Linux-Development-Sys Digest #768, Volume #6 Wed, 2 Jun 99 19:14:17 EDT
Contents:
Re: Might Linux SMP write memory out of order? (Joe Pfeiffer)
firewall transparent proxy using ipfwadm ("Cliff")
Re: .h files missing for Make Config in Redhat (Alex Rhomberg)
Re: advance power management (Stewart Perry)
Re: Might Linux SMP write memory out of order? ("Stefan Monnier "
<[EMAIL PROTECTED]>)
MTRR problem on cyrix in 2.2.9 ("Clifford Smith")
Re: TAO: the ultimate OS ("G. Sumner Hayes")
Re: What are the differences between mySQL and mSQL? (Frank)
Re: What are the differences between mySQL and mSQL? (Don Baccus)
Re: Linux Device Driver Overflow? ("J�rgen Hermanrud Fjeld")
mkfs for UFS? (Paul Gray)
Re: RAID-1 and 5 broken in 2.2.9? (Az0th)
Re: lifting limit of four md devices ind 2.2.x kernels? (bill davidsen)
Re: What are the differences between mySQL and mSQL? (Philip Brown)
Re: What are the differences between mySQL and mSQL? (Paul D. Smith)
----------------------------------------------------------------------------
From: Joe Pfeiffer <[EMAIL PROTECTED]>
Subject: Re: Might Linux SMP write memory out of order?
Date: 02 Jun 1999 12:44:44 -0600
"JDonner" <[EMAIL PROTECTED]> writes:
> I'm wanting to use simple flags to coordinate between threads. I'm using
> well known algorithms, but these rely on writes being completed in the order
> they're executed. I've been told that with some multi-processor systems
> writes get written out of order, ie a thread may write memory location A
> then
> B, but the system ends up actually writing B before A. Does Linux SMP do
> this?
That's an architecture question, not an OS question. I don't know
whether multiprocessor PCs maintain sequential consistency, but I
would expect it.
--
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: "Cliff" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking,linux.redhat.development
Subject: firewall transparent proxy using ipfwadm
Date: Wed, 02 Jun 1999 18:48:37 GMT
The ipfwadm program has the ability to redirect incoming traffic to a
socket on the firewall, even if the traffic is not for the firewall.
Redirection works fine. How do you determine what the original destination
was? Example; 3 machines - client (inside the firewall), firewall, and
distant.host. Client attempts to access a web page on distant.host,
firewall redirects the request to a socket on the firewall. How does the
program accept()'ing on that firewall socket determine that the request is
_to_ distant.host? I can get the client with a getpeername() call. But
what about the destination? I'm using RedHat 5, kernel 2.0.36. Thanks.
--
-Cliff
Views expressed are my own and not necessarily those of my employer
Concordia Net, Inc. When replying via email please use; cwheat at concordia
dot net not
root@localhost
------------------------------
Date: Wed, 02 Jun 1999 13:29:06 +0200
From: Alex Rhomberg <[EMAIL PROTECTED]>
Subject: Re: .h files missing for Make Config in Redhat
Jonathan DelStrother wrote:
>
> hiya
>
> First things first - there's probably a post very similar to this about
> to pop up on the server - in a frenzy of Ctrl-V pressing down below I
> must of pressed Ctrl-Enter, which Netscape kindly assumes means that you
> want to post your message...Oooops
Netscape allows you to cancel a post. Hit Meta-D (Alt-D) or select
"Cancel Message" int he Edit Menu
> Anyways
> I've been trying to persuade sound to work in Redhat 5.2 - go to the
> src/linux/drivers/sound dir, type make config...& i get the following
> error:
> configure.c: 19: stdio.h: No such file or directory
> configure.c: 20: unistd.h: No such file or directory
> configure.c: 21: stdlib.h: No such file or directory
> configure.c: 22: fcntl.h: No such file or directory
> configure.c: 23: string.h: No such file or directory
>
> Anyone know where I can find those files? I've looked all over the
> packages in the CD, but haven't found anything...
glibc-devel is the one you need. Try
rpm -Uvh /wherever/your/packages/are/glibc-devel*.rpm
-
_________________________________________________________________
/ \
| Alex Rhomberg Tel: +41 1 632 49 18 |
| Institut fuer Elektronik Fax: +41 1 632 12 10 |
| ETH Zuerich Zentrale: +41 1 632 11 11 |
| Gloriastrasse 35 |
| 8092 Zuerich email: [EMAIL PROTECTED] |
\_________________________________________________________________/
------------------------------
From: [EMAIL PROTECTED] (Stewart Perry)
Subject: Re: advance power management
Date: 2 Jun 1999 13:45:28 +0100
Marius ([EMAIL PROTECTED]) wrote:
: You should check the system shutdown scripts (/etc/rc.d/rc.0 or
: something like that -- for Red Hat it is /etc/rc.d/init.d/halt, IIRC).
