Linux-Development-Sys Digest #716, Volume #7 Wed, 29 Mar 00 16:13:14 EST
Contents:
Re: bootsect.s (Kaz Kylheku)
Re: gethostbyname() returns "unknown host" ("Someone Insignificant")
Has anyone used LINEO ? (Ishikawa)
kernel memcpy (Weiguang Shi)
Re: bootsect.s (Neil Koozer)
Re: kernel memcpy (Fabrice Peix)
Re: Gonna build a computer... (Christian Winter)
HELP: kgdb remote debugging (Hao-hua Chu)
Re: Precision of Linux's libm??? (David Konerding)
Re: 2.3.99: what's next step? ("Jason Nye")
Re: Precision of Linux's libm??? (bill davidsen)
Re: System.map location (bill davidsen)
Re: Reiserfs + Raid? (bill davidsen)
Blocking I/O with ioctl() to network driver do not work !!! ("Beno�t Cousson")
Re: software raid patch (bill davidsen)
Re: 2.3.99: what's next step? (bill davidsen)
Re: Blocking I/O with ioctl() to network driver do not work !!! (Kaz Kylheku)
Problem with my configuration (Modem and Internet Connection - Kppp....Netscape)
Re: Recomendation for a graphical 'diff' style program for Linux? (Jonathan Magid)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Kaz Kylheku)
Subject: Re: bootsect.s
Reply-To: [EMAIL PROTECTED]
Date: Wed, 29 Mar 2000 16:48:31 GMT
On Wed, 29 Mar 2000 16:17:10 +0100, filipesinclair <[EMAIL PROTECTED]>
wrote:
>Who can tell me where can i find more comments on this file
>some portions of the source code doesn�t have comments
>I cannot understand some parts
You probably need reference materials on 8088 assembly language as well as PC
hardware and BIOS interfaces.
--
#exclude <windows.h>
------------------------------
From: "Someone Insignificant" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.development,comp.os.linux.development.apps,linux.dev.c-programming,linux.redhat.development
Subject: Re: gethostbyname() returns "unknown host"
Date: Wed, 29 Mar 2000 17:15:15 GMT
That is the whole program.
The problem, as near as I can tell, is that the compiler is linking with
an old resolver library, and not the newly upgraded BIND-8 resolver
library.
I havent had time to investigate this but if anyone can verify that this
makes sense, please let me know.
mike
>
> That's really strange. Post the whole program, so we can have a more
> in-depth look.
>
> --
> _ _
> o' \,=./ `o Daniel Beer <[EMAIL PROTECTED]>
> (o o) dlbeer.freeshell.org
> ooO--(_)--Ooo-
>
>
------------------------------
From: Ishikawa <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.liux.questions
Subject: Has anyone used LINEO ?
Date: Thu, 30 Mar 2000 02:26:31 +0900
www.lineo.com distributes "embedix" which
is an embedded version of linux.
Has anyone used it and would like to share the
"feel" of the distribution?
Since the download requires input private information before we can
do so, I would rather find out how the distribution works.
One article found via deja.com search mentioned it lacked
DiskOnChip driver.
But other than that, I can't find any mention of the
size, ease of installation, functionality, etc..
Anyone?
------------------------------
From: Weiguang Shi <[EMAIL PROTECTED]>
Subject: kernel memcpy
Date: Wed, 29 Mar 2000 10:29:30 -0700
Hi, folks:
I know this question may seem silly, but can anyone please tell me if
there exists a memory copy function in the kernel? I don't mean the
functions copying to/from userspace, but a kernel to kernel copy function.
I just need it for my course project.
Thanks.
Weiguang
------------------------------
From: Neil Koozer <[EMAIL PROTECTED]>
Subject: Re: bootsect.s
Date: Wed, 29 Mar 2000 09:41:41 -0800
filipesinclair wrote:
>
> Who can tell me where can i find more comments on this file
> some portions of the source code doesn�t have comments
> I cannot understand some parts
>
> thank you .
The file bootsect.S is much better (maybe that's the one you mean
anyway)
I've done some work in *reading* that file, so maybe I can answer some
questions (if they're easy enough).
