shoelace to try & make linux boot. after no success,
i want to go back. HOW DO I GET RID OF THIS BOOT THING THAT KEEPS
ASKING WHICH PARTITIAN TO BOOT?!?!?!?!?
I almost messed up my Dos partition when installing Shoelace. In the
course of restoring it, I installed Dos, and the installation restored
both the boot sector on the Dos partition (which I expected it to do),
and also the boot sector of the whole hard disk (which I did not
expect). This was Dos 5.0 installed with the setup.exe (or something
similar) that came with it. Hope it works for you.
--
Pertti Kellom\"aki (TeX format) # These opinions are mine,
Tampere Univ. of TeXnology # ALL MINE !
Software Systems Lab # (but go ahead and use them, if you like)
[next article]
From: [email protected] (Zeyd M. Ben-Halim)
Subject: Re: Yet another X question
Keywords: X Windows
Message-ID: <[email protected]>
Date: 23 Apr 92 08:35:38 GMT
References: <[email protected]> <[email protected]
.edu> <[email protected]> <[email protected].
ac.uk>
Sender: [email protected] (netnews admin account)
Organization: University of Denver, Dept. of Math & Comp. Sci.
Lines: 21
[email protected] (Steve Sykes) writes:
>I'm very keen to get something like X up. At the moment my biggest
>problem is a lack of knowledge on how to implement i/o mapping.
>Rather than thinking up some obscure and probably bad way to allow
>user processes access to io-space, I was hoping someone could point me
>in the right direction. I think it's normally done as an ioctl call
>but have now idea how to implement it.
Well, there are patches to add mmap/munmap system calls to linux. Other patches
exist to access i/o ports, to switch to various VGA modes. Unfortunately, these
are not coordinated and no docs. Using these I can switch to graphic modes and
write to the screen. I never got a consistant result, with random appearing
pixels on the screen. I understand that 256-color modes are easier to write to.
One missing addition is code for bank-switching. The code that comes with djgpp
graphics libraries could be cannibalized to achieve this purpose.
I know that several people are working on graphics support, but none are sharing
much of their efforts or results (no flames intended).
Zeyd
[next article]
From: [email protected] (Pat Duffy)
Newsgroups: comp.os.os2.misc,comp.os.linux
Subject: Re: Any comments on Quantum 240A ??
Keywords: OS/2, Linux, 240A, Quantum
Message-ID: <[email protected]>
Date: 21 Apr 92 16:47:44 GMT
References: <[email protected]>
Sender: [email protected] (Usenet News Maintenance)
Organization: University of British Columbia, Vancouver, B.C., Canada
Lines: 16
Nntp-Posting-Host: theory.chem.ubc.ca
In article <[email protected]> [email protected]
is.edu (David A. Sinclair) writes:
>Hello, world.
>
>I need to buy anew hard drive for my system so I can install Linux and
>OS/2 2.0. I am considering the Quantum 240A, (240 meg IDE) and would
>like to hear from anyone who is either using this drive with either
>OS, or has tried and failed to do so.
Hi!
I am using the Quantum 240A with OS/2 V2 GA with no problems at all.
--
Patrick Duffy, E-Mail: [email protected]
"I wouldn't believe in anything if it weren't for my lucky astrology mood
watch." - Steve Martin
[next article]
From: pmacdona@sanjuan (Peter MacDonald)
Subject: fully broiled ideas
Message-ID: <[email protected]>
Date: 23 Apr 92 07:22:56 GMT
Sender: [email protected]
Organization: University of Victoria, Victoria, BC, CANADA
Lines: 33
Nntp-Posting-Host: sanjuan.uvic.ca
Two weeks in the Maui sun, and some half baked ideas are now fully broiled.
Here they are, in no particular order.
I note BSDI is promising binary compatibility with SysV 3.2. I imagine
then that if this happens, 386BSD will eventually follow suit if possible.
Soooo, anyone know if there are any steps that could be taken to move Linux
part way towards it now? Like perhaps, adopting the same system call
interface, if possible? I know that the current shared libraries
interface is incompatible with that goal, but ignore that for the moment.
The primary issue here is to strengthen Unix vis-a-vis other commercial
operating systems.
BTW: has anyone tried the OLEO spreadsheet on tsx yet. Bug reports or
"it works for me" reports would be appreciated.
Is anyone working on a fully automated install program like the one
for SUN and DELL?
I would like to see a layered installation available with perhaps the
following options:
1) Just the runtime system
2) Development (includes, kernel source, compiler, yacc, etc)
3) X-windows runtime.
4) X-windows development.
This would hit most of the interest groups, while remembering the
frugel (space conscious) nature that draws us all to systems like Linux.
Please reply to:
[email protected]
if at all.
[next article]
From: [email protected] (Erik Green)
Subject: Request for repeat info
Message-ID:
Date: 23 Apr 92 09:13:52 GMT
Distribution: comp.os.linux
Organization: Long polymers in double-helix formation
Lines: 23
Nntp-Posting-Host: att2.cs.mankato.msus.edu
Will sonebody please mail me the info on what to do if you have a
parse error in stdio.h(line 101)? I know this has been gone over
a zillion times before, but being a normal, fallible human being
I didn't save it. While you're at it(whoever), send me the address
of the site where these newsgroups are being archived (I'm pretty
sure I heard there was one.)
Thanks,
Erik Green
Thx,Lns
Erik "Longshot" Green
longshot@(vax1 | att2.cs | krypton | theory.cs).mankato.msus.edu
This is a .signature virus scanner. Stop the .sig virus!
Disclaimer: I said the above. So sue me. I'm broke anyway.
--
Erik "Longshot" Green
longshot@(vax1 | att2.cs | krypton | theory.cs).mankato.msus.edu
This is a .signature virus scanner. Stop the .sig virus!
Disclaimer: I said the above. So sue me. I'm broke anyway.
[next article]
From: [email protected] (Steve Robbins)
Subject: News under Linux
Summary: It works a little...
Message-ID: <[email protected]>
Date: 23 Apr 92 02:04:38 GMT
Organization: The chiral cage of chaos
Lines: 20
Howdy,
I finally got C-news to work well enough to read and post news!
Unfortunately, the dbz routines are still screwing up - so the active
file is always growing longer rather than being overwritten.
If someone out there has this working flawlessly, PLEASE tell me what
else I have to do to it. I also found that newsbin/inject/tear didn't
work as distributed -- this may be a problem with awk or with my shell.
I found out the extremely hard way that linux or gnu thinks that fseek()
should return the new offset if sucessful, rather than zero, like all
other stdio's appear to. Is this some new ANSI thing, or is libc screwed
up?
Also strcasecmp() is the same as strcmp(). I got the gcc2.1 distribution
from tsx-11 somewhere around April 4 -- is there a newer version that
has corrected these things?
Steve Z
[next article]
From: [email protected] (Rogier Wolff)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: all kinds of questions (autoparking)
Message-ID: <[email protected]>
Date: 23 Apr 92 08:44:02 GMT
References: <[email protected]> <1992Apr13.223428.29499@colo
rado.edu> <[email protected]> <[email protected]
sinki.FI> <[email protected]>
Sender: [email protected] (UseNet News System)
Organization: Delft University of Technology, Dept. of Electrical Engineering
Lines: 18
Nntp-Posting-Host: neuron.et.tudelft.nl
[email protected] (Rick Kelly) writes:
>These days, most hard disks greater than 40 megs will park themselves on
>power down.
>Rick Kelly [email protected] unixland!rmkhome!rmk [email protected]
Yes. Right. I have a broken 40M disk, and when you power down, it still
discharges a capacitor through the voice coil, which jams the head
behind a latch. (wouldn't autoparking be hard for stepper motor controlled
heads? at 24 ms average seek time, you'd need power for 0.1 seconds
to step the motor back home, worst case)
Roger.
--
If the opposite of "pro" is "con", what is the opposite of "progress"?
(stolen from [email protected])
EMail: [email protected] ** Tel +31-15-783644 or +31-15-142371
[next article]
From: [email protected] (Roberto Favaro)
Newsgroups: comp.os.linux,alt.os.linux
Subject: printer on parallel port
Keywords: printer's device
Message-ID: <[email protected]>
Date: 23 Apr 92 10:55:47 GMT
Sender: [email protected] (Netnews Administrator)
Organization: D.E.I. Universita' di Padova - ITALY
Lines: 16
Hi to all,
I have a little problem with my printer,that has a parallel port
:It doesn't work with Linux.
I have tryed to type 'mknod /dev/lp0 c 6 0' and then
'echo "\n Hello \n" > /dev/lp0 ' but nothing has happened.
Which arguments with mknod have i to use ?
If anybody wants help me ....
Thanks in advance.
---------------------------------------------
| [email protected] | Linux -> 1.0 !WOW!
| [email protected] | I hope. 🙂
---------------------------------------------
[next article]
From: [email protected] (Joerg Pommnitz)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: Miscellaneous questions.
Message-ID:
Date: 23 Apr 92 10:32:12 GMT
References: <[email protected]>
Sender: [email protected] (Owner of all binaries)
Organization: tu-chemnitz
Lines: 34
[email protected] (Rogier Wolff) writes:
>>[email protected] (Scott Dunn) writes:
>>I'd like to ask the same question:
>>>2. What is a working version of tar and un/compress?
>>> A lot of problems seem to be solved by getting new versions of the
>>> aforementioned. However, nobody has indicated how to identify which
>>> are the dodgy copies.
>>> Sometimes when uncompressing it seems to get stuck in an infinite
>>> loop which takes all system resources. ie I can't login to another
>>> terminal and kill the process. Is this what is known as a 'race
>>> condition' ? Is this indicate a dodgy copy of un/compress ?
>>How can I fix these problems ?
>>What's the reason ?
>Compress starts to use memrory that it hasn't allocated when it gets
>invalid input. Most unices will then give it a "segment violation" and
>kill it. Linux currently simply allocates the memory, and lets the
>application continue. This will be fixed in the future. (if nobody
>else does it, I will.)
> Roger.
But my input is valid !!!
I checked it by decompressing the files on a sun !
Then I compressed it again using -B13 ,but compress (on LINUX) still hangs.
That's why I think there must be another problem.
BTW: The same compress with the same files works on another machine.
Joerg
[next article]
From: [email protected] (Linus Benedict Torvalds)
Subject: Re: How to pronounce "Linux"?
Message-ID: <[email protected]>
Date: 23 Apr 92 12:32:16 GMT
References: <[email protected]>
Organization: University of Helsinki
Lines: 25
In article <[email protected]> [email protected] (Ng Ph
eng Siong) writes:
>I apologize if this is considered trivial, but what is the
>proper pronunciation for "Linux"?
>
>Li lee or lie?
>Nux nuks or nerks? Nooks?
Oh god. We just hashed this through on alt.os.linux. It /definitely/
needs to go into the FAQ.
'li' is pronounced with a short [ee] sound: compare prInt, mInImal etc.
'nux' is also short, non-diphtong, like in pUt. It's partly due to
minix: linux was just my working name for the thing, and as I wrote it
to replace minix on my system, the result is what it is... linus' minix
became linux.
I originally intended it to be called freax (although buggix was one
contender after I got fed up with some of the more persistent bugs 🙂
and I think the kernel makefiles up to version 0.11 had something to
that effect ("Makefile for the freax kernel" in a comment). But arl
called the linux directory at nic.funet.fi pub/OS/Linux, and the name
stuck. Maybe just as well: freax doesn't sound too good either (freax
is obviosly free + freak + the obligatory -x).
Linus
[next article]
From: [email protected] (C. G. Albone)
Subject: Dominion 1.06 et. al.
Message-ID: <[email protected]>
Date: 24 Apr 92 13:36:47 GMT
Sender: [email protected]
Organization: Uni Computing Service, Uni of Sydney, Australia
Lines: 17
Hello netters,
I have just recently ported dominion 1.06 to linux, and would
be interested in any comments by people who have done the same. One thing
that i have noticed is that it is so *slow*. It really sucks the power
out of my machine (a 386 25MHz with 4M RAM). It is bad enough when trying
to run the clients, but when updating the machine seems to grind to a halt.
On a related note, after running dominion, the machine seems to remain
somewhat sluggish.. an example being the reboot program posted to the net
recently taking 10+ seconds to run on occasion. Also, it bombs when coming
out of the mail procedure (using it's internal mail system).
On a related note, has anyone checked out conquer5 beta?
well, enuff for now... back to world domination 🙂
chris
[next article]
From: [email protected] (Lance Pickup)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: all kinds of questions (autoparking)
Message-ID: <[email protected]>
Date: 23 Apr 92 13:32:02 GMT
References: <[email protected]> <1992Apr13.223428.29499@colo
rado.edu> <[email protected]> <[email protected]
sinki.FI> <[email protected]> <[email protected].
nl>
Sender: @watson.ibm.com
Reply-To: [email protected]
Organization: IBM Technology Products
Lines: 20
In article <[email protected]>, [email protected]
.nl (Rogier Wolff) writes:
|> Yes. Right. I have a broken 40M disk, and when you power down, it still
|> discharges a capacitor through the voice coil, which jams the head
|> behind a latch. (wouldn't autoparking be hard for stepper motor controlled
|> heads? at 24 ms average seek time, you'd need power for 0.1 seconds
|> to step the motor back home, worst case)
I could be wrong, but I thought that the way auto-parking worked was
that the head was spring loaded such that whenever power was removed,
it would just spring back into a safe location (i.e. you need power
to un-park it).
--
...Lance
-----------------------------------------------------------------------
Lance Pickup VNET/IBM Internet: lpickup@btv
IBM/Vendor Systems Internet: [email protected]
Technology Products
Burlington, VT Phone: (802) 769-7104 (tie 446)
-----------------------------------------------------------------------
[next article]
From: [email protected] (Bennett E. Todd)
Subject: Porting to Linux
Message-ID: <[email protected]>
Date: 23 Apr 92 14:30:30 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Bennett E. Todd)
Organization: The Internet
Lines: 35
I've brought up Linux 0.95c+ on my box. This box is an original AT with
a motherboard change. Original IBM Monochrome Display Adaptor (MDA),
original 30M MFM HD, etc. I swapped out the motherboard for a supercheap
386SX, stuffed 6M RAM in it, and away I go. The installation was a piece
of cake --- nice work, folks.
I started trying to port my favorite stuff to it. Rc(1) was easy, and
elle(1) wasn't too hard, but jove(1) has me stalled out.
In sysdep.h, the jove release has support for two styles of signals. It
looks like Linux implements a third sort. Is there a good book to get
for docs on the Posix interfaces that Linux supports?
Here's the relevent patch of code from the jove release:
#if defined(BSD_SIGS)
extern long SigMask;
# define SigHold(s) sigblock(SigMask |= sigmask((s)))
# define SigRelse(s) sigsetmask(SigMask &= ~sigmask((s)))
#else
# define SigHold(s) sighold(s)
# define SigRelse(s) sigrelse(s)
# define killpg(pid, sig) kill(-(pid), (sig))
#endif
I tried recoding these in terms of sgetmask and ssetmask, from the man
page on signal(2), but these don't seem to be available. When I did an
nm(1) on libc.a, I found ___sgetmask() and ___ssetmask(); however,
calling them didn't seem to do the trick.
Any ideas?
-Bennett (curmudgeon)
[email protected]
[next article]
From: [email protected] (Bill Jones)
Subject: Software index
Message-ID: <[email protected]>
Date: 23 Apr 92 15:43:20 GMT
Organization: 'fp
Lines: 31
Thnks to everyone who has responded concerning my idea of a software
index. There were no responses against it and most of the mail I received
thought that it would be an extremely useful idea. Therefore, I'll go
ahead and begin compiling an index of all of the software that I can
find.
Someone mentioned that this could be laid out in the same fashion as the
Minix index so I'll try to obtain a copy of that and see how the style
looks.
One thing, if you are a sysadmin at a site which is providing ftp access
to linux material and you are NOT at tsx-11.mit.edu, banjo.concert.net,
mcc.uk, or funet.fi, please mail me and let me know of your archive site
(particularly if you are not on the FAQ) and perhaps mail me a copy of
your linux file list.
If you are currently porting a software package or have ported one, let
me know as well so that I can give you credit (and allow others to send
flames if the package doesn't work 🙂 ).
I'll try to have version one out in a couple of weeks and then perhaps we
can try to tie it into the FAQ and the beginners guide.
------------------------------------------------------------------------
Bill Jones 1-614-785-0358
[email protected] Oh, to be in Montana now
[email protected] that spring is here!
[email protected]
------------------------------------------------------------------------
[next article]
From: [email protected] (Bill Jones)
Subject: Everex modems and linux
Message-ID: <[email protected]>
Date: 23 Apr 92 15:47:54 GMT
Organization: 'fp
Lines: 23
A while back I had posted two calls for help concerning my modem (an
Everex Evercom II) and linux. Linux did not seem to even know that
the modem existed, either through kermit, uucp, or directly.
I borrowed an old 1200 bps TEAM modem from a friend, installed it,
changed nothing at all in my .kermrc or /etc/rc and all of a sudden
linux recognized the modem and I had communication with the outside
world. Since then, the 1200 bps modem has been replaced with a Cardinal
9600 v.32, v.42bis modem and everything still works well. When the
Everex is tried again, linux still fails to notice its existance.
The moral of the story seems to be: Don't use an Everex Evercom II
modem under linux and expect it to work.
Hope that this helps other people with the same problem.
------------------------------------------------------------------------
Bill Jones 1-614-785-0358
[email protected] Oh, to be in Montana now
[email protected] that spring is here!
[email protected]
------------------------------------------------------------------------
[next article]
From: [email protected] (Kevin Cummings)
Newsgroups: alt.os.linux,comp.os.linux
Subject: Re: help, cannot mount file system...
Keywords: mount umount /dev
Message-ID: <[email protected]>
Date: 23 Apr 92 15:34:48 GMT
References: <[email protected]>
Sender: cummings@hammer (Kevin Cummings)
Organization: Prime Computer R&D
Lines: 29
In article <[email protected]>, [email protected] (gary a moyer) writes:
> Hey netters! I _accidentally_ messed up my ability to mount
> a file system! I did it inadvertantly while trying to incorporate
> some _protection_. Anyhow, what other files does mount access besides
> the actual device in /dev ? I am getting the message:
> mount: unable to open lock-file
>
WHY ISN'T THIS A FAQ?????????? Linus, PLEASE change the mount
message to write out the lock file pathname! And possibly another
message about what to do about it.
In the meantime, if you are getting this problem, you have most likely
editted your /etc/mtab file. rm /etc/mtab~ That's the file mount
tries to create to ensure no other process is updating the mount
table at the same time.
=================================================================
Kevin J. Cummings Prime Computer Inc.
20 Briarwood Road 500 Old Connecticut Path
Framingham, Mass. Framingham, Mass.
Work: [email protected]
Home: [email protected]
Std. Disclaimer: "Mr. McKittrick, after careful consideration,
I've come to the conclusion that your new
defense system SUCKS..." -- War Games
=================================================================
[next article]
From: [email protected] (Kevin Cummings)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: all kinds of questions (autoparking)
Message-ID: <[email protected]>
Date: 23 Apr 92 15:38:25 GMT
References: <[email protected]> <1992Apr13.223428.29499@colo
rado.edu> <[email protected]> <[email protected]
sinki.FI> <[email protected]>
Sender: cummings@hammer (Kevin Cummings)
Organization: Prime Computer R&D
Lines: 19
In article <[email protected]>, [email protected] (Rick Kelly) writes:
> These days, most hard disks greater than 40 megs will park themselves on
> power down.
ANd when I can find a list of which ones do, and which ones don't,
I'll beleive it. In the meantime, better safe than sorry.
=================================================================
Kevin J. Cummings Prime Computer Inc.
20 Briarwood Road 500 Old Connecticut Path
Framingham, Mass. Framingham, Mass.
Work: [email protected]
Home: [email protected]
Std. Disclaimer: "Mr. McKittrick, after careful consideration,
I've come to the conclusion that your new
defense system SUCKS..." -- War Games
=================================================================
[next article]
From: [email protected] (Matthias Lautner)
Subject: dos emulator
Message-ID: <[email protected]>
Date: 23 Apr 92 15:29:54 GMT
Sender: [email protected] (Matthias Lautner)
Organization: University of Wuerzburg, Germany
Lines: 21
Hi!
I'm working on a dos emulator for linux. I created a new system call for
this purpose. It devotes a 4 MB memory area of virtual process memory to dos
and keeps the 8086 register set in a struct. Starting the vm86 system call,
the registers are transferred to the cpu and the process is running in
virtual 8086 mode using the specified memory area until a fault (eg. general
protection) happens or a signal is caught. Then the registers are stored back
to the struct and the system call returns as usual. The result of the call
shows up the reason for interrupting.
Whith this call it is possible to write a dos emulator. Any interrupt
statement or io machine code will result in a general protection trap an can
be emulated.
Very simple masm generated dos programs are allready running.
If someone is interested in helping me to implement the functions of the
dos int 0x21 it would be grate. I'm currently working with 0.95c.
Matthias
--
*****************************************************************************
Matthias Lautner Carl-Vornberger-Str. 21, 8701 Randersacker, Germany
[email protected]
*****************************************************************************
[next article]
From: [email protected] (Jim Burns)
Subject: scsi setup?
Keywords: minor #s
Message-ID: <[email protected]>
Date: 23 Apr 92 16:04:29 GMT
Organization: Georgia Institute of Technology
Lines: 19
I got the scsi stuff from tsx-11 in pub/linux/scsi - it boots fine, but no
joy from the modified fdisk yet. My Dos driver says this when I boot from
my AHA1542 (floppy disabled):
Host Adapter #: 0 Host Adapter SCSI ID: 7
I/O Port Address: 330 DMA Channel: 5
Interrupt Level: 10 VDS Support LevelL MultiSegRW
Host Adapter # 0 - SCSI ID 0 - Lun 0: Maxtor LXT-213S
After creating dummy /dev/hd[15] on the 95a rootimage to make fdisk happy,
I get sd0 not found. Doing a 'mknod /dev/sd0 b 8 n' where n=0,16,32,48
still gets no joy from fdisk - something to the effect of trying to read
from non-existent block device. Any ideas? Thanx.
