Linux-Development-Sys Digest #429, Volume #6 Thu, 25 Feb 99 17:14:04 EST
Contents:
Re: Why I'm dumping Linux, going back to Windblows (George Bynum)
Re: Has anybody compiled a 100% glibc-2.1 and kernel2.2 based system?
([EMAIL PROTECTED])
Re: Will 2.2.x support removable medias better? (Kalle Olavi Niemitalo)
Advanced block driver question (Felix Rauch)
Re: Overclocking (was: Re: K6-2 and Linux, Are there any Bug?) (BL)
Re: Overclocking (was: Re: K6-2 and Linux, Are there any Bug?) (Cooper)
Re: Kernel 2.2.2 installation (Aaron Faby)
System.map/vmlinuz (Aaron Faby)
Re: Why I'm dumping Linux, going back to Windblows (Julian Robert Yon)
Possible new hardware platform (ErikLBrown)
Re: System.map/vmlinuz (Martin Recktenwald)
Re: "RPM's harmful to Linux" harmful to Linux (Bill Anderson)
Re: Will 2.2.x support removable medias better? (Bill Anderson)
Re: Raw writing to PCMCIA SRAM cards (David Hinds)
Re: System.map/vmlinuz (Chun-Chung Chen)
Re: Why I'm dumping Linux, going back to Windblows (jimmy)
help!! AF_PACKET socket program example ... ("À̱¤ÁØ")
Re: Java 2 (Reality is a point of view)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (George Bynum)
Crossposted-To: alt.os.linux,comp.os.linux.development.apps,comp.os.linux.setup
Subject: Re: Why I'm dumping Linux, going back to Windblows
Date: Thu, 25 Feb 1999 13:57:22 GMT
On Wed, 24 Feb 1999 16:01:34 -0500, Rick Onanian
<[EMAIL PROTECTED]> wrote:
>My personal opinion: Currently, Linux is the best choice for servers
>of all sorts. It's good for those who like to get their hands dirty,
>the computer equivalant of the guy who won't buy a new car, who instead
>rebuilds a 1968 camaro by hand. It's even good for an office environment
>where the users have specific purposes, and there's a good IS guy
>available at all times, although it seems to not have ever been done.
I've purchased RH 5.2 but haven't begun playing with it yet. I'm
comfortable with computers, and prefer DOS to Windows (and use 4DOS
for command line sessions) ... and expect little trouble. My employer
runs SVR4.2, but our only interface to it is one application that
autoruns on login, and ftp.
Now for the question ...
As a server, will it be relatively easy (with LRP preferably) to have
a RAS give me access to an IP network? I've played a little with
Win95's DUS which will "route" (bridge is probably a better term for
the action) IPX, but not IP. My thinking here is that I hope to get a
cable modem for my home network (me, the wife, and 3 children in the
summer), and I'd love to be able to dial in too ...
My employer's dialup, and my present ISP use CHAP authentification;
will my (Windows 9x) clients setup be similar to Linux?
George
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Has anybody compiled a 100% glibc-2.1 and kernel2.2 based system?
Date: 25 Feb 1999 08:14:34 -0500
If I were you I'd wait for the next egcs. It looks like they're
working on a fix for the glibc2.1 problems.
------------------------------
From: Kalle Olavi Niemitalo <[EMAIL PROTECTED]>
Subject: Re: Will 2.2.x support removable medias better?
Date: 25 Feb 1999 13:48:48 +0200
[EMAIL PROTECTED] (Peter Samuelson) writes:
> In some cases I think you can poll for that information, one
> way or another -- Windoze seems to for CD's.
This Microsoft document might be relevant:
Media Status Notification for SCSI and ATAPI Devices
http://microsoft.com/hwdev/respec/scsimed.htm
------------------------------
From: [EMAIL PROTECTED] (Felix Rauch)
Subject: Advanced block driver question
Date: 25 Feb 1999 00:40:29 +0100
I'm not sure if it is possible to handle multiple requests (in the
request()-function of a block driver) in parallel before calling
end_request().