Neil.
------------------------------
From: Fabrice Peix <[EMAIL PROTECTED]>
Subject: Re: kernel memcpy
Date: Wed, 29 Mar 2000 19:53:37 +0200
Weiguang Shi wrote:
>
> Hi, folks:
> I know this question may seem silly, but can anyone please tell me if
> there exists a memory copy function in the kernel? I don't mean the
> functions copying to/from userspace, but a kernel to kernel copy function.
> I just need it for my course project.
>
> Thanks.
> Weiguang
Yop,
I think you can use memcpy or memmove.
This function are define in asm/string.h
bye and Ooops.
------------------------------
From: Christian Winter <[EMAIL PROTECTED]>
Subject: Re: Gonna build a computer...
Date: Wed, 29 Mar 2000 19:25:00 +0200
Anders Larsen schrob:
> Christian Winter wrote:
> > Regarding Video cards, I would neither buy a new ATI card as their
> > driver support for linux leaks, nor any clone of a well known card.
> > You should have a look at the websites of big distributors like RedHat
> > oder SuSE, who both have compatibility databases online.
>
> Funny, others explicitly recommend ATI cards (bought one last month, works
> like a dream), and http://www.xfree86.org/ certainly leaves the impression
> that the ATI drivers are being maintained and improved...
Hi,
yes, seems like things have changed over the last months.
Not bad news :)
I just have had some problems when I bought a Xpert@Work a
few months after it was released and constantly had Xserver-
crashes due to memory leaks, and again there where problems
with the newly introduced RageFury. That time I read some
criticising articles on ATI's driver policy, maybe media
influence has changed their opinion on this topic.
Regards
Christian
--
|~-_ /~~~~~ Free Linux Portal: http://www.linux-config.de ~~~~~\ _-~|
| // de.etc.schreiben.* - Usenet-Literatur im www: \\ |
| // http://www.usenet-autoren.de \\ |
|_||[EMAIL PROTECTED]http://www.thepoet1.de__||_|
------------------------------
From: Hao-hua Chu <[EMAIL PROTECTED]>
Crossposted-To: intel.linux.kernel-dev
Subject: HELP: kgdb remote debugging
Date: Wed, 29 Mar 2000 11:13:59 -0800
I am trying to setup kgdb for kernel debugging without success.
I am using Redhat 6.1 with 2.2.12 kernel.
This is what I did and where I got stuck:
-Downloaded kgdb patch + applied the patch.
- Build kernel
- Copied the kernel to target machine
- Rebooted target machine , and the target machine said
Waiting for connection from remote gdb
- Started gdb on the host machine. I tried "remote target
/dev/ttyS0" and "remote target /dev/ttyS1". And I got the following
error message:
(gdb) target remote /dev/ttyS1
Remote debugging using /dev/ttyS1
Ignoring packet error, continuing...
I would greatly appreciate your help on this.
Hao
------------------------------
From: [EMAIL PROTECTED] (David Konerding)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.misc
Subject: Re: Precision of Linux's libm???
Date: Wed, 29 Mar 2000 19:19:54 GMT
On Tue, 28 Mar 2000 19:01:19 -0800, Tom Mitchell <[EMAIL PROTECTED]> wrote:
>
>On Thu, 17 Feb 2000, Johan Kullstam wrote:
>> From: Johan Kullstam <[EMAIL PROTECTED]>
>> Subject: Re: Precision of Linux's libm???
>>
>> chad <[EMAIL PROTECTED]> writes:
>>
>> > I am looking into doing so molecular modeling on Linux platforms because
>> > they are cost effective. After running some tests, I find errors,
>> > albeit small, in the results as compared to similar tests on SGIs or
>> > SUNs.
What do you mean errors? All floating point calculations on CPUs
involve some sort of approximation. There is round-off error and
truncation error. Each CPU implements floating point a little bit
different internally,
We get the same #'s in our molecular modelling code, on the various
CPUs: HP PA-RISC, MIPS, Alpha, ix86. A lot of work has gone into the
calculation of the potential, its derivative, and doing the timestep
integration to make sure that the calculation is accurate as possible.
