Linux-Development-Sys Digest #776, Volume #6 Fri, 4 Jun 99 07:14:27 EDT
Contents:
Re: kernel vs egcs vs PentiumPro/II (Matt Bartley)
Re: Terabite Plus Filesystems (Rick Jones)
Re: Strange partial write on sockets problem (Dave Platt)
Re: TAO: the ultimate OS (Vladimir Z. Nuri)
Re: What are the differences between mySQL and mSQL? ("Jon Smirl")
Re: How to get all PID ? (Arun Sharma)
Re: What are the differences between mySQL and mSQL? (Paul D. Smith)
How to determine tick frequency at runtime? (Mark Frazer)
Re: the ultimate OS (Vladimir Z. Nuri)
Re: What are the differences between mySQL and mSQL? (Don Baccus)
Re: TAO: the ultimate OS (Mario Klebsch)
Re: What are the differences between mySQL and mSQL? ("Jon Smirl")
Re: What are the differences between mySQL and mSQL? (Don Baccus)
Re: bootinfo.h missing for make dep (Andreas Schwab)
Re: Redhat 6 & NFS (Chris Raper)
Re: Terabite Plus Filesystems (Phillip Fayers)
How to serialize read operations in char driver? (Grant Edwards)
Re: TAO: the ultimate OS (James Lee)
Re: TAO: the ultimate OS (Larry Blanchard)
Re: Might Linux SMP write memory out of order? (David Wragg)
Re: Terabite Plus Filesystems (Tor Arntsen)
Re: A simple question... (Igor Zlatkovic)
Help - MAX size of tarfile ?? Is a tarfile a TAR or a FILE?? (Dave Bynion)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Matt Bartley)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: kernel vs egcs vs PentiumPro/II
Date: 3 Jun 1999 09:39:37 -0700
In article <[EMAIL PROTECTED]>,
Paul Kimoto <[EMAIL PROTECTED]> wrote:
>In article <7j55l2$6g9$[EMAIL PROTECTED]>, Matt Bartley wrote:
>> Some time ago I read about a command line option to gcc and/or egcs,
>> similar to "-v" but more verbose, which among other things shows all
>> the the "-f<optimisation>" flags which are in effect at the current
>> "-O" level.
>This (undocumented) option is "-Q -v".
Thanks. Added to a random notes file so hopefully I won't forget again.
>(I have read [maybe on the egcs mailing list?] that the "-O" options
>turn on some optimizations beyond those specifiable by "-f" options.)
I tried compiling a hello.c program with egcs, and got a weird combination
of options:
gcc -v -Q -o hello hello.c
[...]
options passed:
options enabled: -fpeephole -ffunction-cse -fkeep-static-consts
-fpcc-struct-return -fcommon -fgnu-linker -fargument-alias -m80387
-mhard-float -mno-soft-float -mieee-fp -mfp-ret-in-387 -mschedule-prologue
-mcpu=i486 -march=pentium
Aren't -mcpu=i486 and march=pentium mutually exclusive? This is on a
Pentium MMX system with egcs-1.1.2.
--
"When PCs run new applications successfully, most people feel relief
and almost pathetic gratitude - a standard of reliability tolerated in
no other consumer product."
_Economist_, Sept. 12 1998
------------------------------
From: [EMAIL PROTECTED] (Rick Jones)
Crossposted-To:
comp.os.ms-windows.nt.admin.misc,comp.sys.sun.admin,comp.sys.hp.misc,comp.os.linux.hardware
Subject: Re: Terabite Plus Filesystems
Date: 3 Jun 1999 17:57:53 GMT
Just a couple of pseudo-random observations:
*) one does not require a 64 bit CPU, nor a 64 bit OS to have "64 bit
or terabyte filesystems." it might however make it easier/more
convenient.
*) if the application(s) can be broken-up into discrete tasks with little
or no interaction among the tasks, one can basically take N systems
running just about any OS and joined with just about anything
including a length of wet string and create a "system." you will
then also be able to report how it is M times faster than a Cray
and how it is "one of the world's largest supertcomputers."
there are a number of scientific/numerical simualtions where this
is possible, one example currently expending its 15 minutes is the
SETI@home work
"scaling" can be a very, um, elastic, term. especially in the hands
of the media and marketing make sure to ask for and read the
fine-print...
rick jones
Someone creating "the world's largest super computer" is much more
likely to get press than "a bunch of folks were able to string a
zillion systems together to run a really parallelizable application."
