Linux-Development-Sys Digest #115, Volume #7 Fri, 27 Aug 99 06:14:16 EDT
Contents:
Re: The optimization debate (was: why not C++?) (David Fox)
Re: I HATE LT WIN MODEM!!!!!!!!!!!!!!! ([EMAIL PROTECTED])
Re: Jesus: the ultimate OS (Miles Bader)
Re: kernel and -m486 (Daniel Robert Franklin)
Re: glibc-2.1.1 problems (Andreas Jaeger)
Semaphore/Thread/Socket synchronization ("Christian Weisz")
Re: Semaphore/Thread/Socket synchronization ("Scott Simpson")
Re: where are packets created? (Ken Booth)
bzImage, where is it? (Brian Lepore)
Re: Newbies Programming course (Tristan Wibberley)
Struct SK_BUFF(Networking code) (Fernando Ortega)
Re: does linux have a package for drawing block diagrams ? (Peter Pointner)
Re: strcmp in kernel (Peter Pointner)
CDROM: Illegal Request (Dr. Peer Griebel)
Re: why not C++? (Tristan Wibberley)
Re: glibc-2.1.1 problems (Mike Dowling)
----------------------------------------------------------------------------
From: d s f o x @ c o g s c i . u c s d . e d u (David Fox)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.misc
Subject: Re: The optimization debate (was: why not C++?)
Date: 26 Aug 1999 21:58:59 -0700
[EMAIL PROTECTED] (Paul D. Smith) writes:
> likely to lead to more efficient code! I, myself, quite often use
> temporary pointers to walk through arrays rather than incrementing a
> counter and using array indexing. I actually do it because I find the
> code simpler to understand that way, but it's probably faster, too.
I believe this is faster on a sparc, slower on an x86. Or perhaps its
the other way around...
--
David Fox http://hci.ucsd.edu/dsf xoF divaD
UCSD HCI Lab baL ICH DSCU
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.hardware,redhat.hardware.arch.intel
Subject: Re: I HATE LT WIN MODEM!!!!!!!!!!!!!!!
Date: Thu, 26 Aug 1999 18:57:25 GMT
On Sun, 22 Aug 1999 22:40:15 -0400, george <[EMAIL PROTECTED]> wrote:
Win modems are also known as controllerless modems. They rely on the
system CPU to run the phone - computer interface. Traditional modems
have an on board processor to handle this task. Depending on the
machine and the amount of data being transferred, the system can take
a significant performance hit when managing the modem. This is also
true for certian sound cards which also rely on the system processor
to do the work. Beware of inexpensive products. It likely doesn't
perform as well as it's more expensive competitors.
>Not to be rude, but Win modems do not work under Linux. Hence the name "Win"
>modem - they only work under Windows. This is because the hardware
>manufacturer was very cheep and decided that rather than add the extra
>components to make it a real modem, they would supply drivers - for windows
>only - to make the modem work.
>
>Sorry.
>
>liuyb wrote:
>
>> I have a lt win modem, but I can't let it work in linux. I tried all means
>> and spend a lot of time , the result is the same. So I hate it.
>> I HATE WIN MODEM!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
>
>--
>George Nimmer - [EMAIL PROTECTED]
>Linux - The choice of a GNU generation
>
>
>
------------------------------
From: Miles Bader <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: Jesus: the ultimate OS
Date: 27 Aug 1999 12:58:44 +0900
[EMAIL PROTECTED] (Peter Samuelson) writes:
> [Miles Bader <[EMAIL PROTECTED]>]
> > Love is a snowmobile racing across the tundra. Suddenly it flips over,
> > pinning you underneath. At night the ice weasels come. --Nietzsche
>
> Someone else's sig, somewhere, attributes this quote or one very like
> it to Matt Groening [of Simpsons fame]. Who's right?
It was attributed to Nietzsche in the original (Matt Groenig) cartoon
(the newspaper strip, `Life in Hell', not The Simpsons, BTW).
