Linux-Development-Sys Digest #404, Volume #6 Mon, 15 Feb 99 22:14:30 EST
Contents:
Re: 2.2 and Binutils (BL)
Re: SMP Support (Daniel Robert Franklin)
Re: 2.2 and Binutils (Peter Mardahl)
2.2 and Binutils ("garry flood")
Kernel Errors under 2.2.1 ("Geoff Thornton")
Re: How to detect CPU clock speed? (Usseglio_Gaudi Francesco)
Re: "No Available PTYS" with 2.2.0 (Tor Arntsen)
Re: Internal PCI modem (Julian Robert Yon)
Re: Ncurses with g++ ("T.E.Dickey")
Re: Internal PCI modem (Johan Kullstam)
kernel 2.2.1 spurious APIC interrupt, ayiee, should never happen (Buffat Marc)
Re: Can I extend struct kernel_stat? (Kalle Olavi Niemitalo)
Re: how to write linux device driver (Sumanth Sukumar)
Re: SMP Support (Adam P. Jenkins)
Bypassing Telnet limit on SGI (John Simpson)
how to create new ptys and ttys? (Bob Berman)
spurious inter-package dependencies (Stefan Monnier)
Re: Ncurses with g++ (Bryan Hackney)
Re: Linux developers working with AIC7890 ? ([EMAIL PROTECTED])
Re: Ncurses with g++ (Don Waugaman)
cold.apps or cold.system for libraries? (Phil Howard)
Re: spurious inter-package dependencies (Emile van Bergen)
Re: "No Available PTYS" with 2.2.0 (H. Peter Anvin)
----------------------------------------------------------------------------
From: BL <[EMAIL PROTECTED]>
Subject: Re: 2.2 and Binutils
Date: 15 Feb 1999 08:22:13 GMT
Reply-To: no.spambots.please
use bison.
garry flood <[EMAIL PROTECTED]> wrote:
: I just upgraded to 2.2 and decided to upgrade binutils to 2.1.123 (sp)..
: You know the one advertised as required for 2.2. WHen I tried to build ld,
: it required flex. When I tried to build flex, it required YACC. I have
: been trying for 2 days to find YACC on the net. Does anybody klnow where to
: find it?
: The reason I think I need it, I run ./configure and configure bombs in the
: flex directory complaining about missing YACC.
: Thanks in advance
: GF
--
AntiSpam: For email, change all 'zero' chars to letter 'o' chars.
------------------------------
From: [EMAIL PROTECTED] (Daniel Robert Franklin)
Subject: Re: SMP Support
Date: 15 Feb 99 06:47:27 GMT
"garry flood" <[EMAIL PROTECTED]> writes:
>Now that we know how to tell if the new 2.2 kernel knows about more than 1
>CPU, does anybody know of what types of applcations are written w/threads
>and can take advantage of more than one CPU. ie compilers, linkers,
>Powertools. I heard rumors that Adobe Photoshop in the NT environment is
>thread aware, and runs nicely with SMP.
Well, one thing you can do is this:
make -j N
where N is the number of jobs you want to run in parallel... speeds up
kernel compiles a bit :-) I think Gimp 1.1.x is multithreaded.
- Daniel
--
******************************************************************************
* Daniel Franklin - Postgraduate student in Electrical Engineering
* [EMAIL PROTECTED]
******************************************************************************
------------------------------
From: [EMAIL PROTECTED] (Peter Mardahl)
Subject: Re: 2.2 and Binutils
Date: 15 Feb 1999 00:46:42 -0800
In article <BSPx2.6701$[EMAIL PROTECTED]>,
garry flood <[EMAIL PROTECTED]> wrote:
>I just upgraded to 2.2 and decided to upgrade binutils to 2.1.123 (sp)..
>You know the one advertised as required for 2.2. WHen I tried to build ld,
>it required flex. When I tried to build flex, it required YACC. I have
>been trying for 2 days to find YACC on the net. Does anybody klnow where to
>find it?
