Linux-Development-Sys Digest #626, Volume #7 Fri, 25 Feb 00 18:13:12 EST
Contents:
Re: RedHat 6.0 (Dr H. T. Leung)
DAQ HELP Needed... (Mustafa Ihsan Ecemis)
Re: interrupt handler not being called . ("Norm Dresner")
Re: Binary compatibility: what kind of crack are they smoking? (JEDIDIAH)
How to open/read/write files inside kernel? (Jun Xu)
NIC's promisc mode (=?iso-8859-1?Q?=C0=EE=CE=E4?=)
list exported symbols from dll? (Scott Dudley)
Re: Does linux support DIRECT I/O? (Erik de Castro Lopo)
Re: list exported symbols from dll? (Nate Eldredge)
Re: Binary compatibility: what kind of crack are they smoking? (Adam Ierymenko)
complex math in a device driver ("Dan Miller")
framebuffer programming ("Jens Skakkebaek")
Re: complex math in a device driver (Grant Edwards)
Re: Binary compatibility: what kind of crack are they smoking? (J Bland)
Re: complex math in a device driver (Andreas Jaeger)
Re: changing glibc (Juergen Heinzl)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Dr H. T. Leung)
Subject: Re: RedHat 6.0
Date: 25 Feb 2000 19:11:56 GMT
I think there is a redhat newsgroup (or a few of them), BTW. I know for certain
that there is one for slackware and debian :-).
This is an FAQ. Put append "mem=256Mb" at the top of /etc/lilo.conf and rerun
/sbin/lilo . Read "man lilo.conf" for details.
In article <894onm$plc$[EMAIL PROTECTED]>, "Sake" <[EMAIL PROTECTED]> writes:
|> Hi there,
|>
|> Given the popularity of RedHat Linux, there probably should be a RedHat
|> group.
|> Anyway, I'll use this for now.
|>
|> Problem that I have is that, cat on /proc files and 'dmesg' both show 64MB
|> of RAM although
|> my machine is equipted with 256MB of RAM. I'm using RedHat 6.0 fresh out of
|> box.
|>
|> Where is my memory ?? :(
|>
|> Any hints are appreciated.
--
--------------------------------------------------
"What you don't care cannot hurt you." Chap. 7a, AMS-NS
------------------------------
From: Mustafa Ihsan Ecemis <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux,comp.os.linux.misc,comp.os.linux.hardware,alt.comp.hardware,de.comp.os.unix.linux.hardware
Subject: DAQ HELP Needed...
Date: Fri, 25 Feb 2000 14:23:09 -0500
Hello,
I'm looking for a PCMCIA DAQ Card with the following specs,
but I couldn't found any yet. The OS I'm using is Linux,
Kernel 2.2. Does anyone know any card matching these specs?
Sampling rate : at least 500kHz
Input : at least 1 differential analog (0-5V)
Output : at least 1 TTL digital
Input res. : 12 bits or higher
You can write me directly at [EMAIL PROTECTED]
Any help is greatly appreciated, and thanks in advance.
Ihsan Ecemis.
------------------------------
Reply-To: "Norm Dresner" <[EMAIL PROTECTED]>
From: "Norm Dresner" <[EMAIL PROTECTED]>
Subject: Re: interrupt handler not being called .
Date: Fri, 25 Feb 2000 19:24:22 GMT
Do you know, in fact, that the kernel is assigning the interrupt to you and
that it is seeing the interrupt? Try
cat /proc/interrupts
and see who's got which interrupt and how many have occurred.
Norm
Nilesh Patel <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> hi ,
> there is an ISR which is being registered for a specific IRQ .
> The interrupt has been registered and a
> shared IRQ is being used . however there is no other device using this
> irq . The hardware has been properly
> initialised , we are getting an irq line , but the interrupt handler is
> not being called . The interrupt handler
> is to be called when a certain buffer gets full . but once the buffer
> gets full , the kernel hangs . Please let know
> the cause for this . Again , do we have to call a enable_irq explicitly
> ..
>
> summing up ,
> why is the handler not being called ?
>
> do reply back at the earliest .
>
> thanks,
> raj.
>
------------------------------
From: [EMAIL PROTECTED] (JEDIDIAH)
Crossposted-To: comp.os.linux.advocacy,comp.os.linux.setup
Subject: Re: Binary compatibility: what kind of crack are they smoking?
Date: Fri, 25 Feb 2000 19:24:56 GMT
On Fri, 25 Feb 2000 18:34:33 +0100, Mario Klebsch <[EMAIL PROTECTED]> wrote:
>Albert Ulmer <[EMAIL PROTECTED]> writes:
>
>>Mario Klebsch wrote:
>>> >If your talking about an operating system,
>>> >you should be calling it GNU/Linux, like i.e. Debian does.