-Miles
--
Love is a snowmobile racing across the tundra. Suddenly it flips over,
pinning you underneath. At night the ice weasels come. --Nietzsche
------------------------------
From: [EMAIL PROTECTED] (Daniel Robert Franklin)
Subject: Re: kernel and -m486
Date: 27 Aug 99 03:42:11 GMT
[EMAIL PROTECTED] (Jacek =?iso-8859-2?Q?Pop=B3awski?=) writes:
> In every kernel I compile it's -m486 option. My question is - is it good
>to compile kernel with that option? Why not -m586 ? I thought, that 486
>optimisation for pentium is worse than no optimisation...
gcc 2.7.x, which is probably what you have, only had 386 and 486
optimisations. For gcc-2.95 (the latest) you should be able to specify
-march=pentium (I think). Check the documentation.
- Daniel
--
******************************************************************************
* Daniel Franklin - Postgraduate student in Electrical Engineering
* [EMAIL PROTECTED]
******************************************************************************
------------------------------
From: Andreas Jaeger <[EMAIL PROTECTED]>
Subject: Re: glibc-2.1.1 problems
Date: 27 Aug 1999 07:38:18 +0200
>>>>> Mike Dowling writes:
>> __setfpucw is an internal symbol of glibc, none of your binaries
>> should use it (see FAQ 3.12). Please check which binary/library wants
>> __setfpucw -- and why -- and rebuild those first.
Mike> Unfortunately, the vast majority of binaries want __setfpucw. I suppose
Mike> that this means that my gcc (egcs-1.1.2) is stuffed. It was compiled
Mike> with a binary release of egcc-1.0 I think.
Mike> Even helloworld.c wants __setfpucw! In short, every binary produced
Mike> with the egcs-1.1.2 compiler wants __setfpucw, and every binary produced
Mike> with the binary distribution of egcs-1.0 also. In fact, the only
Mike> binaries that don't requite __setfpucw are the file-utils compiled in
Mike> August last year! There might be more; probably X11 stuff is all right,
Mike> as that was also a binary XFree release.
If even helloworld.c wants __setfpucw, your environment is totally
broken. You should check which header/object file is the culprit.
Mike> Even assuming that I could find or compile a compiler that no longer
Mike> produces binaries that want __setfpucw, I would have to re-compile
Mike> almost every binary on the system.
Mike> I could do this over time, a very long time during which I would have to
Mike> continue to use glibc-2.0.6, but it would at least eventually solve the
Mike> problem. The only thing stopping me from doing that is that I have no
Mike> idea as to how I could obtain or compile a version of gcc that does not
Mike> exhibit the problem. I could perhaps use gcc-2.95.1.tar.gz instead, but
Mike> the FAQ advises egcs. Perhaps its worth a try anyway.
If you like to use gcc 2.95.1, wait for glibc 2.1.2.
Andreas
--
Andreas Jaeger [EMAIL PROTECTED] [EMAIL PROTECTED]
for pgp-key finger [EMAIL PROTECTED]
------------------------------
From: "Christian Weisz" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking,comp.os.linux.misc
Subject: Semaphore/Thread/Socket synchronization
Date: Fri, 27 Aug 1999 07:40:48 +0200
Hi all ,
i have the following design problem.
My application have to wait for some events, that are
Thread terminations ,
Socket traffic ,
Semaphores (which synchronizing a communication via shared memory)
I want to implement a method where all of the events are handled
sequentially in a democracy order. In case of socket (file descriptor based)
events only, I think I can use the select call. But how can I include
semaphores and threads in such a process ?
Suggestions are very appreciated.
Thanks in advance ,
===========================================================================
Christian Wei�
===========================================================================
Segue Softwareentwicklung Ges.m.b.H
Tel: ++43-732-336694-62
[EMAIL PROTECTED]
===========================================================================
------------------------------
From: "Scott Simpson" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking,comp.os.linux.misc
Subject: Re: Semaphore/Thread/Socket synchronization
Date: Thu, 26 Aug 1999 22:31:50 -0700
Christian Weisz <[EMAIL PROTECTED]> wrote in message
news:7q58cp$h0i$[EMAIL PROTECTED]...