--
these opinions are mine, all mine; HP might not want them anyway... :)
feel free to email, or post, but please do not do both...
my email address is raj in the cup.hp.com domain...
------------------------------
From: [EMAIL PROTECTED] (Dave Platt)
Subject: Re: Strange partial write on sockets problem
Date: 3 Jun 1999 17:11:49 GMT
>The system we get is that the tail of the print job is sometimes lost
>(not for every job, but if a particular job fails, it always fails)..
>tracing the code, it appears that the write of the data to the socket
>returns the correct number of bytes written, but only the 1st so many
>bytes appear to arrive at the remote socket...!?
>
>Example, write writes 1338 bytes, remote socket receives only 684, and
>this is consistent... We can trace on the terminal server how many bytes
>it sends down the serial wire (and thus how many it has received on the
>socket)
I've seen occasional print failures on Red Hat Linux, when printing
PostScript from the "Executor" Mac emulator. My PostScript printer
would report an error - illegal command, or something like that. When
I looked at the affected file in the print queue, I found a bunch of
data missing out of the middle of the file.
According to Cliff at ARDI, Executor is simply writing into a pipe,
directed at a child process running /bin/lpr. It's as if data is
being lost as it's written through the pipe.
I've seen the problem with both the 2.0 and 2.2 kernel series.
I've upgraded the affected machine recently, from the old BSD version
of the lpr/lpd suite to the current version of LPRng. Haven't run
enough tests yet to know whether the problem went away (which would
suggest that it might be the old lpr program) or whether it's still
present (which would tend to point a finger at the kernel's
socket/pipe code).
--
Dave Platt [EMAIL PROTECTED]
Visit the Jade Warrior home page: http://www.radagast.org/jade-warrior/
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
------------------------------
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
From: [EMAIL PROTECTED] (Vladimir Z. Nuri)
Subject: Re: TAO: the ultimate OS
Date: Thu, 3 Jun 1999 17:28:30 GMT
Alexander Viro ([EMAIL PROTECTED]) wrote:
: [Tons of marketspeak]
"design document"
: Where is your code? Talk is cheap
I agree, so are tiresome and predictable objections all
along the same lines.
: - unless you can demonstrate a code your
: words are worth nothing.
"call me when you have a complete system and I will
take it from you for free, and possibly not call you
a bozo anymore"
: Sorry, but you sound like a cross between manager and
: salesweasel and those animals are, erm, not too good in producing things.
: Write something that would work and demonstrate it.
I produced an essay far longer than your predictable flame.
write something that challenges the essay rather than
my manhood or hacker-esque geek quotient.
--
~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^
"in theory, there's no difference [EMAIL PROTECTED]
between theory and practice, mad genius research lab
but in practice there is!" http://www8.pair.com/mnajtiv/
------------------------------
From: "Jon Smirl" <[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: Thu, 3 Jun 1999 12:34:28 -0400
Don Baccus <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Meanwhile, MySQL seems like a good choice if you don't
> need the transaction model.
> --
You would have to be insane to implement any significant database app
without transaction support. I spent weeks trying to recover data out of an
older non-transaction based system when the system failed. Transactions are
critical to ensuring the integrity of your data.
Apache also causes databases to be heavily multi-user even if they are
running and the same machine with Apache. Each of Apache's multiple
processes (there may be hundreds) looks like a separate user to the database
engine. Transactions are a necessity for sorting this out.
--
Jon Smirl
[EMAIL PROTECTED]
------------------------------
Subject: Re: How to get all PID ?
From: Arun Sharma <[EMAIL PROTECTED]>
Date: Fri, 04 Jun 1999 05:35:13 GMT
"Soohyung Lee" <[EMAIL PROTECTED]> writes:
> How can I get all Process ID running at any instance.
> Not by script (ps aux), But in the C program .
> I really need your help.
Oops, I didn't read your post properly. One possible way of doing it
is parsing /proc. This is what ps aux seems to do.
-Arun
------------------------------
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: 03 Jun 1999 13:50:47 -0400
Reply-To: [EMAIL PROTECTED]
%% [EMAIL PROTECTED] (Don Baccus) writes:
db> Paul D. Smith <[EMAIL PROTECTED]> wrote:
>> I'm talking about "free speech", not "free beer". MySQL is much more
>> free than Sybase, since it comes with complete source code, permission
>> to modify it, etc.
db> On the other hand, Sybase is a proven enterprise-class database
db> engine...