In the past, the code was not "conserative" (it didn't compute the same
value across platforms), but with general improvement of the calculation
algorithm, it became the same across all platforms. I attribute this
to pretty much IEEE compliance on the part of these CPUs.
have you verified that what you ascribe to "errors" is not a value which
is "correct" within the limits of typical CPU precision? I assume you
are using double precision #s, right?
Note there is one small point I need to mention here: one of the
bug-fixes we made was to replace the Linux/GNU libc implementation of
the erfc() (error function, complementary) because we found an error
in the libc implementation. The new code has been submitted to the
libc programmers, but I don't know if it's been implemented there. So,
indeed, if your modelling code uses the erfc() function, then you may
see an error.
PS: even if there is a small difference in the values, you're not really
concerned by that. Due to the chaotic nature of molecular dynamics
simulations, a tiny difference *will* accumulate over time to be a
large difference. But you're not so concerned with that, since you'll
usually end up averaging a property over the whole time course of the
simulation, and that value can be measured to be the same even if the
two simulations do deviate on two architectures with slightly different
FPU implementations.
Dave
------------------------------
From: "Jason Nye" <[EMAIL PROTECTED]>
Subject: Re: 2.3.99: what's next step?
Date: Wed, 29 Mar 2000 19:20:40 GMT
>Well, it's probably _pretty_ close, but IIRC the 2.1 series went up to
>2.1.131 (which is not to say the 2.3's will necessarily go that high).
No, Giampaolo is right -- it looks like we're *really* close! The version
before 2.3.99 was 2.3.51 -- this is something significant, I think.
Cheerio
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.misc
Subject: Re: Precision of Linux's libm???
Date: 29 Mar 2000 19:23:37 GMT
In article <[EMAIL PROTECTED]>,
Tom Mitchell <[EMAIL PROTECTED]> wrote:
|
| On Thu, 17 Feb 2000, Johan Kullstam wrote:
| > From: Johan Kullstam <[EMAIL PROTECTED]>
| > Subject: Re: Precision of Linux's libm???
| >
| > chad <[EMAIL PROTECTED]> writes:
| >
| > > I am looking into doing so molecular modeling on Linux platforms because
| > > they are cost effective. After running some tests, I find errors,
| > > albeit small, in the results as compared to similar tests on SGIs or
| > > SUNs.
| >
| > most CPUs use IEEE 64 bit floating point (or 32 bit sometimes). intel
| > uses 80 bit floating point registers.
| ....
| yes, internal registers....
[ much discussion of numerical analysis deleted ]
I will just note here, gcc supports "long double" to allow use of 80
bit results stored in variables. This is not (a) portable nor (b) a
substitute for good numerical analysis, but it does allow things which
just would not produce any significant digits otherwise.
It also can make some things converge a *lot* faster!
Example:
#include <stdio.h>
#define dodiff(t,x) x = 1 - 3 * x;\
if (x < 0) x = -x;\
printf("%20s %.4Le\n", t, one/x)
main() {
long double foo, one = 1;
double bar;
float zot;
foo = 1; foo /= 3;
bar = 1; bar /= 3;
zot = 1; zot /= 3;
printf("Values:\n%35.30f\n%35.30f\n%35.30Lf\n", zot, bar, foo);
printf("\nError, 1 part in:\n");
dodiff("float", zot);
dodiff("double", bar);
dodiff("long double", foo);
exit(0);
}
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
The hardest test of maturity is knowing the difference between
resisting temptation and missing a once-in-a-lifetime opportunity.
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Subject: Re: System.map location
Date: 29 Mar 2000 19:29:23 GMT
In article <[EMAIL PROTECTED]>,
D. Stimits <[EMAIL PROTECTED]> wrote:
| I'm trying to find out where in the kernel it decides that it must
| read /boot/System.map, so I can create alternate locations and names.
| I'd like to boot multiple kernels without having to repoint sym links
| first.
There's a map= control in lilo.conf, I have not looked at the
implementation in detail, but perhaps that will be useful. You can
define the root partition on a per-stanza basis, perhaps the map
location can be set that way as well.
Look if yu care.
