800 lines
30 KiB
Plaintext
800 lines
30 KiB
Plaintext
From: Digestifier <Linux-Development-Request@senator-bedfellow.mit.edu>
|
||
To: Linux-Development@senator-bedfellow.mit.edu
|
||
Reply-To: Linux-Development@senator-bedfellow.mit.edu
|
||
Date: Sat, 26 Feb 94 14:13:07 EST
|
||
Subject: Linux-Development Digest #501
|
||
|
||
Linux-Development Digest #501, Volume #1 Sat, 26 Feb 94 14:13:07 EST
|
||
|
||
Contents:
|
||
Re: Specialix driver (Chris Royle)
|
||
Re: GOD SPEAKS ON LINUX! (Rikhardur Egilsson)
|
||
Re: PLEASE use the GPL (Jon Brawn)
|
||
Re: Specialix driver (Robert Sanders)
|
||
Re: accessing BIOS
|
||
Getting nuts... (Philippe Steindl)
|
||
XRN with TERM support? (Gregory Larkin)
|
||
Log Structured File System: Is there one ? (Amrik Thethi)
|
||
Re: undefined symbols in modules (Erik Troan)
|
||
Re: effectiveness of cache ram? (William Henning)
|
||
Re: PLEASE use the GPL (Kai Henningsen)
|
||
Re: Specialix driver (Kai Henningsen)
|
||
Re: Context switch for pthreads (Alfred Longyear)
|
||
RF Info on pty handling (Garrett D'Amore)
|
||
kmalloc fails on 4096 bytes... what is the max now? (Theodore Ts'o)
|
||
stty() and gtty() in C libs... (Theodore Ts'o)
|
||
Serial problem with 0.99.15: tty65: input overrun (Theodore Ts'o)
|
||
Bug in pl15f serial code (Theodore Ts'o)
|
||
Re: linux.cf missing in X11/lib/config ? (R.C.Van-Den-Bergh)
|
||
Re: How to create shared libs ? (R.C.Van-Den-Bergh)
|
||
|
||
----------------------------------------------------------------------------
|
||
|
||
Crossposted-To: gnu.misc.discuss
|
||
From: Chris Royle <car1002@cus.cam.ac.uk>
|
||
Subject: Re: Specialix driver
|
||
Date: Sat, 26 Feb 1994 14:43:15 GMT
|
||
|
||
OK, here's a thought for you all. Linux is wonderful (that wasn't it..).
|
||
However, if it is ever to get anywhere in the world on a serious foot,
|
||
people *must must must* be able to provide systems for it which use source
|
||
which they do not wish to disclose -- eg Motif. [And don't mail me and tell
|
||
me the code concered is non-Linux - I know. And people who insist that the
|
||
whole source to the microcode should be provided because of the GPL are off
|
||
their metaphorical trolleys].
|
||
|
||
In arguing that Specialix cannot implement this driver for the reasons
|
||
discussed, you are hindering the progress and usefulness of Linux. This
|
||
applies not only to the Specialix case, but for others in the future, too.
|
||
If Specialix won't do their driver for this reason, Linux will be seen by
|
||
the world as something which could have been really excellent, but was
|
||
hampered by the daft license it was issued under. If, on the other hand,
|
||
Specialix (and anyone else in the same boat) are allowed to do their driver,
|
||
then Linux will end up being seen as something very useful and going places.
|
||
|
||
I'm disgusted by all this prevaricating that's going on. I suggest Specialix
|
||
get on and do it.
|
||
|
||
Think on it.
|
||
|
||
Chris.
|
||
|
||
--
|
||
Chris Royle "Come rest your head on these two"
|
||
G&CC Undergraduate (Author: E. Hamlin)
|
||
0223 335436 car1002@cus.cam.ac.uk / c@royle.org (Internet)
|
||
0850 668151 car1002@uk.ac.cam.cus (JANET)
|
||
|
||
------------------------------
|
||
|
||
From: rikardur@rhi.hi.is (Rikhardur Egilsson)
|
||
Crossposted-To: comp.os.linux.admin,comp.os.linux.help,comp.os.linux.misc
|
||
Subject: Re: GOD SPEAKS ON LINUX!
|
||
Date: 26 Feb 1994 14:12:14 GMT
|
||
|
||
Plesae correct me if I'm wrong but the method god used was something like:
|
||
|
||
1 ) Joined a university in heaven with access to internet.
|
||
|
||
2 ) Learned how to spoof email by telnetting to port 25.
|
||
|
||
3 ) Wrote the whole message following the outlines of rfc1036.