>>
>>> Well, that probably would be Debian GNU/Linux, and there is RatHad,
>>> SUSE, Caldera,... They all claim to be Linux, but in fact are
>>> different OSes.
>
>>Nonsense, they all cater the diverse needs of various users. In my view
>>that's the main point about the whole open source movement: CHOICE! It
>>is good to be different!
>
>The absence of the ABI destroys the choice. If you are not running
>RedHat or SUSE Linux, you really are in trouble, today. However, the
>degree of trouble varies from distribution to distribution.
No, you just need a package made for your particular iteration
assuming that you are unwilling to build your own software or
your distributor is incapable of doing it for you.
>
>If we want to keep the choice, we'd better create a well defined ABI
>yesterday!
Well, if you avoid C++, that's not really a problem.
This issue is more a 'Why not C++' thing than a 'Why not Linux'.
--
|||
Resistance is not futile. / | \
Need sane PPP docs? Try penguin.lvcm.com.
------------------------------
From: Jun Xu <[EMAIL PROTECTED]>
Subject: How to open/read/write files inside kernel?
Date: Fri, 25 Feb 2000 13:59:36 -0600
Maybe this is a stupid question.
I want to open a file and write stuffs into that from time to time in the
TCP/IP part inside the kernel. I tried sys_open, sys_write, but sys_open
always always return <0 value. Can anyone tell me how to do that?
Appreicate it.
Jun
------------------------------
From: =?iso-8859-1?Q?=C0=EE=CE=E4?= <[EMAIL PROTECTED]>
Subject: NIC's promisc mode
Date: Fri, 25 Feb 2000 22:11:03 -0500
Reply-To: [EMAIL PROTECTED]
hi, all.
there is a new way to set network interface card to promisc mode
in the linux kernel 2.2 and above. but my program looks did not work.
why? maybe you can help me.
#include <stdio.h>
#include <sys/ioctl.h>
#include <sys/socket.h>
#include <linux/if_arp.h>
main(){
int s, ret;
struct ifreq ifr;
struct packet_mreq mr;
socklen_t mr_len; /*typedef unsigned int socklen_t*/
mr_len = sizeof(mr);
s = socket(PF_PACKET, SOCK_DGRAM, htons(ETH_P_IP));
strcpy(ifr.ifr_name, "eth0");
if(ioctl(s, SIOCGIFHWADDR, &ifr) == -1){
printf("error\n");
exit(1);
}
memset(&mr, 0, sizeof(mr));
memcpy(&mr.mr_address[0], ifr.ifr_hwaddr.sa_data, 6);
mr.mr_alen = 6;
mr.mr_ifindex = 1;
mr.mr_type |= PACKET_MR_PROMISC;
ret = setsockopt(s, SOL_PACKET, PACKET_ADD_MEMBERSHIP, &mr, mr_len);
close(s);
}
i called setsockopt to set , and it return 0, meanes ok. but when i use
ifconfig command , it is still in MULTICAST mode.
------------------------------
From: Scott Dudley <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: list exported symbols from dll?
Date: Fri, 25 Feb 2000 13:15:06 -0700
This is a multi-part message in MIME format.
==============4387B5BF190080D65304E587
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
How do I obtain a list of exported symbols (functions) from a
dynamically loadable module (.so file)?
Thanks.
--
Microsoft gives you Windows but Unix gives you the whole house!
==============4387B5BF190080D65304E587
Content-Type: text/x-vcard; charset=us-ascii;
name="scott.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Scott Dudley
Content-Disposition: attachment;
filename="scott.vcf"
begin:vcard
n:Dudley;Scott
tel;fax:(602) 308-1300
tel;work:(602) 308-1115
x-mozilla-html:FALSE
url:www.telesoft.com
org:Telesoft Corp.;Research & Development
version:2.1
email;internet:[EMAIL PROTECTED]
title:Developer
adr;quoted-printable:;;3443 N Central Ave=0D=0ASuite 1800;Phoenix;AZ;85016;US
x-mozilla-cpt:;-25280
fn:Scott Dudley
end:vcard
==============4387B5BF190080D65304E587==
------------------------------
From: Erik de Castro Lopo <[EMAIL PROTECTED]>
Subject: Re: Does linux support DIRECT I/O?
Date: Fri, 25 Feb 2000 20:48:03 +0000
Marc SCHAEFER wrote:
>
> Hong Shen <[EMAIL PROTECTED]> wrote:
> : I am working on a video editing program on linux. Because of the high
> : bitrate of the video file, the caching seems no use and may add some
> : overhead.
>
> You could:
> use the raw device (the direct I/O) one
Correct me if I'm wrong, but doesn't a raw device have to be a whole
partition? That may not be what you want.
