Linux-Development-Sys Digest #613, Volume #6 Sun, 11 Apr 99 21:14:42 EDT
Contents:
Re: Idea: Make a seperate "i686" tree for Redhat Linux 6.0 (Fred Flatstone)
Re: Trusted Linux (Christopher B. Browne)
Re: Linux NFS server, Solaris cient, bad news ... (David T. Blake)
Re: module-info (Gary Momarison)
ELKS status? ("Daniel Pietsch")
STREAMS Mythology (David Grothe)
Re: Idea: Make a seperate "i686" tree for Redhat Linux 6.0 (Andrew Comech)
JOBS? ("Scott Davis")
Re: Writers wanted for zine articles. (Don Baccus)
Re: Problems with 2.2.5 & sound & apm (Roope Anttinen)
Help with Linux Schedulers Pease (Major's Room)
Debuging system calls (Wlmet)
Redhat Linux L10N (Song Yang)
Re: ISR: Kernel->User space call (Joe Pfeiffer)
Re: You can now use Winmodems in Linux!!!!!!! (Viljo Hakala)
----------------------------------------------------------------------------
From: Fred Flatstone <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.misc,linux.redhat.misc,alt.os.linux,comp.os.linux.hardware
Subject: Re: Idea: Make a seperate "i686" tree for Redhat Linux 6.0
Date: 11 Apr 1999 08:14:14 -0700
John Thompson <[EMAIL PROTECTED]> writes:
> Andrew Comech wrote:
>
> > On 08 Apr 1999 14:45:41 -0400, Johan Kullstam wrote:
> > >[EMAIL PROTECTED] (who?) writes:
> >
> > >not if you're from a commonwealth country - which includes new zealand
> > >(where is the old zealand btw?). in *english* (as opposed to american
> > >english) redhat is a group entity and considered plural. therefore,
> > >redhat do.
>
> > If you were to say "redHat sucks", you'd say "redHat sucks", though.
> > Cheese,
> > Andrew
>
> And if I wanted to say "you're a funny lad with cheese in
> your nose" I'd say "you're a funny lad with cheese in your
> nose." Your point?
I thought his point was that "redHat sucks" would refer to a sucky
distribution (singular), not a sucky company of people (plural
in British). We all know the Red Hat guys don't suck. (The brims
get in the way.)
------------------------------
From: [EMAIL PROTECTED] (Christopher B. Browne)
Subject: Re: Trusted Linux
Reply-To: [EMAIL PROTECTED]
Date: Tue, 06 Apr 1999 13:58:53 GMT
On 6 Apr 1999 06:16:19 -0700, Tim Smith <[EMAIL PROTECTED]> posted:
>ACL's only require significant user space support if you implement them
>incorrectly. (They also don't need special support in the file system
>if implemented correctly).
>
>For the right approach, study FILDAE on TOPS-10 (the idea, not the
>implementation--the implementation was a kludge leading to security
>problems, but the idea is sound). I've posted descriptions of how
>that scheme would work on Unix to usenet and to the kernel mailing
>list in the past, so a search should find them if anyone is curious.
Not documented "on the web," but locatable at DejaNews as:
<http://x16.dejanews.com/[ST_rn=ps]/getdoc.xp?AN=435914490.1&CONTEXT=923406353.1141047422&hitnum=0>
Quite a good presentation of the idea, to be sure.
--
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 free software today?..."
------------------------------
From: [EMAIL PROTECTED] (David T. Blake)
Crossposted-To: matrix.lists.linux.kernel
Subject: Re: Linux NFS server, Solaris cient, bad news ...
Date: 11 Apr 1999 10:00:14 -0700
mlw <[EMAIL PROTECTED]> writes:
>Sorry guys, 2.2.5 doesn't work either.
>I have seen this very problem. It seems as though when a file is
>created,deleted, then mkdir creates a directory, on NFS fom solaris. The
>Solaris box refuses to think the directory is a directory.
>
>I would love to see this fixed, I got people giving me sh$t about it.
http://sunsolve.Sun.COM/pub-cgi/us/pubpatchpage.pl
might help. All releases of Solaris have nfs patches
available. The standard Solaris release nfs has bugs
that make Sun server:linux client not work quite
right.
Note: This may or may not solve your problem, but it
would be the first thing to verify - that the Sun NFS
patches are applied.
--
Dave Blake
[EMAIL PROTECTED]
------------------------------
From: Gary Momarison <[EMAIL PROTECTED]>
Subject: Re: module-info
Date: 11 Apr 1999 08:20:25 -0700
Svein K Svendsen <[EMAIL PROTECTED]> writes:
> I have upgraded my rh5.2 with the new 2.2.5 kernel and can't figure out
> how to
> generate a new module-info-2.2.5 file. Hope somone will help me out.
