618 lines
22 KiB
Plaintext
618 lines
22 KiB
Plaintext
From: Digestifier <Linux-Development-Request@senator-bedfellow.mit.edu>
|
|
To: Linux-Development@senator-bedfellow.mit.edu
|
|
Reply-To: Linux-Development@senator-bedfellow.mit.edu
|
|
Date: Thu, 15 Sep 94 21:13:15 EDT
|
|
Subject: Linux-Development Digest #179
|
|
|
|
Linux-Development Digest #179, Volume #2 Thu, 15 Sep 94 21:13:15 EDT
|
|
|
|
Contents:
|
|
Re: A thought to improve security (Jim Finnis)
|
|
Re: Driver for NCR 53C825 on the horizon? (Janne Sinkkonen)
|
|
Re: AX25 & KISS Amateur Radio Protocols in Linux?? (Peter Onion)
|
|
Re: need developer! (Don Garrett)
|
|
Re: HELP: How to patch Slackware 2.0 to 1.1.44 Kernel (Mark P. Nelson)
|
|
Re: DOSEMU 0.53 notes (Dan Maynard)
|
|
Re: inb(),outb()--help? (Wolfgang Jung)
|
|
Re: Linux v1.0 SMAIL problem (Neal Becker)
|
|
Re: IDEA: Energy saving features for harddisks (Pete Deuel)
|
|
Need a hand from C programmer... Please? (Bart Kindt)
|
|
Status of Token Ring? (Todd B. Templeton)
|
|
Re: RFD: Linux and end-users (Was: Don't use Linux?!) (Alan Cox)
|
|
Re: 320x200 X resolution? (Andreas Matthias)
|
|
Re: Alpha Linux (Mike Haertel)
|
|
Re: Multiprocessing Pentium Systems (Hugh Emberson)
|
|
Re: ET4000 and X-Windows (Jan Nicolai Langfeldt)
|
|
Re: Don't use Linux?! (Jeff Kesselman)
|
|
Re: File locking--gurus please read. :) (Alan Braggins)
|
|
Re: Future of linux -- the sequel (Ken Latta)
|
|
|
|
----------------------------------------------------------------------------
|
|
|
|
From: clef@aber.ac.uk (Jim Finnis)
|
|
Subject: Re: A thought to improve security
|
|
Date: Thu, 15 Sep 1994 13:57:31 GMT
|
|
|
|
In article <Cw62Cs.8o8@info.swan.ac.uk>
|
|
iialan@iifeak.swan.ac.uk (Alan Cox) burbled:
|
|
|
|
>
|
|
>Presumably the pyramid ones or the draft Posix (did these end up the same
|
|
>?). B2 security would be fun too.
|
|
>
|
|
|
|
I'd love to see GECOS ACL's..
|
|
|
|
Jim
|
|
|
|
|
|
------------------------------
|
|
|
|
From: janne@oliivi.pc.helsinki.fi (Janne Sinkkonen)
|
|
Crossposted-To: comp.os.linux.help
|
|
Subject: Re: Driver for NCR 53C825 on the horizon?
|
|
Date: 15 Sep 1994 16:54:24 +0300
|
|
|
|
In article <357av4$o67@csnews.cs.colorado.edu>,
|
|
Drew Eckhardt <drew@frisbee.cs.Colorado.EDU> wrote:
|
|
|
|
>>The NCR53c810 driver works after one (minor) bug is corrected,
|
|
>>namely the removal of the [2] which had been used to dimmension
|
|
>>the pci_chip_ids[] array -
|
|
|
|
I started the driver included in 1.1.49, and the following was
|
|
required in addition to what you told:
|
|
|
|
> To: Drew Eckhardt <drew@frisbee.cs.Colorado.EDU>
|
|
> Subject: Re: Driver for NCR 53C825 on the horizon?
|
|
> Date: Wed, 14 Sep 94 22:10 EET DST
|
|
>
|
|
>
|
|
> I got it working! A little fiddling was required, however:
|
|
>
|
|
> - in bios32.c, bios32_init():
|
|
>
|
|
> for (check = (union bios32 *) 0xe0000; check <= (union bios32 *) 0xefff0; ++check) {
|
|
>
|
|
> The change of the upper address limit was necessary to prevent
|
|
> it finding double signature.
