Linux-Development-Sys Digest #261, Volume #6 Tue, 12 Jan 99 04:14:12 EST
Contents:
Re: Where is top source? (Peter Samuelson)
Re: 2.2.0pre4 dual boot with 2.0.36 question (mlw)
Re: lp0 on fire in 2.1.131 (bill davidsen)
Re: 2.2.0pre6 booting errors (Nathan Myers)
Re: Acessing binary file from the code without open(argv[0] ..) (John Reiser)
Re: Adopting COM? (Christopher B. Browne)
Re: silly question (Joe Pfeiffer)
Re: How to get Disk I/O rate (bps/sec) per disk ? (JiSook Kim)
Re: lp0 on fire in 2.1.131 (bill davidsen)
Re: Open Configuration Storage - was Registry for Linux (Leslie Mikesell)
Re: Open Configuration Storage - was Registry for Linux (George MacDonald)
Re: How to run Windows Applications on Linux (Keven R. Pittsinger)
Re: Rebuilding kernel help (Villy Kruse)
Re: Open Configuration Storage - was Registry for Linux (George MacDonald)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Peter Samuelson)
Subject: Re: Where is top source?
Date: 11 Jan 1999 22:55:33 -0600
Reply-To: Peter Samuelson <[EMAIL PROTECTED]>
[JiSook Kim <[EMAIL PROTECTED]>]
> I'd like to get top source.
> Where is top source?
For Linux? Look under `procps'.
--
Peter Samuelson
<sampo.creighton.edu!psamuels>
------------------------------
From: mlw <[EMAIL PROTECTED]>
Subject: Re: 2.2.0pre4 dual boot with 2.0.36 question
Date: Tue, 12 Jan 1999 04:47:17 +0000
Mark Jeacocke wrote:
>
> Hi I'd like to help out a bit test out 2.2.0pre4 on my RedHat 5.2
> system.
>
Peice of cake. I am using pre6, look at my mail headers.
RedHat 5.2 works with the 2.2 kernel.
--
Mohawk Software
Windows 95, Windows NT, UNIX, Linux. Applications, drivers, support.
Visit the Mohawk Software website: www.mohawksoft.com
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Subject: Re: lp0 on fire in 2.1.131
Date: 12 Jan 1999 00:11:37 GMT
In article <[EMAIL PROTECTED]>,
Peter Pointner <[EMAIL PROTECTED]> wrote:
| I'd guess the printer was connected to lp1. lp0 is on io port 0x3bc, I think
| that was the one on the ancient graphic cards. In any case you can check with
| cat /proc/ioports and man lp.
parport doesn't work that way. The first printer is lp0, like the first
IDE disk is hda. The io/irq set is no longer mapped to name, although
you can force it when you load the module.
BTW: with 2.2.0pre6 when I loaded the driver it told me the printer was
online, ready and out of paper. Next stuff, since it was connected to
another system at the time ;-)
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
"Too soon we grow old, and too late we grow smart" -Arthur Godfrey
------------------------------
From: [EMAIL PROTECTED] (Nathan Myers)
Subject: Re: 2.2.0pre6 booting errors
Date: 11 Jan 1999 15:17:48 -0800
Mumit Khan<[EMAIL PROTECTED]> wrote:
>Frank Hale <[EMAIL PROTECTED]> wrote:
>>Okay I have 2.2.0pre6 installed and when I boot it I get the following
>>errors
>>
>>depmod: error in loading shared libraries
>>: undefined symbol : __bzero
>>By the way I am on a RedHat 5.2 machine.
>
>Trying to get my dual-PII 450Mhz box running, I ran into the same problem.
> - install 5.2 + all the updates
> - get 2.2.0pre6 + Alan Cox patch and build from source
> - update the following from RawHide (rawhide.redhat.com)
> - glibc
> - modutils
> - util-linux
> - net-tools
>
>In that order. Upgrading glibc should fix your particular problem (not
>finding __bzero).