|
||
|
||
4 ) Sent the fake message to news@localhost by using ( 2 )
|
||
|
||
Correct ?
|
||
|
||
_______________________________________________________
|
||
| Rikhardur Egilsson | |
|
||
| rikardur@rhi.hi.is | the void (shrunk 2x) |
|
||
| TF3RET | |
|
||
=======================================================
|
||
|
||
|
||
|
||
|
||
------------------------------
|
||
|
||
From: jonb@specialix.com (Jon Brawn)
|
||
Subject: Re: PLEASE use the GPL
|
||
Date: Fri, 25 Feb 1994 23:45:55 GMT
|
||
|
||
kenney@u.washington.edu (Michael Kenney) writes:
|
||
|
||
>In article <PCG.94Feb23162638@oracle.aber.ac.uk>,
|
||
>Piercarlo Grandi <pcg@aber.ac.uk> wrote:
|
||
>>>>> On 23 Feb 1994 10:47:18 GMT, kevinl@bruce.cs.monash.edu.au (Kevin
|
||
>>>>> Lentin) said:
|
||
>>
|
||
>[ stuff deleted ]
|
||
>>Kevin> I am also having a hard time coming to terms with your statements
|
||
>>Kevin> that a GPL somehow assures you of support
|
||
>>
|
||
>>Under the GPL anybody can provide commercial value added services (from
|
||
>>distribution to support) for a piece of software, not just the authors.
|
||
>>
|
||
>>If the authors of a "non commercial use" piece of sw are no longer
|
||
>>interested in distributing and supporting it, then it's basically dead
|
||
>>in the water, as the only recourse is for each user to do their own
|
||
>>support; but most companies would rather rely on professional support
|
||
>>services, even for GPL software.
|
||
>>
|
||
>>If the software is GPL'ed, anybody can support it commercially, even if
|
||
>>the original authors have decided to become jewellers or nuns or
|
||
>>whatever. From the user's point of view ther GPL is a guarantee that the
|
||
>>sw need not become orphaned.
|
||
>>
|
||
|
||
>Just because anybody *can* support it doesn't mean anybody *will*. The GPL
|
||
>is no guarantee that there will be anyone supporting the software years from
|
||
>now.
|
||
|
||
BUT there is more CHANCE of it being supported.
|
||
|
||
>If a piece of software becomes useful or popular enough it will stick
|
||
>around and be supported *regardless* of the copyright/license. Look at X
|
||
>and BSD.
|
||
|
||
Ah, but for a piece of ``non-commercial-use-only'' software, this may well
|
||
be ILLEGAL!!!!
|
||
|
||
>----
|
||
>Mike Kenney
|
||
>UW Applied Physics Lab
|
||
>mikek@apl.washington.edu
|
||
|
||
|
||
------------------------------
|
||
|
||
From: gt8134b@prism.gatech.EDU (Robert Sanders)
|
||
Subject: Re: Specialix driver
|
||
Date: 26 Feb 94 00:40:29 GMT
|
||
|
||
dholland@husc7.harvard.edu (David Holland) writes:
|
||
|
||
>gt8134b@prism.gatech.EDU's message of 23 Feb 94 23:59:42 GMT said:
|
||
|
||
> > I would posit that the Linux kernel is the only available implementation
|
||
> > of the Linux kernel interface. The Linux kernel is GPL'ed, therefore
|
||
> > any driver written to interface with the kernel is GPL'ed. Russ's
|
||
> > assertion is correct in the eyes of the FSF.
|
||
|
||
>All right. The Linux kernel is the only available implementation that
|
||
>will run Linux binaries. Therefore Linux binaries have been created to
|
||
>interface with the Linux kernel. The Linux kernel is GPL'd. Therefore,
|
||
>everything which has ever been compiled to run under Linux is
|
||
>automatically GPL'd.
|
||
|
||
Sigh. PLEASE move this to gnu.misc.discuss. Better yet, drop it
|
||
altogether.
|
||
|
||
But since it's here, I don't think that follows at all. Linux binaries
|
||
aren't linked into the kernel. Most Linux binaries are written to a more
|
||
general interface, e.g. POSIX. The FSF position obviously wouldn't say
|
||
that any programs that use GCC-specific extensions are auto-GPL'ed, nor
|
||
that ELISP files are auto-GPL'ed, nor gawk..., etc.