|
|
>
|
|
> - And in 53c7,8xx.c, the "case 825" line was missing:
|
|
>
|
|
> switch (hostdata->chip) {
|
|
> case 810:
|
|
> case 815:
|
|
> case 825:
|
|
> hostdata->dstat_sir_intr = NCR53c8x0_dstat_sir_intr;
|
|
> hostdata->init_save_regs = NULL;
|
|
> hostdata->dsa_fixup = NCR53c8xx_dsa_fixup;
|
|
>
|
|
> This was with an Inter Premiere 2, P5/90
|
|
|
|
--
|
|
Janne
|
|
|
|
------------------------------
|
|
|
|
From: onion_p_j@bt-web.bt.co.uk (Peter Onion)
|
|
Subject: Re: AX25 & KISS Amateur Radio Protocols in Linux??
|
|
Date: 15 Sep 1994 16:15:18 GMT
|
|
|
|
In article <Cw4KF3.LBn@aston.ac.uk>, evansmp@mb4715.aston.ac.uk (Mark
|
|
Evans) says:
|
|
|
|
>Have you tried recompiling tin with the NNTP extensions disabled?
|
|
|
|
The problem is a bit more fundamental than that! The WNOS NNTP server
|
|
seems incapable of dealing with some basic requests. I cant remember
|
|
exactly the commands but it revolved around WNOS not taking any notice of
|
|
parameters to command like HEAD, TAIL, BODY etc......
|
|
|
|
I don't think its the NNTP extensions that are the problem (but I may be
|
|
wrong!)
|
|
|
|
|
|
Peter Onion G0DZB.
|
|
|
|
------------------------------
|
|
|
|
From: dgarrett@orbit.cs.engr.latech.edu (Don Garrett)
|
|
Subject: Re: need developer!
|
|
Date: 14 Sep 1994 04:20:17 GMT
|
|
|
|
Humenberger Edmund (k3076e5@cxmeta.edvz.uni-linz.ac.at) wrote:
|
|
: You should implement for a phone interface card
|
|
: an device driver.
|
|
: You will get the card, all support from the
|
|
: manufactor and money from me. the driver will
|
|
: be free.
|
|
|
|
: email your answer. ed
|
|
|
|
Can you nail down just what you mean by 'phone interface card'? Do
|
|
you mean voice mail? And what card? Do you get to pick our favorite,
|
|
or do you have a particular one in mind?
|
|
|
|
I don't have much free time. But if you're not in a big hurry, I'm
|
|
game.
|
|
|
|
--
|
|
Don Garrett Louisiana Tech
|
|
dgarrett@engr.latech.edu University
|
|
http://info.latech.edu/~dgarrett/
|
|
|
|
------------------------------
|
|
|
|
From: mpn@AlleleB.Berkeley.EDU (Mark P. Nelson)
|
|
Subject: Re: HELP: How to patch Slackware 2.0 to 1.1.44 Kernel
|
|
Date: 15 Sep 1994 21:33:20 GMT
|
|
Reply-To: mpn@alleleb.berkeley.edu
|
|
|
|
Chris Wong (chris@helser54) wrote:
|
|
|
|
: Is the following right?
|
|
|
|
: cd /usr/src/linux
|
|
: patch < $patchFileDir/patchfile
|
|
|
|
No.
|
|
|
|
cd /usr/src
|
|
|
|
patch -s -p0 </patchpath/patchfile
|
|
|
|
--
|
|
Mark P. Nelson (mpn@alleleb.berkeley.edu)
|
|
While I'll admit that anyone can make a mistake once,
|
|
to go on making the same lethal errors century after
|
|
century seems to me nothing short of deliberate.--V.
|
|
|
|
------------------------------
|
|
|
|
From: dcm@omega.csuohio.edu (Dan Maynard)
|
|
Subject: Re: DOSEMU 0.53 notes
|
|
Date: Wed, 14 Sep 1994 01:59:48 GMT
|
|
|
|
|
|
: >: I don't think so. Only ET4000, S3, trident for now.
