Linux-Development-Sys Digest #418, Volume #6 Sun, 21 Feb 99 02:14:41 EST
Contents:
How to I make a Screenshot in X. ("Robert Acklin")
Make extra $$$ by using of your pc at home! ("Ben Tan")
Re: memcpy from process to process in module (Sam Roberts)
lwp library for linux (Moti Gindi)
Re: Linking externals w/ ld on GNU/Linux (Kip Rugger)
Re: Really slow tar (Siegfried Pohl)
Re: memcpy from process to process in module (Sam Roberts)
Re: tpconfig and gps/X (Henry)
Re: Linux programming jobs? (John S. Dyson)
Re: 2.2/Red Hat 5.2 upgrades disaster! (Peter Samuelson)
Re: memcpy from process to process in module (Sam Roberts)
----------------------------------------------------------------------------
From: "Robert Acklin" <[EMAIL PROTECTED]>
Subject: How to I make a Screenshot in X.
Date: Sat, 20 Feb 1999 20:30:43 -0500
How to I make a Screenshot in X.
/*----------------oOOo--(- *)--Oooo------------------*/
while (( Win95 > Msdos ) && ( Linux != FreeBSD ))
if (( Asm < Pascal ) && ( C++ > C ))
email("[EMAIL PROTECTED]");
else
email("[EMAIL PROTECTED]");
/*---------------------------------------------------*/
------------------------------
From: "Ben Tan" <[EMAIL PROTECTED]>
Subject: Make extra $$$ by using of your pc at home!
Date: Sun, 21 Feb 1999 09:34:34 +0800
Make extra $$$ by using of your pc at home!
Work from home with your PC distributing software and
information all over the world via E-mail and the Internet.
Our exclusive software PC-SUPER-MLM will give you all the
information and tools you need to start your own Computer
Based Business as a Software Distributor.
Make thousands of dollars doing simple and pleasant work
from home with your computer using our proven method.
Reply to this article and I will send you a FREE copy of
the PC-SUPER-MLM software for your evaluation, which runs
under DOS or Windows.
The software will be sent to you via E-mail as an
attachment. (147kb)
Happy Computing
------------------------------
From: Sam Roberts <[EMAIL PROTECTED]>
Subject: Re: memcpy from process to process in module
Date: 19 Feb 1999 18:51:04 -0500
[EMAIL PROTECTED] (Dave Platt) writes:
> > I'm writing a kernel module that implements Send/Receive/Reply
> > message passing similar to that of the QNX Real-Time O/S.
>
> Another option is to set up a shared-memory pool which is mapped into
> both (or all) processes' memory spaces. This can be done using the
> System V shmem features, or by using mmap() on a shared file.
> Construct the messages in the shared memory area (using either
> user-mode spinlocks or kernel-mode locks, or both, to avoid
> contention), and use signals, or I/O to a trivially-small driver, as a
> way of informing the other processes that messages are ready.
>
> Do this right, and you may be able to go down from a two-memcopy
> approach (user1->kernel, kernel->user2), bypassing entirely the QNX
> approach (direct user1->user2), and go to a zero-copy message passing
> system.
Sorry, I should also say that existing code depends on sending
arbitrary data, located anywhere, so while S/R/R could be implimented
with shared memory, it would still involve two copies in the common
case, once into the shared memory, and once out of it, the same
as doing it as a kernel module. Otherwise, yes, this could be done.
Sam
> Dave Platt [EMAIL PROTECTED]
--
Sam Roberts ([EMAIL PROTECTED]), Cogent Real-Time Systems (www.cogent.ca)
"News is very popular among its readers." - RFC 977 (NNTP)
------------------------------
From: Moti Gindi <[EMAIL PROTECTED]>
Subject: lwp library for linux
Date: Sat, 20 Feb 1999 00:52:28 +0200
I am searching for a library for light-weight-processes (threads)
called "LWP". I know of a version of this library for sun-os machines,
but I cant't find any files for linux.
Does anyone know where I can find it?
Thanks, moti.