I do have an idea, but I'm not sure if it is the right thing. This is
important for a project in our lab where we would like to access
multiple remote machines in parallel over a highspeed network. I asked
the question already two weeks ago, but didn't get an answer :-(
A student in our lab is writing a block driver for Linux which
transfers blocks from/to remote machines over Myrinet (a gigabit/s
SAN/LAN networking technology). His driver seems to work fine, he can
access a remote machine by mounting his device (e.g. by executing
"mount /dev/mdfsa /import/mdfsa", where "mdfs" stands for "Myrinet
Distributed File System"). He can even use the MD-driver to "pack" 4
of his devices into a raid0 device. The remaining problem is that the
"raided" devices are not faster than a single device :-(
I read "Linux Device Drivers" by Alessandro Rubini (which is a quite
good book btw.!) and <linux/blk.h> and thought about it. So my
question is:
Instead of something like (which is the SOP, AFAIK):
for(;;) {
INIT_REQUEST;
look at CURRENT, request block block from remote machine and
wait until it's here;
when the block is here, call end_request();
}
Is the following procedure (or something similar) in a
request()-function possible?
for(;;) {
INIT_REQUEST;
while(CURRENT != NULL) {
send requests for all buffer_heads in CURRENT and remove them
from the request-queue (but we keep a pointer to the
buffer_heads somewhere to be able to unlock them later).
CURRENT = CURRENT->next;
}
while not all requests have been answered {
/* The myrinet-card sends an interrupt to the CPU when a
packet has been received, so: */
wait for next answer with interruptible_sleep_on();
call mark_buffer_uptodate() and unlock_buffer() for the buffer
which has just been filled (do basically the stuff which is
in end_request() ).
}
}
Is this possible or have I missed something? I am especially
interested to know if it is sufficient to call mark_buffer_uptodate()
and unlock_buffer() for every received buffer or if there is something
else.
I would appreciate hints for this as I'm really interested in how much
speed we can get for such a distributed, parallel filesystem.
- Felix
--
Felix Rauch, research assistant @ ETH Zurich, Institute for Computersystems
Homepage: http://nice.ethz.ch/~felix/ (includes PGP public key)
Email: [EMAIL PROTECTED] -> This article contains my personal views only <-
------------------------------
From: BL <[EMAIL PROTECTED]>
Crossposted-To:
alt.os.linux.slackware,comp.os.linux.hardware,comp.os.linux.misc,comp.os.linux.setup
Subject: Re: Overclocking (was: Re: K6-2 and Linux, Are there any Bug?)
Date: 25 Feb 1999 14:45:29 GMT
: the only chips i hear of that can be is celeron but besides that.... the
: chip is not made to be over clocked..... if it was supposed to be over
: clocked it would RUN at THAT SPEED not 50-100 MHz slower
ever hear of "Marketing"?
do you think that simply marking numbers on a chip is enough for gullible
consumers to believe that that's all the chip was -designed- to do?
sometimes the markings are accurate (technically) and sometimes not.
don't accept blanket statements about o/c-ing. it all depends on the
particular chip. check the net first before taking such a hardline on the o/c
issue... a lot of people can o/c the celeron, whereas other chips may
actually be at their marked limits. it all depends.
------------------------------
From: Cooper <[EMAIL PROTECTED]>
Crossposted-To:
alt.os.linux.slackware,comp.os.linux.hardware,comp.os.linux.misc,comp.os.linux.setup
Subject: Re: Overclocking (was: Re: K6-2 and Linux, Are there any Bug?)
Date: Thu, 25 Feb 1999 15:50:31 -0100
Michael Creasy wrote:
>
> If this is the case can someone tell me why my K6-2 350 when overclocked
> to 400 crashes linux on boot, it has a huge fan on it.