|
|
: >I to wish they supported ATI. I have one (weird card) and DOSEMu stops
|
|
: >when I try to use the graphics mode. Do I need to make a copy of the
|
|
: >BIOS for the video card? Any strange thing to get graphics mode. Character
|
|
: >mode works great. The old DOSEMu 52 worked with the ATI. Cursor looked
|
|
: >funny but it worked.
|
|
|
|
with the ATI GUP -
|
|
|
|
ports { 0x1ce 0x1cf }
|
|
|
|
cleared up my lock-up problem in FULL graphics mode
|
|
|
|
|
|
|
|
------------------------------
|
|
|
|
From: woju@keep.in-berlin.de (Wolfgang Jung)
|
|
Subject: Re: inb(),outb()--help?
|
|
Date: 15 Sep 1994 13:37:23 +0200
|
|
|
|
Steve Larsen (slarsen@gonix.com) wrote:
|
|
: Hi:
|
|
: I would like to write a modem diagnostic program for Linux of the sort
|
|
: I used to write for DOS in assembler. But I don't want to write in assembler.
|
|
: I want to use C, and am admittedly no wiz. Anyway, I 'ioperm' the correct
|
|
: ports, then go to use inb(), and get the "undefined reference to __inb_p()"
|
|
: message, or something to that affect, so I know it's tracing it back to
|
|
: asm/io.h. Am I leaving out an include file, or do I have to fill out the
|
|
: port call w/some sort of define? A one or 2 line example of code would
|
|
: be greatly appreciated!
|
|
|
|
Do you really need to program the hardware for the modem diagnostics ?
|
|
If you just need to talk to the Modem, use open/read/write , termios, etc
|
|
(You can select everything without programming the hardware (The Kernel does
|
|
this for you....)
|
|
|
|
|
|
|
|
--
|
|
===============================================================================
|
|
| Gruss |ISO-8859-1 Mail: woju@keep.in-berlin.de |
|
|
| Wolfgang | woju@keep.bln.sub.org |
|
|
| Jung | wong@cs.tu-berlin.de |
|
|
| | wojuacac@w250zrz.zrz.tu-berlin.de |
|
|
===============================================================================
|
|
|
|
------------------------------
|
|
|
|
From: neal@ctd.comsat.com (Neal Becker)
|
|
Crossposted-To: comp.os.linux.admin,comp.os.linux.misc,comp.os.linux.prog,dc.org.linux-users
|
|
Subject: Re: Linux v1.0 SMAIL problem
|
|
Date: 15 Sep 1994 03:13:11 GMT
|
|
|
|
Did you really want uucp_neighbors? How about removing the offending
|
|
router?
|
|
|
|
------------------------------
|
|
|
|
From: deuelpm@craft.camp.clarkson.edu (Pete Deuel)
|
|
Subject: Re: IDEA: Energy saving features for harddisks
|
|
Date: Wed, 14 Sep 1994 02:29:11 GMT
|
|
|
|
In article <1994Sep12.232713.11779@ida.liu.se> y93chrwe@ida.liu.se (Christer Weinigel) writes:
|
|
|
|
> And wouldn't it be great to be able to proudly say "Linux supports
|
|
>the EPA (or whatever the latest three letter acronym is) specification."
|
|
|
|
And don't forget the "The EPA doesn't officially endorse any product."
|
|
slogan...
|
|
|
|
Catchy, eh? ;)
|
|
|
|
|
|
Pete
|
|
===================================================
|
|
"Actually, I'm a lab mouse on stilts..."
|
|
|
|
E-mail: deuelpm@craft.camp.clarkson.edu
|
|
===================================================
|
|
|
|
------------------------------
|
|
|
|
From: bart@dunedin.es.co.nz (Bart Kindt)
|
|
Subject: Need a hand from C programmer... Please?
|
|
Date: Fri, 16 Sep 1994 16:53:16 GMT
|
|
|
|
Hi. I need a bit of help from somebody with C (and Kernel) experience.
|
|
For the last 3 months I have been trying to fix a serious (for me) problem
|
|
with a tiny wee program called "sliplogin". The program is used here for a
|
|
multiline dial-in server. It works fine, *except* that it sometimes does not
|
|
*kill* itself after the dial-in user has disconnected. This especially seems
|
|
to happen when there is still traffic going over the SLIP line. The problem
|
|
is, that the sliplogin program 'hangs' in the " close(0) " function.