: Add the `-p' option to the shutdown command in the script (or,
: if the script calls `halt' command, use `poweroff' instead). This
: should work (it does for me).
: I have seen a very nice explaination of this `problem' (or should I say
: `new feature'?), unfortunately, I do not recall the source now.
Yes, the key is that the shutdown command behaves differently
if the system is in runlevel 0 or 6 - see the manual page.
Unfortunately, poweroff still doesn't seem to work on mine
and other bioses - apparently nothing to do with user-level
processes. The apm_bios_call_simple() function in apm.c
crashes with a segv when it calls the bios interface. I will
post the panic text when I get home if anyone is interested.
I downloaded the apm bios 1.2 spec from intel, and as far as
I can tell, linux follows it to the letter. However,
power-off works fine under another popular "operating
system". I suspect that this system calls the bios function
with some different register values - and as the bios works
with this system the manufacturer is satisfied that it is apm
1.2 compliant. Hmm. I guess the answer would be to re-flash
the bios, if I could work out who manufactured the board...
Cheers,
Stewart.
--
High Performance Systems Group Tel: +44 (0)1203 522485
Department of Computer Science Fax: +44 (0)1203 525714
University of Warwick
CV4 7AL (UK) www.dcs.warwick.ac.uk/~scp
------------------------------
From: "Stefan Monnier <[EMAIL PROTECTED]>"
<[EMAIL PROTECTED]>
Subject: Re: Might Linux SMP write memory out of order?
Date: 02 Jun 1999 15:58:42 -0400
>>>>> "Joe" == Joe Pfeiffer <[EMAIL PROTECTED]> writes:
> "JDonner" <[EMAIL PROTECTED]> writes:
>> I'm wanting to use simple flags to coordinate between threads. I'm using
>> well known algorithms, but these rely on writes being completed in the order
>> they're executed. I've been told that with some multi-processor systems
>> writes get written out of order, ie a thread may write memory location A
>> then
>> B, but the system ends up actually writing B before A. Does Linux SMP do
>> this?
> That's an architecture question, not an OS question. I don't know
> whether multiprocessor PCs maintain sequential consistency, but I
> would expect it.
Indeed, I seem to remember that the x86 architecture is supposed to
guarantee some reasonable kind of consistency. But beware the compiler:
many compilers will happily reorder memory accesses unless specifically asked
not to (with something like C's "volatile" or via synchronization primitives).
Stefan
------------------------------
From: "Clifford Smith" <[EMAIL PROTECTED]>
Subject: MTRR problem on cyrix in 2.2.9
Date: 2 Jun 1999 20:51:49 GMT
Hi. I know cyrix processors use ARRs which emmulate MTRRs. However, one
of my MTRRs seem to be missing.... here's a cat of /proc/mtrr:
r46h31:~$ cat /proc/mtrr
reg00: base=0x000a0000 ( 0MB), size= 128kB: write-combining, count=1
reg01: base=0x000c0000 ( 0MB), size= 256kB: uncachable, count=1
reg03: base=0x000a8000 ( 0MB), size= 32kB: write-combining, count=1
reg07: base=0x00000000 ( 0MB), size= 32MB: write-back, count=1
reg06, which is used to setup some kind of a video buffer(?) is not
present. My processor is a cyrix 6x86MX PR200+. If you have any ideas,
please email me as well as following up. Or don't email me at all i don't
care =] I'm willing to provide as much information as i can to anyone
interested....
--Clifford Smith
------------------------------
From: "G. Sumner Hayes" <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: TAO: the ultimate OS
Date: Wed, 02 Jun 1999 16:13:22 -0400
[Followups to comp.os.misc since this is off-topic on most of these
groups]
"Vladimir Z. Nuri" wrote:
[Lots snipped]
You may want to read Gelernter's lifestream papers. Some of his
ideas seem applicable. I don't know if they're online;
a search for "lifestreams" "gelernter" and "mirrorworlds" is
probably in order. Being familiar with OpenDoc is probably also
a good idea. I presume that you know all about CORBA and COM (the
theory, not the practice).
None of the things that I mention are identical with what you
have, but they're some of the approaches people have tried to
get results similar to what you propose. None of them have
succeeded 100%. May you have better luck...
--Sumner
------------------------------
From: Frank <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.development.apps,comp.os.linux.misc,comp.os.linux.setup,comp.lang.java.databases
Subject: Re: What are the differences between mySQL and mSQL?
