Linux-Development-Sys Digest #398, Volume #6 Sat, 13 Feb 99 05:16:10 EST
Contents:
Re: Shared library programming (Stephen Gennard)
Re: Help: Debug buggy dma in kernel module? (Tony Hoyle)
Re: Glibc compiled but fails test (Andreas Jaeger)
Re: Modest next goal for Linux (Bill Anderson)
Re: linux-2.2.1 swapper oops on Alpha (Martin Kahlert)
SMP Support (Bucky4me)
Re: cobalt and linux (Bob Tennent)
Re: glibc 2.1 ;) (Paul D. Smith)
Re: LYNX Benchmarking Commands (Rohit Sharma)
Re: SMP Support (Rob Komar)
/usr/bin/ld: warning: cannot find entry symbol _start; ("Bob Harmon - Eng.")
Re: glibc 2.1 ;) (Andreas Jaeger)
Re: LYNX Benchmarking Commands (David T. Blake)
Re: Q]Linux device driver ([EMAIL PROTECTED])
Re: Will 2.2.x support removable medias better? (mkk)
----------------------------------------------------------------------------
From: Stephen Gennard <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Shared library programming
Date: Wed, 10 Feb 1999 16:23:32 +0000
Hi,
For dllMain() functionality on UNIX use: init() and fini().
For thread startup notification then you will have to keep a first
time list using pthread_self() as the thread id, for thread
termination you push a cleanup handler eg: pthread_cleanup_push.
Hope this helps.
Karl Heyes wrote:
>
> XuYifeng wrote:
>
> > Hi,
> >
> > I am relate new to linux programming. currently I
> > want to build a shared library, problem is the library
> > need automatic init and finish function like DllMain
> > function in WindozeNT which will be invoked at library
> > loaded or freed ?
> >
> > Is there a shared library programming guide avail?
> >
> > any help will be appreciated,
> >
> > XuYifeng
>
> If it's C++ lib just have a dummy class which has it's constructor
> invoked. Not sure about just a C library.
>
> karl
--
Stephen Gennard, Micro Focus Ltd, Newbury, Berkshire
Unix Runtime Development, N.O.W. Products Group, Email: [EMAIL PROTECTED]
------------------------------
From: Tony Hoyle <[EMAIL PROTECTED]>
Subject: Re: Help: Debug buggy dma in kernel module?
Date: Thu, 11 Feb 1999 18:57:18 +0000
"Albrecht Dreß" wrote:
>
> Hi all,
> o Where should I look in the kernel sources (= insert some debug cmnds) to
> find the source of the error?
Have you tried running ksymoops? That should at least tell you which
routine it
is crashing in.
Tony
====================================================================================
If Bill Gates had a dime for every time a Windows box crashed...
.... Oh, wait a minute, he already does.
====================================================================================
[EMAIL PROTECTED] http://falklands.magenta-logic.com
====================================================================================
------------------------------
From: Andreas Jaeger <[EMAIL PROTECTED]>
Subject: Re: Glibc compiled but fails test
Date: 11 Feb 1999 23:06:56 +0100
>>>>> Juergen Koslowski writes:
> Hi there,
> Glibc-2.1 seemed to compile fine with egcs-1.1.1 (not the latest
> snapshots, which also produced buggy kernels), however, during
> "make check" I got the following problem in the nss directory:
Please read the glibc 2.1 FAQ entry about `make check' failing.
