Linux-Development-Sys Digest #18, Volume #7 Tue, 3 Aug 99 21:14:12 EDT
Contents:
Questions about java benchmarks under Linux ("Darrel Davis")
Re: RH upgrade over non RH system? (Steve Hier)
Re: Q: setserial with multiport option (Holger Petersen)
Qt. about Shared libraries (Anand Paka)
Re: Video mode switching with asm (Tristan Wibberley)
Redhat boot initrd (Hugh Sparks)
Need kernel compile help ([EMAIL PROTECTED])
Re: 2.0.37 cold start while booting (Heinz Guenter Scharf)
Re: RH upgrade over non RH system? (Robin Becker)
Warning! The eclipse approaches... {6.0z} (reject)
Re: outlook can't send to qmail using pine (Raymonds Doetjes)
Re: Linux Journal - worth or not?
Re: RH upgrade over non RH system? (Johan Kullstam)
Re: GCC Cross Compiling (Victor Wagner)
Re: c++ grammer (Victor Wagner)
Re: struct byte alignment ??? (Victor Wagner)
Java Colors on Red Hat Linux 6.0 (Robert S Laramee)
Linking for libc5 under libc6 (Hugh Sparks)
Java Colors on Red Hat Linux 6.0 (Robert S Laramee)
Re: address? (Konrad Mierendorff)
----------------------------------------------------------------------------
From: "Darrel Davis" <[EMAIL PROTECTED]>
Subject: Questions about java benchmarks under Linux
Date: Tue, 3 Aug 1999 14:53:10 -0400
I read the benchmarks for the various JVMs at Javaworld
http://www.javaworld.com/javaworld/jw-03-1999/jw-03-volanomark_p.html
and was wondering about the 1024 limit on open file descriptors that
the author talks about. Has this been fixed in the 2.2.x kernel? I am
running the 2.2.15 (RH6.0) kernel and am producing a server app that
needs to accept alot of concurrent connections and, between the numbers
shown in the benchmarks above and the 1024 socket limits, I am a bit
concerned.
Also, has anyone seen any benchmarks on IBM's 1.1.6 JVM that has
been released under Linux? The 1.1.7 JVM on Win32 is very efficient.
TIA,
-darrel
------------------------------
From: Steve Hier <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: RH upgrade over non RH system?
Date: Tue, 03 Aug 1999 07:47:09 -0700
Robin Becker wrote:
>
> problem is I'm not sure whether I can get my cd-rw to work with the
> current setup so maybe backup is problematic. I certainly wouldn't want
> RH to touch the first partition which is win-95. I want to get into
> libc6 and the amount of work required to rebuild is huge. According to
> the manual it will destroy all information on the selected partitions
> during an install. An upgrade requires an RPM database so howmuch of an
> upgrade must I build ahead of time to get it all to work. Presumably I
> would need the RH password scheme etc etc and all that it implies.
>
I've installed RH into existing windoze systems
(that had free partitions) without any problems.
RH does not "destroy" existing Windoze partitions
unless you tell it to (hey -- maybe that could be
a new feature?).
Basically, RTFM and follow the prompts. Not a problem.
--
------------------------------
From: [EMAIL PROTECTED] (Holger Petersen)
Subject: Re: Q: setserial with multiport option
Date: Tue, 3 Aug 1999 13:14:19 GMT
[EMAIL PROTECTED] (bryant h marc) writes:
>Hi,
Ho!
>appreciated :
>1. The Serial HOWTO says that only Kernels 2.2 and up
>support IRQ sharing for serial ports.
With just those words, it is definitely false.
Ever since I use Linux (0.99pl6) there was support for the AST-Fourport-
Card AND for plain IRQ-Sharing vie an ORing Irq-lines by simple diodes.