Overclocking speeds up more than just the CPU. If the rest of your system such
as your memory can't cope, it still won't run.
Cooper
--
Linux: Proof of intelligent life on earth
------------------------------
From: Aaron Faby <[EMAIL PROTECTED]>
Subject: Re: Kernel 2.2.2 installation
Date: Thu, 25 Feb 1999 09:58:01 -0500
==============38971E22BA5D7987D1C2BE84
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Surfer wrote:
> Hi Aaron,
> I wonder how you compiled 2.2.2 - make depend failed
>
> -----------------------------------------------
> "
> surfer@ml01:/usr/src/linux-2.2.2 > make depend
> gcc -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -o scripts/mkdep
> scripts/
> mkdep.c
> scripts/mkdep.c: In function `do_depend':
> scripts/mkdep.c:446: `PROT_READ' undeclared (first use in this function)
> scripts/mkdep.c:446: (Each undeclared identifier is reported only once
> scripts/mkdep.c:446: for each function it appears in.)
> scripts/mkdep.c:446: `MAP_PRIVATE' undeclared (first use in this
> function)
> make: *** [scripts/mkdep] Error 1
> "
> ------------------------------------------------
>
> Can you help me?
> thanks,
> Martin.
I just ran the default commands as usual :
make xconfig
make dep
make clean
make zImage
make modules
make modules_install
make zlilo
than it compiled without problems. The installation is where i had
trouble. Maybe you have a conflicting configuration. Go through config
again with different options and see if you can get any further.
--
Aaron Faby
[EMAIL PROTECTED]
System Administrator/Technical Support
Yourlink, Inc.
==============38971E22BA5D7987D1C2BE84
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
Surfer wrote:
<blockquote TYPE=CITE>Hi Aaron,
<br>I wonder how you compiled 2.2.2 - make depend failed
<p>-----------------------------------------------
<br>"
<br>surfer@ml01:/usr/src/linux-2.2.2 > make depend
<br>gcc -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -o scripts/mkdep
<br>scripts/
<br>mkdep.c
<br>scripts/mkdep.c: In function `do_depend':
<br>scripts/mkdep.c:446: `PROT_READ' undeclared (first use in this function)
<br>scripts/mkdep.c:446: (Each undeclared identifier is reported only once
<br>scripts/mkdep.c:446: for each function it appears in.)
<br>scripts/mkdep.c:446: `MAP_PRIVATE' undeclared (first use in this
<br>function)
<br>make: *** [scripts/mkdep] Error 1
<br>"
<br>------------------------------------------------
<p>Can you help me?
<br>thanks,
<br>Martin.</blockquote>
I just ran the default commands as usual :
<p>make xconfig
<br>make dep
<br>make clean
<br>make zImage
<br>make modules
<br>make modules_install
<br>make zlilo
<p>than it compiled without problems. The installation is where i had
<br>trouble. Maybe you have a conflicting configuration. Go through config
<br>again with different options and see if you can get any further.
<pre>--
Aaron Faby
[EMAIL PROTECTED]
System Administrator/Technical Support
Yourlink, Inc.</pre>
</html>
==============38971E22BA5D7987D1C2BE84==
------------------------------
From: Aaron Faby <[EMAIL PROTECTED]>
Subject: System.map/vmlinuz
Date: Thu, 25 Feb 1999 12:06:31 -0500
I installed 2.2.2 and for some reason System.map, vmlinuz, and
vmlinuz.old
showed up in / as well as /boot. Anyone know why and how I can avoid the
files being dropped in / ? Thanks.
--
Aaron Faby
[EMAIL PROTECTED]
System Administrator/Technical Support
Yourlink, Inc.
------------------------------
From: Julian Robert Yon <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.development.apps,comp.os.linux.setup
Subject: Re: Why I'm dumping Linux, going back to Windblows
Date: Thu, 25 Feb 1999 17:11:34 +0000
Sniper wrote:
> Hit the nail straight on the Head my friend !