> make -C nss tests
> make[1]: Entering directory `/root/view/src/glibc-2.1/nss'
> gcc test-netdb.c -c -O -Wall -Winline -Wstrict-prototypes -Wwrite-strings -g
>-I../include -I. -I.. -I../libio -I../sysdeps/i386/elf -I../crypt/sysdeps/unix
>-I../linuxthreads/sysdeps/unix/sysv/linux -I../linuxthreads/sysdeps/pthread
>-I../linuxthreads/sysdeps/unix/sysv -I../linuxthreads/sysdeps/unix
>-I../linuxthreads/sysdeps/i386 -I../linuxthreads/sysdeps/pthread/no-cmpxchg
>-I../sysdeps/unix/sysv/linux/i386 -I../sysdeps/unix/sysv/linux -I../sysdeps/gnu
>-I../sysdeps/unix/common -I../sysdeps/unix/mman -
> I../sysdeps/unix/inet -I../sysdeps/unix/sysv/i386 -I../sysdeps/unix/sysv
>-I../sysdeps/unix/i386 -I../sysdeps/unix -I../sysdeps/posix -I../sysdeps/i386/i586
>-I../sysdeps/i386/i486 -I../sysdeps/i386/fpu -I../sysdeps/libm-i387 -I../sysdeps/i386
>-I../sysdeps/wordsize-32 -I../sysdeps/ieee754 -I../sysdeps/libm-ieee754
>-I../sysdeps/generic/elf -I../sysdeps/generic -D_LIBC_REENTRANT -include
>../include/libc-symbols.h -o test-netdb.o
> gcc -nostdlib -nostartfiles -o test-netdb
>-Wl,-dynamic-linker=/usr/local/lib/ld-linux.so.2 ../csu/crt1.o ../csu/crti.o `gcc
>--print-file-name=crtbegin.o` test-netdb.o
>-Wl,-rpath-link=..:../math:../elf:.:../nis:../db2:../rt:../resolv:../linuxthreads
>../libc.so.6 ../libc_nonshared.a -lgcc `gcc --print-file-name=crtend.o` ../csu/crtn.o
> ../elf/ld-linux.so.2 --library-path
>..:../math:../elf:.:../nis:../db2:../rt:../resolv:../linuxthreads ./test-netdb >
>test-netdb.out
> The program was just hanging, and upon pressing CTR-C produced the
> following output:
> make[1]: *** Deleting file `test-netdb.out'
> make[1]: *** wait: No child processes. Stop.
> make[1]: *** Waiting for unfinished jobs....
> make[1]: *** wait: No child processes. Stop.
> make: *** [nss/tests] Error 2
> While hanging, I got the following from ps:
> 15492 p1 S 0:02 make -C nss tests
> 15502 p1 S 0:00 /bin/sh -c ../elf/ld-linux.so.2 --library-path ..:../math:..
> 15503 p1 S 0:00 ../elf/ld-linux.so.2 --library-path ..:../math:../elf:.:../n
Please call the program by hand and send me the results:
cd netdb;
../elf/ld-linux.so.2 --library-path
..:../math:../elf:.:../nis:../db2:../rt:../resolv:../linuxthreads ./test-netdb
> Any ideas where to look for a misconfiguration? This is a stand-alone
> machine not connected to the net, but with a network card. I didn't
> see an option to disable nss during configuration, but there may be
> such a possibility. --disable-nls wasn't listed either, but didn't
> produce complaints.
Have a look at nss/test-netdb.c. The tests are rather basic and if
they fail something is really broken. Please analyse the output (see
above), and look at the sources.
Andreas
--
Andreas Jaeger [EMAIL PROTECTED] [EMAIL PROTECTED]
for pgp-key finger [EMAIL PROTECTED]
------------------------------
From: Bill Anderson <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.advocacy
Subject: Re: Modest next goal for Linux
Date: Thu, 11 Feb 1999 19:45:23 +0000
Earl Malmrose wrote:
>
> [EMAIL PROTECTED] wrote in message ...
> >I dunno...my NT installation here at work takes 33 Mb just "resting".
> >
> >Of course, that's virtual memory, not physical; one could probably run
> >NT in 20 megs, but one wouldn't enjoy it very much ... swap swap ...
> >swap swap ...