--
BURNS,JIM (returned student)
Georgia Institute of Technology, 30178 Georgia Tech Station,
Atlanta Georgia, 30332 | Internet: [email protected]
uucp: ...!{decvax,hplabs,ncar,purdue,rutgers}!gatech!prism!gt0178a
[next article]
From: [email protected] (Peter Thomson)
Subject: free (memory allocation table)
Message-ID: <[email protected]>
Date: 23 Apr 92 12:40:18 GMT
Sender: [email protected]
Reply-To: [email protected] (Peter Thomson)
Organization: Laboratory for the Foundations of Computer Science, Edinburgh U
Lines: 8
Hi can some one tell me if the command free reports back any more than
1Mbyte of memory. There is 2Mbytes on my machine but free states there is
only one. running multiple tasks and watching ps confirms this and swapping
starts at the 1Mbyte limit
Thanks in advance Peter.
[next article]
From: [email protected] (Rainer Kirsch)
Subject: Questions: Large SCSI Partitions, native SCSI partitioning
Keywords: linux SCSI partitions partitioning
Message-ID:
Date: 23 Apr 92 16:27:12 GMT
Sender: [email protected]
Organization: Inst. for General Electrics & Electronics, TU Vienna
Lines: 20
Nntp-Posting-Host: eaep30.tuwien.ac.at
Hello Linux Fans!
1.I would like to know, if the development of a SCSI software, which is able
to handle large partitions (>100 MB) already has been started. If yes, tell
me about the estimated release date.
2.Is it correct, that the existence of one large (about 180MB) partition,
inhibits the usage of any other partition (on the same drive) to Linux?
3.Is there some software available to partition SCSI disks under Linux?
Many thanks in advance!
Rainer Kirsch
+------------------------------------------------------+
| [email protected] |
+------------------------------------------------------+
| Institute of Applied Electrotechnics and Electronics |
| TU-Vienna / Austria / Europe |
+------------------------------------------------------+
[next article]
From: [email protected] (Drew Eckhardt)
Subject: Re: dos emulator
Message-ID: <[email protected]>
Date: 23 Apr 92 16:33:51 GMT
References: <[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 51
Nntp-Posting-Host: kinglear.cs.colorado.edu
In article <[email protected]> [email protected]
formatik.uni-wuerzburg.de (Matthias Lautner) writes:
>Hi!
>
>I'm working on a dos emulator for linux. I created a new system call for
>this purpose. It devotes a 4 MB memory area of virtual process memory to dos
>and keeps the 8086 register set in a struct. Starting the vm86 system call,
>the registers are transferred to the cpu and the process is running in
>virtual 8086 mode using the specified memory area until a fault (eg. general
>protection) happens or a signal is caught. Then the registers are stored back
>to the struct and the system call returns as usual. The result of the call
>shows up the reason for interrupting.
>
>Whith this call it is possible to write a dos emulator. Any interrupt
>statement or io machine code will result in a general protection trap an can
>be emulated.
>
>Very simple masm generated dos programs are allready running.
>
>If someone is interested in helping me to implement the functions of the
>dos int 0x21 it would be grate. I'm currently working with 0.95c.
>
The following questions are hypothetical :
1. How many documented DOS calls are there?
2. How many undocumented DOS calls are there?
The answer is "many". The DOS kernel is ~50K of machine code. You
don't want to duplicate that. Instead, you emulate what DOS runs on,
and catch all of the programs that run there too, and run a *REAL*
dos.
DOS runs on BIOS. Question : how many BIOS calls are there?
Again, the answer is "many", so you don't emulate BIOS, use the
real thing, and emulate the underlying hardware where
possible.
Something that runs simple MASM programs is sort of interesting,
and lets you see how 386 virtual mode works. Something that is 100%
MSDOS compatable is like a Tandy 2000 or DEC rainbow - it runs
MSDOS - but it's not compatable. A few small programs will
run, but none of the programs that make me keep DOS on my
hard disk. Something that is 100% BIOS compatable is getting
there, and I can run many programs - but anything that does
direct screen I/O is out - which rules out MOST of
the programs out there.
Certain BIOS calls *should* be emulated, such as disk accesses,
which should be redirected out to unix devices. Others, such
as most video routines, which read/write to your remapped
video buffer are OK.
[next article]
From: [email protected] (Joel M. Hoffman)
Subject: Is Shoelace permanent? [was HELP! SHOELACE IS STRANGLING ME!!!]
Message-ID: <[email protected]>
Date: 23 Apr 92 17:12:32 GMT
References: <[email protected]> <[email protected]
ime.COM> <[email protected]>
Sender: [email protected] (USENET News system)
Organization: University of Maryland, College Park
Lines: 23
Nntp-Posting-Host: next.wam.umd.edu
In article <[email protected]> [email protected] (A. V. Le Blanc) wr
ites:
>In article <[email protected]> [email protected] (Kevin
Cummings) writes:
>>In article <[email protected]>, [email protected] (
Greenup) writes:
>>>
>>> I recently used shoelace to try & make linux boot. after no success,
>>> i want to go back. HOW DO I GET RID OF THIS BOOT THING THAT KEEPS
>>> ASKING WHICH PARTITIAN TO BOOT?!?!?!?!?
>>
>>Just a wild guess, but boot from your ORIGINAL BOOT floppy using your
>>[etc.]
>
>I suspect it isn't that simple. Installing the shoelace menu program
>in the boot partition overwrites the boot sector. You cannot replace
>it as it was unless you have saved a copy of it, or unless you do a
>low-level format of your hard disk! Norton's Disk Doctor, as I recall,
>can write a pretty standard boot sector to your hard disk and save
>having to boot from a floppy.
Does this mean that for IDE drives, shoelace is a permanent step???
If so, a notice to that effect would be welcome in the FAQ!
-Joel
([email protected])
[next article]
From: [email protected] (Nate Williams)
Subject: Re: Is Shoelace permanent? [was HELP! SHOELACE IS STRANGLING ME!!!]
Keywords: n
Message-ID: <[email protected]>
Date: 23 Apr 92 17:49:09 GMT
References: <[email protected]> <[email protected]> <199
[email protected]>
Sender: [email protected] (USENET News System)
Organization: Montana State University, Bozeman MT
Lines: 67
In article <[email protected]> [email protected] (Joel M. Hoffman
) writes:
>In article <[email protected]> [email protected] (A. V. Le Blanc) w
rites:
>>>In article <[email protected]>, [email protected]
(Greenup) writes:
>>>>
>>>> I recently used shoelace to try & make linux boot. after no success,
>>>> i want to go back. HOW DO I GET RID OF THIS BOOT THING THAT KEEPS
>>>> ASKING WHICH PARTITIAN TO BOOT?!?!?!?!?
>>
>>I suspect it isn't that simple. Installing the shoelace menu program
>>in the boot partition overwrites the boot sector. You cannot replace
>>it as it was unless you have saved a copy of it, or unless you do a
>>low-level format of your hard disk! Norton's Disk Doctor, as I recall,
>>can write a pretty standard boot sector to your hard disk and save
>>having to boot from a floppy.
>
>Does this mean that for IDE drives, shoelace is a permanent step???
>If so, a notice to that effect would be welcome in the FAQ!
>
>-Joel
>([email protected])
Nope,
I talked to John, so now I will (hopefully) explain what should
happen when you install shoelace.
1) Try it out on floppies first!!!
Read the docs, it explains what is going on.
2) Try it out on floppies first!!!
Read the docs, it explains what is going on. (Get my drift 🙂
3) When you finally install shoelace onto your hardrive, install it in a
partition. DON'T install it in on the whole drive. This means don't
install it on /dev/hda, but on /dev/hda[1234]. On an IDE drive you
can't low-level format to get rid of winiboot w/out re-writing a new one
on top of the old one.
4) If you want to install it on your WHOLE drive, copy the old version
of the master boot sector to a floppy so you can go restore it if you need
to.
Something like :
$ dd if=/dev/hda of=/dev/floppy
Do a
floppy. Someone with a bit more knowledge than I can give you the block
count and block size to be exact.
5) When you want to use shoelace, use fdisk and set the partition that
contains the shoelace bootblok to active. If you don't like, don't set that
partition to active.
6) Even if you don't install shoelace on the whole drive, it still might be a
good idea to save the old boot sector to a floppy, before you install
shoelace.
This posting probably was more confusing than helpful, so email me
if you have questions/need more information.
Nate
--
:-):-):-):-):-):-):-):_):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-)
:-)Inet: [email protected] 🙂 Sys. Mgr. - GIS lab : Earth Sci.:-)
:-)Bnet: osynw@MTSUNIX (Painfully slow):-) The guy blamed for everything 🙂
:-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-):-)
[next article]
From: [email protected] (Lance Pickup)
Subject: Re: dos emulator
Message-ID: <[email protected]>
Date: 23 Apr 92 17:29:14 GMT
References: <[email protected]>
Sender: @watson.ibm.com
Reply-To: [email protected]
Organization: IBM Technology Products
Lines: 29
In article <[email protected]>, [email protected]
nformatik.uni-wuerzburg.de (Matthias Lautner) writes:
|> If someone is interested in helping me to implement the functions of the
|> dos int 0x21 it would be grate. I'm currently working with 0.95c.
This may sound like an outrageous amount of work, but in the long
run it may prove to be worth it: I think commercially available
DOS emulators (well, I can really only speak for the ones available
in AIX because that's all I'm familiar with) use DOS itself to provide
INT 21h support. Provided you supply all the BIOS and I/O support
(this is the really outrageous part!--big IF!), DOS itself should
be able to run providing INT 21h support. I have seen this approach
recommended before.
The reason it would be well worth the time to develop is that you
are more assured that ALL DOS software will run, and it should be
easier to support follow-on versions of DOS as it becomes available.
However, the magnitude of the work might be overwhelming for one
person, so keep up the good work however you decide to go! It
sounds real promising!
--
...Lance
-----------------------------------------------------------------------
Lance Pickup VNET/IBM Internet: lpickup@btv
IBM/Vendor Systems Internet: [email protected]
Technology Products
Burlington, VT Phone: (802) 769-7104 (tie 446)
-----------------------------------------------------------------------
[next article]
From: [email protected] (LCDR Michael E. Dobson)
Newsgroups: comp.os.linux,comp.os.minix
Subject: Re: HELP! SHOELACE IS STRANGLING ME!!!
Message-ID: <[email protected]>
Date: 23 Apr 92 12:01:03 GMT
References: <[email protected]> <[email protected]
ime.COM>
Organization: Naval Medical Research & Development Command
Lines: 25
In article <[email protected]> [email protected] (Kevin
Cummings) writes:
>In article <[email protected]>, [email protected] (G
reenup) writes:
>>
>> I recently used shoelace to try & make linux boot. after no success,
>> i want to go back. HOW DO I GET RID OF THIS BOOT THING THAT KEEPS
>> ASKING WHICH PARTITIAN TO BOOT?!?!?!?!?
>
>Just a wild guess, but boot from your ORIGINAL BOOT floppy using your
>ORIGINAL ROOT floppy (you do still have those, don't you?). And mount
>your current hard disk on /mnt. Now you should be able to "cd /mnt/etc"
>and undo whatever you did with shoelace.
>
That won't work, the tools that come with shoelace have no way to remove
the shoelace bootstrap program that was written into the Master Boot Record
on the hard drive. To get the standard DOS bootstrap back you have to use
something like Norton Utilities. If you have completely backed up your
hard drive, you can also perform a low level format of the hard drive
followed by a high level format and reinstallation. Maybe someone else
knows an easier way.
--
LCDR Mike Dobson, Sys Admin for | Internet: [email protected]
nmrdc1.nmrdc.nnmc.navy.mil | UUCP: ...uunet!mimsy!nmrdc1!rdc30
AT&T 3B2/600G Sys V R 3.2.3 | BITNET: dobson@usuhsb or nrd0mxd@vmnmdsc
WIN/TCP for 3B2 R3.2 | MCI-Mail: 377-2719 or [email protected]
[next article]
From: [email protected] (Doug Mayfield)
Subject: Math Co-Processor library
Message-ID: <[email protected]>
Date: 23 Apr 92 18:54:28 GMT
Sender: [email protected]
Reply-To: [email protected] (Doug Mayfield)
Organization: UC Davis, EECS Division of Computer Science
Lines: 15
I was wondering if anyone knew if the libm.a that comes with
GCC could be recreated for a system with a Co-Processor or 486.
If such a monster exists, could someone point me to it. And
maybe some test code with binary file sizes to check if it works
right.
---As a side note, how does Linux deal with sharing the co-processor
among various users. Or does it???
------------------------------------------------------------
Douglas B. Mayfield
[email protected]
(Computer Science Dept.)
(University of California at Davis)
[next article]
From: [email protected]
Subject: Shells and Shared libraries
Message-ID: <[email protected]>
Date: 23 Apr 92 18:16:46 GMT
Organization: Penn State University
Lines: 11
First question, does anyone know where I can get hold of any shells for
Linux other than bash and ash, particularlly csh?
Second, I have the source to the kernel and a bunch of utilities, plus GCC
2.1. How do I tell it to activate shared libraries, or does it just do it?
--
Steven A. Bairstow * "There ain't no rules around here!
InternNet : [email protected] * We're trying to accomplish something!"
or : [email protected] * - Thomas Edison
WWIV-Net : 216 @8450 *
[next article]
From: [email protected] (PTed Garvin)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: all kinds of questions (autoparking)
Message-ID: <[email protected]>
Date: 23 Apr 92 17:38:48 GMT
References: <[email protected]> <1992Apr13.223428.29499@colo
rado.edu> <[email protected]> <[email protected]
sinki.FI> <[email protected]> <[email protected].
nl> <1992Apr23.
133202.65382
Organization: Discordian Information Production and Distribution, Ltd
Lines: 35
Xref: nimbus comp.os.linux:749 alt.os.linux:1140
[email protected] (Lance Pickup) writes:
>In article <[email protected]>, [email protected]
t.nl (Rogier Wolff) writes:
>|> Yes. Right. I have a broken 40M disk, and when you power down, it still
>|> discharges a capacitor through the voice coil, which jams the head
>|> behind a latch. (wouldn't autoparking be hard for stepper motor controlled
>|> heads? at 24 ms average seek time, you'd need power for 0.1 seconds
>|> to step the motor back home, worst case)
>I could be wrong, but I thought that the way auto-parking worked was
>that the head was spring loaded such that whenever power was removed,
>it would just spring back into a safe location (i.e. you need power
>to un-park it).
>--
> ...Lance
>-----------------------------------------------------------------------
>Lance Pickup VNET/IBM Internet: lpickup@btv
>IBM/Vendor Systems Internet: [email protected]
>Technology Products
>Burlington, VT Phone: (802) 769-7104 (tie 446)
>-----------------------------------------------------------------------
None of the heads are spring loaded. It is possible that some
manufacturers could use some kind of cap to discharge but I have never
seen one large enough to do that or even enough room inside a drive
to mount a cap that large (I used to build hard drives).
They have a nice big spindle motor to spin the media, which I believe
is being used as a generator when the power is shut off. Not only
would that be able to generate power to park the heads but it would serve
as a brake to help it spin down faster. Notice they usually spin down
pretty quick... that isn't bearing friction 🙂
- Vincent
[next article]
From: [email protected] (Bennett E. Todd)
Subject: Re: compress doesn't work
Message-ID: <[email protected]>
Date: 23 Apr 92 19:47:15 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Bennett E. Todd)
Organization: The Internet
Lines: 24
Folks who are having trouble, how are you reading and/or writing the
floppy? dd(1) really loves to drop out chunks of data, because it treats
each inbound chunk of data as a block; if it doesn't get as much on a
read as its blocksize, it calls it a short block. If the floppy is
correctly written then the sequence
cat /dev/PS0 | compress -d | tar xvf -
should process correctly, and it will only blow up after you get past
all the good data in the archive (and extract it); then compress will
proceed to go nuts on the extra garbage past the end of the archive.
Harmless, though annoying.
Likewise, use cat(1) to write the floppy. If you need to split up a big
archive, use split(1). If you don't have enough disk to hold a copy of
the whole archive (split wants to write its pieces to xaa, xab, ...)
then you could try something like what I did in my wrflops script
(recently posted here in floptools.sh), where I used a separate little C
program (I called it catbytes) to bust up the input stream into
floppy-sized wads, and separate invocations of cat(1) to deposit them
onto the floppies.
-Bennett (curmudgeon)
[email protected]
[next article]
From: [email protected] (Theodore Ts'o)
Subject: Re: dos emulator
Message-ID: <[email protected]>
Date: 23 Apr 92 20:20:06 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 27
From: [email protected] (Drew Eckhardt)
Date: 23 Apr 92 16:33:51 GMT
1. How many documented DOS calls are there?
The answer is "many". The DOS kernel is ~50K of machine code. You
don't want to duplicate that. Instead, you emulate what DOS runs on,
and catch all of the programs that run there too, and run a *REAL*
dos.
Actually, what I would do is emulate DOS by mapping the DOS filesystem
calls into Unix filesystem calls. That should simply things quite a
bit. Of course, you will still have to emulate the BIOS calls, since
many programs call direct to the BIOS. But if you emulate the BIOS
calls and the DOS 2.11 calls, you'd be surprised how many DOS programs
will actually run correctly.
One question about your (Matthias's) technique however; as I recall, the
386 virtual memory mode requires that the task use *virtual memory*
addresses from 0k to 640k, where the kernel is currently located. You
can fix this by actually swapping the page maps in and out, but that's
much more significant than just changing the segment registers. Are you
actually playing page map games to make this work? And if so, how are
you doing this?
- Ted
[next article]
From: [email protected]
Subject: random()?
Message-ID: <[email protected]>
Date: 23 Apr 92 20:20:28 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 12
this question is for the c programmers out there. I have a few basic and pascal
programs that I would like to re-write in C, but a lot of them use some type of
random number generator. I once saw a reference to a random() function, and I
am wondering if this (or some other random function)is included in gcc 2.1 .
If not, I have some C source to create a pseudo-random # based on a user inputed
seed. I figure that if I use the date/time info for the seed, it should serve t
he
purpose. However, I don't know the calls for time/date functions. 🙁
(In case you can't tell, I am quite new to C, I am native to Pascal, but I have
a book written to teach C to Pascal programmers called: _C_Programming_in_a_UNIX
_Environment_ I got it for a class I didn't take. I also have: _C_Primer_Plus_
by the Waite Group)
Thanks for your help, please e-mail: [email protected]
[next article]
From: [email protected] (Michael K. Johnson)
Subject: new stuff at tsx-11
Message-ID: <[email protected]>
Date: 23 Apr 92 21:21:42 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Michael K. Johnson)
Organization: The Internet
Lines: 32
OK, there is new stuff at tsx-11. Let's see if I can remember it all.
all this is at tsx-11.mit.edu
NEW FAQ!!!!!!!!!!!
/pub/linux/FAQ
/pub/linux/docs/FAQ
/pub/linux/docs/FAQ_04_21
(all the same)
Apparently, no one knew where bootlin.zip is, and so Fabien uploaded
another copy -- it is
/pub/linux/INSTALL/bootlin.zip
GDB 4.5
/pub/linux/[binaries,source]/usr.bin/gdb-4.5.tar.Z
(both are the same file, containing binary and diffs to the
GNU source)
dvi2xx
/pub/linux/binaries/tex/dvi2xx.tar.Z
And finally, I removed the link from linux-0.95c++.tar.Z to
pre-0.96.tar.Z because it was confusing several people, I believe.
Sorry to anyone I confused.
I would like to apologize for being behind _again_ on the buglist and
slightly behind on the patch directory -- my professors are beginning
to demand that I actually _do_ something in their classes... 😉
michaelkjohnson
[email protected]
[next article]
From: [email protected]
Subject: Re: dos emulator
Message-ID: <[email protected]>
Date: 23 Apr 92 21:35:28 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 24
--------
From: [email protected] (Drew Eckhardt)
Date: 23 Apr 92 16:33:51 GMT
One question about Matthias's technique however; as I recall, the
386 virtual memory mode requires that the task use *virtual memory*
addresses from 0k to 640k, where the kernel is currently located. You
can fix this by actually swapping the page maps in and out, but that's
much more significant than just changing the segment registers.
That's correct - a 386 virtual mode task must start at
virtual address 0, and if you allow the full address space of
a 286 in real mode, up to 1M + ~64K.
Are you
actually playing page map games to make this work? And if so, how are
you doing this?
This is matthias's play toy - ask him.
- Ted
--------
[next article]
From: [email protected] (Hongjiu Lu -- Graduate Student)
Subject: Re: News under Linux
Message-ID: <[email protected]>
Date: 23 Apr 92 16:28:55 GMT
References: <[email protected]>
Sender: [email protected] (USENET News System)
Organization: Washington State University
Lines: 23
In article <[email protected]> [email protected] (Steve
Robbins) writes:
>Howdy,
>
>
>I found out the extremely hard way that linux or gnu thinks that fseek()
>should return the new offset if sucessful, rather than zero, like all
>other stdio's appear to. Is this some new ANSI thing, or is libc screwed
>up?
>
Not totally my fault. It has been fixed in the next release.
>Also strcasecmp() is the same as strcmp(). I got the gcc2.1 distribution
What is your input?
>from tsx-11 somewhere around April 4 -- is there a newer version that
>has corrected these things?
>
>Steve Z
>
H.J.
[next article]
From: [email protected] (gary a moyer)
Subject: RE: cannot mount filesystem
Summary: Thanks for the help!
Message-ID: <[email protected]>
Date: 23 Apr 92 21:58:14 GMT
Sender: [email protected]
Lines: 6
Thanks to all for the input on the inability to mount.
The fix is to remove the file mtab~ in /etc
thx much.