Yours, Holger
PS: (sorry for german)
Einfach-Schaltung:
Pro Quelle eine Diode auf einen gemeinsamen Pull-Down-Widerstand (ca. 0,5
bis 1,5 KOhm).
|\ |
======| >|========*
|/ | |
*-------~~~~-----*-------~~~~~----> ISA-Bus IRQ-Signal
|\ | | |
======| >|========* _
|/ | | |
| |
-
|
v
Masse
Die Dioden sollen eine m�glichst kleine Durchlass-Spannung haben; eventuell
Germanium oder Shottky. Die Bauteile k�nnen sich prinzipiell auch auf drei
Karten verteilen...
<english>
for each source use one diode with a simple pulldown-resistor, 0.5..1.5 KOhm).
use diodes with small voltage-drop (shottky or germanium). You may put all
parts on different cards...
this arrangement is �invented here� but probably at many other places, too.
I�ve seeen an attribute to an american HAM-Operater "a...".
Yours , Holger (dg3lp)
------------------------------
From: Anand Paka <[EMAIL PROTECTED]>
Subject: Qt. about Shared libraries
Date: Tue, 3 Aug 1999 14:36:55 -0400
Hi,
I'm writing a shared library to override the network calls like socket,
connect, recv, send etc. I am able to override some calls like socket,
close and read but I dont seem to be able to override recv, send, connect
etc. All these symbols are defined as weak in the standard library and can
be defined again in a shared librabry. Does anybody have an idea as to why
I am unable to do this?
-Anand
------------------------------
From: Tristan Wibberley <[EMAIL PROTECTED]>
Subject: Re: Video mode switching with asm
Date: Mon, 02 Aug 1999 18:24:23 +0100
Reply-To: [EMAIL PROTECTED]
Jylam wrote:
>
> Hello
> I would like to know how switch to 320x200x256c for example without any
> library, in assembly.
> I can't use any lib because it's for any 4kb demo competition ...
You'll need to use a kernel with framebuffer support. I suggest you
subscribe to the linux framebuffer mailing list (though I don't remember
the address right now).
But then the problem is that the total code to do this is much greater
than 4kb because you're using the kernel.
--
Tristan Wibberley
------------------------------
From: Hugh Sparks <[EMAIL PROTECTED]>
Subject: Redhat boot initrd
Date: 3 Aug 1999 20:58:11 GMT
Why does the Redhat 6.0 want to boot with an
initrd image?
Is there any reason to keep this around?
I compiled my kernel and skipped
building this with no ill effects...
Thanks,
Hugh Sparks, [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED]
Subject: Need kernel compile help
Date: Tue, 03 Aug 1999 17:05:00 GMT
Hi,
HELP! I must be doing something wrong here.
I have red hat 6.0 and a dual processor computer. I can not
get my modules to work after I recompile a new kernel.
When I boot to the new kernel, it will stop at the first instance of a
module. I can not boot at all if I compile my scsi controller as a
module.
If I compile all my critical stuff as non-modules, the non-critical
stuff won't work (like my sound card). Changing these modules
to non-modules and re-compiling fixes the problem.
I have attached a note I wrote myself on compiling a
new kernel. What am doing wrong?
Many thanks,
--Tony
[EMAIL PROTECTED]
Please email me as well as respond to the news group (my nntp server
only gets back about 1 in 10 replies and I have to dig them out out
dejanews).
================== Note on compiling a kernel:============
1) login as root
2) cd /usr/src/linux
3) startx
4) Note: Red Hat 5.2 and lower: If you are compiling an SMP
(multiple processors) kernel, edit "Makefile" and set SMP=1
(note that "1" stands for true not the number of processors)
5) make xconfig
Make desired changes
Red Hat 6.0 and higher: for SMP support is enabled in "Processor
Type and Features". You also must turn off APM support in "General"
AND enable "Enhanced Real Time Support" in Character Devices
6) make dep; make clean
7) "make zImage" or make "bzImage" if the kernel is large
8) make modules
Then rename the old modules direcory, example:
mv /lib/modules/2.2.5-15 /lib/modules/2.2.5-15-old
make modules_install
9) cp /usr/src/linux/arch/i386/boot/zImage /boot/vmlinuz.smp
or
cp /usr/src/linux/arch/i386/boot/bzImage /boot/vmlinuz.smp
Note: change vmlinuz.smp's name as needed. NEVER overwrite the
original kernel.