>
> I've booted it in 8 megs. I was also able to open up Word, Excel, Paint, and
> IE 4. I won't say anything about it's speed. :)
Of course, if you are talking about 4.0, you would have *installed* with
more than 8, then removed RAM until you get down to 8MB after the
install was complete. You may be able to install 3.51, tnen do an
'upgrade' to 4.0, never tried it. The NT installer checks for a minimum
ram (don't recall if it is 12 or 16, but it *is* one of the two) before
installing. IIRC, the registry alone takes 8MB RAM.
Not saying you didn't do it, just calrifying some things for people who
may read your post and think they can take a machine w/8MB RAM, install
and have a truly painful experience. ;^) ( I have done it on 24MB, OS
only ... p75 ... ooh, the painful memories!)
Bill
------------------------------
From: [EMAIL PROTECTED] (Martin Kahlert)
Crossposted-To: comp.os.linux.alpha
Subject: Re: linux-2.2.1 swapper oops on Alpha
Date: Thu, 11 Feb 1999 21:05:53 +0100
Reply-To: [EMAIL PROTECTED]
In article <[EMAIL PROTECTED]>,
Pierre Asselin <[EMAIL PROTECTED]> writes:
> [EMAIL PROTECTED] wrote:
>>
>> starting kswapd v.1.5
>> parport0: PC-style at 0x378 [SPP,PS2,EPP]
>> Unable to handle kernel paging request at virtual address 097ffce0
>> swapper(1): Oops0
>>
>> and then a lot of hex values and a register dump.
>
> This looks similar to a bug described at
> http://www.alphalinux.org/faq/FAQ-12.html .
> But this was for RedHat 5.1 ??
I have a RedHat 5.2 system and 2.0.36 runs extremly stable.
Only kernel 2.2.1 continues giving this Oops. In contrast to
the above mentioned webpage it is very reproducable.
> I also have a 256MB Ruffian, and 2.2.1 runs fine on it. I had
> to copy the kernel to a floppy and boot from there because of
> MILO confusion, but other than that everything seems OK.
Did you support any kernel boot parameters?
Do you have compiled in a Ramdisk?
Thanks for your reply.
Martin.
--
Your mouse has moved, Windows must be restarted for changes
to take affect - restart now?
------------------------------
From: [EMAIL PROTECTED] (Bucky4me)
Subject: SMP Support
Date: 11 Feb 1999 20:45:55 GMT
I recompiled my kernel for smp but do not know how to check if the OS see's
both. How can I do this?
[EMAIL PROTECTED]
------------------------------
From: r d t@c s.q u e e n s u.c a (Bob Tennent)
Subject: Re: cobalt and linux
Date: 11 Feb 1999 20:44:15 GMT
On 11 Feb 1999 19:59:22 GMT, AME wrote:
>
>
>I have a cobalt box running Linux. However when I type commands like man
>or gcc , the system says"command not found". Does this mean there
>programs ae not installed? How do I know what programs are installed and
>what commands I can run? Where do I install such programs such that I can
>access them from every directory just like the case witl ls and cd
>commands?
>
In Unix-like systems, executables are normally stored in bin directories:
/bin, /usr/bin, /usr/local/bin and so on. These directory names are
listed in the PATH environment variable and the shell looks in
these directories for any command you want to execute (that isn't
a shell command). To see which directories are on your path, do echo $PATH
To add a directory to the path, see the man page for your shell to
see how to assign to environment variables.
Bob T.
------------------------------
From: [EMAIL PROTECTED] (Paul D. Smith)
Subject: Re: glibc 2.1 ;)
Date: 13 Feb 1999 00:47:52 -0500
Reply-To: [EMAIL PROTECTED]
%% [EMAIL PROTECTED] (Nathan Paul Simons) writes:
nps> Now, in case you haven't heard, glibc 2.1 has been pulled due to
nps> licensing problems and won't be available for a while.
This is completely wrong. I don't know where this rumor got started,
but the reason glibc 2.1 was pulled from ftp.gnu.org has absolutely
nothing to do with any licensing.
The real reason is even... stranger...