Instead, have a look at mmap(), especially a post I made here a
couple of days ago in a thread titled "Reading files without disk
buffering". Using a fast LVDS SCSI disk I was able to average
a disk transfer rate of 35Meg/sec for 40 seconds with a very low
CPU overhead. The fastest disks available at the moment can only
do about 38Meg/sec.
You will find it on deja.com. Use this to search:
http://www.deja.com/home_ps.shtml
Try searching on my name and the subject line.
Cheers,
Erik
--
+-------------------------------------------------+
Erik de Castro Lopo [EMAIL PROTECTED]
+-------------------------------------------------+
"In civilian equipment, such as computers, the number
of components alone makes miniaturization essential if
the computer is to be housed in a reasonable-sized
building." Electronics Oct. 1, 1957, p. 178
------------------------------
From: Nate Eldredge <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Re: list exported symbols from dll?
Date: 25 Feb 2000 12:59:03 -0800
Scott Dudley <[EMAIL PROTECTED]> writes:
> How do I obtain a list of exported symbols (functions) from a
> dynamically loadable module (.so file)?
man nm
--
Nate Eldredge
[EMAIL PROTECTED]
------------------------------
Date: Fri, 25 Feb 2000 16:12:55 -0500
From: Adam Ierymenko <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.advocacy,comp.os.linux.setup
Subject: Re: Binary compatibility: what kind of crack are they smoking?
JEDIDIAH wrote:
> >If we want to keep the choice, we'd better create a well defined ABI
> >yesterday!
>
> Well, if you avoid C++, that's not really a problem.
>
> This issue is more a 'Why not C++' thing than a 'Why not Linux'.
Yeah, lets go back to the 80s! Lets avoid object oriented
programming, GUIs, and common installers! Terminals and
C all the way! Where's my Tears for Fears CD?
Not that I have anything against plain C... Plain C is better
for some tasks like small utilities and small, highly efficient
servers. But for things like GUI and windowing environments
or any other code that has a lot of state and hierarchy, C++ is
the way to go.
Notice how KDE, developed in C++, is many times more
stable than Gnome which is developed in C (even though it
uses an 'object oriented model' in C). Not that I want to
start a Gnome/KDE holy war, but I think that KDE being
written in a real object oriented language has something
to do with it's stability and rapid development.
The solution is not ludditism but rather making things work.
I think the best solution for C++ would be to statically link
the C++ libraries.
------------------------------
From: "Dan Miller" <[EMAIL PROTECTED]>
Subject: complex math in a device driver
Date: Fri, 25 Feb 2000 13:14:04 -0800
I need to access some math library function in my device driver. I believe
I've read here that one cannot do that, and I don't actually see a way to do
so (-lm isn't an argument to ld). Is there some way I can manage this
anyway?? If not, why is that improper??
The functions I need are pow(), log10(), floor(), ceil(). Of course, the
last two I could create pretty easily, but the first two are less trivial,
and I'd rather not duplicate other peoples' good work... How do I resolve
this dilemma??
Dan Miller
------------------------------
From: "Jens Skakkebaek" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.hardware
Subject: framebuffer programming
Date: Fri, 25 Feb 2000 13:22:49 -0800
Hi,
I need some insights on frame buffers in Linux:
I have a Cyberpro 5050 on another platform (Motorola Coldfire) and need to
port the framebuffer driver. I'm planning on using the Cyber2000fb driver
as the basis.
However, I'm still puzzled about how the this driver fits in with fbcon (if
at all?), how fonts get used, etc. I've looked around. Basically, it's all
uncommented code and hard to find the way around in.
Do you know of any documentation around that explains how to program a frame
buffer device?
The link in section 18 of the framebuffer howto:
http://www.csoft.net/~systems/jim/fb.html
doesn't work anymore.
Any pointers would be greatly appreciated.
Best regards,
Jens
Jens Skakkebaek
[EMAIL PROTECTED]
------------------------------
From: grant@nowhere. (Grant Edwards)
Subject: Re: complex math in a device driver
Date: Fri, 25 Feb 2000 21:25:32 GMT
In article <896rbt$o2f$[EMAIL PROTECTED]>, Dan Miller wrote:
>I need to access some math library function in my device
>driver. I believe I've read here that one cannot do that, and
>I don't actually see a way to do so (-lm isn't an argument to
>ld). Is there some way I can manage this anyway?? If not, why
>is that improper??
Last time I asked (9 months ago), you're not allowed to do
floating point operations in a device driver (at least not on
Intel).
>The functions I need are pow(), log10(), floor(), ceil(). Of
>course, the last two I could create pretty easily, but the
>first two are less trivial, and I'd rather not duplicate other
>peoples' good work... How do I resolve this dilemma??
--
Grant Edwards grante Yow! When this load is
at DONE I think I'll wash
visi.com it AGAIN...