>
>The reason I think I need it, I run ./configure and configure bombs in the
>flex directory complaining about missing YACC.
>
Hmm, I think maybe you use "bison" instead of "yacc". In any case,
RedHat Linux ships YACC as an RPM:
rpm -qf /usr/bin/yacc
byacc-1.9-8
So you can just get a binary of it. Failing that, you can get
source code from an SRPM from RedHat.
PeterM
------------------------------
From: "garry flood" <[EMAIL PROTECTED]>
Subject: 2.2 and Binutils
Date: Sun, 14 Feb 1999 23:19:30 -0800
I just upgraded to 2.2 and decided to upgrade binutils to 2.1.123 (sp)..
You know the one advertised as required for 2.2. WHen I tried to build ld,
it required flex. When I tried to build flex, it required YACC. I have
been trying for 2 days to find YACC on the net. Does anybody klnow where to
find it?
The reason I think I need it, I run ./configure and configure bombs in the
flex directory complaining about missing YACC.
Thanks in advance
GF
------------------------------
From: "Geoff Thornton" <[EMAIL PROTECTED]>
Subject: Kernel Errors under 2.2.1
Date: Mon, 15 Feb 1999 20:06:36 +1000
just upgraded my kernel from 2.0.36 to 2.2.1 on a red hat V 5.2 release
and now I get the following messages in /var/log/messages
error seeking in /dev/kmem
error adding kernel module table entry
can anyone tell me what they are and how to fix them?
also after init tries to boot I get the message:
request_module [net-pf-1]: fork failed, errno 11
what the heck does this mean? the system seems to hang for a little bit then
boots up fine, are these errors serious?
--
Regards Geoff :)
==================
[EMAIL PROTECTED]
------------------------------
From: Usseglio_Gaudi Francesco <[EMAIL PROTECTED]>
Subject: Re: How to detect CPU clock speed?
Date: Mon, 15 Feb 1999 11:42:35 +0100
[EMAIL PROTECTED] wrote:
>
> Hi,
>
> I need to find out the CPU clock speed. cat'ing files in /proc didn't help,
> neither did apropos search on a box. It seems (at least to me) that this value
> must be readily available and I'm simply overlooking something. API call is
> preferred (nlist would do just fine if I knew the symbol to search for).
>
> And no, I cannot use bogomips because of certain restrictions on the
> uniformity of the values, imposed by the project.
>
> TIA
> igor
> --
> Igor Shpigelman "Stand aside in the coming battle and though
> Yet Another UNIX Hacker thou fightest be not thou the warrior." M.C.
>
> -----------== Posted via Deja News, The Discussion Network ==----------
> http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
New kernels (2.2.x) have in /proc/cpu the clock speeds of the cpus.
There must be somewhere in the kernel source...
Hope this helps..
Bye
--
==+==
...era un mondo adulto...
...si sbagliava da professionisti.
Paolo Conte
------------------------------
From: [EMAIL PROTECTED] (Tor Arntsen)
Subject: Re: "No Available PTYS" with 2.2.0
Date: 15 Feb 1999 10:55:10 GMT
In article <[EMAIL PROTECTED]>,
David Moenssen <[EMAIL PROTECTED]> writes:
>I've downloaded kernel version 2.2.0 and have encountered a bothersome
[..]
>xterm: no available ptys
Try <http://roadrunner.swansea.linux.org.uk/clue.html>
for 2.2 migration hints.
- Tor
------------------------------
From: Julian Robert Yon <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Internal PCI modem
Date: Mon, 15 Feb 1999 11:40:26 +0000
Stefan Monnier wrote:
>
> >>>>> "Julian" == Julian Robert Yon <[EMAIL PROTECTED]> writes:
> > petition to request that companies manufacture _real_ modems possible?...
>
> Convince people to stop buying winmodems ?
> Buy real modems ?