> I want to implement a method where all of the events are handled
> sequentially in a democracy order. In case of socket (file descriptor
based)
> events only, I think I can use the select call. But how can I include
> semaphores and threads in such a process ?
You'll have to poll (if using a single thread). This is actually an area
(gasp) where I like Windows better. Windows has a single abstraction for
waiting for something: WaitForMultipleObjects. If any of the above are
ready, you use one call to find out which.
Scott
------------------------------
From: Ken Booth <[EMAIL PROTECTED]>
Subject: Re: where are packets created?
Date: Fri, 27 Aug 1999 01:24:04 -0500
One of the best references is:
TCP/IP Illustrated (2 volumes), by W. Richard Stevens and Gary R. Wright.
Pub. Addison Wesley.
Excellent books, but the pair is going to cost you a hundred bucks or so.
--
Regards, Ken
I AM.
Yung-Hsiang Lu wrote:
> Hi, Everyone,
>
> Where are network packets created? Is this protocol specific?
>
> When I ftp a large file, is it divided into small pieces (I guess so).
> Is this the responsibility of ftp or the device driver? Is there a
> study about optimal packet sizes? What are the typical packet sizes?
> How about http for a large gif file?
------------------------------
From: Brian Lepore <[EMAIL PROTECTED]>
Subject: bzImage, where is it?
Date: 27 Aug 1999 06:56:13 GMT
To all,
I tried to compile a kernel for the first time. it was too big using make xImage.
Then I tried used make bzImage, which seemed to work, but there was no image in the
directory. Any ideas?
also, apparently I don't know how to use my keyboard with tin
D
C
C
A
B
B
Thanks ahead of time for any replys
--
Brian Lepore
Research Assistant - Schwartz Lab, Biotech/Genetics
Work: 608-265-0625
fax: 608-262-6748
email: [EMAIL PROTECTED]
Home: 608-256-8343
------------------------------
From: Tristan Wibberley <[EMAIL PROTECTED]>
Subject: Re: Newbies Programming course
Date: Fri, 27 Aug 1999 14:15:39 +0100
Reply-To: [EMAIL PROTECTED]
Tor Harald Thorland wrote:
>
> Hello,
>
> I'm trying to learn some programming and wonder if someone can tell me
> which programming language too start with, and where to find some
> tutorials/documents for newbies.
> Where to find that specific program would be great too.
You want something pertaining to being new to linux internals? Or to
being new to programming in general. If the latter, don't ask here,
please - that's just silly.
There are many newsgroups in the comp heirarchy that would be far more
appropriate.
--
Tristan Wibberley
------------------------------
From: Fernando Ortega <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking
Subject: Struct SK_BUFF(Networking code)
Date: Fri, 27 Aug 1999 11:25:57 +0200
I need to know how the networking manages to get data from a process
and passes it to the sk_buff struct , so that, this is the struct used to
pass the data to the devices(for example to Ethernet).
I know the last call made , before passing the data to the device is:
dev->hard_start_xmit(skb,dev)
but which is the way data get into skb?
I mean,
How many packets for each skb? ----> It depends? , What about?
How many priority queue buffers per device? ----> 3?
So, as you can see I am a bit lost.
If you know any thing, that helps, please let me know.
Any help will be wellcome.
Thanks in advance
--
Fernando Ortega Bellosta
[EMAIL PROTECTED]
------------------------------
From: Peter Pointner <[EMAIL PROTECTED]>
Subject: Re: does linux have a package for drawing block diagrams ?
Date: Fri, 27 Aug 1999 05:32:18 GMT
Phil Howard <[EMAIL PROTECTED]> wrote:
> On Wed, 25 Aug 1999 23:59:04 GMT Peter Pointner ([EMAIL PROTECTED]) wrote:
> | ... and if not, you will start development of a kernel module that
> | does that?
