Linux-Development-Sys Digest #737, Volume #6 Sat, 22 May 99 08:13:58 EDT
Contents:
Re: never reboot to upgrade ? ("Ty! Boyack")
Re: never reboot to upgrade ? ("G. Sumner Hayes")
is vnc video card independent ? (Matt)
Update: Re: 2.2.9 Hangs after a few hours of uptime (John Ioannidis)
Re: New Project: Linux Upgrade Monitor (upgrademon) (Paul D. Smith)
Re: mapping user space and kernel space (Nitin Malik)
Re: Microsecond resolution timer? (Peter Samuelson)
Promote your site - ONLY $14.95 42400 ([EMAIL PROTECTED])
Went to download Staroffice... Microsoft screwed it up. (Kevin Burton)
Virtual to Physical address translation (Joseph Virzi)
Translation the installation diskette (=?iso-8859-1?Q?Fr=E9d=E9ric?= FAUCOUNEAU)
Re: [Q] sched_setscheduler ? (Roope Anttinen)
Using a shared library... Why won't my code compile. (Kevin Burton)
After X, console is messed up (Timothy J. Lee)
Re: Using a shared library... Why won't my code compile. (Jean Wolter)
Re: Using a shared library... Why won't my code compile. (Bob Withers)
----------------------------------------------------------------------------
From: "Ty! Boyack" <[EMAIL PROTECTED]>
Subject: Re: never reboot to upgrade ?
Date: Fri, 21 May 1999 12:42:08 -0600
Mike Rushford wrote:
>
> I have enjoyed the stability of Linux for years and have only had to
> shut down my Linux boxes to upgrade the next Kernel version.
>
> Is there a way of building a kernel so that it takes the place of a
> currently running one without ever shutting down?
Might be somthing we'll have to do to "keep-up-with-the-jones'"
Sun has said that by Solaris 9 you will be able to do live
version upgrades. We'll have to see if they mean live in
the same sense that we are talking about, but so far it
sounds like it.
-Ty!
--
|==========================||======================================|
| Ty! Boyack || Colorado State University |
| Unix Services Manager || College of Natural Resources |
| [EMAIL PROTECTED] || A105B, Advanced Technology Lab, NESB |
| (970) 491-4966 || |
|==========================||======================================|
------------------------------
From: "G. Sumner Hayes" <[EMAIL PROTECTED]>
Subject: Re: never reboot to upgrade ?
Date: Fri, 21 May 1999 14:39:46 -0400
Tony Smith wrote:
> Tandem's implementation of (a) is seriously good stuff. It's called
> ServerNet and I believe that today it runs at around 50 mega BYTES per
> second so there's plenty of I/O bandwidth between nodes.
How's the latency on that? Do you have any pointers?
FWIW, a couple of years ago a lot of clustering solutions were using
HIPPI at 800-1600 Mb/sec, which is 100-200 megabytes/sec. The bandwidth
is high enough that you can do lots of neat stuff with it. It's not
quite down to the 60 nanosecond latency of local RAM, though. If you're
just migrating processes, of course, latency becomes much less
important.
--Sumner
------------------------------
Date: Fri, 21 May 1999 22:42:21 +0100
From: Matt <[EMAIL PROTECTED]>
Subject: is vnc video card independent ?
Crossposted-To:
comp.os.linux.x,comp.os.linux.networking,comp.os.os-windows.nt.misc,comp.os.linux.development.apps
Hi,
Problem I can get the vnc server running on NT and
Linux but there seems to be another problem.
Question is vnc video card independent ?
Both boxes have diff video cards one Martox G200
and NT 3D savage. Can I use vnc on NT and view
Linux x windows which box has a Martox card
or do I have to load the 3d Savage drives on
the Linux box.
Many thanks
Matt
------------------------------
From: [EMAIL PROTECTED] (John Ioannidis)
Subject: Update: Re: 2.2.9 Hangs after a few hours of uptime
Date: Sat, 22 May 1999 04:25:43 GMT
In article <[EMAIL PROTECTED]>,
John Ioannidis <[EMAIL PROTECTED]> (that's me!) wrote:
>I have 2.2.9 running on Pentium II machines (Tyan S1692DL motherboard,
>256M of ECC SDRAM, 333MHz P-II (id 0x650, stepping 0), 2940UW, generic
>video card, no IDE devices present), and after a few hours of uptime
>they lock up. No crashes, no diagnostics on the console; they don't
>respond to ctl-alt-del, or to sysreq chords; the only thing that
>brings them back is a hard reset. I don't recall seeing this behavior
>in 2.2.5, so something must have gone wrong with 2.2.9. Is anyone
>else experiencing this?