The point is that the modem was not sold as a winmodem. Quite obviously
if I had been told the truth when I asked I wouldn't have bought it. It
is getting increasingly hard to find a high speed modem which is any use
- at least at a sensible price. When building a system I object to
having to buy the modem separately from everything else, because a
particular supplier only sells "new technology".
Julian
~~~~~~~~~~~~~~~~~~~~
[EMAIL PROTECTED]
[EMAIL PROTECTED]
------------------------------
From: "T.E.Dickey" <[EMAIL PROTECTED]>
Subject: Re: Ncurses with g++
Date: 15 Feb 1999 12:27:56 GMT
Bryan Hackney <[EMAIL PROTECTED]> wrote:
> The curses and ncurses APIs are pathetic ancient
> messes. The problems you are having are probably due to one of the
> hundreds of macros in curses.h.
> You need to isolate to the greatest extent possible the curses
> calls, and include curses.h in as few (one) of your sources as
> possible.
well, that certainly is a nonproductive comment.
--
Thomas E. Dickey
[EMAIL PROTECTED]
http://www.clark.net/pub/dickey
------------------------------
Crossposted-To: comp.os.linux.setup
Subject: Re: Internal PCI modem
From: Johan Kullstam <[EMAIL PROTECTED]>
Date: 15 Feb 1999 08:36:34 -0500
Julian Robert Yon <[EMAIL PROTECTED]> writes:
> Stefan Monnier wrote:
> >
> > >>>>> "Julian" == Julian Robert Yon <[EMAIL PROTECTED]> writes:
> > > petition to request that companies manufacture _real_ modems possible?...
> >
> > Convince people to stop buying winmodems ?
> > Buy real modems ?
>
> The point is that the modem was not sold as a winmodem. Quite obviously
> if I had been told the truth when I asked I wouldn't have bought it. It
> is getting increasingly hard to find a high speed modem which is any use
> - at least at a sensible price. When building a system I object to
> having to buy the modem separately from everything else, because a
> particular supplier only sells "new technology".
this is a common problem when buying anything bundled. they lure you
in with one good component, e.g., fast cpu, and then give substandard
components all around. bundling is about reducing choice and
unloading less desirable parts with the desirable.
for example, a balanced system good for most usage would have a modest
speed cpu, scsi disk drives and a real modem. the money saved on the
cpu would go toward the better components. you wouldn't need such a
great cpu since the scsi and real modem would be sharing the load
somewhat.
bundlers do not what to sell you a balanced system. they will sell a
draw item at or below cost and make up the difference by adding low
cost items at a vastly inflated price. the principle applies to
computers, automobiles (great car with overpriced and low quality
radio), mc-meals (burger with drink and fries).
it's all a variation on bait and switch.
--
Johan Kullstam [[EMAIL PROTECTED]] Don't Fear the Penguin!
------------------------------
From: [EMAIL PROTECTED] (Buffat Marc)
Subject: kernel 2.2.1 spurious APIC interrupt, ayiee, should never happen
Date: 15 Feb 1999 14:20:33 GMT
Hello,
I have tried to compile the new Linux kernel 2.2.1 on a dual Dell PPRO
and I got the following kernel message (SMP kernel enable)
"spurious APIC interrupt, ayiee, should never happen"
Is it a bug of the PPRO or of the kernel ?
I was running the kernel 2.1.43 in SMP mode without problem
Thanks in advance
Marc BUFFAT
--
Marc BUFFAT, Prof. Universite Claude Bernard LYON I
tel: (33) 4/72/18/61/6 (LMFA) ou (33) 4/72/43/10/05 (UCBL)
fax: (33) 4/78/64/71/45
[EMAIL PROTECTED] | http://mecapar.univ-lyon1.fr
------------------------------
From: Kalle Olavi Niemitalo <[EMAIL PROTECTED]>
Subject: Re: Can I extend struct kernel_stat?