[email protected]
[next article]
From: [email protected] (Michael Pereckas)
Subject: Re: linux
Message-ID:
Date: 23 Apr 92 21:55:12 GMT
References:
Sender: [email protected] (News)
Organization: University of Illinois at Urbana
Lines: 16
[email protected] (Matt Mosley) writes:
>3) Where can I get troff/groff/nroff for Linux? (or at least workable).
Look on tsx-11.mit.edu in pub/linux/binaries/groff, or some such
directory. It installs easily and works great. For typeset output,
you'll need a postscript printer, ghostscript (which ports fairly
easily), or a TeX dvi driver (and Computer Modern fonts). I usually use
the dvi output, dvips (and metafont), and ghostscript to get output on
my Alps Allegro, which is Epson compatable. I think the CM fonts are
better that the ghostscript fonts, at least at 360x180 dpi, and I use
TeX, so I have dvips and metafont set up.
There is also cawf, a little nroff-like program.
Michael Pereckas
[next article]
From: [email protected] (H. Peter Anvin N9ITP)
Subject: Latin-1 again
Message-ID: <[email protected]>
Date: 23 Apr 92 20:50:17 GMT
Reply-To: [email protected] (Peter Anvin)
Organization: Northwestern University, Evanston, Illinois, USA
Lines: 35
Is there an interest to have Linux use Latin-1? I have been working on a
collection of Latin-1 CRT fonts, and I would be happy to implement it as
soon as I get Linux running on my machine.
My solution is as follows:
EGA/VGA/MCGA/HGC+/InColor:
Download the fonts to the character generator chip, use text mode as usual.
CGA/HGC:
Display characters in graphic mode.
MDA:
Use a translation table.
Keyboard:
Create appropriate keyboard routines, possibly with use of special key
combinations to generate extended characters.
Printer:
Use a customizable code file that contains the necessary translations;
question: should this translation be done by /dev/lp* or by the lp program.
Why latin-1? This code effectively supports just about all Western
European languages (although dropping the early-draft OE ligature was a
mistake) in a much more effective and *standard* way than the PC code
pages. In addition, this is the preferred code standard for most UNIX
systems that support 8-bit characters.
/hpa
--
INTERNET: [email protected] TALK: [email protected]
BITNET: HPA@NUACC HAM RADIO: N9ITP, SM4TKN
IRC: Scalar NeXTMAIL: [email protected]
"Kissing a smoker is like licking an ashtray." -- Anon.
[next article]
From: [email protected] (Al Clark)
Subject: Re: free (memory allocation table)
Message-ID: <[email protected]>
Date: 23 Apr 92 22:49:30 GMT
References: <[email protected]>
Organization: Self
Lines: 25
In article <[email protected]> [email protected] (Peter Thomson) writes:
>
>
>Hi can some one tell me if the command free reports back any more than
>1Mbyte of memory. There is 2Mbytes on my machine but free states there is
>only one. running multiple tasks and watching ps confirms this and swapping
>starts at the 1Mbyte limit
>
> Thanks in advance Peter.
This is my output from free on a 4 MB system.
------------------- start of screen dump ------------
807 # free
total used free shared
memory: 2944 1052 1892 360
swap: 20396 0 20396
808 # free -c
total used free shared
memory: 3014656 1077248 1937408 368640
swap: 20885504 0 20885504
--------------------- end of screen dump ------------
--
Al - [email protected] - My opinions are my own.
*** Practice random kindness and senseless acts of beauty! ***
[next article]
From: [email protected] (Cornell Kinderknecht)
Subject: mcc-interim and shadow /bin/login ignore .cshrc with tcsh
Keywords: passwd login cshrc tcsh
Message-ID:
Date: 23 Apr 92 23:28:47 GMT
Sender: [email protected] (-- cjk --)
Reply-To: [email protected]
Organization: NEC USA, Corporate Network Administration Division
Lines: 10
Nntp-Posting-Host: vivaldi.inoc.dl.nec.com
Both the /bin/login with the mcc-interim 0.95c+ release and the one that
come with the shadow-bin.tar.Z on tsx-11 ignore the ~/.cshrc file
when logging in with tcsh. Any ideas why or a fix?
--- Cornell
--
| Cornell Kinderknecht | Email: [email protected] |
| CNAD/CCSL | UUCP: uunet!necbsd!cornell |
| NEC USA/NEC America, Inc. | Phone: 214-518-3509 |
| Irving, TX | Fax: 214-518-3552 |
[next article]
From: [email protected] (Rik Faith)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Re: all kinds of questions (autoparking)
Message-ID: <[email protected]>
Date: 23 Apr 92 16:22:11 GMT
References: <[email protected]> <[email protected]
.COM> <[email protected]>
Sender: [email protected]
Followup-To: comp.os.linux
Organization: University of North Carolina, Chapel Hill
Lines: 17
In article <[email protected]> [email protected].
FI (Risto Kankkunen) writes:
>In [email protected] (Kevin Cummings) writes:
>> So unless you are shutting down your system to reboot DOS (or something
>> else), your disk heads will not be parked by LINUX.
>
>But isn't head parking suggested only before moving the computer or the
>disk itself?
No. Head parking has been suggested as *absolutely* necessary before
moving a computer, but many people do so before powering down the computer
for any reason. This protects the drive from accidental damage when you
happen to trip over your tower case (for example).
Most (all?) newer drives park heads automagically at power down, but I have
an old MFM drive which does not do this. So, whenever I turn off the
machine after using linux, I reboot DOS and run my disk head parking
program.
[next article]
From: [email protected] (Peter Williams 8169821)
Subject: ispell American dictionary available
Message-ID: <[email protected]>
Date: 23 Apr 92 23:51:01 GMT
Sender: [email protected]
Organization: Design Computing Unit, The University of Sydney
Lines: 10
Nntp-Posting-Host: dean.arch.su.oz.au
The american version of an ispell dictionary is now available at
archsci.arch.su.oz.au (129.78.66.1) if file pub/linux/ispell/yank.hash.Z
It can be used to replace /usr/local/lib/ispell/english.hash or as an extra
dictionary in accordance with the ispell documentation.
--
Peter Williams |e-mail: [email protected]
Key Centre for Design Quality |phone: +61-2-692 2053 or +61-2-660 6156
University of Sydney |+++++++++++++++++++++++++++++++++++++++
[next article]
From: [email protected] (Peter Williams 8169821)
Subject: no German ispell dictionary. Sorry!
Message-ID: <[email protected]>
Date: 23 Apr 92 23:56:25 GMT
Sender: [email protected]
Organization: Design Computing Unit, The University of Sydney
Lines: 10
Nntp-Posting-Host: dean.arch.su.oz.au
It appears that the list of German words that I thought I had was a figment
of my imagination and I, therefore, will be unable to build the German
dictionary as promised.
Sorry!
--
Peter Williams |e-mail: [email protected]
Key Centre for Design Quality |phone: +61-2-692 2053 or +61-2-660 6156
University of Sydney |+++++++++++++++++++++++++++++++++++++++
[next article]
From: [email protected] (Hongjiu Lu -- Graduate Student)
Subject: Re: random()?
Message-ID: <[email protected]>
Date: 23 Apr 92 21:53:56 GMT
References: <[email protected]>
Sender: [email protected] (USENET News System)
Organization: Washington State University
Lines: 16
In article <[email protected]> [email protected] writes:
>this question is for the c programmers out there. I have a few basic and pasca
l
>programs that I would like to re-write in C, but a lot of them use some type of
>random number generator. I once saw a reference to a random() function, and I
>am wondering if this (or some other random function)is included in gcc 2.1 .
>If not, I have some C source to create a pseudo-random # based on a user inpute
d
As you may know, Linux is trying to be ANSI and POSIX compatible. All
the library functions are declared in the header files. Please take a
look at those files. Somebody is working on the list. I hope it will be
out soon.
I believe Linux has most of useful functions in library. Your random is
one of them. Take a look at your book and stdlib.h.
H.J.
[next article]
From: [email protected] (Satish Chittamuru)
Newsgroups: comp.os.os2.misc,comp.os.linux
Subject: Re: Any comments on Quantum 240A ??
Message-ID: <[email protected]>
Date: 23 Apr 92 22:23:48 GMT
References: <[email protected]>
Organization: Microsoft Corp.
Lines: 38
In article <[email protected]> [email protected]
eis.edu (David A. Sinclair) writes:
>
> In a recent post, I asked for people's evaluation of the Quantum
> LPS240AT drive, particularly its useability for OS/2 2.0 and for
> Linux. All of the many replies I received were positive --
> apparently, it's a really good drive. Unfortunately, it is also very
> hard to find in stock!! I did, however, find what I believe is a good
> substitute at a good price:
>
> Seagate ST3283A
> 245 meg
> IDE, 3.5" HH
> 12 (!!) ms access
> 128K cache
>
> The cache is smaller than the Quantum's, but the access speed a bit
> faster. I ordered mine today from Elco computer for a price of $605.
Nice! But what is the transfer rate. The Quantum is a Zone Bit recorded
drive (3 zones) and has a max transfer rate of 3.8MBytes/sec. Contrast
this with the Maxtor LXT213 (15ms, 32K cache, 212meg) which has a Max
transfer reate of 1.96MBytes/sec.
> If there is any interest, I will post a followup after having received
> the drive and installed both OS/2 and Linux on it (and used it a bit!!
Please do. especially numbers for DOS and Linux. I guess a comparision with
OS/2 would be interesting but I want to see the difference between a
dedicated OS (read: simple, single tasking) vs a multi-tasking OS.
> Cheers,
> David Sinclair
> [email protected]
>
--
Satish K. Chittamuru [email protected]
Microsoft Corporation Redmond, WA
=====
Veni! Vidi! Visa! (I came! I saw! I purchased!). -- Julius, Sieze Her!
[next article]
From: [email protected] (Janne Himanka)
Subject: Re: Latin-1 again
Message-ID:
Date: 24 Apr 92 07:18:02 GMT
References: <[email protected]>
Sender: [email protected]
Reply-To: [email protected]
Organization: Tilde's home bakery
Lines: 10
In-reply-to: [email protected]'s message of Thu, 23 Apr 1992 20:50:17 GMT
Supported! That could even make it possible to do something *useful*
in Linux. I don't know if that's good or bad. I mainly consider Linux
as fun.
Janne
"What did you do with the Pharaoh + [email protected]
Great Outhouse 8?" + [email protected]
"We ate him. He was unspeakably + TKOP-HJH@FINOU
toothsome." + Oulu Univ., Dept. of Libr. & Inf. Sc.
[next article]
From: [email protected] (Paul Nulsen)
Newsgroups: comp.os.linux,comp.os.minix
Subject: Re: HELP! SHOELACE IS STRANGLING ME!!!
Message-ID: <[email protected]>
Date: 23 Apr 92 23:35:05 GMT
References: <[email protected]> <[email protected]
ime.COM> <[email protected]>
Organization: University Of Wollongong
Lines: 26
[email protected] (LCDR Michael E. Dobson) writes:
>>In article <[email protected]>, [email protected] (
Greenup) writes:
>>>
>>> I recently used shoelace to try & make linux boot. after no success,
>>> i want to go back. HOW DO I GET RID OF THIS BOOT THING THAT KEEPS
>>> ASKING WHICH PARTITIAN TO BOOT?!?!?!?!?
>
> [suggestion deleted]
>
>That won't work, the tools that come with shoelace have no way to remove
>the shoelace bootstrap program that was written into the Master Boot Record
>on the hard drive. To get the standard DOS bootstrap back you have to use
>something like Norton Utilities. If you have completely backed up your
>hard drive, you can also perform a low level format of the hard drive
>followed by a high level format and reinstallation. Maybe someone else
>knows an easier way.
I read recently that there is an undocumented fdisk option in DOS
fdisk /mbr
which, as I understand it, rewrites the master boot record. I presume this
command preserves the partition table, but I am not game to try it to find
out. This would cure the main problem if it writes a fresh copy of the DOS
MBR. Anyone know?
Paul Nulsen
[email protected]
[next article]
From: [email protected]
Subject: DEL using xcomm XC
Message-ID: <[email protected]>
Date: 24 Apr 92 00:40:15 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 9
How does I send a DEL using xcomm, XC. I know kermit had trouble with
VT100 termcap and somehow used backspace BS for DEL and ^H for
backspace. Not sure how and didn't really like doing ^H for
backspacing.
BTW: I assume e-mail to linux-activists not posted to digest within a
day has gone to the bit bucket. Ted, you didn't mention that you
did/didnot fix something.
John
[next article]
From: [email protected] (H.H. Bergman)
Subject: Re: IDE drive performance with linux
Message-ID: <[email protected]>
Date: 24 Apr 92 00:00:41 GMT
References: <[email protected]>
Organization: Faculty of Arts & Letters, Groningen University, The Netherlands
Lines: 75
In article <[email protected]> [email protected] (
Drew Eckhardt) writes:
>In article
.rutgers.edu (Charles Hedrick) writes:
>>If you're getting one block per rotation, that means that the system
>>can't keep up with the rotation speed of the disk. This is not at all
>>unusual. Classically what you do is arrange the files on disk using
>>every other block or every third block, or whatever you need in order
>>that the system can keep up. With system v, there's an argument to
>>mkfs to specify this interleave. With the Berkeley Fast File System,
>>it's specified to mkfs or tunefs. Since folks are busy implementing
>>an improved file system for Linux, I'm going to hope that this feature
>>is included. It would presumably be easy to put into the current one
>>as well, but I haven't looked at the code to see what would be
>>involved.
>
>
>You don't want to do this - you'll take X (where X is the interleve)
>revolutions to read anything. Instead, you want to speed up reads
>(switch to DMA), and read an entire track at once. This will take the
>same amount of time as reading one sector.
>
I agree, with interleave you won't get the max bandwidth of the drive.
With an os that depends heavily on disk access, this seems wasteful
to me. I'd rather see the effort needed for implementing interleaving
spent on improving the hd driver and/or the file system. With full track
reads, my drive performance should (theoretically) go from 60kB/sec to
1MB/sec (max). 🙂 With interleaving it could 'only' reach 0.5MB/sec.
(But either would be a significant improvement. 🙂
>
>You can read into a secondary buffer, or increase the buffer cache size
>to be the same size as a track.
>
>The problem with the first method is that you still have to do a copy for
>each transfer to buffercache, which will be a performance hit. The
>second method will have problems if you are using lots of small files,
>as you will use more buffer cache than you would with the current
>system, and stand a better chance of knocking something out of cache
>and having to actually read it. Both have a problem in that track size
>varies for different disks, and you can have a zone recorded device -
>but anything that goes through the trouble of being zone
>recorded, or has a really high density probably also has at least a track
>buffer (some of the newer SCSI disks have 256K buffers), and getting the
>whole thing in one shot so you don't get hit by rotational latency
>isn't as much of an issue.
>
My disk has zoning, but appearently no proper track buffering. 🙁
I suggested to put a program in /etc/rc to inform the hd driver of the
zoning info. This way you can still use the system if you don't have
the info about your drive. [Also usful for ditribution disks :-)] I'm
not sure if we should follow the 'Unix way' of doing that with
'disktab's, though.
Once the driver has got the zone-info it can start reading full tracks.
The kernel should try to group requests for the block devices as much
as possible, so that the block drivers can use large transfers, saving
overhead (like rotational delays ...).
>
>
>which will have some penalty want to buffer at the track level -
>if you want, you can even DMA into it
Theodore Ts'o suggested the following:
>(BTW, a very simple change to the block allocation algorithm for the
>Minix fs would be to pass it the location of the previous block in the
>file, and for it to try to allocate a block in the same "zone". If
>there is no previous block, it should allocate a block in the zone with
>the highest amount of free space, to spread out the allocated blocks
>around. Such a change would give you a large part of the speed
>advantages of the Berkeley FFS, with relatively little work. Of course,
>it still wouldn't give you long filenames...)
I don't know the current allocation scheme, but this sounds good!
-Hennus Bergman
[email protected] [until July 3, 1992]
[email protected] [keep trying 'till it doesn't bounce :-(]
[next article]
From: [email protected] (A. V. Le Blanc)
Subject: MCC 'interim' version of Linux (was Re: two questions)
Message-ID: <[email protected]>
Date: 23 Apr 92 13:21:07 GMT
References:
aremont.edu>
Followup-To: [email protected]
Organization: Computing Centre, University of Manchester
Lines: 146
In article <[email protected]> [email protected]
nt.edu (Jim Winstead Jr.) writes:
>In article
Wetmore) writes:
>>2) What is the MCC release of Linux?
>
>It's the Manchester Computing Center (?) release, which serves as a
>sort of extended-release, as I see it. It includes many more
>utilities than the basic release from Linus and myself, and as a
>result comes on three (four?) floppies.
It comes on two floppies, but there are a few others available.
The Manchester ComputING CentRE was once Manchester University Regional
Computer Centre (pronounced 'murk') and later UMRCC. The change from
'computer' to 'computing' was made (supposedly) to mark a shift in
emphasis from supporting machines to supporting the people who use
them. (Some of the people who use them complain that the shift in
emphasis has not yet been implemented.) The University of Manchester
has been known to claim that computers were invented here, which they
were if you define 'computer' properly. We are a Centre, not a
Center, because (as someone pointed out in a recent note) that is
how it is spelled in 'proper' English (English + Irish + Australian +
Indian + Canadian + ...) as opposed to the American dialects. I try
to mention MCC from time to time, since they pay me and supply me
with equipment. I am in fact supposed to be doing other things for
MCC, and, also in fact, I do.
The MCC 'interim' releases of Linux are unofficial experiments.
They vary depending on my whims and on the time I have to give them,
usually not enough, I am afraid. The latest so far, 0.95c+,
had (or has?) the following goals:
(a) To provide a simple installation procedure.
(b) To provide a more complete installation procedure.
(c) To provide a backup/recovery service.
(d) To backup my (then) current system!
When I first put Linux on a PC, back at 0.11, I got the standard
boot and root floppies, found there was no working fdisk, tried
edpart, which made a mess of my partition table by sorting it,
used the MINIX fdisk, and finally got a system going. Then I
started trying to get other bits to it. I decided this was a
bit awkward, and hoped someone would do something about it.
I also hate having to search through 3 ftp sites for useful bits,
the fragmented nature of a system which, I believe, was one of
the serious problems with MINIX, and having to compile everything
again and again because it was originally linked with some
defective version of the library, as happens often enough even
with 'mature' commercial Unixes.
Theodore Ts'o wrote the ramdisk code in the Linux kernel. As he
remarked, it was originally designed to make it possible to store
some files on the ramdisk, and so free the disk drive for other
purposes, for example, for creating or modifying a boot floppy
so that it boots using a hard disk partition as the root device.
Both Ted and Linus warned me that the ramdisk code was inefficient,
but I thought this was no problem for an installation/backup/
recovery system. Both also pointed out that the ramdisk uses
much more memory than it should, and this has in fact proven to
be a problem on systems with only 2mb of RAM.
Nevertheless, the latest 'interim' release from MCC does manage to
squeeze quite a lot onto TWO disks: one of which combines the boot
and root disk, and one of which I called the 'utilities' disk.
The boot disk boots, loads its root device from the same disk, and
then starts executing /etc/rc. This runs a little script which
asks for the drive size, and mounts the utilities disk (which you
will of course have placed in the drive when you were instructed to).
The commands available on the combined boot/util combination are
approximately equivalent to those on Jim Winstead's root disk.
Of course, I have a lot more space than he does -- about 500k more --
so I use a lot of this space for tar.Z files which contain all the
usual Unix commands I can find, excluding 'man' and the compilers,
but including, for example, awk (gawk), all of the GNU shell/file/text
utilities, grep and its cousins, sed, vi, more, less, tar, compress,
uuencoding/decoding, the mtools package for reading/writing DOS
files, and make. I also added the joe editor, for those who find
vi too alien. The format was a bit to cramped to try to include
emacs or tex or any other monster utilities, and almost everything
I wanted fit except Kermit, shoelace (which is awful, I admit, but
works if you know how to do it), and man pages.
All of this fit on two disks, the boot disk and the utilities disk,
as I have said before. Now I tried this out on some of my unsuspecting
friends, who made interesting suggestions and complaints. One in
particular asked if I could add a disk containing the C compiler.
So my first additional disk (comp) contains gcc 2.1 plus the include
files and the libraries. I couldn't fit g++ on the same disk, so I
added shoelace and a bit of g++. (I supplied a new README file for
shoelace, explaining how to test it out from a floppy before you
overwrite your primary boot sector). Later I added another disk
(comp2) containing the rest of g++ (include files, binaries, libraries)
and Kermit, which is too big to fit on the utilities disk. When
I released this version (which did not include Kermit at first),
I received some favourable comments, but most people felt that
man pages should be included. I therefore project another disk,
which should include groff, man, and man pages. Note that these
additional disks (comp, comp2, and man) are not really part of the
'interim' version proper, though it is convenient to lump them
together. Not also that the man disk does not yet exist, though
the binaries and patches for groff are in fact available; the
binaries are not really usable, since they include only unlinked
executables, not the groff library stuff. When the man disk comes
out, it will probably include only preformatted man pages, with the
unformatted pages available for ftp in another format. The pages
are collected from the GNU sources and from elsewhere, including,
of course, those from the excellent Linux-man contributors.
The man-1.0 program does not do QUITE what I would like, and so I
am messing with it -- I hate distributing things that don't do what
they should.
Contrary to my expectations, I did in fact create a sixth image,
which is called xdisk, and which -- in an awkward way -- allows
people with less than 4mb of RAM to install the 'interim' version
without using the ramdisk; they don't need my boot disk, but they
do need a standard 0.95c+ boot disk and my xdisk.
I hope this clarifies what the MCC 'interim' version of Linux 0.95c+
is, and why, when it consists of SIX disks, of which one does not
exists, while two come in two versions (US and UK keyboards), etc.,
I would still say that, properly speaking, it consists of TWO disks,
together with a number of optional extra disks.