>
I configured a serial console, and here is the last (only) console
message right before the system crashed:
eth0: Too much work in interrupt, status e081. Temporarily disabling
functions (7f7e).
The ethernet card is a 3Com 3C905-B.
/ji
------------------------------
From: [EMAIL PROTECTED] (Paul D. Smith)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: New Project: Linux Upgrade Monitor (upgrademon)
Date: 22 May 1999 01:30:33 -0400
Reply-To: [EMAIL PROTECTED]
%% Kevin Burton <[EMAIL PROTECTED]> writes:
>> The one thing missing is simple auto-configuration and upgrading. I
>> shouldn't have to know about ftp sites, I shouldn't have to think
>> about software upgrades. It should just work. It's a hard problem.
Not really--Debian's new apt-get tool does it already. Of course,
Debian has a much better structured release environment for this kind of
thing than RedHat.
apt-get has a config file that contains a list of places to look for
Debian packages. The default config looks at all the standard Debian
dists, of course. You can add more for, for example, the KDE site if
you want.
You use "apt-get update" to make sure the database of possible packages
is up-to-date, then you just say "apt-get upgrade" and that's it: it
downloads, unpacks things in the right order, installs, and configures.
Also, Debian has gnome-apt which looks OK--needs some work I think, but
seems like a reasonable tool.
--
===============================================================================
Paul D. Smith <[EMAIL PROTECTED]> Network Management Development
"Please remain calm...I may be mad, but I am a professional." --Mad Scientist
===============================================================================
These are my opinions---Nortel Networks takes no responsibility for them.
------------------------------
Date: Fri, 21 May 1999 19:45:57 -0400
From: Nitin Malik <[EMAIL PROTECTED]>
Subject: Re: mapping user space and kernel space
> > The network devices don't have a mmap routine listed as a possible device
>operations
> > (in netdevice.h). Does this mean it doesn't support it or some one didn't define
> > such an operation for network devices.
> >
> > So what do I do? Do I add the mmap routine to the list of operations defined for
>n/w
> > devices? If this is possible, it would be a great... else any clues?
>
> Current networking has no support for zero copy operation (=receiving
> directly from/to error space). You basically have to rewrite the network
> stack for that. It is a _big_ task.
>
i don't think so... as it is in the existing scheme, most of the time a
pointer to the packet is passed from the driver to the above layers...
this depends on the copybreak value. Hence we effectively have only one
copy... from kernel to user space. If we map this area to the user space,
we avoid this extra copy...
so i come back to the original question... can i do the mmap??
nitin
------------------------------
From: [EMAIL PROTECTED] (Peter Samuelson)
Subject: Re: Microsecond resolution timer?
Date: 20 May 1999 22:24:36 -0500
Reply-To: Peter Samuelson <[EMAIL PROTECTED]>
[Peter Samuelson <sampo.creighton.edu!psamuels>]
> > For that sort of RT you need kernel support. Write your driver
> > partly or completely in kernel space.
[Mark Hahn <[EMAIL PROTECTED]>]
> actually, this is untrue. RT user-space processes work dandy for
> this sort of thing. of course, you wind up busy-waiting for at least
> some time (my code spins looking at parallel/game/serial ports and
> doing rdtsc.)
Yeah, I guess I mean *without* busy-waiting. Sure, you can do anything
in user-space if you don't mind those sorts of tradeoffs. Maybe
busy-waiting a few intervals of 20us once in awhile wouldn't be as bad
as all that. But if the hardware demands constant attention -- well,
your box just won't get much other work done; one just hopes you won't
get upset if your bash prompt seems sluggish....
Depending on how hard your RT needs to be, you will want to schedule
the job with RT priority so it doesn't get preempted at an inconvenient
time, and you may even need to cli(), which (from userspace) is evil.
--
Peter Samuelson
<sampo.creighton.edu!psamuels>
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.multimedia,comp.os.linux.advocacy
Subject: Promote your site - ONLY $14.95 42400
Reply-To: [EMAIL PROTECTED]
Date: Saturday, 22 May 1999 02:12:10 -0600
Promote your website to Over 1300 Search Engines and Directories for
ONLY 14.95
Don't pass this great deal up!!
http://promotesite.webjump.com
F%
------------------------------
From: Kevin Burton <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Went to download Staroffice... Microsoft screwed it up.
Date: Sat, 22 May 1999 02:09:46 -0700
Went to DL the new Staroffice. Look what I got:
[Microsoft][ODBC SQL Server Driver]Unknown token received from SQL
Server
--
Kevin A. Burton
Internet Guy
------------------------------
From: Joseph Virzi <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.development.apps
Subject: Virtual to Physical address translation
Date: Fri, 21 May 1999 18:15:49 -0700
I've a device which DMAs data into memory. Given a buffer, whose pointer
is really a virtual address, how do I get the physical address of that
buffer, so that I can tell the DMA driver where to put the data?