------------------------------
From: [EMAIL PROTECTED] (Kip Rugger)
Subject: Re: Linking externals w/ ld on GNU/Linux
Date: 19 Feb 1999 17:54:07 -0600
In article <[EMAIL PROTECTED]>,
Brent Anderson <[EMAIL PROTECTED]> wrote:
>I have a GNU ld relocatable application (filesys.o) from which I need
>to intercept all the "stat ()" OS calls. I need to link my "stat ()"
>calls to the normal Linux OS library. Then I need to link all the
>"stat" calls in this relocatable application to the "stat" function in
>my module.
If you want to override stat, you should know two things:
-- libc defines the "real stat" as __stat to avoid namespace pollution
-- it also defines "stat" as a weak alias for __stat
This means that you are free to define a "stat" routine yourself: the
weak alias stat defined by libc won't be used in this case because your
symbol exists. The stat that you define can, of course, call __stat to
get the normal libc function.
Linking -r does not let you do a pairwise resolution of external symbols.
You might use -r for some other reason (eg. packaging), but the resolution
of symbols is the same whether you link piecemeal or all at once.
Warning: old versions of libc may not have the weak aliases
done correctly. I am using glibc-2.0.7 which runs the following
program successfully.
#include <sys/stat.h>
int stat(const char * file, struct stat * buf)
{
write(1, "Hi\n", 3);
return __stat(file, buf);
}
int main()
{
struct stat buf;
stat("/etc/passwd", &buf);
printf("%d\n", buf.st_ino);
exit(0);
}
------------------------------
From: [EMAIL PROTECTED] (Siegfried Pohl)
Crossposted-To: comp.os.linux.misc
Subject: Re: Really slow tar
Date: 19 Feb 1999 21:37:04 GMT
Jayasuthan ([EMAIL PROTECTED]) wrote:
> I did something ... dangerous ..
> try some setting on hdparm on you hard disk... It works for me.
> man hdparm
Interesting! I've encountered the same problem at mine since I changed my dis
from SuSe to RH5.0 about a year ago or so.
In the man page about hdparm, they say hdparm is just for IDE Drives, mine
are both SCSI, I dont have any IDE stuff at all, so it cannot be something
to do with IDE drivers in the kernel?!
MfG, Siggi
BTW: I downgraded to tar Version somethingreallyold but there was the same
prob...
> Daren Scot Wilson wrote:
> >
> > tar runs very slowly on my machine. This started about a month ago, after
> > reloading linux (RH5.0) and upgrading the kernel, compilers, etc. Everything
> > else runs normal. There are no symptoms other than that tar runs slow. It's
> > not unzipping - I can run gunzip in a few seconds, then tar -xzvf takes
> > forever. The -v option shows one to maybe six files flying by, then many
> > seconds paused, then one to a few files, the another pause, etc.
> >
> > How slow? A tarball several meg in size used to untar in just a minute or two
> > or three, depending on CPU load. Now i have time to eat lunch, see a movie,
> > see another movie... Linux kernel 2.2.1 took all hours overnight to untar.
> >
> > Once the stuff is untarred, it's good. I'm running a 2.2.1 kernel, 2.1 glibc,
> > and the latest xterm, bash, netscape all having been slowly untarred.
> >
> > Tar, and nearly everything else, was compiled with gcc 2.8.1, using glibc 2.0.6
> > and since today, glibc 2.1.
> >
> > Computer is a PII-400, 96RAM, 6G HD nowhere near full, 100M swap. Plenty of
> > raw material - nothing else runs anomalously slow.
> >
> > Any good diagnostics suggestions? Fixes?
> >
> > --
> > Daren Scot Wilson
> > [EMAIL PROTECTED]
> > www.newcolor.com
> > ----
> > "A ship in a harbor is safe, but that is not what ships are built for"
> > -- William Shedd
--
=====================================================================
Siegfried Pohl
http://www.in-berlin.de/User/czbbs
[EMAIL PROTECTED]
Tel: +49 3381 718056
Fax: +49 3381 718058
> Only wimps use tape backups; real men put their software on <
> ftp-servers and let the rest of the world mirror it. <
> (C) Linus Torvalds <
=====================================================================
------------------------------
From: Sam Roberts <[EMAIL PROTECTED]>
Subject: Re: memcpy from process to process in module
Date: 19 Feb 1999 18:37:13 -0500
Emile van Bergen <[EMAIL PROTECTED]> writes:
> Sam Roberts wrote:
> >
> The following issues come to mind:
>
> * What if a Send() of a large block gets interrupted/preempted? Let the
> app test for this and restart the operation manually (as in qnx<4.25)?