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
The hardest test of maturity is knowing the difference between
resisting temptation and missing a once-in-a-lifetime opportunity.
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Crossposted-To: redhat.config
Subject: Re: Reiserfs + Raid?
Date: 29 Mar 2000 19:37:52 GMT
In article <[EMAIL PROTECTED]>,
Kai Leibrandt <[EMAIL PROTECTED]> wrote:
| I am eagerly awaiting SGI's XFS filesystem (due this summer I think), but
| until then I'm experimeting with Reiserfs. Does anyone have any information
| on combining ReiserFS with a soft- or hardware RAID 1 or 5 array?
I have reason to believe that the abstraction allows the f/s to work
on any partition, however created. Haven't tried ReiserFS to test that,
but I seem to recall running xiafs on raid-0 at one time.
| On a similar note, does anyone know of a site with benchmark results
| comparing different CPU speeds but same hardware otherwise (i.e controllers,
| disks, memory) on a software RAID 5?
Someone in news.software.nntp commented that the BSD software RAID was
faster than the Megaraid controller, but that's hardly a qualtitative
description. I've run sw raid for a long time, and it seems to drive all
disks to hardware limits using raid-0 (122 iops on 9ms drives), or get
usefully close. This with <10% system time.
Raid-5 is another issue, speed goes up as chunks get larger, CPU use
does, too.
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
The hardest test of maturity is knowing the difference between
resisting temptation and missing a once-in-a-lifetime opportunity.
------------------------------
From: "Beno�t Cousson" <[EMAIL PROTECTED]>
Subject: Blocking I/O with ioctl() to network driver do not work !!!
Date: Wed, 29 Mar 2000 19:49:47 GMT
Hi,
Is it possible to implement the blocking I/O mechanism (sleep_on() &
wake_up()) with two IOCTLs (MY_IOCTL_WAIT and MY_IOCTL_WAKEUP) for a network
driver.
I need to send special commands to a specific network board and since these
command may last few seconds, I need a function to cancel the wait from the
user mode.
If I call ioctl(sock, MY_IOCTL_WAIT,.. ) the application is blocked, but
when I call ioctl(sock, MY_IOCTL_WAKE_UP,.. ) the calling application is
also blocked and I have to send a Ctrl-C to the first application to wake up
the two applications.
Is it poosible to implement that or not???
Thanks in advance.
Benoit.
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Subject: Re: software raid patch
Date: 29 Mar 2000 19:47:21 GMT
In article <[EMAIL PROTECTED]>,
D. Stimits <[EMAIL PROTECTED]> wrote:
| It appears that the software raid 0 patch (e.g., raidtools with
| raid0145-19990824-2.2.11) is still not part of the kernel at 2.2.14.
| Anyone know if software raid for striping will be built into the
| kernel soon, since it is a time-consuming to manually adjust these
| patches to work on newer kernels?
RAID-0 is already built into the kernel. You are using the wrong
toolset, which in turn requires the patches. See the "md" stuff.
The only thing which actually does seem to work is RAID-0, I couldn't
ever get mirroring to work, and the RAID-5 seems totally missing. But
what you asked for is there, and works just fine. I've been running it
for several years, before going to newer code to get 1, 0+1, and 5.
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
The hardest test of maturity is knowing the difference between
resisting temptation and missing a once-in-a-lifetime opportunity.
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Subject: Re: 2.3.99: what's next step?
Date: 29 Mar 2000 19:53:24 GMT
In article <cIsE4.6053$[EMAIL PROTECTED]>,
Jason Nye <[EMAIL PROTECTED]> wrote:
| >Well, it's probably _pretty_ close, but IIRC the 2.1 series went up to
| >2.1.131 (which is not to say the 2.3's will necessarily go that high).
|
|
| No, Giampaolo is right -- it looks like we're *really* close! The version
| before 2.3.99 was 2.3.51 -- this is something significant, I think.
Two of the last three "pre" releases failed to even compile without
hacks, that is something significant:-( One had a compile problem in a
driver (from memory) and the other was totally missing autoconf.h,
required by the makefile and menuconfig.