You've obviously made up your mind. Surely there's no point in a
religious war about this. You can live with the consequences of your
decision, and the rest of us will do the same.
Julian
~~~~~~~~~~~~~~~~~~~~
[EMAIL PROTECTED]
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (ErikLBrown)
Subject: Possible new hardware platform
Date: 25 Feb 1999 17:22:24 GMT
Greetings!
For those of you who read both this newsgroup and comp.os.linux.advocacy, I
apologize for the redundant information.
My name is Erik Brown, and I have come up with a computer architecture that I
would like to develop in an open, collaborative fashion. I would like to
explore ways of using the patent on the architecture as a sort of hardware GPL,
and am considering using GNU tools and perhaps the Linux kernel as the basis
for the machine's first O.S. I submit that the asymmetric multiprocessing
nature of the architecture may be of interest to those that read this
newsgroup, and I ask for any comments and suggestions you might have regarding
the project. The project web site is:
http://www.oomecs.org
Thanks for your time!
Erik
------------------------------
From: Martin Recktenwald <[EMAIL PROTECTED]>
Subject: Re: System.map/vmlinuz
Date: 25 Feb 1999 18:57:13 +0100
Aaron Faby <[EMAIL PROTECTED]> writes:
> I installed 2.2.2 and for some reason System.map, vmlinuz, and
> vmlinuz.old
> showed up in / as well as /boot. Anyone know why and how I can avoid the
>
> files being dropped in / ? Thanks.
Don't install them there. Just make the kernel with "make zImage" or
better "make bzImage" and copy the resulting files (System.map in
kernel source dir, vmlinuz in arch/i386/boot for ia86) where you want
them.
"make zlilo" automagically places System.map and vmlinuz in /.
Martin.
--
"The spider at the center of this self-spinning web."
- New York Times about Linus Torvalds and Linux
------------------------------
From: Bill Anderson <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.advocacy
Subject: Re: "RPM's harmful to Linux" harmful to Linux
Date: Thu, 25 Feb 1999 18:02:33 +0000
Peter Mutsaers wrote:
>
> >> "CB" == Christopher Browne <[EMAIL PROTECTED]> writes:
>
> CB> I'm not *specifically* aware that there is an RPM port to
> CB> FreeBSD, OpenBSD, or NetBSD, but would be surprised to find it
> CB> an overly difficult task.
>
> Yes it has been ported (available as a FBSD "port"), though it is only
> used for adding Linux RPM's for use in the Linux emulator, i.e. when
> you want to run Oracle for Linux under FBSD you need some RPM's in
> /compat/linux.
>
> CB> At any rate, the *BSD folk don't use a "pure tarball" approach
> CB> for software, but rather use the "Ports" system.
>
> The distribution itself still is tarball. The ports just get the
> tarball sourcecode, then patch & build it.
>
> I find it pretty stupid and annoying that RH uses rpm's that contain
> full sourcecode instead of only patches and build instructions. For
> every minor update you need to download the whole package again,
> waisting bandwidth and time.
Are you downloading .src.rpm every time and then rebuilding the package?
A non source rpm (unless goofed up by the packager) does not include the
full source. A non-source rpm distributes binaries and
(usually)documentation.
And if you have the source rpm of something, and a patch comes out, you
can update your spec file, put the patch in the source directory, and
the rebuild & install the new rpm, thus saving you the download of
sources/binaries.
What gives me angst is kde.org's use of rpm. They build one big (~21MB)
tarball containg the kde 1.1 rpms for RH, but don't seem to have them
seperate.
Bill
------------------------------
From: Bill Anderson <[EMAIL PROTECTED]>
Subject: Re: Will 2.2.x support removable medias better?