> That wouldn't be nice. I guess you should create a kernel thread for
> each transfer operation to solve this.
I'm wasn't aware that Linux has kernel threads. Does it?
I don't think a system call under linux is going to be preempted. It
should either go to a sleep_on_interruptible(), or until return,
at which point rescheduling may occur. Please correct me if I
misunderstand.
> * The swapping problem mentioned yourself. This could be solved without
> locking BOTH the sender and the receiver in core, because if you
> implement S/R/R using kernel threads, they may block on something (i.e.
> page-ins) while keeping the callers blocked until the transfer is
> complete.
>
> BTW, Memory suffers from a scheme like this because BOTH processes have
> to be in core at the same time, whereas a copy-via-kernel space may have
> the sending process swapped out after it has sent its data while
> swapping in the receiving process...
>
[snip...]
Not being familiar with Linux kernel threads, I don't know if I understand
this. However, I see nothing but good results coming from locking
my (soft) real-time applications into core. If X swaps, fine, but
not my controller!
Thanks for your comments.
Sam
> M.vr.gr. / Best regards,
>
> Emile van Bergen (e-mail address: [EMAIL PROTECTED])
>
--
Sam Roberts ([EMAIL PROTECTED]), Cogent Real-Time Systems (www.cogent.ca)
"News is very popular among its readers." - RFC 977 (NNTP)
------------------------------
From: Henry <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.hardware,comp.os.linux.development.apps,comp.os.linux.misc
Subject: Re: tpconfig and gps/X
Date: Fri, 19 Feb 1999 20:48:20 -0500
I just submitted a change to GPM to support the Synaptics touchpad
(serial and PS/2). It is in release 1.17.5 (ftp.prosa.it/pub/gpm --
just out today). I did this because the tpconfig program couldn't
perform the changes which I wanted against my touchpad. It appears that
the newer touchpads have fewer options and the options are now handled
by the driver. Also I use my touchpad through the serial port and use a
PS/2 mouse at the same time (gives me options for when a touchpad
doesn't seem quite the right option).
Also, I wanted the cursor to continue when I hit the edge of the
touchpad. This wasn't an option to my touchpad and only worked on drag.
This change to GPM supports many of the things which the touchpads used
to do internally. This has the advantage of configuring GPM as you want
(it reads a configuration file to setup options). The configuration can
only be changed by changing the config file and restarting GPM.
Just thought you might want to know.
Henry
Bruce Kall wrote:
>
> I currently support the tpconfig utility (Synaptics
> Linux touchpad driver -- www.compass.com/synaptics). Anyway
> I just released the latest version to fix a problem with
> the 2.2.X kernel. The next enhancement I would like to
> add is to make it work while either gpm and/or X is
> running. From what I know from the initial author of
> tpconfig, the PS/2 port does not support sharing and
> this is the reason why tpconfig hasn't support running
> itself while gpm and/or X is running.
>
> What I'm looking for is suggestions/pointers on how to
> add this functionality to tpconfig (run on top of X).
> Any suggestions would be welcome.
>
> Thanks,
> Bruce
>
> --
> ====================================================================
> Bruce Kall
> Mayo Foundation
> Rochester, MN
>
> e-mail: [EMAIL PROTECTED]
> phone: (507) 255-4768
>
> ====================================================================
------------------------------
From: [EMAIL PROTECTED] (John S. Dyson)
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.networking
Subject: Re: Linux programming jobs?
Date: 21 Feb 1999 03:20:58 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Rich) writes:
> On 19 Feb 1999 11:04:48 -0800, David T. Blake <[EMAIL PROTECTED]> wrote:
>>A decent contract programmer who actually finishes jobs
>>in a respectable time frame makes $50-90/hr, depending
>>on the reputation of the programmer and the sort of work.