--
===============================================================================
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: Rohit Sharma <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.lynx,comp.os.linux.networking,comp.arch.bus.vmebus,comp.dcom.lans.ethernet
Subject: Re: LYNX Benchmarking Commands
Date: Thu, 11 Feb 1999 15:10:33 -0600
Shark wrote:
> All,
>
> I am new to the Lynx OS.
>
> I am going to be implementing and bench marking a Moto MPC8260
> PowerQUICCII microprocessor, VME bus, and 100BaseTX running Lynx OS.
>
> Can someone tell if there are commands that can benchmark CPU
> utilization, all types of I/O, MIPS, and anything else related to bench
> marking?
>
> Is there free source code that can be compiled?
>
> Thanks in advance.
>
> Shark
You could write your own code for benchmarking. When I do comaprisons for
RTOSs, these are the things I look at:
(1) Task or process switch latency
(2) Function context switch time
(3) Interrupt latency (time to start the isr after receiving the interupt)
Other than these, you could also compare the level of optimizations in the
compilers and the facilities offered by the debuggers. These are important
issues too in evaluation. Alongwith customer support and responsiveness
and knowledge of tech support :-)
Get in touch with me and I will give you pointers to some real good stuff
in RTOS benchmarking. You can use my code too. BTW, I am doing the same
thing right now - I am evaluating LynxOS for PowerPC 604e and 750.
I would be interested in knowing your results too.
Thanks !
Rohit
--
+----------------------------------+----------------------------------------+
| ROHIT SHARMA | Software Engineer |
| Email : [EMAIL PROTECTED] | Strategic Features |
| X.400 : [EMAIL PROTECTED] | (ADVANCED PACKET DUPLICATOR project) |
| Phone : (847) 576 - 9864 | iDEN Technology Division |
| Fax : (847) 576 - 6150 | IL02 / Mail Location 3127 |
| Pager : (847) 576 - 0295 #21442 | Motorola CNS Sector |
+----------------------------------+ 1301 E. Algonquin Rd. |
| /|_|\ /|_|\ | Schaumburg, IL 60196 |
| |/ \| |/ \| --- | |
+----------------------------------+----------------------------------------+
------------------------------
From: [EMAIL PROTECTED] (Rob Komar)
Subject: Re: SMP Support
Date: 11 Feb 1999 22:00:40 GMT
Bucky4me ([EMAIL PROTECTED]) wrote:
: I recompiled my kernel for smp but do not know how to check if the OS see's
: both. How can I do this?
cat /proc/cpuinfo
should show two CPUs. Also, the start of the boot-up messages show
if both are recognized (they're probably stored in /var/log/messages
if `dmesg' doesn't show them).
Cheers,
Rob Komar
------------------------------
Date: Thu, 11 Feb 1999 17:28:12 -0500
From: "Bob Harmon - Eng." <[EMAIL PROTECTED]>
Subject: /usr/bin/ld: warning: cannot find entry symbol _start;
I compile and link my (fairly large 1.5M) application but the linker
complains:
/usr/bin/ld: warning: cannot find entry symbol _start; defaulting to
08049620
Then when I type ./mate I get
./mate: Command not found
I checked that . is in my path too!
There is a 'main':
I tried adding /usr/lib/crt1.o to the start of ld line, then it links OK
but still gets
./mate: Command not found
When I run it.
Any help out there?
--
Robert B. Harmon ! [EMAIL PROTECTED]
Osicom Technologies !
9020 Junction Dr ! Any opinions expressed aren't mine...
Annapolis Jct. MD 20701 ! ...flame Don ([EMAIL PROTECTED])
------------------------------
From: Andreas Jaeger <[EMAIL PROTECTED]>
Subject: Re: glibc 2.1 ;)
Date: 13 Feb 1999 08:02:49 +0100
>>>>> Nathan Paul Simons writes:
> This isn't a question, or a request for help - it is a warning.
> glibc 2.1 is a bit dangerous, and until they clear up a few licensing
> issues, it should be avoided (or at least dealt with very carefully).