Your comment is completely orthogonal to mine, as far as I can tell :).
--
===============================================================================
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.
------------------------------
From: Mark Frazer <[EMAIL PROTECTED]>
Subject: How to determine tick frequency at runtime?
Date: Thu, 03 Jun 1999 17:55:09 +0000
Can I determine the value of tick as found in
kernel/sched.c at runtime?
(Yes, I'm using third party drivers they're not
willing to give out the source for. Ug.)
------------------------------
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
From: [EMAIL PROTECTED] (Vladimir Z. Nuri)
Subject: Re: the ultimate OS
Date: Thu, 3 Jun 1999 17:23:24 GMT
Tim Doffing ([EMAIL PROTECTED]) wrote:
: Blah, blah, blah. I intend on inventing cold fusion, and a pill that will
: extend your life indefinitely. No Code=No OS. We all have dreams!
the tiresome hacker position/canard/harangue that without code, you
have nothing. "design is bull****" .. which btw I have some
sympathy for.
excuse me, but I am both a hacker & a designer. and sometimes
design logically comes before hacking. want to overtake win95?
you have to have a vision. you can't just keep writing code
blindly.
I agree, writing code is hard work, and writing design is less
difficult. but there are many hackers with no vision. arguably
linux has very little vision. what is it, a copy of a
OS that existed two decades ago? you want "world domination"?
you think that you will get there by endlessly tweaking the
performance of the kernel?
--
~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^~^
"in theory, there's no difference [EMAIL PROTECTED]
between theory and practice, mad genius research lab
but in practice there is!" http://www8.pair.com/mnajtiv/
------------------------------
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: 3 Jun 1999 11:38:23 PST
In article <[EMAIL PROTECTED]>,
Paul D. Smith <[EMAIL PROTECTED]> wrote:
>%% [EMAIL PROTECTED] (Don Baccus) writes:
>
> db> Paul D. Smith <[EMAIL PROTECTED]> wrote:
>
> >> I'm talking about "free speech", not "free beer". MySQL is much more
> >> free than Sybase, since it comes with complete source code, permission
> >> to modify it, etc.
> db> On the other hand, Sybase is a proven enterprise-class database
> db> engine...
>Your comment is completely orthogonal to mine, as far as I can tell :).
The point is that sometimes getting the job done is more important
than principle, and for some environments a non-transactional db
just ain't the way to go. At the moment, Sybase is really the
only practical free db out there if you need a full-featured
database engine.
I HOPE, as I've said before, that open-source Postgres will grow
into a solid, reliable db and that the folks working on it seem
to be making great progress.
But at the moment I wouldn't trust a system dealing with, say,
people's money to either Postgres or MySQL and Sybase is really
the only free option I know of.
As a total aside, AOLServer, which has great db connectivity
features and runs great on linux, has just annouced their next
version will be Open Source. It's been free thus far, but not
Open Source...
Despite the name, it wasn't developed by AOL (which is why
it works, of course).
--
- Don Baccus, Portland OR <[EMAIL PROTECTED]>
Nature photos, on-line guides, at http://donb.photo.net
------------------------------
From: Mario Klebsch <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: TAO: the ultimate OS
Date: Fri, 4 Jun 1999 10:09:03 +0200
"Shamsuddin, Amir (EXCHANGE:MDN05:7E24)" <[EMAIL PROTECTED]> writes:
>Current hardware, almost without exception, uses 8-bit binary bytes as a
>fundamental unit (call it a char if you like). Thusly, all software uses these to.
>Any attempt whatsoever to hide this from people will limit things, there would
>always be a lower level to go to. This is the reason I like *nix over soley
>GUIfied oses, you can do a majority of things from a text console (Where the text
>characters correspond to the underlying 8-bit bytes).
Have you ever read about early unix systems? In the beginning, it
was not always 8 bit bytes. There where a lot of different word sizes
and several methods of packing characters into words. Some systems
even did not have character addressing.