The topic has been discussed a little recently and I haven't seen
anything helpful about it. I searched my RH52 boot scripts for
it and I don't see the file used anywhere. IIRC, a link was made.
You should probably just ignore it until you have module problems.
You might try removing it from /boot and from the boot script that
links it and if nothing breaks, tell Red Hat to remove it. They have
a pretty good bug reporting system which eventually lets you know
how they handled your bug report.
--
Look for Linux info at http://www.dejanews.com/home_ps.shtml and in
Gary's Encyclopedia at http://www.aa.net/~swear/pedia/index.html
------------------------------
From: "Daniel Pietsch" <[EMAIL PROTECTED]>
Subject: ELKS status?
Date: Sun, 11 Apr 1999 11:41:10 -0400
Does anyone know what the status of the ELKS project is?
The web and ftp sites still exist but nothing seems to have
been added since February of 1998. Has the project been
abandoned?
--
---
Daniel Pietsch
SST (a division of Woodhead Canada Limited)
Windows 95/NT/CE Developer (Applications and Device Drivers)
Remove NOSPAM and its mirror from my email address to reply,
------------------------------
From: David Grothe <[EMAIL PROTECTED]>
Subject: STREAMS Mythology
Date: Sun, 11 Apr 1999 12:55:14 -0500
Reply-To: [EMAIL PROTECTED]
It is now time to debunk some STREAMS mythology that has been floating
around this discussion group for a few years. I do not intend to
address _all_ the misinformation in a single post. In this post I want
to address the following STREAMS myth:
"STREAMS is so slow and clunky that major commercial Unix operating
system vendors have gone to sockets and bypassed the STREAMS
implementation of TCP/IP."
Recently Mike Jagdis wrote:
> It depends what you mean by "STREAMS TCP/IP" :-). Most of the SVR3
> derived systems used the Lachman stack which used a BSD-ish IP
> module with STREAMS wrappers front and back. The socket interface
> was dumped behind a STREAMS message interface and accessed by
> simply setting up the socket call number and arguments in a
> buffer instead of on the stack and passing it down through the
> STREAMS head using an ioctl.
>
Correct, as far as I know. The BSD to STREAMS port of TCP/IP was
particularly ugly and is probably the source of much of the "STREAMS is slow"
idea.
> SCO introduced the system call bypass in, I think, the net100
> patch to 3.2v5.0.0. Even with today's network speeds the overheads
> in dereferencing through STREAMS module stacks and the cache pollution
> implications become significant - especially once you start talking
> about fast pathing IP.
>
Incorrect. On an Intel box running SCO OpenServer 5.0.4, I placed a
breakpoint on the "send" routine in the kernel and single stepped as
follows: send calls sosend; sosend calls allocmsg which calls allocb
which allocates a STREAMS buffer; sosend calls soputmsg; soputmsg calls
putnext which is the STREAMS message passing routine.
This implementation has bypassed the "stream head" which is not the same
thing as bypassing STREAMS. It is passing the data from the socket
"send" call down to TCP using the STREAMS construct of "putmsg".
In another recent post G. Sumner Hayes wrote:
> Solaris (in recent versions, certainly 2.6 and later) uses
> kernel sockets-based TCP/IP in the fast path and has STREAMS relegated
> to secondary status. There is also a TCP/IP-over-streams implementation,
> but the kernel sockets are preferred unless you need STREAMS (and the
> corresponding performance hit). This is fairly transparent to the
> application programmer.
>
On a SPARC running Solaris 2.6 I placed a breakpoint at the "send"
routine in the kernel. The calling chain went as follows: send calls
sendit; sendit calls sosendmsg; sosendmsg calls strwrite.
The routine "strwrite" is actually the STREAMS write entry point. It is
the same routine that gets called when a user program does a "write"
system call on a STREAMS file. So the socket "send" call is just a
filtered call on the STREAMS write procedure. In this case the system
does not even bypass the "stream head".
What is bypassed is the TLI library in user space. The "send" call in
user space ends up calling a library routine so_send which ends up doing
the system call.
To recapitulate: Both SCO OpenServer version 5 and Solaris 2.6 still
have TCP/IP implemented in STREAMS. Socket calls do not go through TLI
in user space; they make direct kernel calls. Once in the kernel the
flow of control enters TCP via its STREAMS interface.
For any of you who are still reading, here is some information that may
help clear up this past confusion.
Both "sockets" and "TLI" are APIs. Each is a "library" (whether
implemented in the kernel or in user space) of routines that allows a
user program to interface with TCP and other networking protocols.