|
|
|
|
That part of the source looks like this:
|
|
|
|
system(.....); execute the slip.logout script; This takes the Route/ Arp out.
|
|
close(0); The program does sometimes *not return* from this function!
|
|
exit(1); This kills the program, but is often not reached!
|
|
|
|
Now I have been told that the close(0) function is not properly implemented in
|
|
Linux. It does not flush the buffers, but only waits 60 seconds before it
|
|
returns. I don't know if this is true, but what I would like to do, is
|
|
manually *flush* all data in the buffers, send reset commands to all open
|
|
TCP/IP connections, and *then* close the port. Is this possible? Or is there
|
|
an other way to solve this problem?
|
|
|
|
If you can help me, please reply by E-Mail, and I will send you the complete
|
|
source (which is only 300 lines).
|
|
|
|
This is very important to me, because every time it happens, the dial-in line
|
|
does not answer anymore, and if it is the 'master' line, due to the Telecom
|
|
Stepping system, my entire Internet Server is down until I mannualy KILL the
|
|
hanging Sliplogin program.
|
|
|
|
If you can do anything to help me, please do!
|
|
|
|
Greetings, Bart.
|
|
|
|
==============================================================================
|
|
Bart Kindt, System Supervisor, Efficient Software NZ LTD, Dunedin, New Zealand
|
|
Amateur Radio: ZL4FOX / PA2FOX
|
|
==============================================================================
|
|
|
|
------------------------------
|
|
|
|
From: toddt@phelpsd.com (Todd B. Templeton)
|
|
Subject: Status of Token Ring?
|
|
Reply-To: toddt@phelpsd.com (Todd B. Templeton)
|
|
Date: Thu, 15 Sep 1994 21:39:14 GMT
|
|
|
|
|
|
|
|
What is the status of the project to support Token Ring NICs? Is this
|
|
progressing? We have just loaded up Linux and are *VERY* impressed but our
|
|
LAN is strictly Token-Ring. We would be very glad to participate in this
|
|
effort as a beta test site--i.e., volunteer--aaaaaghhh! :-). Seriously,
|
|
if test sites are needed, we are more than willing to help out.
|
|
|
|
Our network has approximately 300 nodes, running IPX and TCP/IP with an
|
|
Internet connection (if that's not obvious) via a Livingston router and
|
|
several wide-area network connections via Wellfleet routers. We can
|
|
probably give the driver a fairly good stress test. If interested, please
|
|
E-Mail me at toddt@phelpsd.com
|
|
|
|
Thanks...
|
|
|
|
--
|
|
<------------------------------------------------------------------------------>
|
|
Todd Templeton | All opinions are solely my own and do not reflect the
|
|
Lead Software Engineer | views of Phelps Dodge Corporation or its management.
|
|
<------------------------------------------------------------------------------>
|
|
|
|
|
|
------------------------------
|
|
|
|
From: iialan@iifeak.swan.ac.uk (Alan Cox)
|
|
Subject: Re: RFD: Linux and end-users (Was: Don't use Linux?!)
|
|
Date: Thu, 15 Sep 1994 16:29:23 GMT
|
|
|
|
In article <34uc2aINNj8o@sbusol.rz.uni-sb.de> hightec@sbusol.rz.uni-sb.de (Michael Schumacher) writes:
|
|
>versions of the C library. If a customer purchases your database and is
|
|
>not able to get things running, he will definitely make YOU responsible
|
|
>for that, even though the problems are not caused by your database code.
|
|
|
|
Vendors are used to this. Most are very good. We had numerous no longer
|
|
works with windows 3.1 software (and a load of networking stuff win 3.11
|
|
really took out). Most vendors seem to expect this and were helpful, others
|
|
were not 8). I don't see it as a Linux problem.
|
|
|
|
On stability. Except for a few bugs that were faults in the program
|
|
concerned (old route etc) everything I have from Slackware 1.1 runs
|
|
faultlessly under libc4.5.26, most claims to be relative to libc 4.1. I
|
|
did some tests with even older incompatible stuff using libc2.2.2 and those
|
|
when libc2.2.2 is still installed also work fine.