>>
>
> Yes, but he also gets to pay $1200.00/month for an apartment
> that you couldn't give away for free in other areas of the
> country. :-) ( Unless you want a 2-hour commute to work... )
>
> When people quote salaries, you always have to find out the
> cost of living in the area, if you want to know how much you
> will REALLY end up putting in the bank.
>
I happen to live 2000mi from work :-).
--
John | Never try to teach a pig to sing,
[EMAIL PROTECTED] | it makes one look stupid
[EMAIL PROTECTED] | and it irritates the pig.
------------------------------
From: [EMAIL PROTECTED] (Peter Samuelson)
Subject: Re: 2.2/Red Hat 5.2 upgrades disaster!
Date: 19 Feb 1999 20:21:35 -0600
Reply-To: Peter Samuelson <[EMAIL PROTECTED]>
[Calvin Mitchell <[EMAIL PROTECTED]>]
> I've had difficulties running a 2.0.36 kernel after having built gcc
> 2.8.1 (modprobe/insmod problems). I even applied the recommended
> patch and upgraded the modutils...
What recommended patch? There is *no* officially recommended patch to
linux-2.0.* to make it run with gcc 2.8.x or egcs. Alan Cox (principle
maintainer of the 2.0.x branch) does not even want to hear about
problems you experience when you compile a 2.0 kernel with anything but
gcc 2.7.x. That is why he hasn't fixed ioport.c, even though he knows
very well it is broken. He doesn't want anyone thinking it is now safe
to compile 2.0 with new compilers, because then they would start
getting bitten by the subtler problems.
> is using gcc 2.8.1 that big of a nightmare???
The general consensus is that if you are willing to risk the
instability of compiling 2.0.x with a non-gcc-2.7, why not just risk
using a 2.2 kernel and compile it with anything you want? The other
general consensus (outside the FSF) is that gcc is buggier and has
fewer features than egcs.
--
Peter Samuelson
<sampo.creighton.edu!psamuels>
------------------------------
From: Sam Roberts <[EMAIL PROTECTED]>
Subject: Re: memcpy from process to process in module
Date: 19 Feb 1999 18:48:12 -0500
[EMAIL PROTECTED] (Dave Platt) writes:
> > I'm writing a kernel module that implements Send/Receive/Reply
> > message passing similar to that of the QNX Real-Time O/S.
>
> Another option is to set up a shared-memory pool which is mapped into
> both (or all) processes' memory spaces. This can be done using the
>
> Do this right, and you may be able to go down from a two-memcopy
> approach (user1->kernel, kernel->user2), bypassing entirely the QNX
> approach (direct user1->user2), and go to a zero-copy message passing
> system.
A few people have mentioned this. I should point out that we have
a lot of code that relies on the blocking, synchronous message passing
semantics of S/R/R, which is the main reason I'm implimenting. The
QNX applications I have worked on that use shared memory and semaphores
would port more trivially.
I like the 1-copy S/R/R because it is easier to use than shared
memory and semaphores:
char reply[50];
Send(pid, "hello", reply, strlen("hello"), 50);
is definitely easier then setting up the shared memory region and the
semaphore. Of course, its not as fast as a 0-copy, put the data in
shared memory where everybody can see it and hope everybody uses
the semaphores correctly approach. Nothing beats shared memory for
fast IPC.
Anyhow, there's lots of interesting IPC mechanisms, but I'm working
on getting Send/Receive/Reply working!
So the question remains, can I lock processes or pages down in memory,
and access them both?
This facility is specified by POSIX and SVR4 as the mlockall(), and
munlockall() system calls, so it would be worth having under Linux.
Thanks for your comments.
Sam
> --
> Dave Platt [EMAIL PROTECTED]
--
Sam Roberts ([EMAIL PROTECTED]), Cogent Real-Time Systems (www.cogent.ca)
"News is very popular among its readers." - RFC 977 (NNTP)
------------------------------
** 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
******************************