Strange, I installed vanilla 5.2 on my brother's Libretto (P75/16M)
and then 2.2.0-pre6, and everything (except parport modules) worked.
With static parport everything worked. Is the only difference the
compiler? I built the kernel, modules, and pcmcia stuff with
Egcs-1.1.1.
--
Nathan Myers
[EMAIL PROTECTED] http://www.cantrip.org/
------------------------------
From: [EMAIL PROTECTED] (John Reiser)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.misc
Subject: Re: Acessing binary file from the code without open(argv[0] ..)
Date: Tue, 12 Jan 1999 00:33:36 GMT
Here are some approaches which may be related to what you want to do.
Look at /usr/include/elf.h for the layout of an ELF executable file.
In particular, pay close attention to e_phoff, Elf32_Phdr.p_vaddr,
phdr.p_type == PT_LOAD, and AT_PHDR.
Consult the output of "objdump --all-headers a.out", particularly the
Program Header and Section Headers. [There should be a separate
option "objdump --program-header", but there isn't.]
Look at the output of "cat /proc/self/maps", or in general at
/proc/<pid>/maps which is a pseudo-file of text.
The initial stack has some goodies. The filename of the a.out is
stored with its null terminating byte at 0xbffffffc on x86 systems in
stock (unmodified) popular distributions. The stack itself has
pointers to arguments, then a 0, then pointers to environment
variables, then 0, then the AT_* table, then a 0, then the strings for
arguments and environment variables, ..., then the filename of the
a.out.
Example of looking at initial stack:
objdump --file-headers /bin/date
# look for "start address"
gdb /bin/date
b *0x8048b60 # use the actual start address
run
x/40x $esp
x/40x
x/40x
<etc.>
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Christopher B. Browne)
Crossposted-To: comp.os.linux.advocacy
Subject: Re: Adopting COM?
Date: 12 Jan 1999 05:27:33 GMT
Reply-To: [EMAIL PROTECTED]
On Mon, 11 Jan 1999 16:53:48 GMT, Brett W. McCoy <[EMAIL PROTECTED]>
posted:
>On 10 Jan 1999 21:01:42 GMT, Christopher B. Browne
><[EMAIL PROTECTED]> wrote:
>
>>On the other hand, it is likely that one or another of the ORBs will
>>become a "preferable option."
>>
>>Applications tend to get deployed for a specific ORB; the client and
>>server code *is* ORB-specific to at least some extent. As a result,
>>GNOME apps will almost certainly be ORBit-specific. They may request
>>services that are being provided via another ORB; nothing stops you
>>from installing multiple ORBs.
>
>This is true to some extent, but a client application needs not know the
>platform-specific implementations of a server, and vice versa (this is the
>idea behind CORBA, isn't it?).
That's certainly the intent of "They may request services being provided
via another ORB."
Client and server might use different ORBs, for sure.
>An application using ORBit should be able
>to make the interfaces available to clients (even ones running on Windows
>95, for that matter) that are using compliant ORBs. Can someone write a
>GNOME-compliant distributed application based soleley on IDL?
Presumably some non-IDL code is necessary at some point :-). More
seriously, it is not clear to what extent ORBit has been validated to
"grok" IIOP and its ability to talk to other ORBs. If it can't, that
certainly represents a bug...
--
Those who do not understand Unix are condemned to reinvent it, poorly.
-- Henry Spencer <http://www.hex.net/~cbbrowne/lsf.html>
[EMAIL PROTECTED] - "What have you contributed to Linux today?..."
------------------------------
From: Joe Pfeiffer <[EMAIL PROTECTED]>
Subject: Re: silly question
Date: 11 Jan 1999 23:07:27 -0700
mlw <[EMAIL PROTECTED]> writes:
> Josef Moellers wrote:
> > UN*X' answer to this question is: There's anough stuff given to build it
> > yourself, e.g.