STREAMS is a data communications message passing executive that resides
in the kernel. It is used to build protocol stacks and pass messages
between the protocol layers.
There are other non-STREAMS methods of building TCP/IP/Mac layer
stacks. BSD has one method; Linux has another; Windows still a third.
The "socket" API does not imply that the underlying implementation is
BSD-like (non-STREAMS). Witness SCO and Solaris.
The TLI API does not imply that the underlying implementation is
STREAMS. Linux is the case in point here in which there is a TLI
library available which implements the TLI API in terms of sockets.
The "stream head" is a kernel module at the top of a STREAMS
implementation which bridges between file system calls (read, write,
etc) and the message passing mechanisms of STREAMS drivers.
SCO's socket implementation seems to bypass the "stream head" and pass a
message directly down to the TCP STREAMS driver. Solaris makes the
"send" call look like a "write" from user space and enters the "stream
head" to actually copy the data into the kernel and pass it down to TCP.
If anyone would like some more technical information about STREAMS go to
http://www.gcom.com/#streams.
-- Dave
------------------------------
From: [EMAIL PROTECTED] (Andrew Comech)
Crossposted-To:
comp.os.linux.misc,linux.redhat.misc,alt.os.linux,comp.os.linux.hardware
Subject: Re: Idea: Make a seperate "i686" tree for Redhat Linux 6.0
Reply-To: [EMAIL PROTECTED]
Date: 11 Apr 1999 14:48:49 -0500
On 11 Apr 1999 08:14:14 -0700, Fred Flatstone wrote:
>John Thompson <[EMAIL PROTECTED]> writes:
>> Andrew Comech wrote:
>> > On 08 Apr 1999 14:45:41 -0400, Johan Kullstam wrote:
>> > >[EMAIL PROTECTED] (who?) writes:
>> > >not if you're from a commonwealth country - which includes new zealand
>> > >(where is the old zealand btw?). in *english* (as opposed to american
>> > >english) redhat is a group entity and considered plural. therefore,
>> > >redhat do.
>>
>> > If you were to say "redHat sucks", you'd say "redHat sucks", though.
>>
>> And if I wanted to say "you're a funny lad with cheese in
>> your nose" I'd say "you're a funny lad with cheese in your
>> nose." Your point?
>
>I thought his point was that "redHat sucks" would refer to a sucky
>distribution (singular), not a sucky company of people (plural
>in British). We all know the Red Hat guys don't suck. (The brims
>get in the way.)
Right, that's exactly the way the things are! English language
allows to give distinct characteristics to the group of people and to
the distribution which are referred to with the same name; I should have
noticed it myself. You filled it with sense to the brim. It is such a
pleasure to speak to literate people; thank you Fred!
Best,
Andrew
--
Looking for a Linux-compatible V.90 modem? See
http://www.math.sunysb.edu/~comech/tools/CheapBox.html#modem
Expect to pay below $50.
------------------------------
From: "Scott Davis" <[EMAIL PROTECTED]>
Subject: JOBS?
Date: Sat, 10 Apr 1999 13:39:04 -0500
Reply-To: "Scott Davis" <[EMAIL PROTECTED]>
US Recruiters, www.usrecruiters.com, is presently seeking individuals with
expertise and
experience in networking (NT, Novell, UNIX), programming ( Oracle, Java,
C++, etc ) and
other IT specialities for both domestic and oversears contract placements.
Contract terms range from 6 months to 3 years.
Please visit our website for further information at www.usrecruiters.com.
Thanks,
Scott Davis
US Recruiters
------------------------------
Subject: Re: Writers wanted for zine articles.
From: [EMAIL PROTECTED] (Don Baccus)
Date: 11 Apr 1999 12:09:59 PST
In article <ZdTP2.498$ya.637@amsnews>, Frank v Waveren <[EMAIL PROTECTED]> wrote:
>Why would being freelance rule out pay?
It doesn't. That's probably why this guy's only running an e-zine
rather than a real publication.
>Don't you mean free-ware? or GNU etc?
Something like that, I'm sure.
>A magazine for the 'hack/phreak community' as you call it, coming from an AOL
>user?
Hey, maybe he's not paying for it! :)
--
- Don Baccus, Portland OR <[EMAIL PROTECTED]>
Nature photos, on-line guides, at http://donb.photo.net
------------------------------
From: Roope Anttinen <[EMAIL PROTECTED]>
Subject: Re: Problems with 2.2.5 & sound & apm
Date: 11 Apr 1999 19:15:29 GMT
Reply-To: [EMAIL PROTECTED]
Kent Ritchie <[EMAIL PROTECTED]> wrote:
> I had similar problems upgrading from 2.0.34 to 2.2.x. The APM features
> don't seem to work correctly, at least poweroff on shutdown and screen
> blanking. And I was experiencing problems with my sound card - Real
> Player would complain of a general error... and then would display an
> error about buffers...