10) edit /etc/lilo.conf as needed. Example:
boot=/dev/sda
map=/boot/map
install=/boot/boot.b
default=linux.smp
#default=linux
prompt
timeout=150
image=/boot/vmlinuz.smp
label=linux.smp
image=/boot/vmlinuz-2.0.32
label=linux
root=/dev/sda3
initrd=/boot/initrd-2.0.32.img
read-only
other=/dev/sda1
label=dos
table=/dev/sda
11) fix initrd image
run the following command to creat a new initrd in the boot
directory. Make sure that the lilo.conf entry points to it.
/sbin/mkinitrd /boot/newinitrd-image 2.2.5-15
Where newinitrd-image is the name of the image in lilo.conf
and 2.2.5-15 is the current version of the modules you are
using (see "ls -alg /lib/modules")
12) lilo
Fix any error that pop up with 10 & 11
13) shutdown -r now
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
------------------------------
From: [EMAIL PROTECTED] (Heinz Guenter Scharf)
Subject: Re: 2.0.37 cold start while booting
Date: 3 Aug 1999 21:10:10 GMT
Reply-To: [EMAIL PROTECTED]
I had the same experience, even with a completely stripped down kernel,
I went back to 2.0.36, which is running happily on my machine since it's
appereance.
My machine is a Pentium 100, 32MB, IDE Drives, Flashpoint LT for DAT tape and
Yamaha CDRW, old Mitsumi FX001, RedHat 4.2, never went to glibc.
Guenter
[EMAIL PROTECTED] wrote:
>Greetings,
>
>I just updated by kernel to 2.0.37, recompiled, did the whole LILO bit,
>and now my computer reboots while it is printing all the '...' stuff when
>it is loading the kernel. Fortunately, I made a boot floppy of 2.0.36 so
>nothing is hosed. I'm wondering if anyone else is experiencing this sort
>of thing or is it just me?
>
>Thanks,
>Phil
>
>
--
If a listener nods his head when you're explaining your program, wake him up.
------------------------------
From: Robin Becker <[EMAIL PROTECTED]>
Subject: Re: RH upgrade over non RH system?
Date: Tue, 3 Aug 1999 17:38:37 +0100
In article <[EMAIL PROTECTED]>, Steve Hier
<[EMAIL PROTECTED]> writes
>Robin Becker wrote:
>>
>> problem is I'm not sure whether I can get my cd-rw to work with the
>> current setup so maybe backup is problematic. I certainly wouldn't want
>> RH to touch the first partition which is win-95. I want to get into
>> libc6 and the amount of work required to rebuild is huge. According to
>> the manual it will destroy all information on the selected partitions
>> during an install. An upgrade requires an RPM database so howmuch of an
>> upgrade must I build ahead of time to get it all to work. Presumably I
>> would need the RH password scheme etc etc and all that it implies.
>>
>
>I've installed RH into existing windoze systems
>(that had free partitions) without any problems.
that's the problem, with existing partitions & no RPM db I must backup
which means cdrecord, faking scsi building the loopback fs.
The howto on Slackware+RH might be a way to go, but then I must get the
working glibc, modules etc etc from RH and all the bits init etc etc
must be recompiled before anything will work. I've hated LILO since an
NT box bombed me to ground zero once with mbr tricks etc.
>RH does not "destroy" existing Windoze partitions
>unless you tell it to (hey -- maybe that could be
>a new feature?).
>
>Basically, RTFM and follow the prompts. Not a problem.