|
||
|
||
--
|
||
_g, '96 --->>>>>>>>>> gt8134b@prism.gatech.edu <<<<<<<<<--- CompSci ,g_
|
||
W@@@W__ |-\ ^ | disclaimer: <---> "Bow before ZOD!" __W@@@W
|
||
W@@@@**~~~' ro|-<ert s/_\ nders | who am I??? ^ from Superman '~~~**@@@@W
|
||
`*MV' hi,ocie! |-/ad! / \ss!! | ooga ooga!! | II (cool)! `VW*'
|
||
|
||
------------------------------
|
||
|
||
From: whitney@galileo.Meakins.McGill.CA ()
|
||
Subject: Re: accessing BIOS
|
||
Date: Sat, 26 Feb 1994 00:54:32 GMT
|
||
|
||
Zeyd M. Ben-Halim (zmbenhal@netcom.com) wrote:
|
||
|
||
: It sounds like what you are trying to do is not Linux specific.
|
||
: The details of switching from protected mode to vm86 mode are
|
||
: in any decent 80386 reference.
|
||
|
||
Yup. I am just picking the mind of Linux types that know
|
||
this like the back of their hand.
|
||
|
||
: >So, the question remains : How do I get from protected mode
|
||
: >to real mode and back ?
|
||
|
||
: If you are talking about linux, then there is a vm86() call.
|
||
|
||
I am not intrested in the v86 feature for my experiment - I want
|
||
to know how to switch to real mode.
|
||
|
||
Time to hit the books. Thanks anyway.
|
||
|
||
Whitney
|
||
|
||
------------------------------
|
||
|
||
From: ilg@imp.ch (Philippe Steindl)
|
||
Crossposted-To: comp.os.linux.help
|
||
Subject: Getting nuts...
|
||
Date: 26 Feb 1994 02:00:22 +0100
|
||
|
||
Hi Linuxers,
|
||
|
||
hm .. is this a bug or not? I compiled pl15 with gcc 2.5.7 and libs 4.5.8.
|
||
I am using dip 3.3.4 to get a slip connection working. This worked under
|
||
pl14. Now, since pl15, my modem doesn't autoanswer an incoming call when
|
||
using dip, even if the modem is on autoanswer (S0=2). I switched back to pl14
|
||
to check this and it worked again. Back again to pl15, it didn't. I recompiled
|
||
getty_ps because I thought it could be the getty ... to no avail. The
|
||
recompiled getty (newest I could find + c patches for pl15) didn't do any
|
||
good. Now is this a bug of pl15, dip or something in the libs?
|
||
|
||
thanx for any help
|
||
|
||
Philippe Steindl
|
||
|
||
|
||
|
||
------------------------------
|
||
|
||
From: greg@equipe.viewlogic.com (Gregory Larkin)
|
||
Subject: XRN with TERM support?
|
||
Date: Sat, 26 Feb 1994 15:57:26 GMT
|
||
|
||
|
||
Hi,
|
||
|
||
I checked on sunsite, but it does appear that there
|
||
is a copy of XRN with TERM support. Does anyone know
|
||
if this has been done yet?
|
||
|
||
Is there a FAQ somewhere that tells how applications
|
||
using network code are ported to use TERM instead?
|
||
Does anyone have any insight on how the Mosaic 2.2
|
||
with TERM support was done?
|
||
|
||
Thanks,
|
||
--
|
||
Greg Larkin
|
||
Viewlogic Systems, Inc.
|
||
Marlboro, Massachusetts, USA
|
||
greg@Viewlogic.COM
|
||
|
||
------------------------------
|
||
|
||
From: at@setanta.demon.co.uk (Amrik Thethi)
|
||
Subject: Log Structured File System: Is there one ?
|
||
Date: Fri, 25 Feb 1994 14:27:25 GMT
|
||
|
||
Is there, or will there be a log structured file system for Linux.
|
||
|
||
This seems to be the trend in file systems in general, because of their
|
||
inherent advantages.
|
||
|
||
Is anybody working on a port or designing one from scratch?
|
||
|
||
I thought of writing one, but I need more info on the best way to do it.
|
||
|
||
|
||
--
|
||
Amrik Thethi. Tel. +223 421 008 Fax. +223 421 024
|
||
Setanta Software Ltd. Internet: at@setanta.demon.co.uk
|
||
Cambridge, UK.
|
||
|
||
------------------------------
|
||
|
||
From: ewt@sunSITE.unc.edu (Erik Troan)
|
||
Subject: Re: undefined symbols in modules
|
||
Date: 26 Feb 1994 16:50:55 GMT
|
||
|
||
In article <1994Feb26.105501.5976@pe1chl.ampr.org>,
|
||
Rob Janssen <pe1chl@rabo.nl> wrote:
|
||
>In <2klvfa$gtv@bigblue.oit.unc.edu> ewt@sunSITE.unc.edu (Erik Troan) writes:
|
||
>
|
||
>I think you should put the symbol in kernel/ksyms.S, the list of symbols
|
||
>exported to modules.