The latest 'interim' version, while put together by me, is deeply
indebted to many other people, including Linus himself and Ted Ts'o,
who put up with a lot of hassle while I was working it up, and
Jim Winstead, some of whose bits are included. Also included are
poe-igl-1.2, lots of GNU code, HLU's C, C++, and libraries, all GNU
in origin, but with an awful lot of work involved in porting them,
the UU-, XX- and AtoB encoding/decoding utilities by Konrad Bernloehr,
and all the people who helped with testing and by reporting bugs
and even those who just said 'Thanks'.
I would hope that the 'interim' version would influence other
versions, from which I will of course get new ideas and code.
But it would be very awkward to do the standard distribution this
way. First of all, it is a pain producing a boot disk. If you want
to change a letter in a text file, you have to mount, edit, unmount,
copy the disk to a file, copy the file and the image to another disk,
and then reboot. The present system, with Linus providing boot disks
and Jim providing root disks, is more convenient in many respects --
but much harder for the end user. Perhaps it will change eventually
when Linux becomes stable -- but can you ever believe it will stop
changing?
-- Owen
[email protected]
[next article]
From: [email protected] (A. V. Le Blanc)
Newsgroups: alt.os.linux,comp.os.linux
Subject: Re: help, cannot mount file system...
Message-ID: <[email protected]>
Date: 23 Apr 92 19:38:51 GMT
References: <[email protected]> <[email protected]>
Followup-To: [email protected]
Organization: Computing Centre, University of Manchester
Lines: 29
In article <[email protected]> [email protected] (Kevin
Cummings) writes:
>In article <[email protected]>, [email protected] (gary a moyer) writes:
>> Hey netters! I _accidentally_ messed up my ability to mount
>> a file system! I did it inadvertantly while trying to incorporate
>> some _protection_. Anyhow, what other files does mount access besides
>> the actual device in /dev ? I am getting the message:
>> mount: unable to open lock-file
>>
>
>WHY ISN'T THIS A FAQ?????????? Linus, PLEASE change the mount
>message to write out the lock file pathname! And possibly another
>message about what to do about it.
This is, in fact, a FAQ. (see section V, question 45). The
message comes not from the operating system, but from the mount
command, which is maintained not by Linus but by Jim. The
message is in fact fairly informative, as Unix messages go; it is
far better then the errors produce by unsuccessful attempts to
umount, or to mount, nonexistent devices. I agree that the frequency
of this question indicates that some improvement is necessary, but
we have come an awfully long way since October and November if
this is the worst that there is (it isn't).
Anyway, the solution is this: login as root and
rm /etc/mtab~
-- Owen
[email protected]
[next article]
From: [email protected] (Steve Robbins)
Subject: News 'n' Mail
Message-ID: <[email protected]>
Date: 23 Apr 92 17:25:17 GMT
Organization: The chiral cage of chaos
Lines: 77
Recently, [email protected] (Matt Mosley) asked:
>
>I would like to know how to do bl following:
>
>1) Does Linux have *any* working mail program? I can't seem to find any.
>2) How do I transfer my news and mail feeds over bo Linux (I'd first need
> a mail program, of course, plus readnews and uucp).
>3) Where can I get troff/groff/nroff for Linux? (or at least workable).
>
I've been able to get mail going and almost get news going with
Taylor-UUCP, smail and elm, and c-news and rn.
None of these programs are 100% yet, so I'd really appreciate hearing from
others who have solid versions.
One of the most constant irritations in compiling these and gnu utilties, is
that the code always wants to declare standard library functions, like
"extern long lseek()" or whatever. My include files, on the other hand,
already have these in stdio.h and the like, and gcc complains. Is declaring
the library routines peculiar to linux or gcc or is it ANSI that nobody uses
yet or what? The other major irritation is that almost none of the configure
scripts works right.
The hardest problems to find, though, lie with libc.a. My version is from
early April gcc2.1 from tsx-11, so I'd appreciate knowing if these are fixed,
or better, where I can get source for the library functions. The two I've
found so far is strcasecmp() doesn't ignore case as it should, and fseek()
returns the new position on a successful call, rather than zero.
After sorting all that out, I have the following residual problems:
Elm:
When running as a regular user, I get the error 'problems restoring
permissions of /usr/spool/mail/steve' when I quit. No mail seems to be lost
or mangled, but this is vaugely disturbing nonetheless.
Smail:
When mail is sent to nonexistent local users, it gets bounced correctly, but
two 'general protection: 0000' messages come up, which is also disturbing.
As an aside, I recall someone once saying that they had smail going, but
couldn't get it to bounce mail...could you mail me and tell me how you did
that? I can't get mine to *not* bounce mail!
Also, is it just me, or is the configure/build tangle of shell, sed, awk,
and make scripts incredibly labyrinthine? I still don't even understand the
entire thing, and I can only get it to compile by putting in a dummy Makefile
in the src/lookup directory to build an empty ldrivlib.a. Then I have to
'touch' this makefile when the master make fails in some obscure awk program.
My src/lookup directory came absolutely empty...did I get a bad archive?
Taylor:
(Using taylor-style everything: config, spool, logfile, etc)
After a transfer, uuxqt starts up and works correctly, but since there are other
files in the /usr/spool/uucp directory (lockfiles, logfiles, etc) it complains
about not being able to open each of them `filexxxy/X.' ... TWICE! This is
probably a bug in Taylor's code, but if you know how to fix it...please
enlighten me.
C-news:
The dbz database routines seem to be screwing up in some way, which results
in the active file being appended to, rather than overwritten, among other
things. It also fails the regression tests supplied, so I'm relatively
sure there's a problem here, but haven't found it yet. Also, the
newsbin/inject/tear awk script fails for some reason. Removing the first half
of it makes it work again, but I don't know why.
Rn:
Nothing major yet.
So there you have it -- my answer to your question about a working mail (and
news) is "yeah, sorta".
Steve Robbins -- iconoclast at large
"All generalizations are false"
[email protected]
[next article]
From: [email protected] (Erik Talvola)
Subject: Patches for GCC v2.1 for Linux
Message-ID:
Date: 24 Apr 92 00:47:37 GMT
Organization: /usr/users/talvola/.organization
Lines: 14
NNTP-Posting-Host: qin.berkeley.edu
Has the person who ported GCC v2.1 to Linux given the patches back to
the FSF? It would be nice if GCC as shipped from the FSF had Linux
support installed. They should be posted to the gnu.gcc.bug newsgroup,
or sent to [email protected]. This will make it a lot easier to
grab the sources and build, without worrying about where the patches are.
The same hold true for all other GNU utilities...
- Erik
--
+----------------------------+
! Erik Talvola | "It's just what we need... a colossal negative
! [email protected] | space wedgie of great power coming right at us
! ...!ucbvax!qin!talvola | at warp speed." -- Star Drek
[next article]
From: [email protected]
Subject: Where's outb?
Message-ID: <[email protected]>
Date: 23 Apr 92 17:21:25 GMT
Sender: [email protected]
Organization: Digital Equipment Corporation
Lines: 11
I've spent the last few nights installing 95a, Linus' "second" patch, and
GCC 2.1, then trying to undo the mess. The result is that ld says undef
for outb, get_fs_byte, and similar things when I build the kernel (ps for 2.1
successfully compiled and linked.)
I suppose a file is still in the wrong place or something but I don't
know what and where.
Regards,
David Tucker @ATO / KC4ZGO / [email protected]
[next article]
From: [email protected] (I Reid)
Subject: Reliability of stdio with gcc 2.1 ?
Message-ID: <[email protected]>
Date: 24 Apr 92 03:42:37 GMT
Distribution: comp
Organization: Edinburgh University
Lines: 19
Just a quick posting to ask if anyone is having problem with any of the stdio
stuff (specifically fwrite/fwrite) when using gcc 2.1. I ask this because I
suspect this to be the cause of some ?? memory violations I have been having.
Evidence for:
1. Problems only occur when fread/fwrite
2. Explicit calls to fflush after every write 'cure' the problem (and lose me
the buffering I wanted 🙁
3. No problems with non buffered routines (read/write).
I'm not totally convinced that I have 2.1 set up exactly as it is supposed to
be but it all seems to work without any complaints.
Iain
p.s. Runiing 0.95c+ on a 386-33 with 8Mb Ram.
p.p.s. Any good debuggers out there? I know there is kernel support for one but
the last time I tried one it dies horrible every time 🙁
[next article]
From: [email protected] (Al Clark)
Newsgroups: comp.os.linux,comp.os.minix
Subject: Re: HELP! SHOELACE IS STRANGLING ME!!!
Message-ID: <=n8j+#[email protected]>
Date: 24 Apr 92 04:07:59 GMT
References: <[email protected]> <[email protected]
mrdc.nnmc.navy.mil> <[email protected]>
Organization: Self
Lines: 41
In article <[email protected]> [email protected] (Paul Nulsen
) writes:
>[email protected] (LCDR Michael E. Dobson) writes:
>>>In article <[email protected]>, [email protected]
(Greenup) writes:
>>>>
>>>> I recently used shoelace to try & make linux boot. after no success,
>>>> i want to go back. HOW DO I GET RID OF THIS BOOT THING THAT KEEPS
>>>> ASKING WHICH PARTITIAN TO BOOT?!?!?!?!?
>>
>> [suggestion deleted]
>>
>>That won't work, the tools that come with shoelace have no way to remove
>>the shoelace bootstrap program that was written into the Master Boot Record
>>on the hard drive. To get the standard DOS bootstrap back you have to use
>>something like Norton Utilities. If you have completely backed up your
>>hard drive, you can also perform a low level format of the hard drive
>>followed by a high level format and reinstallation. Maybe someone else
>>knows an easier way.
>
>I read recently that there is an undocumented fdisk option in DOS
>fdisk /mbr
>which, as I understand it, rewrites the master boot record. I presume this
>command preserves the partition table, but I am not game to try it to find
>out. This would cure the main problem if it writes a fresh copy of the DOS
>MBR. Anyone know?
>
>Paul Nulsen
>[email protected]
Since I have programs to save and restore the MBR (which I sent to
John Greenup), I saved the MBR on an XT I have, and tried
"fdisk /mbr" from floppy with the two MSDOS versions I have handy.
Compaq DOS 3.31 (running under DOS 4.01)
Result - "The Master Boot code has been updated" is reported.
Afterwards, partition table was same.
DOS 4.01 - Result - "Invalid switch - /mbr".
--
Al - [email protected] - My opinions are my own.
*** Practice random kindness and senseless acts of beauty! ***
[next article]
From: [email protected] (Medvezhonok)
Subject: Shadow password stuff
Message-ID: <[email protected]>
Date: 23 Apr 92 01:14:27 GMT
Organization: Syracuse University, Syracuse, NY
Lines: 17
Does anyone know where I can get documentation for the shadow.tar programs on
banjo's mit mirror's binaries directory? I tried installing them with the
installation program provided, only to have the %?$@ing things lock me out of
my system. If there were some docs, I could figure out what is wrong. (I
finally had to copy all of the stuff from the /etc and /bin directories on my
backup floppy over the stuff on my hd so that I could login again.). The
programs claim to not be able to find the passwd and group files, and don't
recognize root. sorry if this is a FAQ. Great OS so far, I just got dkbtrace
running on it (had to port it with an axe though, don't know much c).
Thanks,
Louis Ruppert
--
"You can say what you want to Debussey,
but there's not much of him left to hear"
-Monty Python
[next article]
From: [email protected] (Mark Tomlinson)
Subject: Re: HELP! Can't get GCC2.1 to run.
Message-ID:
Date: 2 May 92 11:05:11 GMT
Organization: Not an Organization
Lines: 25
I guess someone's going to try and give me the standard reply
"there's a problem in tar", but I don't think that is my problem.
I uncompressed the archives and then un-tar'd them so I didn't
need to use the 'z' option in tar (this has helped me before).
The compiler (cc1) seems to run fine, the problem I am getting is
in linking. Both the new ld and the ld.old give the same results,
so I guess the problem may be in the library file. The error
message I get is "can't resolve external __cbuff" or something
of that effect. If I remove the "#include
program (forcing the use of the library functions rather than
the macros), the linking works, and so does the program.
Anyone else found this problem? Anyone know what I'm doing wrong?
PS - I'm using the 0.95c+ bootimage. The 0.95c bootimage does the
same thing.
Actually, while I'm here, I may as well report that 'more' doesn't
work properly with 0.95c+, whereas it did with 0.95c. When running
on ttys1, pressing CR causes an incorrect display! (I have seen
it do this on the console too - but not as reliably as it does
on the serial line).
Cheers... Mark
[next article]
From: [email protected] (Zeyd M. Ben-Halim)
Subject: Re: Where's outb?
Message-ID: <[email protected]>
Date: 24 Apr 92 06:44:53 GMT
References: <[email protected]>
Sender: [email protected] (netnews admin account)
Organization: University of Denver, Dept. of Math & Comp. Sci.
Lines: 18
[email protected] writes:
>I've spent the last few nights installing 95a, Linus' "second" patch, and
>GCC 2.1, then trying to undo the mess. The result is that ld says undef
>for outb, get_fs_byte, and similar things when I build the kernel (ps for 2.1
>successfully compiled and linked.)
This happens because gcc fails to expand inline definitions for these functions.
This only happens when you don't use the top level Makefile. It set apprpriate
flags to allow this.
Go to /usr/src/linux and type make.
>I suppose a file is still in the wrong place or something but I don't
>know what and where.
>Regards,
>David Tucker @ATO / KC4ZGO / [email protected]
[next article]
From: [email protected] (Henning Spruth)
Subject: RE: Latin-1 again
Message-ID:
Date: 24 Apr 92 07:04:43 GMT
Organization: Technische Universitaet Muenchen
Lines: 12
I think this would be a good idea.
For the printer, the translation code ought to be done in the
lp/lpr program, and it should be able to disable it in order to
print graphics and such.
Henning
--
------------------------------------------------------------------------
Henning Spruth [email protected]
Institute of Electronic Design Automation
Technical University of Munich, Germany
[next article]
From: [email protected] (cm445a17)
Subject: Bootlin --> help please!!!!!!
Message-ID: <[email protected]>
Date: 24 Apr 92 07:22:39 GMT
Organization: University of Maryland University College
Lines: 7
I just installed bootlin.com on my machine as described on the docs but it
doesn't work! It reports that it loaded the boot sector, setup, and
kernel OK but the it prints some garbish characters and hangs!!!
BTW, my bootimage is set up for the right partion. Any ideas would be
appreciated. Thanks,
Jaime
[next article]
From: [email protected] (Hongjiu Lu -- Graduate Student)
Subject: Re: Patches for GCC v2.1 for Linux
Message-ID: <[email protected]>
Date: 24 Apr 92 06:17:53 GMT
References:
Sender: [email protected] (USENET News System)
Organization: Washington State University
Lines: 11
In article
ik Talvola) writes:
>Has the person who ported GCC v2.1 to Linux given the patches back to
>the FSF? It would be nice if GCC as shipped from the FSF had Linux
>support installed. They should be posted to the gnu.gcc.bug newsgroup,
Yes or No. I did send the patches to FSF just before gcc 2.1 was
released. It was too late to make into 2.1. But since then Linux has
changed a little bit. The one they got is different from what I have. I
want to wait until the multiple shared libs have been tested.
H.J.
[next article]
From: [email protected] (william.j.bitner)
Subject: gdb fails on ioctl??
Keywords: gdb ioctl
Message-ID: <[email protected]>
Date: 24 Apr 92 08:56:16 GMT
Organization: AT&T
Lines: 11
Hi. I just got gcc2.1 up and running, rebuilt the "new"
.96 kernel and I thought I'd throw in gdb. It's giving
me an error "ioctl TIOCSPGRP, not owner???" I tried
this both as root and as a user with no difference, did
I do something wrong.
Could really use some advice. 🙂
Thanks
Bill Bitner
[email protected]
[next article]
From: [email protected] (Paul Nulsen)
Subject: Re: HELP! Can't get GCC2.1 to run.
Message-ID: <[email protected]>
Date: 24 Apr 92 08:39:02 GMT
References:
Organization: University Of Wollongong
Lines: 22
[email protected] (Mark Tomlinson) writes:
>...
>The compiler (cc1) seems to run fine, the problem I am getting is
>in linking. Both the new ld and the ld.old give the same results,
>so I guess the problem may be in the library file. The error
>message I get is "can't resolve external __cbuff" or something
>of that effect. If I remove the "#include
>program (forcing the use of the library functions rather than
>the macros), the linking works, and so does the program.
>Anyone else found this problem? Anyone know what I'm doing wrong?
>...
Having just resolved this one for myself:
__cbuff is used in the kernel include file
does something else. You have managed to get your routine to search the
kernel include directories before searching the gcc ones when looking for
ctype.h. You just need to fix the search order in some manner.
Paul Nulsen
[email protected]
[next article]
From: [email protected] (Dave Stow)
Subject: FAQ answers, please!
Message-ID: <[email protected]>
Date: 24 Apr 92 09:23:05 GMT
Organization: Department of Computing at Lancaster University, UK.
Lines: 13
Hi all you linux people,
Could someone email me the answers to the FAQ's, as the list seems
to have expired/never existed on this machine!
I found a list of queries, which answered most of my immediate questions,
but answers there were none...
email to:
[email protected]
Cheers-a-lot
Dave
[next article]
From: [email protected] (& Stevens)
Subject: Re: Patches for GCC v2.1 for Linux
Message-ID:
Date: 24 Apr 92 10:31:19 GMT
References:
Sender: [email protected]
Organization: Dept. Engineering Science, Oxford University, UK
Lines: 17
In-reply-to: [email protected]'s message of 24 Apr 92 00:47:37 GMT
On the subject of GCC v2.1 on Linux has anyone else notices that fseek
does not seem to follow the ANSI standard.
The ANSI validated compiler and library that I use on another system
specifies:
Errors: fseek returns non-zero on error.
The GCC 2.1 library for linux returns the new file offset, this is
useful but not ANSI unless the validated compiler library is wrong.
Of course I is easy to fix the fseek code, but I don't want to have a
different one from every other Linuxer.
--
Michael Stevens, Robotics Research Group,Dept of Engineering Science,Oxford,UK
JANET: [email protected]
INTERNET: michael%[email protected]
[next article]
From: [email protected]
Subject: What about *other* sets (like Latin-3)?
Summary: Was: Latin-1 again
Message-ID: <[email protected]>
Date: 24 Apr 92 12:42:38 GMT
References: <[email protected]>
Sender: [email protected] (USENET News System)
Organization: Jus' me.
Lines: 3
Personally, I'd find Latin-3 more useful than Latin-1...
as would anyone else who wishes to correspond in Esperanto.
[next article]
From: [email protected] (thomas dunbar)
Subject: Emacs 18.58
Message-ID: <[email protected]>
Date: 24 Apr 92 12:39:44 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 6
i've been regularly using emacs v 18.57 with linux 0.95c+ with no problems
recently i tried using emacs v 18.58 and invariably after using it a
little it would generate a segmentation error. i've never had this
problem with other programs (except kermit 5a).
are others using v 18.58 successfully? any suggestions on whats
causing my problem?
[next article]
From: [email protected] (Stephen Gallimore)
Subject: Re: ln,mv, i_nlinks
Message-ID: <[email protected]>
Date: 24 Apr 92 13:27:14 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 34
A couple of people, including Linus, got back to me to say they did not
have my problem when moving over a file with more than one link. However
Linus took a look at the rename system call anyway and indeed found a bug
and told me to try the following in fs/minix/namei.c (do_minix_rename) :
|Yes, the link is correctly updated, but the inode isn't marked dirty.
|Try this:
|
|change:
|
| if (new_inode)
| new_inode->i_nlink--;
|
|to
|
| if (new_inode) {
| new_inode->i_nlink--;
| new_inode->i_dirt = 1;
| }
|
|and tell me if this corrects the problem? It will be corrected in the
|next release. Thanks - it wasn't an easy bug to spot (well, it was easy
|once I started looking, but...)
I made this change and mv now works correctly. I can only assume that there
are different versions of mv floating around, the one I use is the shared
lib version that came with gcc2.1 . Maybe there are versions (from 0.12?)
that do mv a b by unlink b,ln a b,unlink b rather than calling rename, or
that do unlink b, rename a b. Both of those would hide the bug in rename.
Stephen
:Stephen Gallimore:E-Mail [email protected]:Tel (+44) 532 335730:
[next article]
From: [email protected]
Subject: (none)
Message-ID: <[email protected]>
Date: 24 Apr 92 13:34:02 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 13
|>
|> I recently used shoelace to try & make linux boot. after no suc cess,
|> i want to go back. HOW DO I GET RID OF THIS BOOT THING THAT KEEPS
|> ASKING WHICH PARTITIAN TO BOOT?!?!?!?!?
|>
You may be able to rebuild the MBR using FDISK from DOS 5.0
it has a MBR command line switch. I've used to clean up many
a boot sector viruses.
FDISK /MBR
Hope this helps.
Bruce
[next article]
From: [email protected] (Gary Beihl)
Subject: dos emulation
Message-ID: <[email protected]>
Date: 24 Apr 92 12:51:08 GMT
Organization: NCR Corp
Lines: 23
Originator: beihl@qt1
Nntp-Posting-Host: qt1
I have a PC emulator that I am willing to share. It is not Intel
specific, so it will be significantly slower than v86 mode. It runs
on both big and little endian machines by interpreting each 8086
opcode. Video/keyboard emulation uses X, though there is a half-baked
tty-based mode as well. Only floppy drives are emulated and network
drives (which are probably the natural way to access the native
filesystem) are not supported. Mtools is used to transfer files
to/from the floppy image files. It boots various versions of MSDOS and
DRDOS, and with a little work on the interrupt controller could
probably boot MINIX. It would be nice to eventually boost the
performance by using incremental compilation (ala SoftPC) techniques.
I don't have access to a linux system, so if there is interest from
the linux community, I would ask for 1 or 2 victims (uh, I mean
testers 🙂 to compile and run this before uploading it for general
access. In case you haven't guessed, it is best characterized as
pre-alpha due to its lack of features. Don't think it will let you
run all your favorite DOS programs immediately. It still needs a lot
of work. Volunteers need to have their own copy of MSDOS, be capable
of producing an image file of a (DOS) bootable floppy disk and be able
to use mtools with this file.