Date: 15 Feb 1999 19:10:35 +0200
Kalle Olavi Niemitalo <[EMAIL PROTECTED]> writes:
> unsigned int irqs[NR_CPUS][NR_IRQS];
> +#ifdef CONFIG_8259A_IRQGLITCH
> + unsigned int irqglitches[NR_CPUS][2];
> +#endif
> unsigned int ipackets, opackets;
I've now done this (with the #include <linux/config.h>) and it
works fine. I'll test the new version of the patch some more and
publish it.
------------------------------
From: Sumanth Sukumar <[EMAIL PROTECTED]>
Subject: Re: how to write linux device driver
Date: Mon, 15 Feb 1999 16:31:19 -0500
Buy LINUX Device Drivers by Alessandro Rubini. Good book/Bad Translation.
Sumanth
On Mon, 8 Feb 1999, xiaodong wrote:
> does anyone know a good web site where I can learn how to write linux
> device drivers? please send you reply to my email, thank you very much!
>
> --
> Phillip Zhang
> [EMAIL PROTECTED]
>
>
>
>
>
------------------------------
Subject: Re: SMP Support
From: [EMAIL PROTECTED] (Adam P. Jenkins)
Date: 15 Feb 1999 17:25:17 -0500
[EMAIL PROTECTED] (Rob Komar) writes:
> Bucky4me ([EMAIL PROTECTED]) wrote:
> : I recompiled my kernel for smp but do not know how to check if the OS see's
> : both. How can I do this?
>
> cat /proc/cpuinfo
>
> should show two CPUs. Also, the start of the boot-up messages show
> if both are recognized (they're probably stored in /var/log/messages
> if `dmesg' doesn't show them).
Also, try starting two CPU-intensive tasks at once when the system is
idle, and use top to see if they're both getting close to 100% CPU
time. This is what should normally happen with dual CPUs.
--
Adam P. Jenkins
[EMAIL PROTECTED]
------------------------------
From: John Simpson <[EMAIL PROTECTED]>
Subject: Bypassing Telnet limit on SGI
Crossposted-To: comp.sys.sgi.admin,comp.unix.programmer
Date: Mon, 15 Feb 1999 23:15:11 GMT
I'm doing some performance analysis involving large numbers of remote
Telnet users and have run into the limit of 999 simultaneous Telnet
sessions on SGI's IRIX 6.4.
Is there any way to get around this limit? Does it involve the telnetd
server, the pty device, or are the limits in the kernel itself? Are
there any work-arounds, such as a third party telnetd or tweaking some
kernel parameters? I'm most interested in finding a solution, but I'd
also like to understand the problem better. Is this limit unique to
IRIX or is it common with other UNIXes?
Thanks in advance,
John Simpson
--
John R. Simpson [EMAIL PROTECTED]
During the third century there was considerable strife and political upheaval
in the Roman Empire. Claudius II was emperor, and decreed that there would be
no engagements or marriages. He believed a man's happiness at home had a
direct impact on the fierceness of the empire's army. Without a wife, Claudius
reasoned, his men would have no reason to stay home. Thus he could increase
the size of his troops and their hunger to fight. Despite his decree, the
Italian Bishop Valentine clandestinely married young lovers. When Claudius
discovered Valentine's secret nuptials on February 14, 270 A.D., he had him
clubbed to death and beheaded.
Happy Valentine's Day!
------------------------------
From: Bob Berman <[EMAIL PROTECTED]>
Subject: how to create new ptys and ttys?
Date: Mon, 15 Feb 1999 15:25:45 -0500
According to the documentation with the new 2.2.1 kernel, support for
major 4 ttyp and ptys is gone. Now you need major 3 tty and major 2 ptys.
How exactly do I go about creating them? Do I just do a mknod /dev/ptyp0-f
c 2 or what? Is there a shell script around I could download? Does this mean
my old kernel 2.0.35 will definitely not work anymore? I don't trust 2.2.1
yet (especially since I've yet to have a succesful boot!) and would like to
be able to boot back into 2.0.35 if necessary.