|
|
|
|
That to mean says stable enough to work with. There is one counter example
|
|
which is the massive changes to the C++ libraries, which coupled with the
|
|
fact that g++ 2.5.8 is a bit of a bug hazard broke the parc place stuff
|
|
as Warner Losh has explained.
|
|
|
|
Alan
|
|
--
|
|
..-----------,,----------------------------,,----------------------------,,
|
|
// Alan Cox // iialan@www.linux.org.uk // GW4PTS@GB7SWN.#45.GBR.EU //
|
|
``----------'`----------------------------'`----------------------------''
|
|
|
|
------------------------------
|
|
|
|
Crossposted-To: comp.os.linux.misc
|
|
From: andy@titan.central.de (Andreas Matthias)
|
|
Subject: Re: 320x200 X resolution?
|
|
Date: Thu, 15 Sep 1994 02:25:43 GMT
|
|
|
|
Stephen Collyer (stephen@dogmatix.inmos.co.uk) wrote:
|
|
|
|
: |> : "320x200" 25 320 344 376 400 200 204 206 225
|
|
|
|
: I don't see how these values work. With a 25 MHz clock, with total V. lines =
|
|
: 400, and total H. lines 225, we're looking at a refresh rate of roughly
|
|
: 25E6 / (400 x 1.1 x 225 x 1.05) = 240
|
|
: This is way above the spec of any monitor I'm familiar with. Or am I missing
|
|
: something obvious ?
|
|
|
|
I posted a similar mode line a few days ago, and got similar comments.
|
|
|
|
I don't know why such modes work, but obviously they _DO_ on some monitors.
|
|
My AOC 17'' let me play for a whole night in that resolution, without
|
|
complaining or showing discomfort. It still works perfectly, and the AOC
|
|
is a very cheap monitor, nothing special.
|
|
|
|
I suppose that the monitor hardware detects the undisplayable refresh
|
|
rate and perhaps displays only every fourth frame or so.
|
|
|
|
Ciao,
|
|
|
|
Andreas
|
|
|
|
--
|
|
Andreas Matthias <andy@titan.central.de>
|
|
|
|
Zehntenstr.9
|
|
D-37120 Bovenden
|
|
|
|
Voice: +49/551/81377
|
|
|
|
------------------------------
|
|
|
|
From: mike@ichips.intel.com (Mike Haertel)
|
|
Subject: Re: Alpha Linux
|
|
Date: 15 Sep 1994 20:22:45 GMT
|
|
|
|
In article <tmCw3ytC.3D0@netcom.com> tm@netcom.com (Toshiyasu Morita) writes:
|
|
>Nope, if I remember correcty the 68000 and 68008 technically have a 16-bit
|
|
>ALU. If you don't believe me, look at the cycle counts for the 68000 when
|
|
>doing 8/16 bit adds versus 32 bit adds...the 32-bit add takes an extra
|
|
>4 clock cycles because it has to run the data the high word through the
|
|
>ALU as well, even on register-to-register operatoions.
|
|
|
|
Sorry, this is bogus, you are wrong.
|
|
|
|
The 68000 and 68008 both have a 32 bit ALU.
|
|
|
|
The extra 4 cycles (on the 68000) come when loading 32-bit values
|
|
from memory, where the cost is 4 cycles per 16-bit word loaded.
|
|
I assume on a 68008 the cost would be an extra 8 cycles.
|
|
|
|
Thus: a 32-bit register->register add is just as fast as a 16-bit one.
|
|
But a 32-bit memory->register or immediate->register is 4 (or 8)
|
|
cycles slower.
|
|
|
|
Why the heck are we talking about this in comp.os.linux?
|
|
--
|
|
Mike Haertel <mike@ichips.intel.com>
|
|
Not speaking for Intel.