>
>--
--
Robin Becker
------------------------------
Date: Tue, 3 Aug 1999 05:40:05 -0500
From: reject <Use-Author-Address-Header@[127.1]>
Subject: Warning! The eclipse approaches... {6.0z}
Crossposted-To:
comp.sys.hp48,alt.appalachian,rec.music.rock-pop-r+b.1950s,be.comp.os.linux,tw.bbs.soc.chiayi,ont.general
That "monstriferous" Comet Lee will be seen during the
solar eclipse this August 11th, followed by WWIII, the
1300-meter "King of Terror" meteoroid impact before 10
October 1999, and *many* catastrophic events, including
the >20 degree shifting of the polar axis before 2002!
The Tribulation prophesied even by our Lord and Savior
Jesus Christ is begun. ANSWER THIS: Are you prepared?
Godspeed,
Daniel
------------------------------
From: Raymonds Doetjes <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking,redhat.networking.general
Subject: Re: outlook can't send to qmail using pine
Date: Wed, 04 Aug 1999 00:34:09 +0200
For sending mail, you should add your the same SMTP server as is configured
in your pine config.
Raymond
[EMAIL PROTECTED] wrote:
> i've setup qmail and used pine in the linux box. i can send and received
> mail using pine. i setup another machine using outlook express 2000. i can
> received mail but unable to send back or reply. kinddly advise or help.
> thank you in advance.
> error msg from outlook :
>
> Your server has unexpectedly terminated the connection. Possible causes
> for this include server problems, network problems, or a long period of
> inactivity. Account: 'rsmail1.com', Server: 'rsmail1.com', Protocol: SMTP,
> Port: 25, Secure(SSL): No, Error Number: 0x800CCC0F
>
> p/s : searchlinux really cool and all you guys 2. Thank you.
>
> ------------------ Posted via SearchLinux ------------------
> http://www.searchlinux.com
------------------------------
From: [EMAIL PROTECTED] ()
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Linux Journal - worth or not?
Reply-To: [EMAIL PROTECTED]
Date: 3 Aug 1999 13:08:43 -0500
On Fri, 30 Jul 1999 09:16:17 -0500, Steven J. Hill <[EMAIL PROTECTED]> wrote:
>I would tend to agree with Mr. Rempt's initial opinion that the Linux
>Journal is becoming more commercially oriented. I just recently started
>to subscribe to Linux Magazine which IMHO is has good technical content
>and is not so full of advertisements. I have the first three issues and
>have read them all cover to cover. Here is the URL:
>
> http://www.linux-mag.com/
>
>
>-Steve
Cool, just subscribed. Thanks for the advice. Hopefully this one will
last a while before becoming heavily commercial-oriented.
Email: [EMAIL PROTECTED]
ICQ: 19668081
-----------== Posted via Newsfeeds.Com, Uncensored Usenet News ==----------
http://www.newsfeeds.com The Largest Usenet Servers in the World!
======== Over 73,000 Newsgroups = Including Dedicated Binaries Servers =======
------------------------------
Subject: Re: RH upgrade over non RH system?
From: Johan Kullstam <[EMAIL PROTECTED]>
Date: 03 Aug 1999 18:42:26 -0400
Robin Becker <[EMAIL PROTECTED]> writes:
> I've purchased RH 6.0. It won't let me upgrade over my non RH existing
> system as it cannot find an RPM data base. Is there any way to fake the
> install into not deleting my existing system or of getting the upgrade
> to see an RPM database?
not really.
if you have multiple partitions, you can tell it, e.g., to leave /home
alone. i usually resort to tar-ing up /etc /home and /var/qmail and
stuffing that into a spare partition (it's one good use for a windows
partition). then i install fresh. expand the tarballs under /tmp and
restore what configuration you need.
--
J o h a n K u l l s t a m
[[EMAIL PROTECTED]]
Don't Fear the Penguin!