|
||
>
|
||
|
||
Thanks to both of the repliers. I didn't see this file mentioned anywhere,
|
||
but it is what I need.
|
||
|
||
Thanks,
|
||
|
||
Erik
|
||
|
||
--
|
||
========================================================================
|
||
"Could I bend yer ear fer a tick?" ewt@sunsite.unc.edu = Erik Troan
|
||
sasewt@unx.sas.com
|
||
- Strictly Ballroom
|
||
|
||
------------------------------
|
||
|
||
From: bhenning@bhami.wimsey.com (William Henning)
|
||
Subject: Re: effectiveness of cache ram?
|
||
Date: Sat, 26 Feb 1994 05:32:19 GMT
|
||
|
||
In article <2khf69$8o9@jake.esu.edu> tbriggs@myhost.subdomain.domain (Tom Briggs) writes:
|
||
>I believe, from my experience, that if you are talking 386, you want
|
||
>as much cache as you can get. If you are talking 486, RAM cache is not
|
||
>really all that extravagent. I have seen 486/33 DX with 64k, 128, and 256k
|
||
>and all of them performed the same.
|
||
|
||
The above is absolutely true for trivial benchmarks such as SI, Landmark,
|
||
etc., that fit into the 8k on-chip cache. The above is FALSE for large
|
||
programs.
|
||
|
||
>Try this: If you can get access to a 486/33dx, run some benchmark program
|
||
>and determine some reasonable measure of the system speed. Then, go into
|
||
>your CMOS and disable the EXTERNAL ram cache, and run your benchmark again.
|
||
>IF you see more than 1-2% difference, please let me know, and I'll gladly
|
||
>eat these words. Now, if you want to see more, disable the INTERNAL &
|
||
>EXTERNAL cache, and run your benchmark - your 486/33 is now performing
|
||
>about like a 386/25... Now, for your final test, disable INTERNAL and
|
||
>enable EXTERNAL, and see what it does to your performance... it helps,
|
||
>but it does nowhere nearly as much as the 8k internal ram cache built
|
||
>into your 486... thats the key...
|
||
|
||
I did this a few months ago to prove the value of an external cache to some
|
||
skeptics at work.
|
||
|
||
Test setup:
|
||
|
||
VLB Intel 486DX2-66, 256k cache, 16Mb ram, WD250 mb HD, OS/2 v2.1
|
||
|
||
Test:
|
||
|
||
Compilation/linking of ~100k of C source code with icc.
|
||
|
||
With internal/external caches enabled: about 2 minutes
|
||
With internal cache enabled, external disabled: about 6 minutes
|
||
With internal and external caches disabled: about 8 minutes
|
||
With internal cache disabled, external enabled: about 3 minutes
|
||
|
||
(Numbers from memory)
|
||
|
||
>So, no, I would say that more memory is a better thing than more cache.
|
||
|
||
More memory is ALLWAYS a good thing. If you swap a lot, cache efficiency
|
||
gets lost in the time spent swapping. If you are not swapping, the more
|
||
cache the better.
|
||
|
||
>There are many many things that are better than having more cache. I've
|
||
>got a 486/33 board with 64k cache -- I used to have 486/33 256k and I
|
||
>never missed it at all.....
|
||
|
||
Depends on what you are working on - if you are working on I/O bound
|
||
programs you will not notice as big a difference.
|
||
|
||
> Tom Briggs
|
||
|
||
Bill
|
||
|
||
--
|
||
----
|
||
|
||
bhenning@bhami.wimsey.com - Linux & OS/2 user at home, OS/2 developer at work
|
||
|
||
------------------------------
|
||
|
||
Date: 26 Feb 1994 14:44:00 +0100
|
||
From: kai@khms.westfalen.de (Kai Henningsen)
|
||
Subject: Re: PLEASE use the GPL
|
||
|
||
becker@super.org wrote on 24.02.94 in <1994Feb24.051058.1397@super.org>:
|
||
|
||
> There is another reason for using the GPL, even when you don't directly
|
||
> profit from the code. It doesn't have the holes that a shorter copyright
|
||
> notice has. A few days ago I ran across an error message that looked very
|
||
> familiar. Verrryy familiar. I wrote it. I had distributed that code over
|
||
> the net, the code *was* copyrighted, and it had my name and my employer's
|
||
> name in the source, binary and documentation. The binary-and-documentation
|
||
> package I got had no mention of the code's origin or original copyright. If
|
||
> I had used the GPL instead of the "no-brand-name" copyright, I could point
|
||
> out that removing the original copyright notice was an explicit violation of
|
||
> the license. Without the GPL they'll probably claim that my code was
|
||
> essentially public domain, and that they can distribute copies without
|
||
> giving credit :-<.