Gary Beihl ([email protected])
[next article]
From: [email protected] (Kees J. Bot)
Subject: Re: Is Shoelace permanent? [was HELP! SHOELACE IS STRANGLING ME!!!]
Keywords: n
Message-ID: <[email protected]>
Date: 24 Apr 92 14:40:47 GMT
References: <[email protected]> <[email protected]> <199
[email protected]> <[email protected]>
Sender: [email protected]
Lines: 42
[email protected] (Nate Williams) writes:
>4) If you want to install it on your WHOLE drive, copy the old version
>of the master boot sector to a floppy so you can go restore it if you need
>to.
> Something like :
> $ dd if=/dev/hda of=/dev/floppy
> Do a
>floppy. Someone with a bit more knowledge than I can give you the block
>count and block size to be exact.
$ dd if=/dev/hda of=bootcode bs=446 count=1
This will save the bootstrap code without the partition table. If you
want the save the partition table too, then use 'bs=512'.
The layout of the master boot sector:
bytes 0 - 445: bootstrap code.
446 - 509: four 16 byte partition table entries.
510 - 511: boot sector signature: 0x55 0xAA.
The boot procedure of a PC is really simple, the BIOS loads the very
first sector from either a floppy or the hard disk at address 0x07C00
and jumps to it with the code for the boot device in the 'dl' register.
The hard disk master bootstrap replaces itself with the boot code found
in the first sector of the active partition, it then calls it with the
'dl' register still holding the boot device code, and registers 'es:si'
pointing to the partition table entry of the partition just booted.
The code now at 0x07C00 loads the operating system from either the floppy
or the hard disk partition. (Or just a program that is smart enough to
do that, like shoelace.)
By the way, some partition bootstraps use 'ds:si' to find the partition
table entry and others think that si offsets from segment 0. The
system V bootstrap ignores 'es:si' and assumes that the active partition
must be its boot partition. Egocentric junk.
--
Kees J. Bot ([email protected])
Systems Programmer, Vrije Universiteit Amsterdam
[next article]
From: [email protected] (Al Clark)
Subject: Re: HELP! Can't get GCC2.1 to run.
Message-ID: <[email protected]>
Date: 24 Apr 92 14:51:25 GMT
References:
Organization: Self
Lines: 50
In article
mlinson) writes:
>I guess someone's going to try and give me the standard reply
>"there's a problem in tar", but I don't think that is my problem.
>I uncompressed the archives and then un-tar'd them so I didn't
>need to use the 'z' option in tar (this has helped me before).
>
>The compiler (cc1) seems to run fine, the problem I am getting is
>in linking. Both the new ld and the ld.old give the same results,
>so I guess the problem may be in the library file. The error
>message I get is "can't resolve external __cbuff" or something
>of that effect. If I remove the "#include
>program (forcing the use of the library functions rather than
>the macros), the linking works, and so does the program.
>
>Anyone else found this problem? Anyone know what I'm doing wrong?
>
Just worked this problem this morning. I found that the 'ctype.h'
used by linux is different from the gnu ctype.h (I think that's it).
Anyway, I have the following two ctypes.h installed, and I can
compile the kernel, and also other programs (such as ps).
-rw-r--r-- 2 root other 4594 Feb 29 09:29 /usr/include/ctype.h
-rw-r--r-- 1 root root 1049 Nov 7 17:30 /usr/src/linux/include/ct
ype.h
Note the Linux Makefile uses -nostdinc (no standard include files) and
-I/usr/src/linux/include (effectively), so the latter one is used for the
kernel. The former one is used for other programs, such as ps. At least
it's working that way for me. Incidentally, can anyone tell me if there
are any other include files with two versions like this, which are waiting
to bite me???
>PS - I'm using the 0.95c+ bootimage. The 0.95c bootimage does the
>same thing.
>
>Actually, while I'm here, I may as well report that 'more' doesn't
>work properly with 0.95c+, whereas it did with 0.95c. When running
>on ttys1, pressing CR causes an incorrect display! (I have seen
>it do this on the console too - but not as reliably as it does
>on the serial line).
>
Don't use more, I much prefer less, so I can't help you.
>Cheers... Mark
>
--
Al - [email protected] - My opinions are my own.
*** Practice random kindness and senseless acts of beauty! ***
[next article]
From: [email protected] (Al Clark)
Subject: Re: Bootlin --> help please!!!!!!
Message-ID:
Date: 24 Apr 92 14:54:33 GMT
References: <[email protected]>
Organization: Self
Lines: 16
In article <[email protected]> [email protected] (
cm445a17) writes:
>I just installed bootlin.com on my machine as described on the docs but it
>doesn't work! It reports that it loaded the boot sector, setup, and
>kernel OK but the it prints some garbish characters and hangs!!!
>BTW, my bootimage is set up for the right partion. Any ideas would be
>appreciated. Thanks,
>
>Jaime
>
Please post the solution, I have the IDENTICAL problem. The only detail
I might add is that I couldn't unzip under Linux, I had to mwrite the
bootlin.zip file to MSDOS and unzip it there.
--
Al - [email protected] - My opinions are my own.
*** Practice random kindness and senseless acts of beauty! ***
[next article]
From: [email protected] (LCDR Michael E. Dobson)
Subject: Re: free (memory allocation table)
Message-ID: <[email protected]>
Date: 24 Apr 92 16:47:40 GMT
References: <[email protected]>
Organization: Naval Medical Research & Development Command
Lines: 16
In article <[email protected]> [email protected] (Peter Thomson) writes:
>
>
>Hi can some one tell me if the command free reports back any more than
>1Mbyte of memory. There is 2Mbytes on my machine but free states there is
>only one. running multiple tasks and watching ps confirms this and swapping
>starts at the 1Mbyte limit
>
> Thanks in advance Peter.
Free reports >2 Meg free on my system, 640K conventional, 3 Meg extended.
--
LCDR Mike Dobson, Sys Admin for | Internet: [email protected]
nmrdc1.nmrdc.nnmc.navy.mil | UUCP: ...uunet!mimsy!nmrdc1!rdc30
AT&T 3B2/600G Sys V R 3.2.3 | BITNET: dobson@usuhsb or nrd0mxd@vmnmdsc
WIN/TCP for 3B2 R3.2 | MCI-Mail: 377-2719 or [email protected]
[next article]
From: [email protected] (Joel M. Hoffman)
Subject: Re: Bootlin --> help please!!!!!!
Message-ID: <[email protected]>
Date: 24 Apr 92 16:29:37 GMT
References: <[email protected]>
Sender: [email protected] (USENET News system)
Organization: University of Maryland, College Park
Lines: 36
Nntp-Posting-Host: next.wam.umd.edu
In article
>In article <[email protected]> [email protected]
(cm445a17) writes:
>>I just installed bootlin.com on my machine as described on the docs but it
>>doesn't work! It reports that it loaded the boot sector, setup, and
>>kernel OK but the it prints some garbish characters and hangs!!!
>>BTW, my bootimage is set up for the right partion. Any ideas would be
>>appreciated. Thanks,
>>
>>Jaime
>>
>
>Please post the solution, I have the IDENTICAL problem.
I just started using bootlin.com. I also got gibberish, but only after
the Kernel was loaded. I was told:
[gibberish here] Virtual Consoles
The gibberish is IBM ASCII and not reproducable here. However,
everything seems to work.
But I just had a stranger thought. When I saw the nonsense, I hit
SHIFT-PRTSCN, and it worked (almost). I have the printout in front of
me that has on it:
BootSector: ... loaded
LINUX Setup: ... loaded
LINUX Kernel: ............done
[gibberish]
I know I didn't press PRTSCN until after I saw a message about VC's,
so shouldn't Linux already have been running, and PRTSCN have been
disabled??
-Joel
([email protected])
[next article]
From: [email protected] (Jim Winstead Jr.)
Subject: Re: Bootlin --> help please!!!!!!
Message-ID: <[email protected]>
Date: 24 Apr 92 16:52:13 GMT
References: <[email protected]>
Sender: [email protected] (The News System)
Organization: Harvey Mudd College, WIBSTR
Lines: 13
In article <[email protected]> [email protected] (
cm445a17) writes:
>I just installed bootlin.com on my machine as described on the docs but it
>doesn't work! It reports that it loaded the boot sector, setup, and
>kernel OK but the it prints some garbish characters and hangs!!!
Are you sure it hangs? On my system, it prints the gibberish
characters instead of the "Press
but I can just press a key and it boots regardless....
--
Jim Winstead Jr. (CSci '95) | "Catch a fish!"
Harvey Mudd College | -Geddy Lee,
[email protected] | San Diego Sports Arena
Disclaimer: Mine, not theirs! | January 20, 1992
[next article]
From: [email protected] (Joel M. Hoffman)
Subject: Re: Dos emulator (tried this by mail; didn't work)
Message-ID: <[email protected]>
Date: 24 Apr 92 17:39:29 GMT
Sender: [email protected] (USENET News system)
Organization: University of Maryland, College Park
Lines: 10
Nntp-Posting-Host: next.wam.umd.edu
In article <[email protected]> you write:
>I have a PC emulator that I am willing to share. It is not Intel
I'd be willing to give it a shot. Is the code FTP'able? Or is it
small enough for you to mail?
Let me know.
-Joel
([email protected])
[next article]
From: [email protected] (Rajat Datta)
Subject: Re: gdb fails on ioctl??
Keywords: gdb ioctl
Message-ID: <[email protected]>
Date: 24 Apr 92 16:35:55 GMT
References: <[email protected]>
Sender: [email protected] (NNTP News Poster)
Organization: IBM T.J. Watson Research Center
Lines: 14
Disclaimer: This posting represents the poster's views, not necessarily those of
IBM
Nntp-Posting-Host: shravani.watson.ibm.com
I'm getting the same problem. It's related to the under debugging
process doing a read from the tty. I guess gdb is doing TIOCSPGRP to
set the tty process group to be the same as the process under debug to
avoid the stopped input message, but the tty ioctl code is rejecting
it because the current process (gdb) does not have the same process
group as the one we're setting the tty to (the under debug process).
I would appreciate a solution. Except for this, gdb works fine under
95c++ and is an indispensable tool.
--
rajat ([email protected])
[next article]
From: [email protected] (gary a moyer)
Newsgroups: alt.os.linux,comp.os.linux
Subject: roff or the likes?
Summary: Sources anywhere via anonymous ftp?
Keywords: roff troff nroff
Message-ID: <[email protected]>
Date: 24 Apr 92 17:39:56 GMT
Sender: [email protected]
Lines: 4
I am looking for roff (like groff less all the printer specific stuff)
or something that will do simple document formatting (paragraph formatting,
on-the-fly column width adjustment, etc). Either binaries or sources.
thx. [email protected]
[next article]
From: [email protected] (Joel M. Hoffman)
Subject: curses problems (fwopen and vsscanf)
Message-ID: <[email protected]>
Date: 24 Apr 92 18:21:09 GMT
Sender: [email protected] (USENET News system)
Organization: University of Maryland, College Park
Lines: 15
Nntp-Posting-Host: next.wam.umd.edu
While trying to compile gnuchess, I ran into the following undefined
symbols in libcurses.a:
_fwopen (from printw)
_vsscanf (from scanw)
I build libcurses.a from the curses.tar.Z file on tsx. I also
commented out the ``HZ'' bit from the termcap capabilities; I think
it's defined as a macro somewhere else. I doubt the two are related,
though.
Any ideas?
-Joel
([email protected])
[next article]
From: [email protected] (Rik Faith)
Subject: Re: Graphics and IPC questions...
Message-ID: <[email protected]>
Date: 24 Apr 92 14:23:13 GMT
References:
uiuc.edu>
Sender: [email protected]
Organization: UNC-Chapel Hill Computer Science
Lines: 23
In article <[email protected]>, chojnows@osiris (Brian Cho
jnowski) writes:
> I guess I havent made myself clear. I would like to do something like this,
> where I would create a superset of the vt100. But the mode as it is, is
> insufficient for the kind of graphics I am talking about. I want to also add
> this new mode to the telnet program (when it gets written), so that the
> graphics get displayed remotely.
Instead of creating a new vt100 superset, have you considered implementing a
tektronix emulation mode on top of the vt100 emulator? Kermit for the PC has
this feature, as does xterm for X Windows (Tektronix 4014). Several other
programs which emulate the vt100 also have tektronix emulation.
Many graphics programs (like gnuplot and matlab [which is commercial]) support
tektronix emulation. There should be a well documented description of this
mode and how it interacts with the vt100 mode. This way, you wouldn't have to
re-invent the wheel, and there is a base of programs already out there which
could take advantage of the new feature. [Actually, I've thought about doing
this myself, and would be interested in joining the project if you decide to
take this route.]
--
Rik Faith: [email protected]
Paradox is the question of Chaos.
[next article]
From: [email protected] (Scott R. Johnson)
Subject: Linux: Multi-terminal support??
Message-ID: <[email protected]>
Date: 24 Apr 92 18:20:10 GMT
Organization: Key Computer Labs, Fremont, CA
Lines: 19
Folks,
I'm a Linux newbie. (gawd, they're hauling out the tomatoes now...)
Can Linux be set up to allow multiple terminals, say, VT100s, to log in
off the serial ports?
If so, how hard would it be to hack up, say 4 to 6 ports (I know that the PC
spec only allows 2 ports, at 1F0 and 1F8, but the two additional at 1E0 and
1E8 are pretty much de-facto now; they have to share the interrupts with
the first two ports) so a PC could be used as a mini timeshare system?
Thanks a lot,
--
--srj Scott R. Johnson [email protected]
o_O Amdahl Advanced Systems Development
=( ) = 46525 Landing Parkway, M/S 580
U Ack! Fremont, CA 94538 510-623-2333
[next article]
From: [email protected]
Subject: (pick one)
Message-ID: <[email protected]>
Date: 24 Apr 92 20:14:22 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 17
I would like to help contribute some, but my C sucks. I have tried porting
things, but when they work, they don't do anything useful, and when they don't
it is beyond me. 🙂
I think I would like to (try to) keep track of who is porting what, and results
(how it is coming along....) If this is being done already let me know, if not
let me know. (I would still like to know who is porting what, in either case)
I figure that if at least one person knows who is doing what, then before starti
ng
from scratch again, someone who wants to port something can check with the perso
n
who knows. That way might avoid some frustration, and maybe bring porters
together on projects.
I would also like to volunteer to test any pre-alpha software anyone might have
ported but would like tested before beta release.
Thanks, Jim Gifford
Eastern Shore of VA.
[email protected] ...!uunet!attmail.com!jgifford
[next article]
From: [email protected] (Hongjiu Lu -- Graduate Student)
Subject: README: fseek
Message-ID: <[email protected]>
Date: 24 Apr 92 17:39:27 GMT
References:
Sender: [email protected] (USENET News System)
Organization: Washington State University
Lines: 16
In article
ots (& Stevens) writes:
>On the subject of GCC v2.1 on Linux has anyone else notices that fseek
>does not seem to follow the ANSI standard.
>
It is a bug and has been fixed in the next release, which will be out
shortly after 0.96 is released.
If you find a bug in libc.a, please send your bug report to
"[email protected]", along with a short prograam to show the bug. If you
can provide a fix, that will be great, but not required.
BTW, the source code is in libc-2.1.tar.Z, which can be found where
you got gcc 2.1.
H.J.
[next article]
From: [email protected] (Dennis Flaherty)
Subject: new SVGA_MODE patch again
Message-ID: <[email protected]>
Date: 24 Apr 92 21:03:35 GMT
Organization: Miles, Inc. Diagnostics Divis.
Lines: 131
OK, here we go again: :-,
Here's version 3 of my SVGA_MODE patch. In order to support
presetting BOTH the EGA/VGA mode AND ALL SVGA modes, I had to
make the table 1-based. This means that if you are used to
typing 2, you'll have to type 3 now. Sorry.
I've extensively tested my code on my machine, including all
boundary conditions, and it works for me. Let me know if it
doesn't work for you.
Dennis
*** Makefile~ Tue Apr 21 08:27:33 1992
--- Makefile Fri Apr 24 13:54:09 1992
***************
*** 44,49 ****
--- 44,56 ----
AS86 =as86 -0 -a
LD86 =ld86 -0
+ #
+ # If you want to preset the SVGA mode, uncomment the next line and
+ # set SVGA_MODE to whatever number you want. The table is 1-based.
+ # Set it to -DSVGA_MODE=VGA if you just want the EGA/VGA mode.
+ #
+ #SVGA_MODE= -DSVGA_MODE=3
+
AS =as
LD =ld
#LDFLAGS =-s -x -M
***************
*** 130,136 ****
$(LD86) -s -o boot/setup boot/setup.o
boot/setup.s: boot/setup.S include/linux/config.h
! $(CPP) -traditional boot/setup.S -o boot/setup.s
boot/bootsect.s: boot/bootsect.S include/linux/config.h
$(CPP) -traditional boot/bootsect.S -o boot/bootsect.s
--- 137,143 ----
$(LD86) -s -o boot/setup boot/setup.o
boot/setup.s: boot/setup.S include/linux/config.h
! $(CPP) -traditional $(SVGA_MODE) boot/setup.S -o boot/setup.s
boot/bootsect.s: boot/bootsect.S include/linux/config.h
$(CPP) -traditional boot/bootsect.S -o boot/bootsect.s
*** boot/setup.S~ Tue Apr 21 08:24:53 1992
--- boot/setup.S Fri Apr 24 14:00:00 1992
***************
*** 245,250 ****
--- 245,251 ----
mov es,ax
lea si,msg1
call prtstr
+ #ifndef SVGA_MODE
flush: in al,#0x60 ! Flush the keyboard buffer
cmp al,#0x82
jb nokey
***************
*** 256,264 ****
--- 257,268 ----
ja nokey
cmp al,#0x9c
je svga
+ #endif
+ #if !defined(SVGA_MODE) || SVGA_MODE == VGA
mov ax,#0x5019
pop ds
ret
+ #endif
svga: cld
lea si,idati ! Check ATI 'clues'
mov di,#0x31
***************
*** 478,483 ****
--- 482,488 ----
push bx
mov al,bl
sub al,cl
+ inc al
call dprnt
call spcing
lodsw
***************
*** 496,512 ****
lea si,msg3
call prtstr
pop si
! add cl,#0x80
nonum: call getkey
cmp al,#0x82
jb nonum
- cmp al,#0x8b
- je zero
cmp al,cl
ja nonum
! jmp nozero
! zero: sub al,#0x0a
! nozero: sub al,#0x80
dec al
xor ah,ah
add di,ax
--- 501,517 ----
lea si,msg3
call prtstr
pop si
! add cl,#0x81
! #if defined(SVGA_MODE) && SVGA_MODE != VGA
! mov al,#SVGA_MODE ! Preset SVGA mode
! #else
nonum: call getkey
cmp al,#0x82
jb nonum
cmp al,cl
ja nonum
! sub al,#0x81
! #endif
dec al
xor ah,ah
add di,ax
--
Dennis T. Flaherty [email protected]
Miles, Inc. [email protected]
Diagnostics Division
My doctor says Mylanta, but my boss says Alka-Seltzer!
[next article]
From: [email protected]
Subject: Mail Woes
Message-ID: <[email protected]>
Date: 24 Apr 92 21:16:52 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 10
X-Mn-Key: normal
Hi all:
I see that another person is also having mysterious mail problems.
When I post to the list, I see it come out on alt.os.linux but
not the digest! This may be related to switching the newsgroups.
This problem started two weeks ago.
Anyone else caught in a black hole?
jim wiegand [lost in space]
[next article]
From: chojnows@osiris (Brian Chojnowski)
Subject: Re: Graphics and IPC questions...
Message-ID: <[email protected]>
Date: 23 Apr 92 17:44:02 GMT
References:
uiuc.edu> <[email protected]>
Sender: [email protected] (Net Noise owner)
Organization: University of Illinois at Urbana
Lines: 22
Ok, maybe no one understands me. By characterset graphics, I mean that I
can have a group of 1-256 images that can be linearly accessed. (a simple
job if they are al the same size). The charset data sits on the disk, and
when requested, the terminal would open up a number of these. This has one
big advantage, in that you do not have to represent letters with the
charset. You could write some pretty fancy grphics games (even something
like say Bards Tale) just by creating the wall types in a charset, and
then printing them out. Remotely, the charset would have to be downloaded as
it is opened, but most of them will likely be <10k bytes. Then the remote
terminal becomes responible for putting the graphics on the screen.
You can also create a program based on say a 80x40 screen, and then load an
appropriate charset based on the screen type you have. 4x5 for CGA, 8x8 for
ega, 8x12 for vga, 10x15 for 800x600, etc. I have already done all this
under borland C, and I am in the middle of writting a very large real time
rolmaster based multi-player dungeon game. This is what I am trying to do.
ps. 'someone mentioned extending the vt100 codes to do this, that is what "A
superset of" means.
Brian Chojnowski
[email protected]
[next article]
From: [email protected] (Joe Smith)
Subject: has anyone compiled pre-0.96 with ps patches?
Message-ID:
Date: 24 Apr 92 18:52:56 GMT
Sender: [email protected]
Distribution: comp
Organization: University of Pennsylvania, Philadelphia, PA
Lines: 22
Nntp-Posting-Host: mbio.med.upenn.edu
I tried it last night: all but one of the patches succeeded (and I'm
not sure the failed one is needed anymore), and the ps, etc. binaries
seem to work ok, but the resulting kernel won't run binaries that use
the shared library (they die immediately with an 'invalid operand'
exception, unless you are root, in which case they work fine! ???).
The stock pre-0.96 kernel compiled and runs fine.
On a related note, what are the arguments to 'ps U'? I can guess
about /usr/src/linux/tools/system, but what's /dev/swap? My swap
partition is /dev/hda3 - which ps won't accept ('bad swap signature').
It accepts /dev/swap, but there is no such file. I'll take a look at
the ps source tonight - is it described in there somewhere?