> >
> > find . -name '*.cpp' -print | cpio -pdl ../anotherdir
> >
>
> See, this is my point, there is no command to do it. You can even use
> tar cv ... | tar xv *.cpp (as long as you are carefull about paths) My
> point, again, there is no program to do it.
If you want to make a sufficiently narrow point, you can always
succeed. If your question was, ``can it be done,'' then the answer
would have been yes. But you are limiting acceptable answers to
``able to do it in a single command'' rather than ``able to do it'' or
even ``able to do it in fewer than three commands.'' Trying to claim a
deficiency in Unix based on that seems pretty narrow... especially
when an alias can be defined to make the sequence above useable as a
single command.
--
Joseph J. Pfeiffer, Jr., Ph.D. Phone -- (505) 646-1605
Department of Computer Science FAX -- (505) 646-1002
New Mexico State University http://www.cs.nmsu.edu/~pfeiffer
------------------------------
From: JiSook Kim <[EMAIL PROTECTED]>
Subject: Re: How to get Disk I/O rate (bps/sec) per disk ?
Date: Tue, 12 Jan 1999 15:35:12 +0900
I'm writing a program that monitors disk I/O.
I'd like to measure the using disk I/O.
Richard Jones wrote:
> JiSook Kim <[EMAIL PROTECTED]> wrote:
> : hi!
>
> : How to get Disk I/O(bps/sec) per disk?
>
> You can't read this from the kernel (AFAIK ...)
> but you certainly can measure it approximately
> using the /sbin/hdparm tool (see the -t and -T
> options). If you want a more accurate benchmark
> you might also want to do an Altavista search
> for `bonnie'.
>
> : functions or data file...
> : or
> : I want more information for /proc/stat file structure.
>
> ???
>
> Rich.
>
> --
> - Richard Jones. Linux contractor London and SE areas. -
> - Very boring homepage at: http://www.annexia.demon.co.uk/ -
> - You are currently the 1,991,243,100th visitor to this signature. -
> - Original message content Copyright (C) 1998 Richard Jones. -
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Subject: Re: lp0 on fire in 2.1.131
Date: 12 Jan 1999 00:08:22 GMT
In article <[EMAIL PROTECTED]>,
Peter Pointner <[EMAIL PROTECTED]> wrote:
| Btw, talking about good error messages: I recently wanted to do a screen dump
| with Win95 on a HP DeskJet 6something. The printer did nothing, but 3 windows
| popped up: The first told me the printer is ready. The second told me there
| is a problem with the bidirectional communication. The third told me to
| restart windows and try again. I'm not sure if that is better than
| "lp0 on fire".
Early DEC C compiler on VMS:
"warn-w-warners$fatal: fatal error normal termination"
( when a C program terminated with exit(0) )
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
"Too soon we grow old, and too late we grow smart" -Arthur Godfrey
------------------------------
From: [EMAIL PROTECTED] (Leslie Mikesell)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.setup
Subject: Re: Open Configuration Storage - was Registry for Linux
Date: 12 Jan 1999 01:54:43 -0600
In article <[EMAIL PROTECTED]>,
George MacDonald <[EMAIL PROTECTED]> wrote:
>
>Well when you install your system the install process should know
>if you have an ethernet card or not. If so the install procedures
>should ask you if you are setting up for DHCP, if so then it
>should ask if you want to configure the config service to
>get config info from the network exclusively. If so it configures as
>you suggest.
>
>If there is a bad connection the timeout's on DNS are
>a problem, the solution seems obvious.
I think you are missing the point if you are still thinking about
an 'install process' and answering questions per machine. You
could just copy the same image onto hundreds of machines, plug them
into a network and come up running with everything necessary coming
from the DHCP server plus the initial config setup download. If
you move the machine to a different network, it should reconfigure
itself.