|
|
|
|
------------------------------
|
|
|
|
From: hugh@hugh.cosc.canterbury.ac.nz (Hugh Emberson)
|
|
Subject: Re: Multiprocessing Pentium Systems
|
|
Date: 14 Sep 1994 02:30:56 GMT
|
|
Reply-To: hugh@cosc.canterbury.ac.nz
|
|
|
|
>>>>> "Don" == Don Waugaman <dpw@cs.arizona.edu> writes:
|
|
|
|
Don> Spinlocks typically involve busywaiting - tying up the CPU that
|
|
Don> is spinning on the lock for an (often prolonged) period of time.
|
|
Don> This is why spinlocks are typically used for situations where the
|
|
Don> lock will be held for a short period of time.
|
|
|
|
Don> A mutex, on the other hand, guarantees mutual exclusion by
|
|
Don> placing threads waiting on the lock on a queue, then starting
|
|
Don> another thread that is ready to run. Thus, useful work can be
|
|
Don> done rather than having the thread waiting on the lock taking up
|
|
Don> useless CPU (and bus) cycles by spinning.
|
|
|
|
Don> I would guess that Sun used a mutex (as you said) and not a
|
|
Don> spinlock (as Scott said).
|
|
|
|
Just a clarification:
|
|
I said "spinlock".
|
|
Scott said "no can't be".
|
|
Local SunOS expert said "mutex",
|
|
so I silently changed it to "mutex" in my second post.
|
|
|
|
Actually I think that a mutex is anything that guarantees mutual
|
|
exclusion. So a spinlock which busy waits, and a semaphore which
|
|
queues processes are both mutexes. Thats my definition anyway, until
|
|
I find a better one :-)
|
|
|
|
I think 4.1.3 actually uses a spinlock, but I could be wrong.
|
|
|
|
Cheers,
|
|
Hugh
|
|
--
|
|
Hugh Emberson | ... from the end of the Information
|
|
hugh@cosc.canterbury.ac.nz | Super-four-wheel-drive-track.
|
|
|
|
------------------------------
|
|
|
|
From: janl@ifi.uio.no (Jan Nicolai Langfeldt)
|
|
Crossposted-To: comp.os.linux.help
|
|
Subject: Re: ET4000 and X-Windows
|
|
Date: 15 Sep 1994 19:46:43 +0200
|
|
|
|
|
|
In article <3591hr$gtc@hpsystem1.informatik.tu-muenchen.de>, ertl@Informatik.TU-Muenchen.DE (Hubert Ertl) writes:
|
|
> Knows anyone out there how to set up a ET4000 card for X-Windows?
|
|
> I get max. 1024x768 at 60 Hz out of it in Xconfig.
|
|
>
|
|
>
|
|
> How can i get the 1024x768 at 70 Hz mode?
|
|
|
|
This depends on what clocks your card has. It would help
|
|
if you told us that.
|
|
|
|
BTW. This definitly does not belong in comp.os.linux.development,
|
|
followups directed to comp.os.linux.help.
|
|
|
|
Nicolai
|
|
|
|
|
|
------------------------------
|
|
|
|
From: jeffpk@netcom.com (Jeff Kesselman)
|
|
Subject: Re: Don't use Linux?!
|
|
Date: Thu, 15 Sep 1994 03:00:56 GMT
|
|
|
|
In article <MIKE.94Sep14121941@moocow.math.nat.tu-bs.de> on.dowling@zib-berlin.de writes:
|
|
>>>> On 13 Sep 1994 14:08:55 -0700, bhogan@crl.com (Bill Hogan) said:
|
|
>
|
|
>
|
|
>I also suspect that there is yet another reason why DOS software is not ported
|
|
>to unices; such software is designed usually for naive users who have no
|
|
>inkling as to how their computer works, and would like to turn it on and off
|
|
>like a household appliance. Linux is currently only available for a PC. Most
|
|
|
|
In general Mike, I agreed with you. it is not true, necessarily, though
|
|
that DOs software doesn't get ported to Unices (pl. of Unix? neat!).
|
|
I have a terrific version of MS-WORD for SCO. My copy is a little old,
|
|
and i don't know if there are newer version, but it still has alot to
|
|
recommend it. It runs off of curses so any dial in user can use it, runs
|
|
multi-user no problem, talks to the Unix print queue nicely, and otherwise
|
|
is an almost exact replica of the same vintage OS version.