------------------------------
From: [EMAIL PROTECTED] (J Bland)
Crossposted-To: comp.os.linux.advocacy,comp.os.linux.setup
Subject: Re: Binary compatibility: what kind of crack are they smoking?
Date: 25 Feb 2000 20:57:17 GMT
On Fri, 25 Feb 2000 19:24:56 GMT, JEDIDIAH <[EMAIL PROTECTED]> wrote:
>On Fri, 25 Feb 2000 18:34:33 +0100, Mario Klebsch <[EMAIL PROTECTED]> wrote:
>>Albert Ulmer <[EMAIL PROTECTED]> writes:
>>
>>>Mario Klebsch wrote:
>>>> >If your talking about an operating system,
>>>> >you should be calling it GNU/Linux, like i.e. Debian does.
>>>
>>>> Well, that probably would be Debian GNU/Linux, and there is RatHad,
>>>> SUSE, Caldera,... They all claim to be Linux, but in fact are
>>>> different OSes.
>>
>>>Nonsense, they all cater the diverse needs of various users. In my view
>>>that's the main point about the whole open source movement: CHOICE! It
>>>is good to be different!
>>
>>The absence of the ABI destroys the choice. If you are not running
>>RedHat or SUSE Linux, you really are in trouble, today. However, the
>>degree of trouble varies from distribution to distribution.
>
> No, you just need a package made for your particular iteration
> assuming that you are unwilling to build your own software or
> your distributor is incapable of doing it for you.
>
>>
Or just try to always avoid binary releases in favour of source ones. In my
experience with ARM Linux (a redhat derivative), various SuSE distros, an
oldish Debian and Solaris, most source code releases will compile if you have
the right libraries kicking about. Admittedly, the older setups have more
and more trouble but the major stuff compiles and works.
Things are changing and I expect to see an increased number of binary+source
or source-only applications. It is usually in everyone's interest to have a
release with the source.
Support source and try to avoid binaries I say. Most source tarballs only
need a ./configure; make; make install and this is hardly beyond even a newbie
to do.
I have a pipe-dream; that all companies and coders will make software that's
distributed by source and uses standard interchangable file formats. This
*must* happen to prevent the sort of absolute insanity that goes on today
with multitudinous fileformats, filters, "ABI"s, converters and emulators
and god knows what else we use to try to integrate various platforms and
their OSes.
Compatibilty goes both ways; why should I be compatible with closed-source,
constantly changing standards while they won't be compatible with my (often
superior) open-source and fixed formats? Why have only one package for one
platform and one cpu architecture when I can have the source and use it on
the majority of them? It just doesn't make sense (except to the people
making a killing out of doing so).
You can stick your ABIs as long as people are willing to give me their
source to compile on *my* platform.
Shrike (regularly using 3 hardware platforms and 5 OSes and fed up with only
being able to fully support 1 because of binary-only releases)
------------------------------
From: Andreas Jaeger <[EMAIL PROTECTED]>
Subject: Re: complex math in a device driver
Date: 25 Feb 2000 23:06:04 +0100
>>>>> Dan Miller writes:
> I need to access some math library function in my device driver. I believe
> I've read here that one cannot do that, and I don't actually see a way to do
> so (-lm isn't an argument to ld). Is there some way I can manage this
> anyway?? If not, why is that improper??
> The functions I need are pow(), log10(), floor(), ceil(). Of course, the
> last two I could create pretty easily, but the first two are less trivial,
> and I'd rather not duplicate other peoples' good work... How do I resolve
> this dilemma??
You can't use any floating point in the kernel at all. No chance
AFAIK. You have to solve this differently. Are you sure that this
cannot done in userspace?
Andreas
--
Andreas Jaeger
SuSE Labs [EMAIL PROTECTED]
private [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Subject: Re: changing glibc
Date: Fri, 25 Feb 2000 23:00:02 GMT
In article <[EMAIL PROTECTED]>, Mario Klebsch wrote:
>[EMAIL PROTECTED] (Juergen Heinzl) writes:
>
>>Try the update .. 4.4.1 would work mostly,, 4.4.2 does work without
>>problems for me (glibc-2.1.2 && glibc-2.1.3pre4). For me it was
>>actually the only real troublesome thing but I used 2.0.110 - 2.0.112
>>too, so some things updated on the fly beforehand..
>
>I have 4.4.1 :-)
>
>My problem is libstdc++-2.8.1. It does not exist on my system, so I
>got the libstdc++-2.8.1.1.tar.gz and tried to comple it. The
>compilatioon failed, but after some modifications (that were not that
>easy for a X expert not being a C++ expert) it succeded.
>
>But applix refused to run with the just greated library.
You might try some binary. I�ve no distribution at all but as some
binaries require(d) an old libstdc++-2.8.x plugged in one from
S.u.S.E.
Cheers,
Juergen
--
\ Real name : J�rgen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
** 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
******************************