73, Mario
--
Mario Klebsch [EMAIL PROTECTED]
------------------------------
From: "Jon Smirl" <[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: Thu, 3 Jun 1999 15:16:56 -0400
If you have one writer and many readers, and that writer is updating several
tables in a dependent manner, how do you keep the many readers from getting
partial results? How about cursor stability for the readers? Are the
readers sorting the data; what about an insert happening in the middle of
the sort?
In a lightly loaded server your probability of encountering these problems
is low but not zero. Stress the server with lots of activity and these
problems will show themselves. Transactions and isolation levels address
these issues.
--
Jon Smirl
[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?
From: [EMAIL PROTECTED] (Don Baccus)
Date: 3 Jun 1999 12:12:37 PST
In article <ViA53.1666$[EMAIL PROTECTED]>,
bryan <[EMAIL PROTECTED]> wrote:
>again, like I said, with multiple writers contending for common
>resources, yes you're right. for the "one writer, many readers" you
>do NOT need xactions.
Not necessarily true. If you have several related tables
that need to be logically updated at once, the atomicity of
the transactional model is, well, useful if there's a crash
while records are being inserted or updated.
Even if you don't have a crash, when you have such related
tables it's generally kinda nice if "select" statements are
guaranteed to return information in a consistent state.
Without transactions, in this scenario a "select" can pick
up stuff where only a subset of the tables have been
updated.
Of course, even with transactions the programmer can screw
things up :)
It sounds like your application's really simple, and MySQL
sounds like it's great for you.
--
- Don Baccus, Portland OR <[EMAIL PROTECTED]>
Nature photos, on-line guides, at http://donb.photo.net
------------------------------
From: Andreas Schwab <[EMAIL PROTECTED]>
Subject: Re: bootinfo.h missing for make dep
Date: 04 Jun 1999 11:23:34 +0200
Jonathan DelStrother <[EMAIL PROTECTED]> writes:
|> Hiya - got an all new problem for ya....
|> OK, I'm attempting to get sound support in the Redhat 5.2 kernel, so, go
|> to src/linux/drivers/sound, make config, go through all of that stuff...
|> Right, that seemed OK, so - make dep. I get an error message:
|> dmasound.c : 81 : asm/bootinfo.h : no such file or directory
make dep is stupid. asm/bootinfo.h is only used on m68k.
|> O yeah - while I'm here, say I want to create a script....
|> create a new file, add the #!bin/sh bit at the top (I've forgotten
|> exactly what it is, but anyway-) add in some commands (eg echo 'hi'),
|> save the file, do 'chmod +x thisfile', type in thisfile at the console,
./thisfile
--
Andreas Schwab "And now for something
[EMAIL PROTECTED] completely different"
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Chris Raper)
Subject: Re: Redhat 6 & NFS
Date: Fri, 04 Jun 1999 09:38:28 GMT
On Thu, 03 Jun 1999 13:44:14 -0600, Bill Anderson
<[EMAIL PROTECTED]> wrote:
>> I had to enable CONFIG_EXPERIMENTAL to get CONFIG_NFSD. Once
>> CONFIG_NFSD was set (as a module) the nfsd fired up and seemed to
>> work. So, it looks like redhat has things setup to use the kernel
>> nfsd.
>Which,IMNSHO, is a mistake in that it only supports linux nfs clients.
>At least they could have shipped with a non-linux-only nfs daemon, even
>if they wanted to default to knfs.
Hi Bill,
Do you know if this is also true of Caldera 2.2? I only ask because I
just can't seem to get NFS working between Linux and SCO. Linux will
mount one of the SCO exports but SCO keeps saying that Linux isn't
running NFS - and nfsd isn't running :-(
TIA
Chris R.
------------------------------
From: [EMAIL PROTECTED] (Phillip Fayers)
Crossposted-To:
comp.os.ms-windows.nt.admin.misc,comp.sys.sun.admin,comp.sys.hp.misc,comp.os.linux.hardware
Subject: Re: Terabite Plus Filesystems
Date: 4 Jun 1999 09:40:47 GMT
Reply-To: [EMAIL PROTECTED]
In article <[EMAIL PROTECTED]>, Bob Hoekstra <[EMAIL PROTECTED]> writes:
>gus wrote:
>> Gene Heskett wrote:
>> > ... IBM, SCO, and another that slips my mind, have made a
>> > first port of Unix to the ia64(Merced)? Somewhat based on the AIX core
>> > according to the half pager....