Date: Wed, 02 Jun 1999 15:29:41 +0200
Ruiming Chen wrote:
> The Subject askes its all. Are they the same free database software with
> two names?
> Or they are two different free database software?
> Are they both run on Linux?
>
> Thank you!
> --Raymond
> --
> RC Square Team.
This are two different products:
mSQL: www.hughes.au
mySQL: www.tcx.se
They are pretty similar and it is hard to say, which one is better.
Have a look at there websites, they can tell you why they are better.
They both are available on lots of unix, as well as linux.
If you intend to use java ore html-stuff, they are both well equiped.
I tried both of them and brought mySQL easier to run with Java using the
twz1-Driver, then with mSQL . But this is a question of the
JDBC-Driver.....
Both are great and usefull tools.
Fredy
------------------------------
Crossposted-To:
comp.os.linux.development.apps,comp.os.linux.misc,comp.os.linux.setup,comp.lang.java.databases
Subject: Re: What are the differences between mySQL and mSQL?
From: [EMAIL PROTECTED] (Don Baccus)
Date: 2 Jun 1999 14:16:08 PST
In article <nNd53.584$[EMAIL PROTECTED]>,
bryan <[EMAIL PROTECTED]> wrote:
>it has this nasty bug about the last column being a string - I forget the details but
>I ended up going with mysql
>since it has worked for me bug-free for over 2 yrs in a production env.
>but it lacks bells and such. no stored procs and no triggers.
And MySQL isn't transaction-based, which some would claim isn't
simply a lack of a bell or whistle...
Postgres is getting better, I've been using the 6.5 beta and it's
a VAST improvement over 6.4.2 (which I removed from my system
a few hours after installation). The "real" 6.5 will be released
next Monday, apparently.
The core development group seems committed to improving reliability and
functionality. And they have a good handle on the code. They took
it over about two years ago, IIRC.
--
- Don Baccus, Portland OR <[EMAIL PROTECTED]>
Nature photos, on-line guides, at http://donb.photo.net
------------------------------
From: "J�rgen Hermanrud Fjeld" <[EMAIL PROTECTED]>
Subject: Re: Linux Device Driver Overflow?
Date: Thu, 03 Jun 1999 00:14:37 +0200
Reply-To: [EMAIL PROTECTED]
Joe Pfeiffer wrote:
> "J�rgen Hermanrud Fjeld" <[EMAIL PROTECTED]> writes:
> > Could a more microkernel like path of development open for more and
> > easier device drivers for Linux?
>
> I'm of the opinion that loadable modules meet many of the advantages
> originally ascribed to microkernel OSs -- particularly when it turns
> out that Minix (for example) puts all the drivers in the kernel space
> and don't provide the protection against driver bugs causing OS
> crashes that were originally claimed.
After reading some material it is my impression that Minix has a rather
flawfull implementation of the microkernel design.
How does QNX compare with Linux?
Doesn't QNX show that microkernel has potential, when implemented in a good
manner?
Does QNX implements drivers in user space, thus giving the protection that
the microkernel design was thought to have?
How about the the user space versus kernel space debate?
What advantages do you see with either architecture?
--
Vennlig hilsen
J�rgen Hermanrud Fjeld
[EMAIL PROTECTED]
------------------------------
From: Paul Gray <[EMAIL PROTECTED]>
Subject: mkfs for UFS?
Date: 2 Jun 1999 13:27:10 GMT
I was wondering if there was an experimental version of
mkfs for UFS in the works.
ADVthanksANCE
-Paul
------------------------------
From: [EMAIL PROTECTED] (Az0th)
Subject: Re: RAID-1 and 5 broken in 2.2.9?
Date: Wed, 2 Jun 1999 21:52:15 GMT
Hi Bill,
: I guess the answer is that the RAID-[145] stuff in the 2.2.9 kernel
: doesn't work at all with anything any more (why is it there?). The md
: stuff got broken somewhere along the line, and the raidtools stuff only
: works after adding about 200k worth of patches (not a typo), and those
: are alpha and only available for a few kernels, 2.0.36 and 2.2.6.
: Shame to see stuff which worked broken.
The good news is: what works, works better now than it ever did before,
if you can stand to use a 2.2.6/7 kernel (efficient, threaded resync is
a wonder to behold.)
The current development code is very stable, by all accounts, at least for
RAID-[01], and is only called alpha because there remains some functionality
to be added in support of RAID-[45]. 2.2.8 was a disaster that 2.2.9 mostly
corrected, but at present, not much is gained by going past 2.2.7. I do run
2.2.9 on an SMP box, without RAID, and will probably be falling back to play
with Mosix, which is only available as patches to 2.2.7. No big deal.