|
||
|
||
As I understand copyright (which is not that much), as long as you don't
|
||
grant them any rights, the only rights they have is the fair use stuff.
|
||
|
||
If there was a copyright mentioned, I do not see how someone can possibly
|
||
be justified in removing it.
|
||
|
||
They might not need to give more credit than is already there. But I don't
|
||
see how they can remove any that is there, unless you explicitely allow
|
||
them to.
|
||
|
||
|
||
|
||
Kai
|
||
--
|
||
Internet: kh@ms.maus.de, kai@khms.westfalen.de
|
||
Bang: major_backbone!{ms.maus.de!kh,khms.westfalen.de!kai}
|
||
## CrossPoint v2.93 ##
|
||
|
||
------------------------------
|
||
|
||
Date: 26 Feb 1994 15:04:00 +0100
|
||
From: kai@khms.westfalen.de (Kai Henningsen)
|
||
Subject: Re: Specialix driver
|
||
|
||
becker@super.org wrote on 24.02.94 in <1994Feb24.170824.10684@super.org>:
|
||
|
||
> under the copyright. It *is* a derivative work because it required the
|
||
> unique existing GPL code to develop it. Few readers would disagree that a
|
||
|
||
This can't be. Otherwise, each and every program built with gcc that uses
|
||
*any* gcc extension would fall under the GPL, and the FSF explicitely
|
||
states that this is *not* the case.
|
||
|
||
> (non-trivial) patch file is covered under the terms of the original
|
||
> copyright. Merely collecting such patch into a separate object file does
|
||
> not change its status. (In contrast, if the code in question is a separate
|
||
> and distinct program, then it is not covered by the original copyright.)
|
||
|
||
Now here you argue that the original code is needed to *run* the beast.
|
||
That is a completely different animal.
|
||
|
||
Anyway, my copy of the GPL says
|
||
|
||
> GNU GENERAL PUBLIC LICENSE
|
||
> Version 2, June 1991
|
||
|
||
[...]
|
||
|
||
> The precise terms and conditions for copying, distribution and
|
||
>modification follow.
|
||
>
|
||
> GNU GENERAL PUBLIC LICENSE
|
||
> TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION
|
||
>
|
||
> 0. This License applies to any program or other work which contains
|
||
>a notice placed by the copyright holder saying it may be distributed
|
||
>under the terms of this General Public License. The "Program", below,
|
||
>refers to any such program or work, and a "work based on the Program"
|
||
>means either the Program or any derivative work under copyright law:
|
||
>that is to say, a work containing the Program or a portion of it,
|
||
>either verbatim or with modifications and/or translated into another
|
||
>language. (Hereinafter, translation is included without limitation in
|
||
>the term "modification".) Each licensee is addressed as "you".
|
||
|
||
[...]
|
||
|
||
From this text, I do not see how it would be possible to argue that
|
||
anything that needs GPL'd software to do anything useful falls itself
|
||
under the GPL.
|
||
|
||
On the contrary, it says quite clearly that it has to "contain" the
|
||
original or a portion.
|
||
|
||
In fact, with regard to patches, if patches were GPL'd by default, why
|
||
does the FSF argue that it needs an explicit release for each and every
|
||
patch before it can be integrated?
|
||
|
||
I personally think that whoever argues that working-only-with-GPL'd-soft-
|
||
makes-GPL'd has no leg to stand on.
|
||
|
||
Kai
|
||
--
|
||
Internet: kh@ms.maus.de, kai@khms.westfalen.de
|
||
Bang: major_backbone!{ms.maus.de!kh,khms.westfalen.de!kai}
|
||
## CrossPoint v2.93 ##
|
||
|
||
------------------------------
|
||
|
||
From: longyear@netcom.com (Alfred Longyear)
|
||
Subject: Re: Context switch for pthreads
|
||
Date: Sat, 26 Feb 1994 17:27:46 GMT
|
||
|
||
z1g192@rick.cs.ubc.ca (Christopher Andrew Smith) writes:
|
||
|
||
>My questions are:
|
||
|
||
>1) Since threads are contained within one UNIX process, do I have to
|
||
> think about switching between Task State Segments, or is this only
|
||
> a consideration when switching UNIX processes?
|
||
|
||
No. Each process has one and only one TSS. There is no need to switch TSS
|
||
blocks since you have only one to switch.