>> AFAIK, Linux is ported to 64 bit architectures ...
>> My understanding is that Linux will be on the Merced before NT ... ;-)
>I suspect that Solaris will be there before NT as well. In any event, I douibt
>that it will damage Linux much if it is on this platform later than NT. Let's
>face it, there are very few applications where you *must* have a 64 bit OS.
You don't have to `suspect'. Check out Suns press releases, one from
11th January 1999 mentions `Solaris running on the IA-64 Pre-Silicon
Software Development Environment'. I `suspect' that Sun could produce
a version of Solaris for Merced tomorrow if the hardware was there to
run it on.
>Of far more interest is IBM's support of the Linux community and the
>proclamation that they will be selling RS/6000 machines as well as high-end
>PCs loaded with Linux :-)
And this from a company which is also supposed to be part of a group of
software companies producing a `unified' version of UNIX for merced.
--
Phillip Fayers, SunAdmin/Support/Programming/Postmaster/Webmaster(TM)
Dept of Physics & Astronomy, University of Wales, College of Cardiff.
[EMAIL PROTECTED] Attribute these comments to me, not UWCC.
------------------------------
From: [EMAIL PROTECTED] (Grant Edwards)
Subject: How to serialize read operations in char driver?
Date: Thu, 03 Jun 1999 19:34:53 GMT
I need to serialize read operations in a character driver. Multiple
processes can have the device open, and can issue "simultaneous"
read() requests. My driver needs to serialize the read() operations:
If a read() request is received while another read() is being
processed (data reception is interrupt-driven), the second read()
needs to wait until the first has completed.
I think that the tty device driver behaves this way, but after looking
through the n_tty code, I can't really identify where/how this happens.
I need a mutual-exclusion semaphore type mechanism to serialize access
to a critical region. [Damn, based on the buzz-word density in that
sentence I guess grad school wasn't a complete waste of time!]
Can anybody point me to an example of how this is done?
If multiple read() requests are sleeping on a wait queue, will a
subsequent call to wakeup...() wake up just the first one in the queue
or all of them?
If I use a wait queue to block the read() requests that come in while
a read() is in progress, how do I do an atomic test-availability-and-sleep?
In Linux Device Drivers Rubini discusses one method:
while (test)
{
cli();
if (test)
interruptible_sleep_on();
sti();
/* check signals */
}
/* critical region */
I can see that this guarantees that you won't go to sleep when you
shouldn't, but I can't see that the converse is true.
What if two read requests both hit the first "test" at the same time,
both see that it's false, and both proceed? I don't see how this code
actually serializes access to the critical region.
Any hints?
--
Grant Edwards
[EMAIL PROTECTED]
------------------------------
From: James Lee <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: TAO: the ultimate OS
Date: 3 Jun 1999 13:47:41 -0500
: * a new kind of OS: Tao
I'm sure you have heard of:
Dao ke dao, fei chang dao.
which roughly translates to:
the Tao that can be Tao (told), is not extraordinary Tao.
Kind of suits what you say.
8-)
------------------------------
From: Larry Blanchard <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.misc,comp.unix.advocacy
Subject: Re: TAO: the ultimate OS
Date: 03 Jun 1999 19:41:28 PDT
Reply-To: [EMAIL PROTECTED]
Vladimir Z. Nuri wrote:
>
> for a long time, in C, functions were considered all that
> was necesary. C++ introduced the idea of objects. I am
> suggesting the same may be true for file systems. treating
> a file as "just a stream of bits" is appropriate at
> one level of abstraction, but we need a higher level of
> abstraction in which *every* entity in the OS is simply
> an object. implementing this as "raw bytes" does not
> conern me too much. now, I'm not saying this is gonna
> happen tomorrow. but I do believe in a few years, all
> serious OSes will be object oriented.
>
I've seen a fairly large body of opinion that holds that the usefulness
of OOP is in direct porportion to the percentage of user interface
code. In other words, an inverse relationship to the percentage of
internal calculations.
If everything goes OOP, who'll write the ultimate OS for process
control, real time, scientific/engineering apps, etc.?
I think OOP is another example of the human tendency "if all you've got
is a hammer,...".
Now I'll ignore this newsgroup till the flames die down :-).
--
Larry Blanchard - Old roses, old motorcycles, and old trains
Homo Sapiens is a goal, not a description.