The situation with realplayer is that it's doing something that's not
allowed (don't remember what) and the new sound driver is strictier that the
old one thus the error. For me the sound driver was unusable (GUS PnP) but
after hearing peoples testiomonials I tried alsa and it works perfectly -
Linux has never sounded better :)
No I have a new problem. What's the module char-major-4 which modprobe
doesn't find? Despite that the system seems to work quite well.
> I wanted to move to 2.2.x so I could use dhcp, but I guess I will just
> have to wait until some of these little problems come into better focus.
AFAIK dhcp is available in 2.0 series too.
Roope
--
MicroSoft? is that some kind of a toilet paper?
PS: Look for address here, not from headers. And remove NOSPAM's
___________________________________________________________________________
[EMAIL PROTECTED] / [EMAIL PROTECTED]
+358 9 812 7567 / +358 500 445 565 / +358 49 445 565
http://myy.helia.fi/~anttiner/index.html
===========================================================================
Helsinki Business Polytechnic - Institute of information technology
------------------------------
From: Major's Room <[EMAIL PROTECTED]>
Subject: Help with Linux Schedulers Pease
Date: Sun, 11 Apr 1999 20:49:59 GMT
I am working on a project testing different types of schedulers on a
Linux/Debian system. As I am new to Linux, I do not know where to look
for some of the material and info that I need to get started. First
off, could anyone please tell me what type of scheduler debian 2.0
uses...I believe it is priority scheduling but I am not 100% sure.
Also, if you know where this information can be found on the web, can
you please direct me to it.
I also have been searching for the algorithms for different schedulers,
or kernels that I can install which have are already using different
schedulers. If you know where to find them, or have them please let me
know as soon as possible.
I really need to get started on this project and would appreciate any
help I could get.
Thank you,
Jason
------------------------------
From: [EMAIL PROTECTED] (Wlmet)
Subject: Debuging system calls
Date: 11 Apr 1999 20:59:40 GMT
Let us compile the C library with the -g option, then compile the kernel with
the -g option. Let us then run a program using gdb. Will this allow us to
trace the system calls?
If this is true, is it practical to compile the C library with the -g option.
I understand that it is fairly memory intensive to compile anyway.
Is there any other way to do this? I found that strace was not informative
enough for me in some situations.
------------------------------
From: Song Yang <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.questions
Subject: Redhat Linux L10N
Date: Sun, 11 Apr 1999 15:55:32 -0700
Are there Localized versions/extensions of Redhat distribution in
Chinese, Korean, German or French?
--
Song Yang (650)933-8458(Work)
Localization Department (408)253-4695(Home)
Silicon Graphics (650)932-0317(Fax)
Mailstop 43L-748 [EMAIL PROTECTED]
------------------------------
From: Joe Pfeiffer <[EMAIL PROTECTED]>
Subject: Re: ISR: Kernel->User space call
Date: 11 Apr 1999 17:07:11 -0600
[EMAIL PROTECTED] (David Lee) writes:
>
> How about using the call gate? If i recall correctly, it
> allows calling a procedure in low protection range (process
> space) from higher protection range (kernel space).
It's the other way around -- call gates are what lets low protection
code call high protection code (the confusion may be from the fact
that they use a higher number to mean a lower level of protection).
The call gate hijacks the call, and sends it to an address defined by
the call gate -- this provides the necessary protection, because it
doesn't let the low protection code decide where in the high
protection area it wants to run.
> In fact, a project called GAMMA they use call gate to
> trigger a process asyn. handler.
I'm not familiar with that -- what is 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: [EMAIL PROTECTED] (Viljo Hakala)
Subject: Re: You can now use Winmodems in Linux!!!!!!!
Date: 6 Apr 1999 09:25:09 GMT
On Mon, 05 Apr 1999 23:49:38 GMT, Frank v Waveren <[EMAIL PROTECTED]> wrote:
>You can't use a winmodem with vmware because vmware relies on the host
>OS to use the hardware afaik.
That's true, but perhaps you're forgetting the fact how winmodems work.
Winmodems are seen as dumb com-devices, so yes Linux "sees" the device,
but it isn't initialized as winmodes are setup by software.
In this case running windoze under vmware initialized the
com device, so I believe this should work.
However,if it does work it shouldn't be hard to capture the traffic that
the windows drivers send to the device in order to control the modem.
This way you probably could develop a set of drivers for Linux.
--
vh
------------------------------
** 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
******************************