|
||
|
||
>2) I know that I have to save the general registers. Do I have to save
|
||
> all the segment registers? (ES,CS,SS,DS,FS,and GS) (ie is it possible
|
||
> for one UNIX process to span across multiple memory segments?)
|
||
|
||
Linux uses only a few values for the selector registers. When your user level
|
||
code is executing, there is only one value. (Linux calls it USER_CS and
|
||
USER_DS. They are presently 0x23 and 0x2B, respectively.)
|
||
|
||
Like TSS values, you need only recognize that you will have only one value
|
||
for the code selector and data selector(s). So, no, you do not need to switch
|
||
these registers.
|
||
|
||
Linux uses a "flat 32 bit" address space. The access to memory is controlled
|
||
by the page tables, not by the descriptor tables alone. So, yes, there are
|
||
"multiple" segments, if you consider that a segment is a shared block of memory
|
||
such as "user code" and "shared library code" and "user data" and "shared
|
||
memory data". However, they all fit within the 32 bits which are addressable by
|
||
one selector.
|
||
|
||
(It is the iNTEL processor which requires separate selectors for code/data,
|
||
even if they point to the same address range.)
|
||
|
||
>3) Does the context switching code have to run at a higher processor
|
||
> privelege level than the usual user mode?
|
||
|
||
There is only one privilege level for users processes. All user processes
|
||
run at ring 3. What you do within the user code, you will always be at
|
||
ring 3.
|
||
|
||
>4) What are the calling conventions on the 386? I'm going to be
|
||
> linking the context switch with C code, so I need to know:
|
||
>
|
||
> a) Where are procedure parameters passed? (the stack?)
|
||
|
||
Parameters are pushed on the stack. You may find some additional insight
|
||
if you write a small piece of code and then compile it with the "-S" option
|
||
which leaves the assembly language file.
|
||
|
||
answer = call_proc (p1, p2, p3);
|
||
|
||
results in:
|
||
|
||
pushl _p3 << pushed right-to-left for C language
|
||
pushl _p2
|
||
pushl _p1
|
||
call _call_proc
|
||
add $12,%esp << Discard parameters. Sometimes ignored if optimized.
|
||
movl %eax,_answer << for fixed point scalars, result is in EAX.
|
||
<< for floating point, result is in fp(0)
|
||
|
||
If you have a "structure" returned, the structure is pre-allocated on
|
||
the stack and the address to this structure is passed as an additional
|
||
parameter. Try compiling small code and looking at the assembly output.
|
||
|
||
> b) Where do I put return values? (stack or general register)
|
||
|
||
See above.
|
||
|
||
>I'd appreciate if someone could help me with these questions, or point me
|
||
>to the appropriate reference.
|
||
|
||
There are many references. There are many good books on compiler design.
|
||
For Linux, try Kernel Hackers Guide (khg* on sunsite's doc directory).
|
||
For compiler design, I still like the "dragon" book (second edition).
|
||
|
||
>Once I get the pthreads package working and stable on Linux, I will probably
|
||
>make it public domain, with permission of the author of course, so the sooner
|
||
>I can make it work, the sooner everybody can write threaded applications
|
||
>on top of Linux.
|
||
|
||
I am still of the opinion that threads are best implemented in the kernel
|
||
as they are done in Windows NT, Chicago, and OS/2. However, pthreads is
|
||
better than nothing. :-) Thanks.
|
||
|
||
------------------------------
|
||
|
||
From: garrett@netcom.com (Garrett D'Amore)
|
||
Subject: RF Info on pty handling
|
||
Date: Sat, 26 Feb 1994 17:28:24 GMT
|
||
|
||
Hello Linuxers,
|
||
|
||
I am trying to find information about the correct programming of pty's.
|
||
Essentially, what I want to do is have a program that runs a shell, but
|
||
examines all the data coming to or from the shell for a couple of escape
|
||
sequences, and redirects data to "lpr" between certain escape codes.
|
||
The program will be used as a layer of the linux (or other OS) console
|
||
to allow it to work with my vtprint program.
|
||
|
||
I'd like the shell to be unaware that it's not being run from a "true"
|
||
tty -- e.g. the isatty() call in the shell should return true. So, a
|
||
simple filter is unacceptable.
|
||
|
||
I know it is possible to do what I want, because Xterm's do it, and I
|
||
think even emacs does it. However, I have found it difficult to find
|
||
information on this topic, and the xterm sources I have looked at have
|
||
been difficult to understand. Can anyone out there *please* give me
|
||
pointer to info on pty handling (a book, perhaps, or an on-line source
|
||
of information?)