------------------------------
From: Stefan Monnier
<[EMAIL PROTECTED]>
Subject: spurious inter-package dependencies
Date: 15 Feb 1999 17:18:32 -0500
I'm getting tired of RedHat's packages having unnecessary dependencies
(like `slang < 0.99.38-4 conflicts with ncurses3-1.9.9e-6' although I fail to
see why ncurses should have any relation with slang. Or `xterm-color is needed
by fvwm2-2.0.46-8.1' although I can clearly use fvwm2 without xterm-color).
Does anybody know if/which other distribution has a saner policy that
allows picky people like myself to only install packages they use ?
Stefan
------------------------------
From: Bryan Hackney <[EMAIL PROTECTED]>
Subject: Re: Ncurses with g++
Date: Tue, 16 Feb 1999 02:03:48 +0000
Reply-To: [EMAIL PROTECTED]
"T.E.Dickey" wrote:
>
[...]
>
> well, that certainly is a nonproductive comment.
Sorry it seems that way to you. I have experienced the same
issues, and solved them by doing exactly what I suggested, which
is to isolate the curses calls, and hence the inclusion of all those
damn macros, to a very limited scope.
>
> --
> Thomas E. Dickey
> [EMAIL PROTECTED]
> http://www.clark.net/pub/dickey
--
Bryan Hackney / BHC / bhackneyatexpress-news.net
*
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Linux developers working with AIC7890 ?
Date: Mon, 15 Feb 1999 23:13:59 GMT
Doug L. Monitors the mailing list [EMAIL PROTECTED] (email
[EMAIL PROTECTED] to subscribe)
Martin
In article <[EMAIL PROTECTED]>,
"Andy Mucho" <[EMAIL PROTECTED]> wrote:
> Probably wrong NG posting, but I'm unfortunately from the wrong side of the
> OS planet..
> Hope this is the right NG.
>
> Has anyone got any contact information for anybody who's been developing the
> Adaptec AIC7890+3860 drivers ?
> I've got a major problem with an Asus P2B-DS, and need to find out a few
> things about possible quirks in the 7890 chipset..
>
> TIA, or sorry again if this is the wrong NG..
>
> Andy Mucho
>
>
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: [EMAIL PROTECTED] (Don Waugaman)
Subject: Re: Ncurses with g++
Date: 15 Feb 1999 16:36:46 -0700
In article <[EMAIL PROTECTED]>,
Bryan Hackney <[EMAIL PROTECTED]> wrote:
>
>The curses and ncurses APIs are pathetic ancient
>messes. The problems you are having are probably due to one of the
>hundreds of macros in curses.h.
>
>You need to isolate to the greatest extent possible the curses
>calls, and include curses.h in as few (one) of your sources as
>possible.
>RD Alexander wrote:
>>
>> Is there any reason why ncurses won't work with g++ and libg++?
>>
>> Following the instructions on the man page, I added '#include ncurses.h' to
>> my C++ program. Afterwards it wouldn't compile - gcc complained of "parse
>> error before 'char' on line 15". There isn't a 'char' on line 15. Removing
>> the extra include allowed the program to compile again.
Another way to figure out what caused the problem is to compile with
gcc's -save-temps switch:
-save-temps
Store the usual "temporary" intermediate files per-
manently; place them in the current directory and
name them based on the source file. Thus, compil-
ing `foo.c' with `-c -save-temps' would produce
files `foo.cpp' and `foo.s', as well as `foo.o'.
Thus, by looking at the .cpp file, you will be able to see the source
code after the preprocessor has converted it; at which point the reason
for the parse error (and the use of 'char') should become obvious.
- Don
------------------------------
From: [EMAIL PROTECTED] (Phil Howard)
Subject: cold.apps or cold.system for libraries?