--
Joe Smith
University of Pennsylvania [email protected]
Dept. of Biochemistry and Biophysics (215) 898-8348
Philadelphia, PA 19104-6059
[next article]
From: [email protected] (Linus Benedict Torvalds)
Subject: Re: gdb fails on ioctl??
Keywords: gdb ioctl
Message-ID: <[email protected]>
Date: 24 Apr 92 21:53:31 GMT
References: <[email protected]> <1992Apr24.163555.26570@
watson.ibm.com>
Organization: University of Helsinki
Lines: 27
In article <[email protected]> [email protected] (Rajat D
atta) writes:
>I'm getting the same problem. It's related to the under debugging
>process doing a read from the tty. I guess gdb is doing TIOCSPGRP to
>set the tty process group to be the same as the process under debug to
>avoid the stopped input message, but the tty ioctl code is rejecting
>it because the current process (gdb) does not have the same process
>group as the one we're setting the tty to (the under debug process).
Right. The easy fix if to comment out the session testing in
tty_ioctl.c: right after the "case TIOCSPGRP:" comment out the two
lines
if (session_of_pgrp(pgrp) != current->session)
return -EPERM;
and gdb should be happy. The session_of_pgrp() code has problems (I
think I've fixed them now, though).
>I would appreciate a solution. Except for this, gdb works fine under
>95c++ and is an indispensable tool.
My version already has this fixed (not the above quick hack - another
not-quite-as-ugly hack that might even be correct), and the real 0.96
(out next week? maybe) won't have the above problems. In the meantime,
just dike out the two lines.
Linus
[next article]
From: [email protected] (Oren Kastner)
Subject: Linux Newbie - Advice needed... (Long, sorry)
Summary: Installation
Keywords: n
Message-ID: <[email protected]>
Date: 24 Apr 92 21:47:59 GMT
References: <[email protected]> <[email protected]
ana.edu> <[email protected]>
Sender: [email protected] (USENET News System)
Organization: Brandeis University
Lines: 70
On the third day of my Linux life, I have a (hopefully) trivial question.
I started by testing Linux with boot/root floppies. I liked it so much, I
repartitioned my disk and gave it 15 megs (loosing everything on my 130meg HD
in the process -- I thought it worth it, so I did it 🙂 I installed as much
software that I could find which was relatively small and pain free -- utils,
xcomm, pcomm, fileutils, etc. I was able to use kermit, and was really getting
into the flow of things, so I thought I'd start installing "heavy" stuff --
like GCC 2.1 and some sources. I figgered I'd like more space for Linux, so I
repartitioned my disk again:
DOS primary partion (C:) == 41 megs
DOS extended partion (D:) == 51 megs
Linux partition == 38 megs
I read the FAQ again, and noticed the info about making a swap file/partition.
So, I junked my second installation (the 38 megs one) and planned on repartion-
ing the Linux partition into 8(swap), 8(root), 22(usr) -- trying to stay as
close as possible to the FAQ 🙂 I used edpart, but it could only create 2 other
partition (and I needed three), so I figured I'd create an 8 meg partition for
swap-space (double my 4meg RAM) and keep root and usr together as a 30 meg
partition. Things were peachie. I mkswap'ed and mkfs'ed and matter of factly
I'm *ON* my setup right now (again, using kermit to dialup my account). The
gcc installation, however, has been both painless and painful. I uncompressed
and untarred in and things are sitting in /usr/... (the files were: 2.1libc,
2.1lib, 2.1misc, 2.1shared-A). I also uncompressed/tarred the 2.1ps tar file.
(Note: I got all these from tsx-11.mit.edu).
Now, here is what misunderstanding is of... I thought of using the
mcc-interim installation, as it seemed as though it would be painless. From
what I remember reading about it, though, I would have to have it install
itself on the whole HD, or something like that... My patitions:
/dev/hda3 == 8megs swap-space
/dev/hda4 == 30megs root, etc.
Can I have mcc install the two disks *-US disks and the two compiler disks on
/dev/hda4 painlessly? Or should I stick with what I have right now? That is,
I'd like to do the "mcc thing" :), but if I can't have it install things
correctly on /dev/hda4, I'll have to stick to what I have (I'm sick of restoring
my DOS partitions :). Any help would be appreciated (before I go ahead and
transfer about 5 megs of interim stuff (at 2400 baud, no less!)
Finally, sorry if this seems like an RTFM posting. I've read the FAQs and
everything else, I got things to install... but since I'm not sure if things
are installed in the right place, I'd like to use mcc-interim to do it for me.
I could have picked someone prominent in this news group and mailed him/her the
question, but I thought maybe someone else might be in my position and would be
interested in this (yeah, right! 🙂
Thanks, guys/gals (Linus, Ted, etc...) for Linux and any help would be
appreciated. It has certainly been an eventful 3 days (well... 2.5) days!!!
Great job on Linux.
BTW, if anyone cares (and thinks of a better way to set up Linux, my system
"configuration" (hardware wise) is:
ZEOS 386/33
5.25, 3.5 floppy drives
130 MB HD
SVGA
Thanks again,
Oren.
--
Oren Kastner E-mail address:
Brandeis University [email protected]
[next article]
From: [email protected] (Michael Pereckas)
Subject: put_page: page already exists
Message-ID:
Date: 24 Apr 92 23:25:00 GMT
Sender: [email protected] (News)
Organization: University of Illinois at Urbana
Lines: 30
I keep getting the message
put_page: page already exists
This usually but not always immediately precedes a longer message about
a divide error or a segmentation fault. There are two basic
situations:
1. I use some normal, simple program like ls or less, I get the
put_page message, a divide or segmentation error, and the login:
prompt. Sometimes the program runs, producing the expected output but
formatted incorrectly, like a CR and LF are needed but only the LF is
used, and then I get the login: prompt.
2. I run "make" in /usr/src/linux 🙁
This is a sure way to get the put_page message. Often many put_page
messages will be emitted over a period of time before the
semi-inevitable divide/segmentation error, which leaves be back at the
shell prompt (NOT login) Make sometimes does this on other programs,
but the kernel is the surest source of trouble.
What does put_page: page already exists mean? Does any one else get
it?
I am now running 0.95c+, but I had this problem with 0.95a and 0.95c
also. I have a very old 20 MHz 386 with only 2 MBytes of ram. I have
two MFM disks, with Linux using hdb1 for root, hdb4 for /usr, and hdb3
for an 8 MByte swap partition.
Michael Pereckas
[next article]
From: [email protected] (Ed Carp)
Subject: Re: linux
Message-ID: <[email protected]>
Date: 24 Apr 92 21:18:44 GMT
References:
Organization: Bleeding Edge Technologies - 'Sliding Down The Razor Blade Of Life
...'
Lines: 18
In article
>1) Does Linux have *any* working mail program? I can't seem to find any.
>2) How do I transfer my news and mail feeds over to Linux (I'd first need
> a mail program, of course, plus readnews and uucp).
uucp was posted a couple of weeks ago to tsx-11.mit.edu. I included smail3,
which works for me.
You'll notice a major problem if you try and send or receive files above
2400 baud or so. Linus tells me that this is because he turns off all
other interrupts when a disk access is preformed. Prehaps it's a serial
driver problem as well...
--
Ed Carp N7EKG/6 [email protected] [email protected]
Cupertino, CA 415/336-0797
-- Absolutely unabashed Gates McFadden groupie! --
[next article]
From: [email protected] (n.h.chandler)
Subject: Re: put_page: page already exists
Summary: I get put_page: too!
Message-ID: <[email protected]>
Date: 25 Apr 92 00:21:15 GMT
References:
Organization: AT&T
Lines: 27
In article
writes:
>
> I keep getting the message
> put_page: page already exists
>
...
> What does put_page: page already exists mean? Does any one else get
> it?
I get this message too, but with no adverse effects.
>
> I am now running 0.95c+, but I had this problem with 0.95a and 0.95c
I am also running 0.95c+
> also. I have a very old 20 MHz 386 with only 2 MBytes of ram. I have
> two MFM disks, with Linux using hdb1 for root, hdb4 for /usr, and hdb3
> for an 8 MByte swap partition.
Same here (almost) 20 MHz 386 with 8 MBytes of ram, 20MB on hda2 of ESDI
disk for linux root, and 30MB /dev/hdc7 on /home with 8MB /dev/hdc8 for
swap partition. /dev/hdc is a 105MB Quantum Hardcard II XL installed at
the secondary HD addresses.
Neville H. Chandler
[email protected]
[next article]
From: [email protected] (Anthony Starks)
Subject: MGR and Linux
Message-ID: <[email protected]>
Date: 25 Apr 92 00:35:01 GMT
Organization: Merck Sharp & Dohme Research Laboratories
Lines: 2
Has anyone considered porting the MGR window system to Linux
as a less-bloated alternative to X?
[next article]
From: [email protected] (Ishwar Rattan)
Subject: mcc-interim(0.95c+) help!
Keywords: mcc-interim, vi, '>' at shell level
Message-ID: <[email protected]>
Date: 25 Apr 92 01:03:09 GMT
Followup-To: [email protected]
Organization: Western Michigan Univ. Comp. Sci. Dept.
Lines: 12
i had success in installing mcc-interim on Zenith 386sx(4Mb ram,
40Mb IDE hdisk). i have two minor problems: (both look like
terminal problems):
1. '>' seen on the display iff pressed twice, and
2. in vi the arrow keys only move a one line window at the
bottom of the display!!
any help will be appreciated.
BTW has anybody bale to compile zip/unzip/zoo under linux?
- ishwar([email protected])
[next article]
From: [email protected] (Hongjiu Lu)
Subject: Re: curses problems (fwopen and vsscanf)
Message-ID: <[email protected]>
Date: 24 Apr 92 23:02:39 GMT
References: <[email protected]>
Sender: [email protected] (USENET News System)
Organization: Washington State University
Lines: 22
In article <[email protected]>, [email protected] (Joel M. Hoffm
an) writes:
|> While trying to compile gnuchess, I ran into the following undefined
|> symbols in libcurses.a:
|>
|> _fwopen (from printw)
|> _vsscanf (from scanw)
|>
|> I build libcurses.a from the curses.tar.Z file on tsx. I also
|> commented out the ``HZ'' bit from the termcap capabilities; I think
|> it's defined as a macro somewhere else. I doubt the two are related,
|> though.
|>
|> Any ideas?
|>
|> -Joel
|> ([email protected])
There are bugs in curses.tar.Z. The libcurses.a is part of gcc 2.1. I think
those bugs have been fixed. I hope curses.tar.Z will be removed very soon.
H.J.
[next article]
From: [email protected] (David Fenyes)
Subject: Re: Porting to Linux (Jove question)
Summary: Here's what I did
Keywords: jove porting signal
Message-ID: <[email protected]>
Date: 25 Apr 92 00:39:51 GMT
References: <[email protected]>
Sender: [email protected]
Organization: University of Texas Medical School at Houston
Lines: 39
Nntp-Posting-Host: thesis1.med.uth.tmc.edu
In article <[email protected]> [email protected] (Bennett
E. Todd) writes:
>elle(1) wasn't too hard, but jove(1) has me stalled out.
>
>In sysdep.h, the jove release has support for two styles of signals. It
>looks like Linux implements a third sort. Is there a good book to get
>for docs on the Posix interfaces that Linux supports?
>
>Any ideas?
>
>-Bennett (curmudgeon)
>[email protected]
Maybe one or two 🙂
How about --
# if defined SIG_BLOCK && defined SIG_UNBLOCK
# define SigHold(s) sigprocmask(SIG_BLOCK,s,NULL)
# define SigRelse(s) sigprocmask(SIG_UNBLOCK,s,NULL)
# endif
This is from my sysdep.h file. I've had a working Jove 4.14.7
port for over a month, but it uses screen control maneuvers that
are buggy as of the Linux 0.95 release. I have a Jove on the school
hpux system which I use via modem. I've never had any problems when
using Telix on DOS, but I have the same screen problems using the Jove
with kermit on linux as I have when using Jove on my linux box.
Nevertheless, linuxers with non MDA cards may not have the problems.
I'll follow with a post of the cdiffs to make jove work with Linux.
If anyone wants, I can also post the sources to tsx-11.
Hope this helps :-),
David.
--
David Fenyes [email protected]
University of Texas Medical School Houston, Texas
[next article]
From: [email protected] (David Fenyes)
Subject: Jove diffs listing
Keywords: jove porting signal
Message-ID: <[email protected]>
Date: 25 Apr 92 00:43:44 GMT
References: <[email protected]> <[email protected]>
Sender: [email protected]
Organization: University of Texas Medical School at Houston
Lines: 606
Nntp-Posting-Host: thesis1.med.uth.tmc.edu
Hello,
Here are the cdiffs for making Jove 4.14.7 compile nicely on Linux
(as of 0.95 + a patch or two 🙂 It has all the job control stuff,
etc.
enjoy,
David.
--
David Fenyes [email protected]
University of Texas Medical School Houston, Texas
-------------------- cut here ------------------------------------------
*** ../jove/Makefile Sun Aug 11 03:35:47 1991
--- Makefile Sun Mar 22 18:55:49 1992
***************
*** 30,49 ****
TMPDIR = /tmp
RECDIR = /usr/preserve
! JOVEHOME = /local
SHAREDIR = $(JOVEHOME)/lib/jove
LIBDIR = $(JOVEHOME)/lib/jove
BINDIR = $(JOVEHOME)/bin
MANDIR = $(JOVEHOME)/man/man$(MANEXT)
MANEXT = 1
! DFLTSHELL = /bin/csh
# to install executable files
! XINSTALL=cp
! #XINSTALL=install -c -m 755 # -s
# to install text files
! TINSTALL=cp
! #TINSTALL=install -c -m 644
PROG = jove
VERSION = 4.14
--- 30,51 ----
TMPDIR = /tmp
RECDIR = /usr/preserve
! JOVEHOME = /usr/local
SHAREDIR = $(JOVEHOME)/lib/jove
LIBDIR = $(JOVEHOME)/lib/jove
BINDIR = $(JOVEHOME)/bin
MANDIR = $(JOVEHOME)/man/man$(MANEXT)
MANEXT = 1
! DFLTSHELL = /bin/sh
# to install executable files
! #XINSTALL=cp
! XINSTALL=install -c -m 755 # -s
# to install text files
! #TINSTALL=cp
! TINSTALL=install -c -m 644
! #NROFF=nroff
! NROFF=groff -T ascii
PROG = jove
VERSION = 4.14
***************
*** 270,276 ****
$(TINSTALL) doc/teach-jove $(TEACH-JOVE)
doc/cmds.doc: doc/cmds.doc.nr doc/jove.4 doc/jove.5
! nroff doc/cmds.doc.nr doc/jove.4 doc/jove.5 > doc/cmds.doc
$(CMDS.DOC): doc/cmds.doc
$(TINSTALL) doc/cmds.doc $(CMDS.DOC)
--- 272,278 ----
$(TINSTALL) doc/teach-jove $(TEACH-JOVE)
doc/cmds.doc: doc/cmds.doc.nr doc/jove.4 doc/jove.5
! $(NROFF) doc/cmds.doc.nr doc/jove.4 doc/jove.5 > doc/cmds.doc
$(CMDS.DOC): doc/cmds.doc
$(TINSTALL) doc/cmds.doc $(CMDS.DOC)
Only in doc: cmds.doc
*** ../jove/externs.h Wed Jul 3 06:27:24 1991
--- externs.h Sun Mar 22 16:56:39 1992
***************
*** 132,139 ****
--- 132,141 ----
struct timeval; /* forward declaration preventing prototype scoping */
#endif
+ #ifdef BSD
extern int UNMACRO(select) proto((int /*width*/, long */*readfds*/, long */
*writefds*/,
long */*exceptfds*/, struct timeval */*timeout*/));
+ #endif
extern void UNMACRO(bcopy) proto((UnivConstPtr, UnivPtr, size_t));
extern void UNMACRO(bzero) proto((UnivPtr, size_t));
*** ../jove/iproc-ptys.c Mon Aug 12 13:04:59 1991
--- iproc-ptys.c Sun Mar 22 18:45:58 1992
***************
*** 48,54 ****
}
n = read(fd, (UnivPtr) ibuf, sizeof(ibuf) - 1);
! if (n == -1 && (errno == EIO || errno == EWOULDBLOCK)) {
if (proc_state(p) == NEW)
return;
proc_close(p);
--- 48,58 ----
}
n = read(fd, (UnivPtr) ibuf, sizeof(ibuf) - 1);
! if (n == -1 && (errno == EIO
! #ifdef EWOULDBLOCK
! || errno == EWOULDBLOCK
! #endif
! )) {
if (proc_state(p) == NEW)
return;
proc_close(p);
***************
*** 119,149 ****
--- 123,175 ----
void
ProcEof()
{
+ #ifdef SYSV
+ send_p(sg[NO].c_cc[VEOF]);
+ #else
send_p(tc[NO].t_eofc);
+ #endif
}
void
ProcInt()
{
+ #ifdef SYSV
+ send_p(sg[NO].c_cc[VINTR]);
+ #else
send_p(tc[NO].t_intrc);
+ #endif
}
void
ProcQuit()
{
+ #ifdef SYSV
+ send_p(sg[NO].c_cc[VQUIT]);
+ #else
send_p(tc[NO].t_quitc);
+ #endif
}
void
ProcStop()
{
+ #ifdef SYSV
+ send_p(sg[NO].c_cc[VSUSP]);
+ #else
send_p(ls[NO].t_suspc);
+ #endif
}
void
ProcDStop()
{
+ #ifdef SYSV
+ # ifdef VDSUSP
+ send_p(sg[NO].c_cc[VDSUSP]);
+ # endif
+ #else
send_p(ls[NO].t_dsuspc);
+ #endif
}
private void
***************
*** 203,209 ****
--- 229,239 ----
#ifdef BRLUNIX
struct sg_brl sgt;
#else
+ # ifdef SYSV
+ struct termios sgt;
+ # else
struct sgttyb sgt;
+ # endif
#endif
#ifdef TIOCGWINSZ
***************
*** 322,336 ****
#endif
sgt = sg[NO];
sgt.sg_flags &= ~(ECHO | CRMOD | ANYP | ALLDELAY | RAW | LCASE |
CBREAK | TANDEM);
(void) stty(0, &sgt);
!
{
int on = 1;
(void) ioctl(0, TIOCREMOTE, (UnivPtr) &on);
}
!
i = getpid();
(void) ioctl(0, TIOCSPGRP, (UnivPtr) &i);
#ifdef SYSV
--- 352,371 ----
#endif
sgt = sg[NO];
+ #ifdef SYSV
+ sgt.c_iflag = sgt.c_oflag = 0;
+ (void) ioctl(0,TCSETAW, &sgt);
+ #else
sgt.sg_flags &= ~(ECHO | CRMOD | ANYP | ALLDELAY | RAW | LCASE |
CBREAK | TANDEM);
(void) stty(0, &sgt);
! #endif
! #ifdef TIOCREMOTE
{
int on = 1;
(void) ioctl(0, TIOCREMOTE, (UnivPtr) &on);
}
! #endif
i = getpid();
(void) ioctl(0, TIOCSPGRP, (UnivPtr) &i);
#ifdef SYSV
*** ../jove/jove.c Sat Dec 28 05:30:17 1991
--- jove.c Sun Mar 22 18:45:58 1992
***************
*** 67,72 ****
--- 67,76 ----
private char switchar proto((void));
#endif
+ #ifdef ffs
+ #undef ffs
+ #endif
+
private void
ttyset proto((bool n)),
UnsetTerm proto((char *)),
***************
*** 268,274 ****
char *Inputp = NULL;
! #ifndef SYSVR4 /* declared (differently!) in
extern int UNMACRO(ffs) proto((long));
#endif
--- 272,278 ----
char *Inputp = NULL;
! #if !defined SYSVR4 && !defined linux /*declared (differently!) in
/
extern int UNMACRO(ffs) proto((long));
#endif
***************
*** 906,912 ****
tc[YES] = tc[NO];
tc[YES].t_intrc = IntChar;
tc[YES].t_quitc = (char) -1;
! if (OKXonXoff) {
tc[YES].t_stopc = (char) -1;
tc[YES].t_startc = (char) -1;
}
--- 910,916 ----
tc[YES] = tc[NO];
tc[YES].t_intrc = IntChar;
tc[YES].t_quitc = (char) -1;
! if (!OKXonXoff) {
tc[YES].t_stopc = (char) -1;
tc[YES].t_startc = (char) -1;
}
***************
*** 951,957 ****
TABS = !((sg[NO].c_oflag & TAB3) == TAB3);
ospeed = sg[NO].c_cflag & CBAUD;
! if (OKXonXoff)
sg[YES].c_iflag &= ~(IXON | IXOFF);
sg[YES].c_iflag &= ~(INLCR|ICRNL|IGNCR);
/* sg[YES].c_lflag &= ~(ICANON|ECHO); */
--- 955,961 ----
TABS = !((sg[NO].c_oflag & TAB3) == TAB3);
ospeed = sg[NO].c_cflag & CBAUD;
! if (!OKXonXoff)
sg[YES].c_iflag &= ~(IXON | IXOFF);
sg[YES].c_iflag &= ~(INLCR|ICRNL|IGNCR);
/* sg[YES].c_lflag &= ~(ICANON|ECHO); */
***************
*** 962,968 ****
--- 966,974 ----
* We should perhaps handle them; for now, we suppress them.