|
||
|
||
Thank you very much.
|
||
|
||
=================================================================
|
||
Garrett D'Amore | garrett@netcom.com
|
||
Network Programmer/Analyst | Available for hire !!
|
||
SDSU Computer Science Major | Go Aztecs!
|
||
=================================================================
|
||
|
||
Ask me about Linux, the free 32-bit Unix-like OS for PCs!
|
||
|
||
Also, ask for my resume if you are an employer in San Diego
|
||
seeking a qualified network programmer or administrator.
|
||
|
||
------------------------------
|
||
|
||
From: tytso@ATHENA.MIT.EDU (Theodore Ts'o)
|
||
Subject: kmalloc fails on 4096 bytes... what is the max now?
|
||
Date: 26 Feb 1994 13:54:30 -0500
|
||
Reply-To: tytso@ATHENA.MIT.EDU (Theodore Ts'o)
|
||
|
||
From: jmv@Rabbit.edu (jmv)
|
||
Date: 25 Feb 1994 05:24:28 GMT
|
||
|
||
The max kmalloc you can do is 4080. Why ? because there is a 16 bytes
|
||
header. As far as I could tell (tell me if I am wrong), this is only defined
|
||
in linux/include/linux/msg.h:
|
||
#define MAXMSG 4080
|
||
|
||
This is only true if you have the Kernel debugging malloc option turned
|
||
on. If it is off, then there is no 16 byte header, and you can malloc
|
||
up to 4096 bytes.
|
||
|
||
- Ted
|
||
|
||
------------------------------
|
||
|
||
From: tytso@ATHENA.MIT.EDU (Theodore Ts'o)
|
||
Subject: stty() and gtty() in C libs...
|
||
Date: 26 Feb 1994 13:54:30 -0500
|
||
Reply-To: tytso@ATHENA.MIT.EDU (Theodore Ts'o)
|
||
|
||
From: pat@garion.it.com.au (Pat Mackinlay)
|
||
Date: 23 Feb 1994 02:55:33 +0800
|
||
|
||
Hi all. Just a quick question to HJ or whoever's looking after the
|
||
C library these days. The kernel would appear to have stty() and gtty()
|
||
entry points, but these are not exported from any of the standard
|
||
libraries (as of 4.5.19, anyhow). Up until now, I've been inserting:
|
||
|
||
_syscall2(int,stty,int,fd,void *,sg);
|
||
_syscall2(int,gtty,int,fd,void *,sg);
|
||
|
||
into code that wants to use these BSD things, and it seems to do the
|
||
trick. Could we perhaps see these in the BSD library at some stage?
|
||
|
||
.... and this is how they are implemeneted in the kernel, which is why I
|
||
suspect they haven't been implemented in the library. You're probably
|
||
better off replacing them with the appropriate POSIX calls.
|
||
|
||
asmlinkage int sys_stty(void)
|
||
{
|
||
return -ENOSYS;
|
||
}
|
||
|
||
asmlinkage int sys_gtty(void)
|
||
{
|
||
return -ENOSYS;
|
||
}
|
||
|
||
- Ted
|
||
|
||
|
||
|
||
------------------------------
|
||
|
||
From: tytso@ATHENA.MIT.EDU (Theodore Ts'o)
|
||
Subject: Serial problem with 0.99.15: tty65: input overrun
|
||
Date: 26 Feb 1994 13:54:30 -0500
|
||
Reply-To: tytso@ATHENA.MIT.EDU (Theodore Ts'o)
|
||
|
||
From: steve@thelake.mn.org (Steve Yelvington)
|
||
Date: Wed, 23 Feb 1994 20:06:14 GMT
|
||
|
||
Benjamin Ryzman (zarkdav@eddy.frmug.fr.net) wrote:
|
||
|
||
> I'm using an IDE disk, and I get this message (a lot of them actually)
|
||
> while running ifcico (Internet to Fidonet Copy In/Copy Out) with Zmodem
|
||
> protocol, while I don't get them when running Taylor's uucico with the
|
||
> 'i' protocol...
|
||
|
||
This thread is interesting; I'm having a similar problem, but it affects
|
||
only the mouse port (cua0). It works fine until my modem port (cua2 or 3,
|
||
interrupts 5 and 9, I've tried them both) is accessed. At that point the
|
||
modem (internal DSI 14.4+) works fine, but the 1200bps serial mouse just
|
||
freezes. Later I find the tty64 error message on a spare virtual screen. The
|
||
mouse is an Identity 3-button model; I've tried it in both Microsoft and
|
||
Mouse Systems modes.