------------------------------
From: [EMAIL PROTECTED] (Victor Wagner)
Crossposted-To: comp.lang.c++,comp.os.linux.development.apps,hp.os.linux
Subject: Re: GCC Cross Compiling
Date: 3 Aug 1999 23:03:32 +0400
Matt Zinkevicius <[EMAIL PROTECTED]> wrote:
: Hi,
: Our team would like to build a verison of GCC that cross-compiles
: to x86 linux hosted on x86 Windows NT. Has anyone tried this? Anything
: I should be aware of? Is it even possible?
It is possible. Start with installing cygnus tools. But probably, better
preformance can be achived by setting up real Linux system on some
outdated 486 without monitor and keyboard and running real Linux gcc on
it from telnet session. Obvoisly, your team have somewhere 486 or old
pentium motherboard, which just collects dust in dark corner, but it is
perfectly suitable as Linux workstation without GUI just for compiling
and debugging (if your program needs GUI you can still use headless box
installing some X server on NT)
: Thanks for any help you can lend!
: --Matt Zinkevicius
--
========================================================
Victor Wagner @ home = [EMAIL PROTECTED]
I don't answer questions by private E-Mail from this address.
------------------------------
From: [EMAIL PROTECTED] (Victor Wagner)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.misc
Subject: Re: c++ grammer
Date: 3 Aug 1999 22:23:52 +0400
In comp.os.linux.misc jievis <[EMAIL PROTECTED]> wrote:
: Hi, All:
: Where can I find the grammer for C++ writen in lex( or flex) and yacc
: (or bison),
: Thanks in advance
Isn't g++ written on bison?
--
========================================================
Victor Wagner @ home = [EMAIL PROTECTED]
I don't answer questions by private E-Mail from this address.
------------------------------
From: [EMAIL PROTECTED] (Victor Wagner)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: struct byte alignment ???
Date: 3 Aug 1999 22:51:38 +0400
Bayee <[EMAIL PROTECTED]> wrote:
: Hi all,
: A small C question.....
: // start of my_test.c
: int main() {
: struct s1 {
: char field1;
: unsigned short field2;
: } my_test;
: printf("sizeof my_test: %d\n", sizeof(my_test));
: printf("sizeof my_test: %d + %d = %d\n",
: sizeof(my_test.field1), sizeof(my_test.field2),
: sizeof(my_test.field1)+sizeof(my_test.field2));
: }
: // end of my_test.c
: On a Intel machine, with SuSE 6.1 Linux installed,
: % make my_test
: % ./my_test
: sizeof my_test: 4
: sizeof my_test: 1 + 2 = 3
:
: From what I understand, this is due to some alignment issuse. Is
: there any compilation option that we can turn off the "auto-align"
: feature ? (Means that the first printf will return 3 instead of 4).
There is some options on gcc which alows to play with it, but basically
it is wrong style. Intel is not only processor in the world. On Sparc,
for example, you'll get bus error trying to access short int by odd
address and int by address which is not multiple of four. On intel it
would just take ten times longer. That is why compilier aligns data for
you.
Only place when you do really need packed data structure is when you are
reading some binary data format from disk. But in this case there is
also byte-order issue - on sparc and powerpc ints are stored from higher
byte to lower (same way as we write them on paper) and on intel and
alpha - from lower to higher (which allows to access int as
*((short int *)&intvar) if we know that int doesn't exceed limits of
short int).
So, in case of binary format reading, you have to read data in char
array and reconstruct structure byte by byte, especially if format is
originated on Intel. If format uses MSB-first byte order, function
htons htonl etc can be used to convert data from internal representation
to "network byte order", which is accidentally same as on Sparc and
PowerPC.
Note - you are writing for Linux, and Linux runs on all processors
mentioned in my letter.
: Thanks.
: rgds,
: Bayee
--
========================================================
Victor Wagner @ home = [EMAIL PROTECTED]
I don't answer questions by private E-Mail from this address.