> | Or why else do you post this question to comp.os.linux.development.system?
> Unfortunately, lots of people tend to think of everything that was installed
> when they installed the "system" as part of the "system". Which means they
> may not think of things like Netscape or The Gimp as applications. Only what
> is added on would be an application to them.
> Combine that with the somewhat misleading name of this group, and there is
> bound to be some level of confusion. And people already disagree over many
> things rightly classified as the gray area, such as libraries, loaders,
> sysinit scripts, etc. If the group was for the kernel only, it would have
> been called comp.os.linux.development.kernel, right?
> Certainly, a drawing or CAD type tool is an application. But be prepared for
> people that may not even realize that.
Thank you for your explanation. But my question was more rhetorical.
I thought the original is totally off topic (_looking_ for an _application_),
but OK: At least he is looking for a Linux application, and maybe he plans
to use that drawing tool for the development of a driver :-)
And probably my try to remind people of the topic of this group is
wasted bandwidth anyway.
Peter
------------------------------
From: Peter Pointner <[EMAIL PROTECTED]>
Subject: Re: strcmp in kernel
Date: Fri, 27 Aug 1999 05:06:59 GMT
Peter Samuelson <[EMAIL PROTECTED]> wrote:
> [[EMAIL PROTECTED]]
>> > I can't use strcmp in my device driver even if i include
>> > <linux/string.h>. The compiler says implicit declaratiion of
>> > function.
>> > How do i use strcmp a device driver?
> [Peter Pointner <[EMAIL PROTECTED]>]
>> strcmp might be defined as inline function. Use at least gcc -O when
>> compiling your driver to make the inline functions work.
> Yes, you do need -O or -O2 due to inlines, but if that were the problem
> it wouldn't show up until link time (i.e. insmod time), I believe.
Thank you for correction. I always think people should read the question
before they answer. Apparently 4 lines were to much for me.
Sorry,
Peter
[better answer snipped]
------------------------------
From: [EMAIL PROTECTED] (Dr. Peer Griebel)
Subject: CDROM: Illegal Request
Date: Fri, 27 Aug 1999 09:43:41 GMT
Hi,
I just got my new notebook. And I have problems accessing my cdrom
drive from linux-2.2.10
Linux says it is a UJDA150 ATAPI CDROM.
Whenever I try to acces it I get the following message in the log
file:
Aug 27 13:22:54 Gaukler kernel: hdc: ATAPI 24X CD-ROM drive, 128kB
Cache
Aug 27 13:22:54 Gaukler kernel: Uniform CDROM driver Revision: 2.55
Aug 27 13:22:54 Gaukler kernel: hdc: packet command error: status=0x51
{ DriveReady SeekComplete Error }
Aug 27 13:22:54 Gaukler kernel: hdc: packet command error: error=0x50
Aug 27 13:22:54 Gaukler kernel: ATAPI device hdc:
Aug 27 13:22:54 Gaukler kernel: Error: Illegal request -- (Sense
key=0x05)
Aug 27 13:22:54 Gaukler kernel: Invalid field in command packet --
(asc=0x24, ascq=0x00)
Aug 27 13:22:54 Gaukler kernel: The failed "Start Stop Unit" packet
command was:
Aug 27 13:22:54 Gaukler kernel: "1b 00 00 00 03 00 00 00 00 00 00 00
"
Aug 27 13:22:54 Gaukler kernel: Error in command packet byte 254 bit
0
Aug 27 13:22:54 Gaukler kernel: cdrom: open failed.
What's going on here? On all the other machines I installed already
Linux I didn't see this problem.
Thank you for your help.
------------------------------
From: Tristan Wibberley <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.misc
Subject: Re: why not C++?
Date: Fri, 27 Aug 1999 14:43:34 +0100
Reply-To: [EMAIL PROTECTED]
Phil Hunt wrote:
>
> In article <[EMAIL PROTECTED]>
> [EMAIL PROTECTED] "Tristan Wibberley" writes:
> > Johan Kullstam wrote:
> > > one thing that bothers me in C++ is the lack of additional operations.