Les Mikesell
[EMAIL PROTECTED]
------------------------------
From: George MacDonald <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.setup
Subject: Re: Open Configuration Storage - was Registry for Linux
Date: Tue, 12 Jan 1999 07:33:39 GMT
Leslie Mikesell wrote:
>
> In article <[EMAIL PROTECTED]>,
> George MacDonald <[EMAIL PROTECTED]> wrote:
>
> >> I wouldn't question your capabilities. You are just attempting something
> >> that needs transactioning and it can't be done right unless the
> >> back end supports it - or you are willing to accept occasional errors.
> >> Perhaps you should start with postgresql (which supports transactions
> >> and has an open license) as the preferred network store. Even then
> >> you will have problems if you allow settings stored at different
> >> levels in the hierarchy to be used if they come from different databases
> >> or files.
> >
> >There are a couple of different ways to go about this. The first is to
> >simply say the underlying mechanisms limitations flow through. i.e.
> >if it's possible to corrupt a LDAP directory via multiple updates
> >simultaneously from different locations on the same shared data then
> >that is a limitations of that storage/access "module". The same
> >can be said for flat files accessed by root, it is possible for
> >two root users to be modifying a set of files and end up corrupting
> >them instead. Yet it is a practical solution in many situations.
>
> That's not the situation I had in mind, although it poses a problem
> as well. What I meant was that if your API allows get/set of
> individual values as separate operations and two values affect
> each other, it will be possible for a single person doing a
> single update to cause simultaneous get operations to see a
> mix of old and new values. This generally doesn't happen
> with root editing files because you normally have to do
> something to tell the running daemons to re-read the file when
> you are done.
It depends somewhat on the implementation. If you are talking about
updating one of the current system config files then the simplest way
to do it would be to copy the file,
make the changes on the copy, then when the program issues the
"end of transaction", mv the old file to a back up and then mv the newfile
into place. For that particular file a special end of transaction method
may be configured(i.e. to signal the daemon). This opens a small window in
which the file is not there, in which case programs should do a retry on open,
and at boot time a script should be run to see if any "transactions" were in
the middle of the two steps. In which case just complete the transaction.
This is one kind of simplistic way to do it. Another way is to just keep
the new changes in a journal and apply them to the base. Another is
to do a version control system like RCS, in which multiple branches(
simultaneous updates) are allowed. In otherwords keep an authoritative
copy seperate from the actual system config file. Checks could be
put in place during a "start transaction" to validate the system
file is what the authoratative source thinks it should be. If not then
the system file could be assumed to be correct and stored automatically
or the "start transaction" could trigger an error.
I found an interesting article that is somewhat related to this topic,
you might find it interesting. See
http://www.zdnet.com/pcweek/stories/news/0,4153,371690,00.html
I suppose whether or not to do transactional mechanisms or do
single call change mechanism would be defined by the module used
for that "object", but this could be overriden and configured differently
if desired.
>
> >We also need to build a module that uses a new flat file format which
> >will handle the transactional mechanism. This can be done using
> >file locking or Copy on write or versioning based repositories. There
> >are ways to do this that are not as efficient as a DBMS and probably
> >not as robust either but they would be practical on small systems.
>
> If you write a tmp file on the same filesystem, rename() is an
> atomic operation on most unix filesystems.
Yes, it provides the basic mechanism. Much slower than a DB, but
it works. DB's come into their own when you have lots of data,
lots of queries and lots of simultaneous updates. In many
situations a good DB is worth it's weight in gold. You
wont see me complaining about a DB module, to me designing
the service to be able to use one is a prime requirement.
Do you agree that small systems with infrequent config changes
most likely would not use a DB module? I'm thinking small
systems that have memory/diskspace or performance limitations
would most likely want to be file based.
--
We stand on the shoulders of those giants who coded before.
Build a good layer, stand strong, and prepare for the next wave.
Guide those who come after you, give them your shoulder, lend them your code.