Date: Thu, 25 Feb 1999 17:48:44 +0000
"Adam P. Jenkins" wrote:
>
> [EMAIL PROTECTED] (Horst von Brand) writes:
> > >The point is that it should be unnecessary to mount/umount a removable
> > >media if the OS is capable of detecting insertion/ejection of media.
> >
> > Can't be done in general, due to the extremely clever PC architecture.
>
> Can you please elaborate? I don't remember ever having to mount or
> unmount a CDROM, Zip, Jaz, or floppy disk under Windows.
>
> --
> Adam P. Jenkins
> [EMAIL PROTECTED]
FWIW, in my windows setup, if I am in the explorer and look at a cd,
change to another area, eject the cd and replace with another, then look
at the explorer again, the changes are not updated. yes, I have auto
insert notification off (cd-burner set-up). To see any changes, I have
to 'open' that drive in the explorer, and wait for it to mount it,
before siing any changes. If I then want to do any 'autoplay' or get the
icon to change, I have to hit the refresh key.
Windows basically seems to be polling the drive, unless you turn it off.
So, yes, under windows you do mount/unmount drives. It is just that you
don't realize it.
Bill
------------------------------
From: [EMAIL PROTECTED] (David Hinds)
Crossposted-To: comp.os.linux.hardware,comp.os.linux.misc,comp.os.linux.portable
Subject: Re: Raw writing to PCMCIA SRAM cards
Date: 25 Feb 1999 18:18:19 GMT
Mark Smith ([EMAIL PROTECTED]) wrote:
: Hello,
:
: With the PCMCIA drivers under Linux is it possible to write a binary to an
: SRAM card ?
Absolutely.
-- Dave Hinds
------------------------------
From: Chun-Chung Chen <[EMAIL PROTECTED]>
Subject: Re: System.map/vmlinuz
Date: Thu, 25 Feb 1999 10:03:16 -0800
On Thu, 25 Feb 1999, Aaron Faby wrote:
> I installed 2.2.2 and for some reason System.map, vmlinuz, and
> vmlinuz.old
> showed up in / as well as /boot. Anyone know why and how I can avoid the
>
> files being dropped in / ? Thanks.
edit /usr/src/linux/Makefile, find INSTALL_PATH and uncomment it.
Chung .
------------------------------
From: jimmy <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.development.apps,comp.os.linux.setup
Subject: Re: Why I'm dumping Linux, going back to Windblows
Date: Thu, 25 Feb 1999 14:48:42 -0500
Reply-To: [EMAIL PROTECTED]
Julian Robert Yon wrote:
> And of course, Office is the only "Productivity Suite" I have used which
> can make a simple spreadsheet, a couple of charts and a 300 word report
> take up 3.5M of disk space.
A while ago I worked on a one-page document--a timesheet. One page with
an empty table on it. Word ran out of memory on a 24Meg machine <g>.
------------------------------
From: "À̱¤ÁØ" <[EMAIL PROTECTED]>
Subject: help!! AF_PACKET socket program example ...
Date: Thu, 25 Feb 1999 12:04:58 +0900
Hi All,
Does anybody know how to write Client/Server code using AF_PACKET family
socket in the LINUX OS ??
Bye.
G.J. Lee.
------------------------------
From: [EMAIL PROTECTED] (Reality is a point of view)
Subject: Re: Java 2
Date: 25 Feb 1999 21:12:13 GMT
+---- [EMAIL PROTECTED] wrote (Thu, 25 Feb 1999 13:51:14 -0600):
| now that java 2 is open source, what does everyone think about more tightly
| integrating it into Linux for enterprise level app development?
+----
Port it and they will come. Maybe. Unless Eiffel, or a
Smalltalk variant, or, um, Python does first.
I believe Java2 is 'Community Source', though it may also pass
whatever 'Open Source' test is in current fashion . . .
--
Gary Johnson [EMAIL PROTECTED]
Privacy on the net is still illegal.
------------------------------
** 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
******************************
