Linux-Development-Sys Digest #7, Volume #7 Thu, 29 Jul 99 11:13:50 EDT
Contents:
Re: Unresolved symbols in module... (Cameron L. Spitzer)
undefined (but called) functions
Re: RedHat 6.0 support for SB AWE64. Netscape, too (Bill Anderson)
Re: hang at "finding module dependencies" (Bill Anderson)
Re: problems with two network cards. Fundamental limitation? (Tony Gale)
Re: Unresolved symbols in module... (Cameron L. Spitzer)
RedHat 6.0: NFS and VERY bad response? (Ole Jacob Taraldset)
Re: dual Celeron MB blows up constantly! (Bill Anderson)
Re: Linux SCSI Performance Issues (Stefaan A Eeckels)
Re: drives mount in linux and dos-like OS's ([EMAIL PROTECTED])
Device Drivers Programming problems in linux (Ivo)
Re: Server linux config
Re: Unresolved symbols in module... (Allen Ashley)
Re: Unresolved symbols in module... ("Alex Abreu")
Re: HELP: how to measure hard disk access performance on Linux? (James Stevenson)
Re: Unresolved symbols in module... (Heeeeeeeez back!)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Cameron L. Spitzer)
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup,linux.redhat.misc
Subject: Re: Unresolved symbols in module...
Date: 29 Jul 1999 07:36:01 GMT
In article <7nnbvl$[EMAIL PROTECTED]>, Alex Abreu wrote:
>After compiling my kernel and rebooting it, I got the message "Unresolved
>symbols in module ..." for a lot of modules.
I got a whole blast of those today. Weeks ago, I made 2.2.10 with
everything a module that isn't in the boot path, and I don't have the
kernel daemon enabled. For the first time since then I tried to mount
a FAT-16 floppy. The messages came when I tried to insmod
msdos.o before fat.o. Inserting fat.o and then msdos.o worked fine.
It seems the symbols that come with one module are used to insert the next.
>I did everything by the book:
>
>make mrproper ; make menuconfig ; make dep ; make clean ; make boot ; make
>modules ; make modules_install
I don't trust any distribution to set up "make boot" correctly.
The rest of the targets are the work of Mr. Torvalds, but "boot"
depends on the packager, Red Hat in your case. I always
make: mrproper, xconfig, dep, clean, modules, bzImage, modules_install
and then copy the bzImage to where it needs to be and add it to my
Lilo setup. Only do mrproper after unpacking a new source tree;
it blows away the work config/menuconfig/xconfig does.
>I am using RH 4.2 w/ kernel 2.0.36 and libc 5.4.46 and the message also
>shows when I issue "depmod -a" when logged as root.
>
>In the kernel config menu, I chose not to have version information on the
>modules.
Oops, if you're making a set of kernel + modules all at once, put the
version info in them.
Cameron
http://judi.greens.org/lilo/
------------------------------
From: <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.m68k,redhat.kernel.general
Subject: undefined (but called) functions
Date: 28 Jul 1999 22:31:18 GMT
I found that in the kernel source code, there are some functions (macros)
never being defined but being called, for example,
__builtin_return_address()
__builtin_constant_p()
Could anybody please explain this? Thanks.
Jiu
================== Posted via SearchLinux ==================
http://www.searchlinux.com
------------------------------
From: Bill Anderson <[EMAIL PROTECTED]>
Subject: Re: RedHat 6.0 support for SB AWE64. Netscape, too
Date: Thu, 29 Jul 1999 02:29:05 -0600
[EMAIL PROTECTED] wrote:
>
> Bill Anderson <[EMAIL PROTECTED]> writes:
> > As root run setup, select Sound Configuration, and follow along. Has
> > always worked very well for my AWE64.
>
> Awesome, possum! :)
> Did I somehow miss this during the CDROM install?
>
Nope.
SoundCard Configuration is _still_ not in the RH install.
Bill
------------------------------
From: Bill Anderson <[EMAIL PROTECTED]>
Subject: Re: hang at "finding module dependencies"
Date: Thu, 29 Jul 1999 02:44:28 -0600
"David T. Blake" wrote:
>
> Yung-Hsiang Lu <[EMAIL PROTECTED]> wrote:
> > Hi, Everyone,
> >
> > Does anyone know what can make Linux hang at "finding module
> > dependencies"? I believe I did not change anything related to
> > modules. I am using Linux 2.2.5 (redhat 6.0).
>
> Calling depmod -a while the root filesystem is mounted read
> only tends to do that. See /etc/rc.d/rc.sysinit
>
>
FYI, if you use LILO, and this module depency search hangs the system,
pass the nomodules option to LILO IIRC:
LILO boot: <imagename> nomodules
ymmv, hth
Bill
------------------------------
From: [EMAIL PROTECTED] (Tony Gale)
Subject: Re: problems with two network cards. Fundamental limitation?
Date: 29 Jul 1999 08:03:57 GMT
In article <7noo3p$dqn$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Chetan Ahuja) writes:
> Steve Hier ([EMAIL PROTECTED]) wrote:
>: Chetan Ahuja wrote:
>
>
>: Ah -- but if your router is trying to do a reverse lookup for the IP
>: that is connecting, you have to wait for that to time out -- which I've
>: seen take up to 5 minutes on some systems.