Code well and live! - [EMAIL PROTECTED] (7th Coding Battalion)
------------------------------
From: [EMAIL PROTECTED] (Keven R. Pittsinger)
Subject: Re: How to run Windows Applications on Linux
Crossposted-To:
comp.os.linux.advocacy,comp.os.linux.development,comp.os.linux.development.apps
Date: 12 Jan 1999 00:28:47 -0500
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Robin V. Stacey) writes:
>> Santa's making a list. If You could have any piece of software ported
>> to Linux, other than Microsoft's what would it be?
>>
>
> Borland Delphi
An X-compliant version of Paradox.
Keven
------------------------------
From: [EMAIL PROTECTED] (Villy Kruse)
Subject: Re: Rebuilding kernel help
Date: 12 Jan 1999 09:22:01 +0100
In article <Rspm2.51$[EMAIL PROTECTED]>,
Tim Underwood <[EMAIL PROTECTED]> wrote:
>
>So, is there an easier way to get dial-on-demand, or can someone help me to
>get the kernel recompiled with ppp 2.3.5?
>
Until redhat officially starts supporting pppd with dial-on-demand,
I would suggest using diald. It should be available on contrib.redhat.com
The ppp module that comes with pppd and the corresponding module that
comes with the kernel has diverged slightly. That is, some patches has
been applied in the kernel sources which are not in the pppd version and
vice versa. It is not trivial to sort this out properly; maybe this is
not even important.
Villy
------------------------------
From: George MacDonald <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.setup
Subject: Re: Open Configuration Storage - was Registry for Linux
Date: Tue, 12 Jan 1999 08:16:53 GMT
Leslie Mikesell wrote:
>
> In article <[EMAIL PROTECTED]>,
> George MacDonald <[EMAIL PROTECTED]> wrote:
> >
> >Well when you install your system the install process should know
> >if you have an ethernet card or not. If so the install procedures
> >should ask you if you are setting up for DHCP, if so then it
> >should ask if you want to configure the config service to
> >get config info from the network exclusively. If so it configures as
> >you suggest.
> >
> >If there is a bad connection the timeout's on DNS are
> >a problem, the solution seems obvious.
>
> I think you are missing the point if you are still thinking about
> an 'install process' and answering questions per machine. You
> could just copy the same image onto hundreds of machines, plug them
> into a network and come up running with everything necessary coming
> from the DHCP server plus the initial config setup download. If
> you move the machine to a different network, it should reconfigure
> itself.
Well installing images on disks as an install process would make
many assumptions. What about machines that have no ethernet cards?
Stand alone systems are perfectly viable.
As far as making things reconfigurable from the network, that
should be configured. i.e. if that's what you want, then define
your image that way, then it should do just as you suggest.
But not all systems will fall into that modality. Lets say I
am a consultant and bring my laptop into a site, perhaps
to do a demo. I may not need to reconfigure or may not want to.
Of course the client may not want to let me use there services
if I don't, but that's a decision that should not be forced
on anybody. The config system should support both models.
Even assuming DHCP is available is a bit risky, no? For
small networks it's a lot of hastle to setup and
it's not really needed! So why burden a network needlessly?
Don't get me wrong, I think plug and go is a great idea,
and a dynamic config service would help a great deal. I
expect this will become a much bigger topic in the next
few years. Tablet's with short haul high speed radio
links are most likely to become popular in the next
few years. Being able to walk into the conference room
and have the meeting adgenda popup on your netPad
would be a rather nice feature. All we gotta do is
get the pad to dynamically detect the new venue,
perhaps getting a new address, then do the reconfig,
which activates the meeting adgenda object ...
Hmm, this is so cool, it makes absolute zero look warm
by comparison!
--
We stand on the shoulders of those giants who coded before.
Build a good layer, stand strong, and prepare for the next wave.
Guide those who come after you, give them your shoulder, lend them your code.
Code well and live! - [EMAIL PROTECTED] (7th Coding Battalion)
------------------------------
** 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
******************************