Rubini states that virt_to_phys() and virt_to_bus() perform these
functions, but after looking at the actual source code for these
functions, they turn out to be the same address. I know this is not
write, that physical and virtual addresses are the same.
I've used vremap() to translate a physical address to a virtual one, so
that I can access PCI devices properly. This time, I have the inverse
problem. I have a buffer whose physical address I need.
This is a common issue, I'm sure. There's something I'm just not
understanding. All these damn trees obscure my view of the forest.
-Joe
------------------------------
From: =?iso-8859-1?Q?Fr=E9d=E9ric?= FAUCOUNEAU <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.misc
Subject: Translation the installation diskette
Date: Sat, 22 May 1999 10:58:09 +0200
Hello,
I would like to know how can I read in correct charset a file
"frx.tr" or "esx.tr" wich are in file "translationx.tr" which is in
installation diskette. I would like modify this file so I would like
using the good charset.
I have tried with recode but I don't know the initial charset.
I would like too add a command line after the installation but I
don't want use kickstart...
Thanks for you help.
Fr�d
------------------------------
From: Roope Anttinen <[EMAIL PROTECTED]>
Subject: Re: [Q] sched_setscheduler ?
Date: 22 May 1999 11:13:17 GMT
Reply-To: [EMAIL PROTECTED]
Soohyung Lee <[EMAIL PROTECTED]> wrote:
> That is ' Can each process belong to different scheduler (policy) ? '
> For example, is it possible that process A belong to SCHED_OTHER,
> process B belongs to SCHED_FIFO, and process C belongs to SCHED_RR ?
Yup. You can use the RT schedulers for some time critical processes while
the rest of processes use the draditional scheduler. Even different threads
inside one process can use different schedulers. See pthread_setschedparam
for that.
Roope
--
MicroSoft? is that some kind of a toilet paper?
PS: Look for address here, not from headers. And remove NOSPAM's
___________________________________________________________________________
[EMAIL PROTECTED] / [EMAIL PROTECTED]
+358 9 812 7567 / +358 500 445 565 / +358 49 445 565
http://myy.helia.fi/~anttiner/index.html
===========================================================================
Helsinki Business Polytechnic - Institute of information technology
------------------------------
From: Kevin Burton <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Using a shared library... Why won't my code compile.
Date: Sat, 22 May 1999 04:13:59 -0700
Warning. I just starting out with C/C++ on the Linux platform... here
goes.
I am trying to use a library: libftp. Only problem is that I can't get
it to work.
Here is my code:
--
extern void ftpInit(void);
int main(){
//ftpInit();
}
--
I am compiling like this:
--
c++ -DHAVE_CONFIG_H -I. -I. -I.. -O0 -g1 -Wall -c cupgrademon.cpp
c++ -DHAVE_CONFIG_H -I. -I. -I.. -O0 -g1 -Wall -c main.cpp
c++ -O0 -g1 -Wall -o upgrademon cupgrademon.o main.o -lftp
--
(-lftp is supposed to link to my shared library)
I do an ldd on my binary...
--
> ldd upgrademon
libftp.so.0 => /usr/lib/libftp.so.0 (0x4001a000)
libstdc++.so.2.8 => /usr/lib/libstdc++.so.2.8 (0x4001e000)
libm.so.6 => /lib/libm.so.6 (0x4005f000)
libc.so.6 => /lib/libc.so.6 (0x4007d000)
/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
--
(well what do you know... it is shared linked)
So it is shared right?
Then I uncomment line 5 (//ftpInit) to start using libftp.so.0 and my
compile fails.
--
c++ -DHAVE_CONFIG_H -I. -I. -I.. -O0 -g1 -Wall -c cupgrademon.cpp
c++ -DHAVE_CONFIG_H -I. -I. -I.. -O0 -g1 -Wall -c main.cpp
c++ -O0 -g1 -Wall -o upgrademon cupgrademon.o main.o -lftp
main.o: In function `main':
/root/upgrademon/upgrademon/main.cpp(.text+0x4): undefined reference to
`ftpInit(void)'
collect2: ld returned 1 exit status
gmake: *** [upgrademon] Error 1
--
My library does have the ftpInit function because I did an "nm" on it
and it was there:
> nm /usr/lib/libftp.so.0 |grep ftpInit
00000e10 T ftpInit
What am I doing wrong? I would really appreciate some assistance here!
Could it be that ftpInit is in a different name space?
Thanks alot!