|
||
|
||
Once the modem port is free, the mouse comes back to life.
|
||
|
||
I suspect it's a hardware problem; your modem is probably nuking IRQ 4,
|
||
even when you have it configured to ise IRQ 5 or IRQ 9. (One
|
||
possibility is that if your internal modem is configured using jumpers,
|
||
that *two* jumpers have been installed, for both IRQ 4 and IRQ 5 or 9.
|
||
If you're really lucky, that's the problem, and you can fix it pretty
|
||
easily.)
|
||
|
||
Try swapping out your internal modem with someone else's; it will
|
||
probably fix the problem.
|
||
|
||
- Ted
|
||
|
||
|
||
------------------------------
|
||
|
||
From: tytso@ATHENA.MIT.EDU (Theodore Ts'o)
|
||
Subject: Bug in pl15f serial code
|
||
Date: 26 Feb 1994 13:54:38 -0500
|
||
Reply-To: tytso@ATHENA.MIT.EDU (Theodore Ts'o)
|
||
|
||
From: kevinl@bruce.cs.monash.edu.au (Kevin Lentin)
|
||
Date: 22 Feb 1994 00:59:01 GMT
|
||
|
||
I discovered a small problem in the serial drivers last night I think. I
|
||
just put my internal modem back in my machine after a small absence (since
|
||
before pl15 I'd say - maybe mid to late 14's) and it is configured on IRQ5.
|
||
Unfortunately, Linux detect it at IRQ4. If I try IRQ2, Linux detects it as
|
||
IRQ3. Needless to say this means that I can't use my modem on that irq
|
||
which I really would like to do.
|
||
|
||
Use the setserial program to set the irq at boot time. The automatic
|
||
IRQ detection does not always work --- it's very dependent on your
|
||
hardware system, and given the brain-damaged ISA bus, there's no way to
|
||
make it dependable.
|
||
|
||
- Ted
|
||
|
||
------------------------------
|
||
|
||
From: rcv@ukc.ac.uk (R.C.Van-Den-Bergh)
|
||
Subject: Re: linux.cf missing in X11/lib/config ?
|
||
Date: Sat, 26 Feb 94 17:26:08 GMT
|
||
|
||
Hi,
|
||
|
||
having installed slackware 1.1.1. a while ago (December), I seem to be
|
||
missing /usr/X11/lib/X11/config/linux.cf (my /usr/X386 is a link to
|
||
/usr/X11). Should I have one ? If so what is supposed to be in there ? At
|
||
the moment, I have created my own file, based on x386.cf. (x386.cf was the
|
||
only one with a reference to Linux in it).
|
||
|
||
The reason I ask, is that I want to recompile InterViews 3.1 on my box, BUT
|
||
I can't get it to work. The problem I have is (among others) that I am
|
||
missing a number of config files. I have found *some* of them, and am
|
||
improvising the others.
|
||
|
||
Anyone with a valid /usr/X11/lib/X11/config/linux.cf, please mail me a
|
||
copy..
|
||
|
||
CEdric
|
||
--
|
||
Zhaumer, High Priest of Amalgaer, the dwarven God of Something.
|
||
|
||
------------------------------
|
||
|
||
From: rcv@ukc.ac.uk (R.C.Van-Den-Bergh)
|
||
Subject: Re: How to create shared libs ?
|
||
Date: Sat, 26 Feb 94 17:26:40 GMT
|
||
|
||
Hi (again),
|
||
|
||
anyone know where I can get some information on creating shared libraries ?
|
||
I would like to limit disk space usage by making a shared version of libIV.a
|
||
and libUnidraw.a (InterViews 3.1).
|
||
|
||
Thanks,
|
||
|
||
CEdric
|
||
--
|
||
Zhaumer, High Priest of Amalgaer, the dwarven God of Something.
|
||
|
||
------------------------------
|
||
|
||
|
||
** FOR YOUR REFERENCE **
|
||
|
||
The service address, to which questions about the list itself and requests
|
||
to be added to or deleted from it should be directed, is:
|
||
|
||
Internet: Linux-Development-Request@NEWS-DIGESTS.MIT.EDU
|
||
|
||
You can send mail to the entire list (and comp.os.linux.development) via:
|
||
|
||
Internet: Linux-Development@NEWS-DIGESTS.MIT.EDU
|
||
|
||
Linux may be obtained via one of these FTP sites:
|
||
nic.funet.fi pub/OS/Linux
|
||
tsx-11.mit.edu pub/linux
|
||
sunsite.unc.edu pub/Linux
|
||
|
||
End of Linux-Development Digest
|
||
******************************
|