You can patch around them, but hopefully someone will do at least a
bit of QA before 2.4 comes out. Like making sure all features at least
compile in kernel and as modules. At least for the Intel and SPARC
platforms, which most users are using.
Doing full functional testing is not feasible, but a full compile test
certainly is.
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
The hardest test of maturity is knowing the difference between
resisting temptation and missing a once-in-a-lifetime opportunity.
------------------------------
From: [EMAIL PROTECTED] (Kaz Kylheku)
Subject: Re: Blocking I/O with ioctl() to network driver do not work !!!
Reply-To: [EMAIL PROTECTED]
Date: Wed, 29 Mar 2000 20:08:27 GMT
On Wed, 29 Mar 2000 19:49:47 GMT, Beno�t Cousson <[EMAIL PROTECTED]> wrote:
>Hi,
>
>Is it possible to implement the blocking I/O mechanism (sleep_on() &
>wake_up()) with two IOCTLs (MY_IOCTL_WAIT and MY_IOCTL_WAKEUP) for a network
>driver.
>I need to send special commands to a specific network board and since these
>command may last few seconds, I need a function to cancel the wait from the
>user mode.
In that case, don't use sleep_on and wake_up, but rather interruptible_sleep_on
and wake_up_interruptible. You can then send the process a signal to wake it
up. Also, you might want to implement a timeout mechanism. To do that, you
``manually'' enqueue the calling thread in a wait queue, change its state to
TASK_INTERRUPTIBLE and then call schedule_timeout() to do a timed wait.
When this function returns, you can then check for pending unblocked signals.
--
#exclude <windows.h>
------------------------------
From: <[EMAIL PROTECTED]>
Subject: Problem with my configuration (Modem and Internet Connection -
Kppp....Netscape)
Date: Wed, 29 Mar 2000 20:30:11 GMT
Hi,
I am a new user of WinLinux 2000. I enjoy to use very much WinLinux 2000
and I will appreciate if I can use my modem correctly.
Presently, I have a problem with my modem. I tried to configure it but I
can't. I submit you my problem and I hope that you will be able to tell me
how to proceed for my configuration.
Thank you
Technical
Information:
Windows version: Windows 98
Amount of RAM (memory): 64 MB
CPU model: Intel Pentium III
Problem
Description:
The problem
occurs : When trying to run a WinLinux application
A short
subject: Problem with configurations (my modem and my
Internet connection)
Describe your
problem:
After I double-clicked on the icon DialUp ... a window appeared to tell
me :
- Error
This kernel has no PPP support, neither compiled in nor via the kernel
module loader
To solve this problem :
* contact your system administrator
or
* install a kernell with PPP support
OK
But I can continue with the window Kppp.
I can click on CONNECT and when I try to connecting to Internet with my
connection I can't and another window appears to tell me :
Connecting to : Internet Videotron
Sorry, the modem is busy
LOG CANCEL
--
Posted via CNET Help.com
http://www.help.com/
------------------------------
From: [EMAIL PROTECTED] (Jonathan Magid)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Recomendation for a graphical 'diff' style program for Linux?
Date: 29 Mar 2000 13:55:44 -0500
In article <[EMAIL PROTECTED]>,
Tim Harvey <[EMAIL PROTECTED]> wrote:
>Greetings,
>
>I'm looking for recomendations for a nice graphical diff program for
>Linux. I'm looking for something very close to 'Beyond Compare' for
>windows or like the diff that comes on an Irix system. I'm aware of
>'kdiff' but it doesn't have the features I'm looking for. I'm looking
>for something that can diff directories as well as files, and that can
>allow you to edit files by moving diffs back and forth between right and
>left panes.
I know of two, tkdiff (which uses tcl/tk) and mgdiff (which requires motif).
Both allow you to graphically select regions of text to merge into a third
document. Neither allow you to compare and merge directories (as far as I
know).
tkdiff is available at http://www.ede.com/free/tkdiff/index.html
mgdiff is available at various places. Do a search for it on www.google.com.
cheers,
jem.
--
[EMAIL PROTECTED]|SEEKING: High level opening in New World Order.
| Assigned black helicopter is a must.
------------------------------
** 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
******************************