------------------------------
From: Robert S Laramee <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Java Colors on Red Hat Linux 6.0
Date: Tue, 03 Aug 1999 20:10:20 -0400
I'm having a problem with drawing (or painting rather) different colors
in
Java on Red Hat Linux 6.0. We're using the Kaffe compiler and the
updated
jdk117_v3 (java version "1.1.7B"). The problem is, the only lines I can
paint are black. I can't get any of the colors to work. I can however,
set the background color of the canvas to any color. Not that it will
help, but i'm using the following bit of code:
import java.awt.*;
...
public void paint(Graphics g) {
g.setColor(Color.orange);
/* map the viewpoint to the current world coordinates */
mappedPoint = tree.mapViewPoint(eyePoint);
/** tell the tree do draw itself here */
tree.displayTree(tree.getRoot(), mappedPoint);
}
Nothing I do with setColor() will work! It used to work on the previous
version of Red Hat Linux (and their window manager). And this has been
the case ever since we installed Red Hat Linux 6.0. Is there any
documentation on this? Does anyone have a clue what the problem might
be?
Is anyone else having this trouble?
Also, is there away to set the amount of virtual memory the JVM may use.
(I know how to set the heap size).
-cheers, bob
--
Robert S Laramee tel: (603) 868-7831
14 McDaniel Drive office: (603) 862-0350
Durham, NH 03824 URL: http://www.cs.unh.edu/~rlaramee
------------------------------
From: Hugh Sparks <[EMAIL PROTECTED]>
Subject: Linking for libc5 under libc6
Date: 3 Aug 1999 20:50:49 GMT
I want to link a program for execution
on a libc5 machine but I'm developing
under Redhat 6.0 with libc6. I've tried
several things, but when copied to the
target system, the program fails: ldd
reveals that it wants to dynamically link
with libc6.
Thanks,
Hugh Sparks, [EMAIL PROTECTED]
------------------------------
From: Robert S Laramee <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Java Colors on Red Hat Linux 6.0
Date: Tue, 03 Aug 1999 20:06:33 -0400
I'm having a problem with drawing (or painting rather) different colors
in
Java on Red Hat Linux 6.0. We're using the Kaffe compiler and the
updated
jdk117_v3 (java version "1.1.7B"). The problem is, the only lines I can
paint are black. I can't get any of the colors to work. I can however,
set the background color of the canvas to any color. Not that it will
help, but i'm using the following bit of code:
import java.awt.*;
...
public void paint(Graphics g) {
g.setColor(Color.orange);
/* map the viewpoint to the current world coordinates */
mappedPoint = tree.mapViewPoint(eyePoint);
/** tell the tree do draw itself here */
tree.displayTree(tree.getRoot(), mappedPoint);
}
Nothing I do with setColor() will work! It used to work on the previous
version of Red Hat Linux (and their window manager). And this has been
the case ever since we installed Red Hat Linux 6.0. Is there any
documentation on this? Does anyone have a clue what the problem might
be?
Is anyone else having this trouble?
Also, is there away to set the amount of virtual memory the JVM may use.
(I know how to set the heap size).
-cheers, bob
--
Robert S Laramee tel: (603) 868-7831
14 McDaniel Drive, #422 office: (603) 862-0350
Durham, NH 03824 URL: http://www.cs.unh.edu/~rlaramee
------------------------------
From: Konrad Mierendorff <[EMAIL PROTECTED]>
Subject: Re: address?
Date: Tue, 03 Aug 1999 21:46:46 +0200
Ann Chen wrote:
>
> We are trying to find out where does the Linux operating system reside??
> Does it live on the lower half of the memory, or the upper half? And how
> big is the Linux kernel (how much memory does it occupy?)
As to the last question, the answer is easy. The amount of memory the
kernel is printed out when you boot your system. Of course it heavily
depends on the drivers you compiled into the kernel.
I don't know were the kernel resides but anyway why is it of interest?
Bye
- Konrad
------------------------------
** 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
******************************