--
Kevin A. Burton
Internet Guy
------------------------------
From: [EMAIL PROTECTED] (Timothy J. Lee)
Subject: After X, console is messed up
Reply-To: see-signature-for-email-address---junk-not-welcome
Date: Sat, 22 May 1999 05:05:07 GMT
On a computer with the following configuration:
Video card: Cirrus Logic GD5465 chipset, 4 MB, AGP
Memory: 64 MB
Motherboard: Asus P5A (aperture size in BIOS set to 64 MB)
Processor: AMD K6-2, 300 MHz
OS: Red Hat 5.1 Linux, updated with patches from Red Hat including
XFree86 3.3.3.1, updated to kernel 2.0.36 from kernel.org
If I configure X using Xconfigurator and then use it from the console,
it works (SVGA server). But after exiting, the console display is
messed up -- the characters displayed are from previously displayed
lines, not what should be displayed (although the cursor is in the
correct place). Other virtual consoles (those reached by Alt-F2,
etc.) are also messed up. Only rebooting fixes it.
If I edit /etc/X11/XF86Config to put "VideoRam 2048" in the appropriate
place, X works and does not mess up the console after exiting X.
--
========================================================================
Timothy J. Lee timlee@
Unsolicited bulk or commercial email is not welcome. netcom.com
No warranty of any kind is provided with this message.
------------------------------
From: Jean Wolter <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Using a shared library... Why won't my code compile.
Date: 22 May 1999 13:42:48 +0200
Kevin Burton <[EMAIL PROTECTED]> writes:
> My library does have the ftpInit function because I did an "nm" on it
> and it was there:
>
> > nm /usr/lib/libftp.so.0 |grep ftpInit
> 00000e10 T ftpInit
>
>
>
> What am I doing wrong? I would really appreciate some assistance here!
> Could it be that ftpInit is in a different name space?
Make an nm on your object file and you will see the reason:
nm test.o
00000000 ? __FRAME_BEGIN__
U ftpInit__Fv
00000000 t gcc2_compiled.
00000000 T main
c++ does name mangeling (coding parameter and return types into the
function name) and your init function is a normal c function. Try
using the header files for your library or if it still doesn't work
use
extern "C" void ftpInit(void);
to tell the compiler, that it should use a C linkage for this
function.
Hope this helps,
Jean
--
I get up each morning, gather my wits.
Pick up the paper, read the obits.
if I'm not there I know I'm not dead.
So I eat a good breakfast and go back to bed. Peete Seeger
------------------------------
From: [EMAIL PROTECTED] (Bob Withers)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Using a shared library... Why won't my code compile.
Date: Sat, 22 May 1999 12:09:12 GMT
[This followup was posted to comp.os.linux.development.apps and a copy
was sent to [EMAIL PROTECTED] if this address didn't require
editing.]
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] says...
> Warning. I just starting out with C/C++ on the Linux platform... here
> goes.
>
> I am trying to use a library: libftp. Only problem is that I can't get
> it to work.
>
> Here is my code:
>
>
> --
> extern void ftpInit(void);
>
> int main(){
>
> //ftpInit();
>
> }
> --
>
> I am compiling like this:
>
> --
> c++ -DHAVE_CONFIG_H -I. -I. -I.. -O0 -g1 -Wall -c cupgrademon.cpp
> c++ -DHAVE_CONFIG_H -I. -I. -I.. -O0 -g1 -Wall -c main.cpp
> c++ -O0 -g1 -Wall -o upgrademon cupgrademon.o main.o -lftp
> --
>
> (-lftp is supposed to link to my shared library)
>
> I do an ldd on my binary...
>
> --
> > ldd upgrademon
> libftp.so.0 => /usr/lib/libftp.so.0 (0x4001a000)
> libstdc++.so.2.8 => /usr/lib/libstdc++.so.2.8 (0x4001e000)
> libm.so.6 => /lib/libm.so.6 (0x4005f000)
> libc.so.6 => /lib/libc.so.6 (0x4007d000)
> /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
>
> --
>
> (well what do you know... it is shared linked)
>
> So it is shared right?
>
> Then I uncomment line 5 (//ftpInit) to start using libftp.so.0 and my
> compile fails.
>
> --
> c++ -DHAVE_CONFIG_H -I. -I. -I.. -O0 -g1 -Wall -c cupgrademon.cpp
Try declaring the function prototype as extern "C", i.e.:
extern "C" void ftpInit(void);
Regards,
Bob
--
============================================================
Bob Withers Do or do not, there is no try
[EMAIL PROTECTED] Yoda
http://www.pobox.com/~bwit
============================================================
------------------------------
** 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
******************************