> > > for example, in matrix math, you might want two kinds of
> > > multiplication. element-wise or as a matrix. matlab uses .* and *
> > > respectively. C++ doesn't offer any new operators. if C++ had a few
> > > unnassigned operations, you could overload them with perhaps less
> > > confusion than redefining something as basic as addition.
> >
> > A general operator syntax, now that would be nice :)
> >
> >
> > C = A `x B; // cross product
> > C = A `. B; // dot product
> >
> > or something like that - What trigraph would you use for that symbol,
> > \\' maybe.
> >
> > C = A \\'x B;
> > C = A \\'. B;
>
> You've got the source to GCC -- perhaps you might like to add it.
:) The problem is the syntax for declaring the operators. Then what
about:
D = A`xBC; is this D = A `x BC; or D = A `xB C;
Perhaps we use the shortest token we can like in the rest of C++, but
does this then create unexpected behaviour (non-obvious bugs - there are
enough causes in C++ already, or would one more not matter :).
A nice OO syntax for declaration and definition would be appropriate for
a feature like this, but C++ is not OO enough to do this nicely (IMHO).
class A {
A &operator`x (A &a, A &b);
A &operator`xb (A &a, A &b);
};
so d = a`xbc; tries to use operator`x with arguments a and bc, whereas d
= a`xb c; uses operator`xb with arguments a and c.
I can think of a more elegent OO way - but this is the C++ way. With
trigraphs it is:
class A {
A &operator??.x (A &a, A &b);
A &operator??.xb (A &a, A &b);
};
d = a??.xbc;
d = a??.xb c;
This all really looks quite nasty, which is why it was humour and not a
serious suggestion.
> BTW, has anyone read Stroustrup's paper where he suggests overloading
> the whitespace operator? It's quite an elegant idea, for example
> mathematicians would be able to write:
>
> v = a x + b y + c z;
>
> instead of the usual:
>
> v = a * x + b * y + c * z;
Aren't we getting close the realms of forth now?
So we are looking at treating everything as operators where they can be
token boundaries, statement boundaries, or functors.
The parsing of this sort of thing will be a computational nightmare :/
--
Tristan Wibberley
------------------------------
From: [EMAIL PROTECTED] (Mike Dowling)
Subject: Re: glibc-2.1.1 problems
Date: 27 Aug 1999 09:50:55 GMT
On 27 Aug 1999 07:38:18 +0200, Andreas Jaeger <[EMAIL PROTECTED]> wrote:
>>>>>> Mike Dowling writes:
>
>>> __setfpucw is an internal symbol of glibc, none of your binaries
>>> should use it (see FAQ 3.12). Please check which binary/library wants
>>> __setfpucw -- and why -- and rebuild those first.
>
>Mike> Unfortunately, the vast majority of binaries want __setfpucw. I suppose
>Mike> that this means that my gcc (egcs-1.1.2) is stuffed. It was compiled
>Mike> with a binary release of egcc-1.0 I think.
>
>Mike> Even helloworld.c wants __setfpucw!
>If even helloworld.c wants __setfpucw, your environment is totally
>broken. You should check which header/object file is the culprit.
Thanks for all your help!
Just one last question:
Where exactly is the problem likely to lie? My first guess was my gcc,
but you suggest that the problem is likely to lie with headers and
object files. If the problem lies with the headers, then surely, only
gcc headers, glibc headers, and kernel headers could have a bearing on
the problem. If it's .o files, then it seems to me that, again, only
those belonging to gcc and glibc could have a bearing.
I ask this in the hope of narrowing the search.
Cheers,
Mike Dowling
--
My email address [EMAIL PROTECTED] above is a valid email address.
It is, in fact, a sendmail alias; the digit 'N' is incremented regularly.
Spammed aliases will be deleted. Currently, mike[5,7-9,12,13,16] have been
deleted. If email to mikeN bounces, try mikeN+1.
------------------------------
** 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
******************************