Crossposted-To: comp.os.linux.development.apps
Date: Mon, 15 Feb 1999 23:43:15 GMT
For discussions about the development of libraries, principly creating
new shared libraries, which newsgroup would be more appropriate?
comp.os.linux.development.apps ?
comp.os.linux.development.system ?
Libraries seem to fit in between applications and systems. To an app,
a library might be considered to be part of the system. To the system,
particularly the kernel, a library is user space executeable, and so
could be considered to be an app.
--
-- *-----------------------------* Phil Howard KA9WGN * --
-- | Inturnet, Inc. | Director of Internet Services | --
-- | Business Internet Solutions | eng at intur.net | --
-- *-----------------------------* philh at intur.net * --
------------------------------
From: Emile van Bergen <[EMAIL PROTECTED]>
Subject: Re: spurious inter-package dependencies
Date: Tue, 16 Feb 1999 00:29:00 +0100
Stefan Monnier wrote:
>
> I'm getting tired of RedHat's packages having unnecessary dependencies
> (like `slang < 0.99.38-4 conflicts with ncurses3-1.9.9e-6' although I fail to
> see why ncurses should have any relation with slang. Or `xterm-color is needed
> by fvwm2-2.0.46-8.1' although I can clearly use fvwm2 without xterm-color).
>
> Does anybody know if/which other distribution has a saner policy that
> allows picky people like myself to only install packages they use ?
It may not be as 'insane' as you think. Slang and ncurses have different
.h's with the same name, if I recall correctly. As to xterm and fvwm; I
guess it's because they assume you need at least a terminal emulator in
X if you don't want any other applications. The decision by whom they
want to be flamed (you or a newbie saying 'why on earth don't they warn
about that' is a political one.
RedHat uses rpm. Rpm has many advantages but at the same time, it's not
perfect. If you don't want rpm to warn you about possible lossage
resulting from missing packages, and you're sure you can fix it
yourself, why not do a rpm -i --nodeps <package>? Wouldn't that solve
your problem? I guess it would be more constructive than to disregard
the system altogether and ask for a more 'sane' packaging system /
packages.
By the way, a possible solution would be dependancy groups, e.g. fvwm2
is dependent on any of { rxvt | xterm | etc. }. Why not talk to the rpm
guys about it? Or design something like that yourself?
Just my 2c...
--
M.vr.gr. / Best regards,
Emile van Bergen (e-mail address: [EMAIL PROTECTED])
This e-mail message is 100% electronically degradeable and produced
on a GNU/Linux system.
------------------------------
From: [EMAIL PROTECTED] (H. Peter Anvin)
Subject: Re: "No Available PTYS" with 2.2.0
Date: 16 Feb 1999 00:39:19 GMT
Reply-To: [EMAIL PROTECTED] (H. Peter Anvin)
Followup to: <[EMAIL PROTECTED]>
By author: James Youngman <[EMAIL PROTECTED]>
In newsgroup: comp.os.linux.development.system
>
> Rainer Scholz <[EMAIL PROTECTED]> writes:
>
> > David Moenssen <[EMAIL PROTECTED]> wrote:
> > %
> > % [...] I am not able to bring up any xterms.
> > % xterm: no available ptys
> > %
> > If you don't understand the options: Don't reconfigure the kernel!
> >
> > Hint: Disable Unix98 ptys again.
>
> Works with them enabled on RH 5.2.
>
Most likely, the fellow has /dev/pmtx but not /dev/pts. Delete
/dev/ptmx; some early versions of glibc tried to use /dev/ptmx with
BSD ptys; this never worked right, and newer versions of glibc
fortunately don't even try.
-hpa
--
"Linux is a very complete and sophisticated operating system. There
are, and will be, large numbers of applications available for it."
-- Paul Maritz, Group Vice President for Platforms And Applications,
Microsoft Corporation [Reference at: http://www.kernel.org/~hpa/ms.html]
------------------------------
** 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: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.development.system) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Development-System Digest
******************************