------------------------------
From: David Wragg <[EMAIL PROTECTED]>
Subject: Re: Might Linux SMP write memory out of order?
Date: 03 Jun 1999 19:21:46 +0000
"Stefan Monnier <[EMAIL PROTECTED]>"
<[EMAIL PROTECTED]> writes:
> Indeed, I seem to remember that the x86 architecture is supposed to
> guarantee some reasonable kind of consistency.
x86 is just as (un)reasonable as Alpha, MIPS, PPC, Sparc, in this
regard: If you access multiple shared variables without the protection
of synchronisation primitives (mutexes etc.), you need memory barriers
to avoid getting bitten.
> 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).
In many ways, the problems of the compiler not literally following the
memory accesses implied by source code has a similar effect to the
processor dynamically not literally following the memory accesses
implied by machine code. The way to tell the compiler what you want is
using volatile (or function calls). The way to tell the processor what
you want is using memory barriers.
David Wragg
------------------------------
From: [EMAIL PROTECTED] (Tor Arntsen)
Subject: Re: Terabite Plus Filesystems
Crossposted-To: comp.os.ms-windows.nt.admin.misc,comp.os.linux.hardware
Date: Fri, 04 Jun 1999 10:02:29 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Christopher B. Browne) writes:
>(Alpha,
>and UltraSparc, are existing 64 bit ports; I suspect that there may
>be some late-breaking PPC or MIPS systems doing 64 bits, but am not
>sure if Linux runs on 'em...)
MIPS processors are 64 bits, even the old R4000 in the old ('93) Indy I
have here. However you're still correct w.r.t. Linux, the Indy port is
32 bit. (IRIX too is 32 bit on the Indy though, IRIX is 64 bit on their
bigger boxes only, doesn't matter which CPU it is.)
-Tor
------------------------------
From: Igor Zlatkovic <[EMAIL PROTECTED]>
Subject: Re: A simple question...
Date: Fri, 04 Jun 1999 10:04:48 +0000
> Hi everybody,
> I have used for years Linux (since 1.2.13 kernel versions) on
>single-processor
> machines, and now I'm thinking on a multiprocessor machine (2, 4 or 8 CPU's),
> but I don't know what motherboards are supported by the Linux kernel, and which
> is their performance. Could you tell me your experiences?
Hi.
Linux runs on every piece of electronics that knows the difference between zero and
one.
As of my knowledge, it runs on any motherboard.
Performance of the motherboard is not an issue anymore. According to the benchmarks in
the net, the fastest one is about two percent faster than the slowest one. Better, look
for the features you need and choose the motherboard that has them (for example,
on-board
SCSI, number of PCI slots, and so on). You can see and read a lot about this at
http://www.tomshardware.com.
IO-APIC (an enhanced interrupt controller) is a nice and important thing for the
interrupt handling on multiprocessor boards. The documentation of the kernel
(IO-APIC.txt) states that there is a whitelist of boards that have a nice one, and a
blacklist that lists boards that have a broken one. Unfortunately, it does not say
where
these lists can be found (maybe someone else knows?) . If you cannot find this list,
dont
despair, the document states that all boards work, you only have to fiddle a bit with
boot parameters to enable IO-APIC on those blacklisted boards.
I use a dual Pentium II motherboard, Gigabyte GA-6BXD. With the newest BIOS, it claims
to
support Pentum III as well. I had no problems with this board so far.
--
Igor Zlatkovic
mailto:[EMAIL PROTECTED]
University of Applied Sciences
Frankfurt, Germany, EU.
------------------------------
From: Dave Bynion <[EMAIL PROTECTED]>
Subject: Help - MAX size of tarfile ?? Is a tarfile a TAR or a FILE??
Date: Fri, 04 Jun 1999 20:02:32 +1000
Does anyone know if there is a maximum size of a tarfile when writing to
a tape? I know that there is a maximum size for ordinary files (I think
its 2G on a Intel based system), so that would also be the maximum size
of a tarball written to a file in the filesystem. But does the same, or
a similar limit, apply to a tarball written directly to a tape?
I am using 4mm DDS2 tapes (native capacity 4G, compressed capacity
approx 8G), and want to know what utilities I can use for backups. No
individual file is > 1G, but total filesystem capacity could be up to
6G.
anticiTHANKSpation
Dave B.
------------------------------
** 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
******************************