RF B
--
"There's certainly a strong case for people really disliking Microsoft,
and most of that is because their operating systems really suck."
Linus Torvalds
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Subject: Re: lifting limit of four md devices ind 2.2.x kernels?
Date: 2 Jun 1999 22:13:42 GMT
In article <[EMAIL PROTECTED]>,
Marc Mutz <[EMAIL PROTECTED]> wrote:
| Is it a trivial task to do the above (e.g. changing constant definition
| in source file) or is it more involved?
Number of RAID devices and number of partitions in a RAID group are in a
header file (maybe md.h) which I don't have handy.
See the threads on RAID and 2.2.9, they contain some generally useful
information about where RAID is going in Linux.
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
The Internet is not the fountain of youth, but some days it feels like
the fountain of immaturity.
------------------------------
From: [EMAIL PROTECTED] (Philip Brown)
Crossposted-To:
comp.os.linux.development.apps,comp.os.linux.misc,comp.os.linux.setup,comp.lang.java.databases
Subject: Re: What are the differences between mySQL and mSQL?
Reply-To: [EMAIL PROTECTED]
Date: 02 Jun 1999 22:07:12 GMT
On 02 Jun 1999 10:56:17 -0400, [EMAIL PROTECTED] wrote:
>%% Frank <[EMAIL PROTECTED]> writes:
>
> f> Ruiming Chen wrote:
>
> >> The Subject askes its all. Are they the same free database software
> >> with two names? Or they are two different free database software?
> >> Are they both run on Linux?
>
>They are two different databases. Neither is 100% free. mSQL is
>_significantly_ less than 100% free. MySQL is free for the large
>majority of uses. They both run on Linux (and most other UNIX
>platforms, as well as Windows).
of course, if you have the disk space, sybase for linux is 100% free, last
time I checked.
--
[Trim the no-bots from my address to reply to me by email!]
[ Do NOT email-CC me on posts. Pick one or the other.]
--------------------------------------------------
The word of the day is sescaquintillion
------------------------------
From: [EMAIL PROTECTED] (Paul D. Smith)
Crossposted-To:
comp.os.linux.development.apps,comp.os.linux.misc,comp.os.linux.setup,comp.lang.java.databases
Subject: Re: What are the differences between mySQL and mSQL?
Date: 02 Jun 1999 10:56:17 -0400
Reply-To: [EMAIL PROTECTED]
%% Frank <[EMAIL PROTECTED]> writes:
f> Ruiming Chen wrote:
>> The Subject askes its all. Are they the same free database software
>> with two names? Or they are two different free database software?
>> Are they both run on Linux?
They are two different databases. Neither is 100% free. mSQL is
_significantly_ less than 100% free. MySQL is free for the large
majority of uses. They both run on Linux (and most other UNIX
platforms, as well as Windows).
f> This are two different products:
f> mSQL: www.hughes.au
f> mySQL: www.tcx.se
f> They are pretty similar and it is hard to say, which one is better.
f> Have a look at there websites, they can tell you why they are
f> better.
It's not that hard to say, really. MySQL is without question better
(obviously in some sense this depends on what you need, but certainly
MySQL meets a much wider spectrum of needs than mSQL).
MySQL is a much more complete implementation of SQL, it's easier to
manage, has much better support, even if you don't buy a support
contract, has a much better license (see below), and is faster than
mSQL, often significantly [*].
mSQL isn't free at all. Although you do get the source code and build
mSQL yourself, the licensing is very strict: unless you are a non-profit
organization or an educational institution, you _must_ buy a license
from Hughes to use mSQL in any capacity.
http://support.Hughes.com.au/cgi-bin/hughes?solution&11-970715-0000&130-868988450&14-0&15-3&25-&3-&30-
MySQL, while not free in some senses of the word, is _almost_ free (on
non-Windows platforms, anyway). The client is GPL'd. The server
license allows you to do anything you want with the server and use it
any way you like without purchasing a license, _except_ you can't resell
the server in a product. For example, you are allowed to create
commercial web sites using MySQL, etc., and you don't need a license for
that. You just can't build a new product around the server and sell
that, without a license.
There's a comprehensive comparison of both features and performance on
the MySQL web site. They have a nice regression and performance test
suite which is free, so you can easily test for yourself if you like.
[*] The mSQL folks have accused the performance statistics in the test
suite as being skewed in MySQL's favor, but I've not seen any
alternate tests or comprehensive benchmarks--I have hoped they'd add
tests to MySQL's crash-me suite to illustrate this.
--
===============================================================================
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.
------------------------------
** 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
******************************