*/
sg[YES].c_cc[VQUIT] = _POSIX_VDISABLE;
+ #ifdef VSWTCH
sg[YES].c_cc[VSWTCH] = _POSIX_VDISABLE;
+ #endif
sg[YES].c_cc[VSUSP] = _POSIX_VDISABLE;
sg[YES].c_cc[VDSUSP] = _POSIX_VDISABLE;
# else
***************
*** 970,976 ****
sg[YES].c_cc[VQUIT] = (char) -1;
# endif /* _POSIX_VDISABLE */
sg[YES].c_cc[VMIN] = 1;
! sg[YES].c_cc[VTIME] = 1;
# endif /* TERMIO || TERMIOS */
# if defined(SGTTY) || defined(BRLUNIX)
--- 976,982 ----
sg[YES].c_cc[VQUIT] = (char) -1;
# endif /* _POSIX_VDISABLE */
sg[YES].c_cc[VMIN] = 1;
! sg[YES].c_cc[VTIME] = 0;
# endif /* TERMIO || TERMIOS */
# if defined(SGTTY) || defined(BRLUNIX)
***************
*** 1260,1266 ****
UpdMesg = YES;
}
rbell();
! longjmp(mainjmp, ERROR);
}
#ifdef STDARGS
--- 1266,1272 ----
UpdMesg = YES;
}
rbell();
! longjmp(mainjmp, J_ERROR);
}
#ifdef STDARGS
***************
*** 1394,1400 ****
pop_env(savejmp);
return;
! case ERROR:
getDOT(); /* God knows what state linebuf was in */
/*FALLTHROUGH*/
case COMPLAIN:
--- 1400,1406 ----
pop_env(savejmp);
return;
! case J_ERROR:
getDOT(); /* God knows what state linebuf was in */
/*FALLTHROUGH*/
case COMPLAIN:
***************
*** 1690,1696 ****
--- 1696,1704 ----
#ifdef UNIX
(void) signal(SIGHUP, finish);
(void) signal(SIGINT, finish);
+ #ifdef SIGBUS
(void) signal(SIGBUS, finish);
+ #endif
(void) signal(SIGSEGV, finish);
(void) signal(SIGPIPE, finish);
(void) signal(SIGTERM, SIG_IGN);
*** ../jove/jove.h Mon Aug 12 01:46:50 1991
--- jove.h Sun Mar 22 17:02:42 1992
***************
*** 154,160 ****
/* setjmp/longjmp args for DoKeys() mainjmp */
#define FIRSTCALL 0
! #define ERROR 1
#define COMPLAIN 2 /* do the error without a getDOT */
#define QUIT 3 /* leave this level of recursion */
--- 154,160 ----
/* setjmp/longjmp args for DoKeys() mainjmp */
#define FIRSTCALL 0
! #define J_ERROR 1
#define COMPLAIN 2 /* do the error without a getDOT */
#define QUIT 3 /* leave this level of recursion */
*** ../jove/mac.c Sun Aug 11 05:53:14 1991
--- mac.c Sun Mar 22 17:03:55 1992
***************
*** 1712,1718 ****
*row = (p.v / HEIGHT);
*col = (p.h / WIDTH );
if ((*row > MAXROW) || (*col > MAXCOL))
! return ERROR;
return 0;
}
--- 1712,1718 ----
*row = (p.v / HEIGHT);
*col = (p.h / WIDTH );
if ((*row > MAXROW) || (*col > MAXCOL))
! return J_ERROR;
return 0;
}
*** ../jove/re.c Sun Aug 11 20:15:40 1991
--- re.c Sun Mar 22 17:04:44 1992
***************
*** 784,790 ****
while (--n >= 0) {
*off++ = *pp++;
if (off >= endp)
! len_error(ERROR);
}
return off;
}
--- 784,790 ----
while (--n >= 0) {
*off++ = *pp++;
if (off >= endp)
! len_error(J_ERROR);
}
return off;
}
***************
*** 827,833 ****
}
*tp++ = c;
if (tp >= endp)
! len_error(ERROR);
}
}
rp = loc2;
--- 827,833 ----
}
*tp++ = c;
if (tp >= endp)
! len_error(J_ERROR);
}
}
rp = loc2;
***************
*** 845,851 ****
loc2 = re_blk->r_lbuf + REeom;
while ((*tp++ = *rp++) != '\0')
if (tp >= endp)
! len_error(ERROR);
}
void
--- 845,851 ----
loc2 = re_blk->r_lbuf + REeom;
while ((*tp++ = *rp++) != '\0')
if (tp >= endp)
! len_error(J_ERROR);
}
void
*** ../jove/recover.c Sat Aug 10 18:16:35 1991
--- recover.c Sun Mar 22 18:51:46 1992
***************
*** 321,327 ****
char *quest,
*answer;
{
! if (stdin->_cnt <= 0) {
printf("%s", quest);
fflush(stdout);
}
--- 321,327 ----
char *quest,
*answer;
{
! if (isatty(0)) {
printf("%s", quest);
fflush(stdout);
}
*** ../jove/sysdep.h Sun Aug 11 05:09:52 1991
--- sysdep.h Sun Mar 22 18:45:20 1992
***************
*** 106,111 ****
--- 106,121 ----
# define POSIX_PROCS 1 /* processes operate as specified by POSIX */
#endif
+ #ifdef linux
+ # define SYSVR3 1
+ # define DIRENT 1 /* Posix style dirent.h */
+ # define TERMIOS 1 /* new Posix terminal mode management */
+ # define POSIX_UNISTD 1 /* prototypes in unistd.h, don't use our own */
+ # define POSIX_PROCS 1 /* processes operate as specified by POSIX */
+ # define JOB_CONTROL 1 /* if you have job stopping */
+ # define PTYPROCS 1 /* use pseudo-ttys */
+ #endif
+
#ifdef SYSVR3
# ifndef SYSVR2
# define SYSVR2 1 /* SYSVR2 is a subset of SYSVR3 */
***************
*** 112,118 ****
# endif
# define SIGRESTYPE void
# define SIGRESVALUE /*void!*/
! # ifndef SYSVR4
# define SIGCHLD SIGCLD
# endif
#endif
--- 122,128 ----
# endif
# define SIGRESTYPE void
# define SIGRESVALUE /*void!*/
! # if !defined SYSVR4 && !defined linux
# define SIGCHLD SIGCLD
# endif
#endif
***************
*** 243,248 ****
--- 253,260 ----
# endif /* MENLO_JCL */
#endif
+ #include
+
#ifdef BSD_SIGS
extern long SigMask;
***************
*** 249,256 ****
# define SigHold(s) sigblock(SigMask |= sigmask((s)))
# define SigRelse(s) sigsetmask(SigMask &= ~sigmask((s)))
#else
! # define SigHold(s) sighold(s)
! # define SigRelse(s) sigrelse(s)
#endif
/* On a system which limits JOVE to a very small data segment,
--- 261,273 ----
# define SigHold(s) sigblock(SigMask |= sigmask((s)))
# define SigRelse(s) sigsetmask(SigMask &= ~sigmask((s)))
#else
! # if defined SIG_BLOCK && defined SIG_UNBLOCK
! # define SigHold(s) sigprocmask(SIG_BLOCK,s,NULL)
! # define SigRelse(s) sigprocmask(SIG_UNBLOCK,s,NULL)
! # else
! # define SigHold(s) sighold(s)
! # define SigRelse(s) sigrelse(s)
! # endif
#endif
/* On a system which limits JOVE to a very small data segment,
*** ../jove/tune.h Sun Aug 11 03:15:15 1991
--- tune.h Sun Mar 22 16:52:31 1992
***************
*** 24,30 ****
#define ABBREV 1 /* word abbreviation mode */
#ifdef UNIX
! # define ID_CHAR 1 /* include code to IDchar */
/* # define WIRED_TERMS 1 */ /* include code for wired terminals */
#endif
--- 24,30 ----
#define ABBREV 1 /* word abbreviation mode */
#ifdef UNIX
! /*# define ID_CHAR 1 /* include code to IDchar */
/* # define WIRED_TERMS 1 */ /* include code for wired terminals */
#endif
***************
*** 38,44 ****
*/
/* # define LOAD_AV 1 */
! # define BIFF 1 /* if you have biff (or the equivalent) */
# define SPELL 1 /* spell words and buffer commands */
#endif
--- 38,44 ----
*/
/* # define LOAD_AV 1 */
! /*# define BIFF 1 /* if you have biff (or the equivalent)
*/
# define SPELL 1 /* spell words and buffer commands */
#endif
*** ../jove/util.c Sun Aug 11 20:17:11 1991
--- util.c Sun Mar 22 17:05:07 1992
***************
*** 464,470 ****
do {
if (onto >= endp)
! len_error(ERROR);
} while ((*onto++ = *from++) != '\0');
}
--- 464,470 ----
do {
if (onto >= endp)
! len_error(J_ERROR);
} while ((*onto++ = *from++) != '\0');
}
[next article]
From: [email protected] (Drew Eckhardt)
Subject: Re: Is Shoelace permanent? [was HELP! SHOELACE IS STRANGLING ME!!!]
Message-ID: <[email protected]>
Date: 25 Apr 92 01:19:46 GMT
References: <[email protected]> <[email protected]> <199
[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 49
Nntp-Posting-Host: ophelia.cs.colorado.edu
In article <[email protected]> [email protected] (Joel M. Hoffman
) writes:
>In article <[email protected]> [email protected] (A. V. Le Blanc) w
rites:
>>In article <[email protected]> [email protected] (Kevi
n Cummings) writes:
>>>In article <[email protected]>, [email protected]
(Greenup) writes:
>>>>
>>>> I recently used shoelace to try & make linux boot. after no success,
>>>> i want to go back. HOW DO I GET RID OF THIS BOOT THING THAT KEEPS
>>>> ASKING WHICH PARTITIAN TO BOOT?!?!?!?!?
>>>
>>>Just a wild guess, but boot from your ORIGINAL BOOT floppy using your
>>>[etc.]
>>
>>I suspect it isn't that simple. Installing the shoelace menu program
>>in the boot partition overwrites the boot sector. You cannot replace
>>it as it was unless you have saved a copy of it, or unless you do a
>>low-level format of your hard disk! Norton's Disk Doctor, as I recall,
>>can write a pretty standard boot sector to your hard disk and save
>>having to boot from a floppy.
>
>Does this mean that for IDE drives, shoelace is a permanent step???
>If so, a notice to that effect would be welcome in the FAQ!
>
You don't have to relowlevel format it, you just need to
write a new block 0.
If you had been smart when you installed shoelace, you would have
done something like
dd if=/dev/hda bs=512 count=1 of=partition.bak
So you could do a
dd if=partition.bak bs=512 count=1 of=/dev/hda
to restore it.
any one else, take note of this : It's IMPORTANT!
To undo it :
Do the above mentioned backup of your partition table.
Find out what's in your partition table, write it down.
Grab someone else's partition table, write it to
absolute sector 0. Use your favorite partitioning
program to write out your partition table onto that
boot block.
[next article]
From: [email protected] (Drew Eckhardt)
Subject: Re: Linux: Multi-terminal support??
Message-ID: <[email protected]>
Date: 25 Apr 92 02:01:16 GMT
References: <[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 35
Nntp-Posting-Host: ophelia.cs.colorado.edu
In article <[email protected]> [email protected] (Scott R. Johnson) writes:
>Folks,
>
>I'm a Linux newbie. (gawd, they're hauling out the tomatoes now...)
>
>Can Linux be set up to allow multiple terminals, say, VT100s, to log in
>off the serial ports?
Yes. I have an ADDS terminal hung off a serial port, and it
works fine.
Just put a getty on the serial port, and somewhere in your
RC do a stty
>If so, how hard would it be to hack up, say 4 to 6 ports (I know that the PC
>spec only allows 2 ports, at 1F0 and 1F8, but the two additional at 1E0 and
>1E8 are pretty much de-facto now; they have to share the interrupts with
The PC will use 4 ports, just not all interrupt driven.
Sharing interrupts is a hardware problem, not a software problem.
You can get around this with a hardware hack if you want. Simply cut the
interrupt lines coming off of the uarts, run them through a 7432 on a
piece of perfboard, and wire the output to the appropriate place.
(don't forget about +5 and ground!)
Additionally, minor changes to the serial driver are required.
A simpler alternative is to modify your serial hardware to use
int 5, etc. If you can't jumper the card for it, you can
always cut and jump traces.
Again, minor changes to the Linux serial drivers are required to
use the serial ports and set up the appropriate interrupts.
[next article]
From: [email protected]
Subject: Re: Shells and Shared libraries
Message-ID: <[email protected]>
Date: 25 Apr 92 02:04:11 GMT
References: <[email protected]>
Organization: Penn State University
Lines: 4
Much thanks to the people who sent mail saying how to switch on and off the
shared libraries, and those who told me about tcsh. I think I have enough
to get everything working correctly. We'll see when I reboot and try to
compile stuff. 😉
[next article]
From: [email protected] (Paul Allen)
Subject: Re: MGR and Linux
Message-ID: <[email protected]>
Date: 25 Apr 92 02:51:14 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 38
[email protected] (Anthony Starks) wrote:
|Has anyone considered porting the MGR window system to Linux
|as a less-bloated alternative to X?
Yes, I've considered it. I think all the necessary parts are there.
They just need to be integrated. I may get to it one of these days,
but it won't be soon. The time just isn't there. Maybe later this
summer.
Here's why I would like to see MGR under Linux:
It's small. You can compile the whole thing in much less than 20Mb.
It has the basic stuff that you need: multiple overlapping windows,
text and graphics in each window, popup menus, multiple fonts, text
cut & paste, etc.
You can run an MGR client on a remote machine with its window on your
local machine, just like with X. It uses whatever network you've got,
including simple async modem connections.
It lacks most of the complications that encumber X: resources, toolkits,
keysyms, 8 volumes (!) of O'Reilly to explain it, etc.
In other words, it does the basic job without requiring you to devote
your life to it in order to use it effectively.
Clearly, we're going to have X for Linux. I'll probably play with it
a little when it appears. But what I really want is a window system
that's small enough that I can afford to keep the sources around.
That's why we'll eventually have MGR for Linux.
If anybody wants to take a crack at MGR, the Minix port is probably
the place to start. It can be found on bellcore.com with all the
other MGR ports.
Paul Allen
[next article]
From: [email protected] (Bennett E. Todd)
Subject: Re: Re: Porting to Linux (Jove question)
Message-ID: <[email protected]>
Date: 25 Apr 92 02:57:36 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Bennett E. Todd)
Organization: The Internet
Lines: 30
So much help, so FAST! Thanks Everyone!
Mitch ([email protected]) pointed me to a newer release of jove (4.14.7),
which was a lot closer to ready for Posix. He also sent me defines that
would build a jove, and Praise Be a working jove binary, except without
interactive processes.
I found a book:
The POSIX.1 Standard: A Programmer's Guide
by Fred Zlotnick
The Benjamin/Cummings Publishing Company, Inc. (1991)
ISBN 0-8053-9605-5
that seemed to really explain everything neatly. And I built a jove that
seems to work completely correctly. I also have it ready for upload.
Hmmm. David, you want me to beam my stuff your way, so you can compare
and see which seems better? I do have one semi-radical difference; I
first defined the SigHold and kin exactly like you showed, and I got
warnings from gcc about casting an int to a pointer. Sho' 'nuff,
sigprocmask wants pointers to signal sets. So I used code like this:
extern sigset_t SigSet;
# define SigHold(s) (SigSet = (s),sigprocmask(SIG_BLOCK, &SigSet, NULL))
# define SigRelse(s) (SigSet = (s),sigprocmask(SIG_UNBLOCK, &SigSet, NULL))
and defined SigSet in proc.c under ifdef LINUX.
-Bennett (curmudgeon)
[email protected]
[next article]
From: [email protected] (Erik Green)
Subject: Re: mcc-interim(0.95c+) help!
Message-ID:
Date: 25 Apr 92 04:35:09 GMT
Distribution: comp.os.linux
Organization: Long polymers in double-helix formation
Lines: 19
Nntp-Posting-Host: att2.cs.mankato.msus.edu
In a previous post, [email protected] (Ishwar Rattan) says:
>BTW has anybody bale to compile zip/unzip/zoo under linux?
I've got zoo 2.1 compiled, but it's really flaky. As to the reason,
my guess is the bug in fseek() in the current version of GCC 2.1.
I'm going to wait for the next release of the GCC libs, recompile
and see if the problem goes away.
I may try zip, unzip and lharc next week.
-Erik
>- ishwar([email protected])
--
Erik "Longshot" Green
longshot@(vax1 | att2.cs | krypton | theory.cs).mankato.msus.edu
This is a .signature virus scanner. Stop the .sig virus!
Disclaimer: I said the above. So sue me. I'm broke anyway.
[next article]
From: [email protected] (no, I don't repeat it!)
Subject: Re: Latin-1 again
Message-ID:
Date: 25 Apr 92 13:17:06 GMT
References: <[email protected]>
Sender: [email protected] (Usenet pseudouser id)
Reply-To: [email protected]
Organization: Helsinki University of Technology, Finland
Lines: 40
In-Reply-To: [email protected]'s message of Thu, 23 Apr 1992 20:50:17 GMT
Nntp-Posting-Host: vipunen.hut.fi
In article <[email protected]> [email protected] (
H. Peter Anvin N9ITP) writes:
Is there an interest to have Linux use Latin-1?
Definitely.
CGA/HGC:
Display characters in graphic mode.
MDA:
Use a translation table.
I think we could as well support only ega/(s)vga at least from the start.
If someone has a [34]86 running linux with HGC or (yuck) CGA/MDA, it would
*definitely* be time to upgrade...
Printer:
Use a customizable code file that contains the necessary translations;
question: should this translation be done by /dev/lp* or by the lp program.
Your list was missing serial terminals/terminal emulators that often use
character sets other than latin-1. A more general and universal approach
would be to use a translation table for each device (a la SCO:s mapchan)
and possibly not touch the device character sets at all.
In the console we could support either a translation table or downloadable
characters on the display. The former way would be an universal way to
support any currently used display adapter and we could still support
latin-1 in "native mode" on ega/(s)vga. The keyboard could be made to
support either direct latin-1 character set or some other set with a
conversion table. Remember, we still need conversion tables for serial
terminals.
I don't like the idea of slowing linux down by using graphics mode to
display text even with CGA/HDC (at least until we get X, that is). The
translation tables probably belong somewhere in the tty driver.
Paulii
--
Disclaimer fault - lawyers dumped
[next article]
From: [email protected] (Paul Nulsen)
Subject: Re: free (memory allocation table)
Message-ID: <[email protected]>
Date: 25 Apr 92 06:41:29 GMT
References: <[email protected]> <[email protected]
avy.mil>
Organization: University Of Wollongong
Lines: 22
[email protected] (LCDR Michael E. Dobson) writes:
>In article <[email protected]> [email protected] (Peter Thomson) writes:
>>
>>
>>Hi can some one tell me if the command free reports back any more than
>>1Mbyte of memory. There is 2Mbytes on my machine but free states there is
>>only one. running multiple tasks and watching ps confirms this and swapping
>>starts at the 1Mbyte limit
>>
>> Thanks in advance Peter.
>
>Free reports >2 Meg free on my system, 640K conventional, 3 Meg extended.
On my system, with 8 Mbytes of memory, free reports 6144 blocks total (or
thereabouts). Is this because the buffer cache is not counted as memory by
free? If not, what is going on here? I presume 384K is missing because of
the way PC harware works, but even allowing for this I don't seem to be able
to get numbers that total to 8 Mbytes when I add in the buffer cache size.
Paul Nulsen
[email protected]
[next article]
From: [email protected] (Steve Frampton)
Subject: QIC Compatible Tape Backup Software.
Keywords: qic tape backup
Message-ID:
Date: 24 Apr 92 18:49:01 GMT
Organization: Vicuna Systems, Kingston, ON (613) 547-5066
Lines: 25
Hi everyone.
I was wondering if any software had been ported over to Linux that would
provide QIC 40 and/or QIC 80 compatible tape backup and file retrieval?
Specifically I am wondering if anything out there supports the Jumbo line
of tape drives from Colorado Memory Systems.
If there aren't any QIC 40/80/112 compatible software packages available
under Linux, is any work being done on this at the moment? If not, does
anyone know where I can find specs on the QIC standard so that software
CAN be written?
Thanks in advance!
Steve
BTW I am trying to make an intelligent decision regarding the purchase of
a tape drive, so if anyone has any experience with Colorado Memory
Systems and their line of drives I'd appreciate hearing from you (however
please do it in e-mail to save on net.bandwidth).
------------------------------------------------------------------------
Steve Frampton ([email protected]), Kingston, Ontario, CANADA
"It only took me 10 years of MS-DOS before I knew I should be using Unix
...am I quick or what?"
[next article]
From: [email protected] (Drew Eckhardt)
Subject: Re: free (memory allocation table)
Message-ID: <[email protected]>
Date: 25 Apr 92 10:32:35 GMT
References: <[email protected]> <[email protected]
avy.mil> <[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 23
Nntp-Posting-Host: ophelia.cs.colorado.edu
In article <[email protected]> [email protected] (Paul Nulsen
) writes:
>[email protected] (LCDR Michael E. Dobson) writes:
>
>>In article <[email protected]> [email protected] (Peter Thomson) writes:
>>>
>>>
>>>Hi can some one tell me if the command free reports back any more than
>>>1Mbyte of memory. There is 2Mbytes on my machine but free states there is
>>>only one. running multiple tasks and watching ps confirms this and swapping
>>>starts at the 1Mbyte limit
>>>
>>> Thanks in advance Peter.
>>
>>Free reports >2 Meg free on my system, 640K conventional, 3 Meg extended.
>
>On my system, with 8 Mbytes of memory, free reports 6144 blocks total (or
>thereabouts). Is this because the buffer cache is not counted as memory by
>free? If not, what is going on here? I presume 384K is missing because of
>the way PC harware works, but even allowing for this I don't seem to be able
>to get numbers that total to 8 Mbytes when I add in the buffer cache size.
Buffer cache is considered kernel memory, and kernel memory is
not included.
[next article]
From: [email protected] (Jim Burns)
Subject: swapfile on scsi?
Keywords: does it work?
Message-ID: <[email protected]>
Date: 25 Apr 92 10:44:43 GMT
Organization: Georgia Institute of Technology
Lines: 14
I got linux installed on my Maxtor/AHA1542, w/only minor bugs cropping up -
like ^Z works under bash, but fg (of a vi/elvis1.4) hangs the vc - but I'm
still getting that 'swap signature not found' error after doing:
dd bs=1024 count=2048 if=/dev/sd3 of=/tmp/swapfile [it does get created]
mkswap /tmp/swapfile 2048 [file access time is updated]
swapon /tmp/swapfile [EINVAL, swap sig. not found, etc.]