That's totally wrong. glibc 2.1 has no licensing problem. The
"political reason" as mentioned in glibc-2.1-README is a discussion
about egcs and gcc.
> Now, in case you haven't heard, glibc 2.1 has been pulled due to
> licensing problems and won't be available for a while. Even so, I
> managed to get a copy of it before they took it down. I also managed
> to get a src.rpm compiled (of my own creation) and let me tell you, it
> breaks a few things.
> 1) Netscape - kinda weird. Netscape won't connect to any
> external sites anymore, but works with loopback.
> 2) StarOffice - They knew that this would have problems,
> it just plain won't run. But then again, StarOffice has other
> problems . . .
StarOffice is broken.
> 3) RPM - How's this for a way to screw yerself over: make a
> src rpm of a library, compile the src rpm, then install/upgrade
> it. Find out you want to remove it because it breaks a few things,
> but you can't remove it because rpm, even though it is a static
> executable, depends on certain libraries - the ones you just
> replaced. My backup of rpm-2.5.5.tar.gz is compiling as I
> type . . .
> The exact error was something like "Bad owner/group" and the only
> reason I knew it was the library (besides the fact that I just
> upgraded it) was that I did an strace to find out that certain
> library files it was looking for weren't missing exactly, just
> the version numbers had changed
> (/lib/libnss_compat.so.1 -> /lib/libnss.so.2).
That's all explained in the glibc 2 FAQ which comes with glibc2.1.
> Needles to say, I'm not waiting out to find out what else it broke,
> and I wouldn't recommend that anyone else try this anytime soon, especially
> because of the licensing problems. Ah, it's done compiling. Time to
> 'rpm -U --force glibc-2.0.7.i386.rpm'.
Time to read the FAQ and the INSTALL file.
Andreas
--
Andreas Jaeger [EMAIL PROTECTED] [EMAIL PROTECTED]
for pgp-key finger [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (David T. Blake)
Crossposted-To:
comp.os.lynx,comp.os.linux.networking,comp.arch.bus.vmebus,comp.dcom.lans.ethernet
Subject: Re: LYNX Benchmarking Commands
Date: 11 Feb 1999 13:26:36 -0800
Shark <[EMAIL PROTECTED]> writes:
>Can someone tell if there are commands that can benchmark CPU
>utilization, all types of I/O, MIPS, and anything else related to bench
>marking?
>
>Is there free source code that can be compiled?
We generally use the matlab bench command if available
just because it involves no work on our part.
But, see www.specbench.org
--
Dave Blake
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Q]Linux device driver
Date: Thu, 11 Feb 1999 23:06:15 GMT
A person that develops UNIX drivers would in theory find it easier to develop
a Linux driver then a Windows programer.
Does that answer your question?
Martin
In article <[EMAIL PROTECTED]>,
"ÀÓÀçȯ" <[EMAIL PROTECTED]> wrote:
> Hi everyone!
> I have a question for Linux.
> Can a Person to have ability to develop a Linux device driver evolve
> UNIX device driver?
> Thank you for reading my question.
>
>
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: mkk <[EMAIL PROTECTED]>
Subject: Re: Will 2.2.x support removable medias better?
Date: Thu, 11 Feb 1999 17:16:33 -0600
Reply-To: [EMAIL PROTECTED]
Mike Dowling wrote:
>
> great advantage of the automounter is that it does absolutely nothing until
> you try to make access to the CD. Then it quietly mounts it, without a
> word, without starting silly programs and icons that pop up in the middle of
> whatever you are doing. After a short while of not accessing the CD, it
> quietly unmounts it. The isofs and CD support then is automagically removed
> from the kernel if you use modules and configure your system to do this.
>
> Perhaps the only hassle is if you want to check out several CDs in quick
> succession. Either you wait for the time out of the automounter, or you
> unmount the CDs manually.
>
Instead of an arbitrary time out, won't it be a better idea mount when
the media is physically loaded and unmount when it's popped out ?
------------------------------
** 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
******************************