|
|
|
|
Its a terrific product, and I intend to install it on my new Linux system
|
|
after IBCS becomes stable.
|
|
|
|
------------------------------
|
|
|
|
From: armb@setanta.demon.co.uk (Alan Braggins)
|
|
Subject: Re: File locking--gurus please read. :)
|
|
Date: Tue, 13 Sep 1994 11:21:34 GMT
|
|
|
|
In article <350f4d$g8k@hermes.achilles.net> ben@dragon.achilles.net (Ben Eng) writes:
|
|
> if the process is able to deal with it properly. Additionally, your
|
|
> processes can always avoid deadlocks by agreeing on an ordering
|
|
> convention when applying locks to multiple files. (Say your
|
|
> relations are identified by an internal numbering scheme, then you
|
|
> would always apply non-blocking locks in ascending or decending
|
|
> relation number order.)
|
|
|
|
This tends to mean either each transaction has to already know the
|
|
results of its queries, or it has to lock more stuff than necessary.
|
|
Optimistic strategies with deadlock detection give better concurrency
|
|
in most situations.
|
|
--
|
|
Alan Braggins armb@setanta.demon.co.uk abraggins@cix.compulink.co.uk
|
|
"Any technology distinguishable from magic is insufficiently advanced"
|
|
|
|
|
|
------------------------------
|
|
|
|
From: Ken Latta <klatta@pkdla5.syntex.com>
|
|
Subject: Re: Future of linux -- the sequel
|
|
Date: Thu, 15 Sep 94 14:13:46 PDT
|
|
|
|
|
|
In article <3554rv$22o@bmerha64.bnr.ca>, <Hamish.Macdonald@bnr.ca> writes:
|
|
> >>>>> shendrix@escape.widomaker.com (Shannon Hendrix) wrote:
|
|
>
|
|
> Shannon> The IDE interface is not good for multiple drives but most of
|
|
> Shannon> my accesses come from one drive. IDE is 16-bits, SCSI is 8
|
|
> Shannon> so unless you have multple drives and put them under a severe
|
|
> Shannon> load, the Sun's I use are slightly slower (a lot slower in
|
|
> Shannon> some cases).
|
|
>
|
|
> Don't confuse the disk <-> controller speed with the controller <->
|
|
> memory/cpu speed.
|
|
>
|
|
> I believe that IDE is *1* bit at a time between the controller and
|
|
> disk.
|
|
|
|
This is true only if you accept the idea that an IDE "controller" is
|
|
actually embedded in the drive electronics. The data bus between the
|
|
controller and the bus adapter is 8 bits wide. In the case of the common
|
|
inexpensive devices, the bus adapter is just that, a bus buffering and
|
|
level translation device.
|
|
|
|
> SCSI is 8 bits at a time.
|
|
>
|
|
> IDE is 16 bits at a time between the controller and memory/cpu.
|
|
>
|
|
> SCSI depends on the controller (my controller transfers 32 bits at a
|
|
> time between the controller and memory).
|
|
>
|
|
|
|
Many IDE adapters support only 8-bit ISA bus transfers, but more expensive
|
|
ones use 16 and 32 bit bus transfers. Some even have logic to support
|
|
multiblock transfers or caching to enhance performance.
|
|
|
|
SCSI can do much better than IDE on multidisk operations because it can
|
|
issue commands to multiple disks for concurrent execution. IDE only allows
|
|
one drive to be active at a time.
|
|
|
|
Ken Latta
|
|
UNIX Systems Consultant
|
|
klatta@pkdla5.syntex.com
|
|
|
|
------------------------------
|
|
|
|
|
|
** 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: Linux-Development-Request@NEWS-DIGESTS.MIT.EDU
|
|
|
|
You can send mail to the entire list (and comp.os.linux.development) via:
|
|
|
|
Internet: Linux-Development@NEWS-DIGESTS.MIT.EDU
|
|
|
|
Linux may be obtained via one of these FTP sites:
|
|
nic.funet.fi pub/OS/Linux
|
|
tsx-11.mit.edu pub/linux
|
|
sunsite.unc.edu pub/Linux
|
|
|
|
End of Linux-Development Digest
|
|
******************************
|