Any ideas? What should the swapfile permissions be? Thanx.
--
BURNS,JIM (returned student)
Georgia Institute of Technology, 30178 Georgia Tech Station,
Atlanta Georgia, 30332 | Internet: [email protected]
uucp: ...!{decvax,hplabs,ncar,purdue,rutgers}!gatech!prism!gt0178a
[next article]
From: [email protected] (Drew Eckhardt)
Subject: pre-.96 SCSI
Message-ID: <[email protected]>
Date: 25 Apr 92 10:41:31 GMT
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 18
Nntp-Posting-Host: ophelia.cs.colorado.edu
Sources and kernel are available at
headrest.woz.colorado.edu under /pub/linux.
Files are
vmunix.0.96-.Z
scsi.0.96-.tar.Z
Supported hardware includes the Adaptec 1542, Seagate ST01/02,
and Ultrastor 14F. With minor code changes, Adaptec 1742,
and Future domain 8 bit controllers could be supported.
Some ioctl support was added, but mostly only changes to use the
file_operations structures were made.
And, like the "normal" binary, it now mounts root off of floppy, so
I don't get any more whiney users who ignore the FAQ complaining that
it panics and says can't mount root 1/2 🙂
[next article]
From: [email protected] (Michael Hamilton)
Subject: Re: Bootlin --> help please!!!!!!
Message-ID: <[email protected]>
Date: 24 Apr 92 22:34:12 GMT
References: <[email protected]> <1992Apr24.165213.24168@mu
ddcs.claremont.edu>
Organization: Wellington City Council, Wellington, NZ
Lines: 17
>In article <[email protected]> [email protected]
(cm445a17) writes:
>>I just installed bootlin.com on my machine as described on the docs but it
>>doesn't work! It reports that it loaded the boot sector, setup, and
>>kernel OK but the it prints some garbish characters and hangs!!!
>
Definitly hangs with me too. Pressing space or return doesn't do
anything. It actually worked once (see gibberish, but press space
anyway) but I've never been able to repeat it.
HP QS20 386DX 20MHz 4Mb 84MB Quantum IDE. Linux 0.95c.
--
________________
Michael Hamilton, Computer Services Section, Wellington City Council, P.O. Box
2199, Wellington, New Zealand. Phone: (64) (4)801-3300 FAX: (64) (4)801-3020
Domain: [email protected] PSImail: PSI%0530147000090::HAMILTON
[next article]
From: [email protected] (n liu)
Subject: Re: dos emulator
Message-ID: <[email protected]>
Date: 25 Apr 92 12:19:33 GMT
References: <[email protected]> <1992Apr23.1729
[email protected]>
Sender: [email protected]
Organization: California Institute of Technology, Pasadena
Lines: 29
Nntp-Posting-Host: mripc
[email protected] (Lance Pickup) writes:
>In article <[email protected]>, lautner@dec16.
informatik.uni-wuerzburg.de (Matthias Lautner) writes:
>|> If someone is interested in helping me to implement the functions of the
>|> dos int 0x21 it would be grate. I'm currently working with 0.95c.
>This may sound like an outrageous amount of work, but in the long
>run it may prove to be worth it: I think commercially available
>DOS emulators (well, I can really only speak for the ones available
>in AIX because that's all I'm familiar with) use DOS itself to provide
>INT 21h support. Provided you supply all the BIOS and I/O support
>(this is the really outrageous part!--big IF!), DOS itself should
>be able to run providing INT 21h support. I have seen this approach
>recommended before.
the down side is that the DOS/BIOS routines are pretty shitty - i.e.
slow and, more importantly, mostly non-reentrant. OS/2's dos box uses
completely rewritten DOS/BIOS calls, so stuff like floppy disk access
doesn't thrash forever in background. Compare to WinDoze, which has to
use DOS's services and suffers accordingly.
So maybe you could ask IBM for a copy of the rewritten code. I'm sure they'd
cough it up for anyone who asks nicely (and monkeys might fly out of my butt :).
As far as getting info on DOS interrupts, Ralf Brown's list is the most
extensive (and frequently updated) of them all...
--
[email protected]
[email protected]
[next article]
From: [email protected] (Oren Kastner)
Subject: Re: Linux Newbie - Advice needed... (Long, sorry)
Keywords: n
Message-ID: <[email protected]>
Date: 25 Apr 92 12:59:13 GMT
References: <[email protected]> <[email protected]> <1992
[email protected]>
Sender: [email protected] (USENET News System)
Organization: Brandeis University
Lines: 14
Thanks a bunch to all who responded so far (and those that I'll probably
get responses from after I post this). I went through with the MCC interim
installation and it went smoothly. Things seem to be poifect! (couldn't
ask for anything more...)
Thanks again to all who cared enuf to send the very best! 🙂
Oren.
--
Oren Kastner E-mail address:
Brandeis University [email protected]
[next article]
From: [email protected] (david nugent)
Subject: Re: fully broiled ideas
Message-ID:
Date: 24 Apr 92 11:42:38 GMT
References: <[email protected]>
Reply-To: [email protected]
Organization: Unique Computing Pty Ltd, Melbourne, Australia
Lines: 60
pmacdona@sanjuan (Peter MacDonald) writes:
> I note BSDI is promising binary compatibility with SysV 3.2. I imagine
> then that if this happens, 386BSD will eventually follow suit if possible.
> Soooo, anyone know if there are any steps that could be taken to move Linux
> part way towards it now? Like perhaps, adopting the same system call
> interface, if possible?
This isn't technically necessary. The executable type should be
determinable by the a.out header, and it is possible to implement
more than one system call interface concurrently.
Three things needed:
a. a.out header to distinguish between a.out types
b. the "executable type" noted for each process
c. different system call tables for the supported executable types
d. various interface routines to convert between non-native
system calls and native ones, and probably some additional
system calls will need to be implemented
e. probably a whole lot of loose ends which fall out of a-c.
I daresay the kernel would get somewhat more bloated. 🙂
I believe both SCO and ISC 386/UNIX's implement something like this
already - or must do. ISC, for example, has both a "sysv" and "posix"
system call interface, and different shared libraries. It doesn't
use the a.out header to distinguish between the types though, but
implements a common __ostype() system call which allows switching
between modes. The crt0p.o (the posix startup code) incorporates a
call to __ostype() to set the posix system call interface, and the
exec call assumes sysv by default.
> I know that the current shared libraries
> interface is incompatible with that goal, but ignore that for the moment.
> The primary issue here is to strengthen Unix vis-a-vis other commercial
> operating systems.
I don't know that binary compatibility is work it, to tell the truth, but
this is very much a "user driven" need. All it would mean is that
some commercial binary only applications can run under Linux, but I
personally don't think that it's worth it. Linux comes with full source
now, and encouraging or supporting other vendors to supply without
sources isn't necessarily within it's scope.
> BTW: has anyone tried the OLEO spreadsheet on tsx yet. Bug reports or
> "it works for me" reports would be appreciated.
Seems to work here, although I haven't used it to any great extent. I
like it a lot better than sc though.
..............................................................................
david nugent Public Access Usenet "Only Nixon can go to China"
[email protected] +61-3-792-3507 - ancient Vulcan proverb
3:632/400@fidonet, 58:4100/1@intlnet, 199:4242/5@rainbownet, 33:300/6@admin
PO Box 260, Endeavour Hills, Victoria, Australia, 3802.
[next article]
From: [email protected] (david nugent)
Subject: Re: DEL using xcomm XC
Message-ID: <[email protected]>
Date: 24 Apr 92 14:48:55 GMT
References: <[email protected]>
Reply-To: [email protected]
Organization: Unique Computing Pty Ltd, Melbourne, Australia
Lines: 42
[email protected] writes:
> How does I send a DEL using xcomm, XC.
Likewise. From what I can figure out, it just can't be done.
Even ^Backspace that works under DOS apparently isn't used by
Linux, so I have no backspace (DEL) at all when logged into a
VAX.
> I know kermit had trouble with VT100 termcap and somehow
> used backspace BS for DEL and ^H for backspace. Not sure
> how and didn't really like doing ^H for backspacing.
This is noted as a 'feature' in the XC docs, but it's one I
can well live without. Changing the termcap or terminal
settings at the other end is always much better in the long
run than kludges like this.
Actually, I've had real problems with XC's tendancy to redefine
*everything*. It's very annoying really. I like the program, but
don't like the effect it has on my function and many other keys,
even after leaving the program (it's not a simple case of remapping
in the terminal emulator). For example, bash's command line
history no longer works with the cursor keys after running XC.
I've now gone back to kermit to call out to the VAX I use. Linux's
emulation is much better for this use, and since DEC invented vt100.. 🙂
This begs the question - I'd like to get hold of the XC sources and
fix this if possible since I prefer it's functionality over kermit.
I've never liked pcomm much and ecu seems like a real job to port.
(I prefer something simpler than ecu anyway). Does anyone know if and
where the sources are available?
..............................................................................
david nugent Public Access Usenet "Only Nixon can go to China"
[email protected] +61-3-792-3507 - ancient Vulcan proverb
3:632/400@fidonet, 58:4100/1@intlnet, 199:4242/5@rainbownet, 33:300/6@admin
PO Box 260, Endeavour Hills, Victoria, Australia, 3802.
[next article]
From: [email protected] (david nugent)
Subject: Re: News 'n' Mail
Message-ID:
Date: 24 Apr 92 15:00:01 GMT
References: <[email protected]>
Reply-To: [email protected]
Organization: Unique Computing Pty Ltd, Melbourne, Australia
Lines: 99
[email protected] (Steve Robbins) writes:
> One of the most constant irritations in compiling these and gnu utilties, is
> that the code always wants to declare standard library functions, like
> "extern long lseek()" or whatever. My include files, on the other hand,
> already have these in stdio.h and the like, and gcc complains. Is declaring
> the library routines peculiar to linux or gcc or is it ANSI that nobody uses
> yet or what?
Bracket with "#if defined(__STDC__)". They should be defined in headers, but
sometimes aren't in some ansi environments and almost always aren't in non-
ansi. What is even more irritating is if the declarations are incorrect and
don't agree with the headers.
Functions in the system library _should_ be declared in a header file somewhere.
Fixing this in an application file is usually a kludge, but there are still
many non-ansi environments (even if one is available, many will not see the
advantages in using it).
Maybe __STDC__ isn't all that good an idea in the real world - although it
should be like that, and may be one day if customers scream enough to their
UNIX vendors. Maybe "NEED_PROTOTYPE_KLUDGE" would be better and compile with
-DNEED_PROTOTYPE_KLUDGE if you need it. 🙂
It's a good opportunity here to congratulate the great job that H.J. is doing.
Building a useable includes set and libraries isn't an easy task at the best
of times, but Linux is becoming a pretty clean environment. I have far less
problems with gcc 2.1 under Linux than I had with ISC's dev package.
> The other major irritation is that almost none of the configure
> scripts works right.
The problem with configure scripts is almost always failures with awk. I think
we need a new one (or maybe a new shared library to link it with) since the
existing one has some serious problems. I haven't spent enough time on this to
track them down, but they are easily reproduced by running up anything which
makes use of Larry Wall's Configure script.
> The hardest problems to find, though, lie with libc.a. My version is from
> early April gcc2.1 from tsx-11, so I'd appreciate knowing if these are fixed,
> or better, where I can get source for the library functions.
The library source is also on tsx-11, although H.J. has been mentioning a
forthcoming update. The sources on tsx-11 do produce exact byte-for-byte
binaries of the libraries though - at least the c libraries. I haven't
built the g++ library though (these aren't on tsx-11, btw).
> Elm:
> When running as a regular user, I get the error 'problems restoring
> permissions of /usr/spool/mail/steve' when I quit. No mail seems to be lost
> or mangled, but this is vaugely disturbing nonetheless.
Ditto. I didn't count it as a problem since it appears harmless. I had
intended to track this down now that gdb is up and running.
> Smail:
> When mail is sent to nonexistent local users, it gets bounced correctly, but
> two 'general protection: 0000' messages come up, which is also disturbing.
I had this initially, but after installing dummy alias and getting a paths file,
the problem vanished. Both deliveries and bounces work correctly.
> Also, is it just me, or is the configure/build tangle of shell, sed, awk,
> and make scripts incredibly labyrinthine? I still don't even understand the
> entire thing, and I can only get it to compile by putting in a dummy Makefile
> in the src/lookup directory to build an empty ldrivlib.a. Then I have to
> 'touch' this makefile when the master make fails in some obscure awk program.
> My src/lookup directory came absolutely empty...did I get a bad archive?
No. Again, this is due to a bad awk. The install is very mangled, and I had
to complete most of it by hand. The process isn't all that mysterious, and
is accomplished by setting shell variables and doing awk substitutions through
various files. It is _very_ dependant on awk working, but if you get the
.defs files in all the directories working, the rest of the process works from
that point.
> Taylor:
> (Using taylor-style everything: config, spool, logfile, etc)
> After a transfer, uuxqt starts up and works correctly, but since there are ot
> files in the /usr/spool/uucp directory (lockfiles, logfiles, etc) it complain
> about not being able to open each of them `filexxxy/X.' ... TWICE! This is
> probably a bug in Taylor's code, but if you know how to fix it...please
> enlighten me.
Sounds like the mktemp() bug someone just reported. I haven't used the uucp
package other than to find out that the serial driver can't cope at all with
PEP speeds.
..............................................................................
david nugent Public Access Usenet "Only Nixon can go to China"
[email protected] +61-3-792-3507 - ancient Vulcan proverb
3:632/400@fidonet, 58:4100/1@intlnet, 199:4242/5@rainbownet, 33:300/6@admin
PO Box 260, Endeavour Hills, Victoria, Australia, 3802.
[next article]
From: [email protected] (david nugent)
Subject: Re: linux
Message-ID: <[email protected]>
Date: 24 Apr 92 11:29:37 GMT
References:
Reply-To: [email protected]
Organization: Unique Computing Pty Ltd, Melbourne, Australia
Lines: 51
[email protected] (Matt Mosley) writes:
> I would like to know how to do the following:
>
> 1) Does Linux have *any* working mail program? I can't seem to find any.
I use smail 3.1.25. The "port" required nothing more than editing the
EDITME file as described in it's configuration documentation. I did
have problems with the scripts it uses to configure the system and
managed to track thos problems down to the Linux awk (which I believe
is a compile of gnu's awk - it seems to have problems). I'll see about
getting a new awk before doing it again, but at the time I completed
it by hand.
I use an lmail I grabbed from some place long ago; I don't recall the
source (probably a usenet posting). It is quite portable and works
flawlessly under Linux. If anyone is interested, I'll be happy to post
the binary and code.
As for a UA, I compiled elm 2.3 pl#11 as have several other people
judging by some earlier posts here. Once you get smail working, elm
is no problem at all. I heard of one case of someone having problems
compiling it with gcc 2.0, but I used HJ's 2.1 beta and it went
smoothly, again with no significant patching other than the usual
configuration required for a make on any system.
> 2) How do I transfer my news and mail feeds over to Linux (I'd first need
> a mail program, of course, plus readnews and uucp).
You'll need the news system, of course. I haven't tried to get cnews
going yet, but I suspect some interesting script file changes will be
necessary. Ian Taylor's uucp is already ported, but the serial driver
needs a serious overhaul (I've already started on that, but it may be
a week or two before I have time enough to attack it with any real
gusto).
> 3) Where can I get troff/groff/nroff for Linux? (or at least workable).
I grabbed groff from tsx-11.mit.edu. It seems to work well enough for
what I've been using it for - only man pages. Haven't tested with printing
yet, nor played with (g)troff.
..............................................................................
david nugent Public Access Usenet "Only Nixon can go to China"
[email protected] +61-3-792-3507 - ancient Vulcan proverb
3:632/400@fidonet, 58:4100/1@intlnet, 199:4242/5@rainbownet, 33:300/6@admin
PO Box 260, Endeavour Hills, Victoria, Australia, 3802.
[next article]
From: [email protected] (david nugent)
Subject: Re: gdb fails on ioctl??
Keywords: gdb ioctl
Message-ID:
Date: 25 Apr 92 11:37:43 GMT
References: <[email protected]>
Reply-To: [email protected]
Organization: Unique Computing Pty Ltd, Melbourne, Australia
Lines: 26
[email protected] (william.j.bitner) writes:
> Hi. I just got gcc2.1 up and running, rebuilt the "new"
> .96 kernel and I thought I'd throw in gdb. It's giving
> me an error "ioctl TIOCSPGRP, not owner???" I tried
> this both as root and as a user with no difference, did
> I do something wrong.
>
> Could really use some advice. 🙂
I had meant to mention it in the readme, but forgot.
This is a kernel problem that Linus tells me will be fixed
in the final 0.96 release. In the meantime, simply ignore
it.
Regards,
david
..............................................................................
david nugent Public Access Usenet "Only Nixon can go to China"
[email protected] +61-3-792-3507 - ancient Vulcan proverb
3:632/400@fidonet, 58:4100/1@intlnet, 199:4242/5@rainbownet, 33:300/6@admin
PO Box 260, Endeavour Hills, Victoria, Australia, 3802.
[next article]
From: [email protected] (david nugent)
Subject: Re: Linux: Multi-terminal support??
Message-ID:
Date: 25 Apr 92 12:24:00 GMT
References: <[email protected]>
Reply-To: [email protected]
Organization: Unique Computing Pty Ltd, Melbourne, Australia
Lines: 30
[email protected] (Drew Eckhardt) writes:
> Just put a getty on the serial port, and somewhere in your
> RC do a stty
Shouldn't stty work on stdin? I wasn't aware that Linux's was
different - is this the case, and if so why?? I certainly would
explain a few things... 😉
> Sharing interrupts is a hardware problem, not a software problem.
> You can get around this with a hardware hack if you want. Simply cut the
> interrupt lines coming off of the uarts, run them through a 7432 on a
> piece of perfboard, and wire the output to the appropriate place.
> (don't forget about +5 and ground!)
>
> Additionally, minor changes to the serial driver are required.
Could you please let me know what these are (if there is anything other
than the usual when dealing with multiple ports on the same interrupt).
I'm currently hacking on a sio driver for Linux just to support higher
speed coms, hardware handshaking and 16550A's on my own machine.
..............................................................................
david nugent Public Access Usenet "Only Nixon can go to China"
[email protected] +61-3-792-3507 - ancient Vulcan proverb
3:632/400@fidonet, 58:4100/1@intlnet, 199:4242/5@rainbownet, 33:300/6@admin
PO Box 260, Endeavour Hills, Victoria, Australia, 3802.
[next article]
From: [email protected] (Peter Thomson)
Subject: Re free
Message-ID: <[email protected]>
Date: 25 Apr 92 13:46:09 GMT
Sender: [email protected]
Reply-To: [email protected] (Peter Thomson)
Organization: Department of Computer Science, University of Edinburgh
Lines: 29
>Hi can some one tell me if the command free reports back any more than
>1Mbyte of memory. There is 2Mbytes on my machine but free states there is
>only one. running multiple tasks and watching ps confirms this and swapping
>starts at the 1Mbyte limit
In relation to my first posting about free above. Linus mailed me the
following that explains everything. Thanks to all who helped clear
this up.
Peter.
======================================================================
"free" doesn't concern itself with the memory the kernel has allocated
for itself: so what you see is the /user/ memory available. The kernel
has already taken the low 1M for it's own use, so all you memory is
being used, never fear.
The low meg is used for the kernel itself (about 250kB) and for
buffercache and various kernel data structures. On a 8MB machine the
kernel gets even more: 2MB is used for the kernel.
Even though I could rewrite linux to use a smaller buffer cache, and
give the user more mem, it probably wouldn't pay: performance would
probably generally go /down/ instead of up - it would swap a bit more,
but IO would be slower..
Linus
======================================================================
[next article]
From: [email protected] (n.h.chandler)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Display of current directory in bash prompt
Keywords: bash ksh prompt PS1 PWD HOSTNAME
Message-ID: <[email protected]>
Date: 25 Apr 92 16:21:35 GMT
Followup-To: poster
Organization: AT&T
Lines: 5
Is it possible to display the PATH of the current directory
in the bash prompt? If so, how is it done?
Neville Chandler
[email protected]
[next article]
From: [email protected] (Michael K. Johnson)
Subject: Latin-1 again
Message-ID: <[email protected]>
Date: 25 Apr 92 16:47:12 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Michael K. Johnson)
Organization: The Internet
Lines: 16
From: [email protected] (no, I don't repeat it!)
In article <[email protected]> [email protected]
u (H. Peter Anvin N9ITP) writes:
Is there an interest to have Linux use Latin-1?
Definitely.
CGA/HGC:
Display characters in graphic mode.
True hercules monographics cards have downloadable charsets. herc
calls this "RAMFONT"
michaelkjohnson
[email protected]
[next article]
From: [email protected] (Gregory R. Weiss)
Newsgroups: comp.os.linux,alt.os.linux
Subject: Do not post/crosspost to alt.os linux anymore
Summary: I'm a hypocrite. 🙂
Message-ID: <[email protected]>
Date: 25 Apr 92 16:31:49 GMT
Sender: [email protected] (USENET News System)
Followup-To: comp.os.linux
Organization: not!
Lines: 7
Originator: [email protected]
Nntp-Posting-Host: bonnet.princeton.edu
Alt.os.linux posts have dwindled dramatically recently, after the
creation of comp.os.linux. Now, 95+% of the posts are crossposts. I think
it is safe to stop crossposting, guys. 🙂
--Greg Weiss
[email protected]
End of newsgroup comp.os.linux.
Very nice! Thank you for this wonderful archive. I wonder why I found it only now. Long live the BBS file archives!
This is so awesome! 😀 I’d be cool if you could download an entire archive of this at once, though.
But one thing that puzzles me is the “mtswslnkmcjklsdlsbdmMICROSOFT” string. There is an article about it here. It is definitely worth a read: http://www.os2museum.com/wp/mtswslnk/