>
> I don't see why it should do a reverse IP lookup seeing that the IP
> that I am connecting from is on the same subnet which is one of the "internal"
> IP's anyway ( 10.0.0 series). I do hope that the networking code
> is not THAT brain-dead.
>
TCP wrapper (and other such *application* level tools) do reverse DNS
lookups.
-tony
--
E-Mail: Tony Gale <[EMAIL PROTECTED]>
The views expressed above are entirely those of the writer
and do not represent the views, policy or understanding of
any other person or official body.
------------------------------
From: [EMAIL PROTECTED] (Cameron L. Spitzer)
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup,linux.redhat.misc
Subject: Re: Unresolved symbols in module...
Date: 29 Jul 1999 08:01:58 GMT
In article <7nogir$[EMAIL PROTECTED]>, Alex Abreu wrote:
>>did you rerun lilo after the build, but before the reboot? You will need
>>to modify /etc/lilo.conf, if you haven't already done so, to point to
>>the new kernel, and then type `lilo' at the prompt.
>
>Yes, I did everything by the book. /boot/vmlinuz is a symlink to the new
>kernel zImage and is also referenced in /etc/lilo.conf
Why does Red Hat insist on this "/boot/vmlinuz" thing? It must be some
historical relic. Make bzImage (zImage is obsolete) and copy it
to whereever you keep LILO files. Give it a mnemonic name
such as bz2.2.10. No symlinks. Make sure it's in a partition your
BIOS can see. (For most BIOSes, that means it ends below cylinder 1024.)
>What other files to I have to copy to /boot?
The LILO-related files are:
/etc/lilo.conf Tells /sbin/lilo what you want
bzImage Bootimage file: loadable by BIOS, LILO, Syslinux, or Loadlin.
boot.b The loader, also (more or less) known as LILO.
chain.b The chain loader, capable of launching a bootable partition.
message The text you want to see at boot time, preceeding LILO prompt.
map The list of block pointers /sbin/lilo creates
>DO I have to move the new System.map to /boot ?
No! It's a historical relic of the days before the ps(1) command used
the /proc "filesystem." The old ps(1) needed a symbol table to
rummage around in /dev/kmem. It was slow.
I suppose System.map is also used to build symbols into the modules,
but once you have a bzImage and modules you don't need it any more.
It doesn't have anything to do with the LILO boot sequence.
Cameron
------------------------------
From: Ole Jacob Taraldset <[EMAIL PROTECTED]>
Subject: RedHat 6.0: NFS and VERY bad response?
Date: Thu, 29 Jul 1999 10:55:11 +0200
Hi,
I have BIG problems with response on my RedHat 6.0 box and I'm pretty
sure it's a problem with NFS. When I used RH 5.2 I had no problems and
it felt like sitting on a Pentium II 450 MHz, 512 MB RAM, which it is,
but now it feels like sitting on a 386 with 4 MB.
The box has at the moment 2 NFS mounted file systems and mount gives
this:
mailhost:/var/spool/mail on /var/spool/mail type nfs
(rw,hard,intr,rsize=8192,wsize=8192,nfsvers=3,addr=129.177.31.53)
ojt:(pid515) on /u type nfs
(intr,rw,port=1023,timeo=8,retrans=110,indirect,map=amd.u,dev=00000004)
ojt:(pid515) on /net type nfs
(intr,rw,port=1023,timeo=8,retrans=110,indirect,map=amd.net,dev=00000005)
jupiter:/jupiter/home2 on /a/jupiter/jupiter/home2 type nfs
(grpid,rsize=8192,wsize=8192,noconn,rw,intr,hard,dev=00000006)
where "mailhost" is an old sun4. This mount is "fixed". "jupiter" is a 2
processor SGI running IRIX 64 6.2. It is mounted through AMD. Other file
systems can be mounted depending on who's logged in.
When _~90% of the CPU is idle_ the _load average is ~2_ as it is right
now. I'm the only user logged in and the only thing that is happening is
that I'm writing this in Communicator. Star Office 5.1 is also open, but
is not doing anything. I also have ~10 terminal windows open.
I try to dump a 64 MB file to the NFS partition and it takes around 1
minute => 8mbit/s so I don't think the net is saturated, at least the
speed is sufficient.
When I dump 64 MB the machine grinds to a halt. I do a "man tcsh" and it
takes 15 sec to get the page up, even though the man page is on local
disk.
I had none of these problems when I was running RH 5.2. Can it be that
the NFS response is the same, but since NFS has been moved from user
space to kernel space the machine is unable to do anything else when is
occupied with NFS? I have looked in /var/log/messages, but there is
nothing of interest as far as I can see.
Is there a switch I need to set, or an extra mount option that needs to
be included? Please help me with this as I'm going crazy and my
productivity is 50% of what is should be...
Please email me as well as post to this NG.
Thanks,
Ole Jacob
--
GexCon AS, Bergen, Norway <http://www.gexcon.com>
Tel : +47 55574334 (office) +47 55558650 (home)
Mob.tel. : +47 95080525 Fax : +47 55574331
PGP key : <http://home.c2i.net/ojt/pgp.txt>
------------------------------
From: Bill Anderson <[EMAIL PROTECTED]>
Subject: Re: dual Celeron MB blows up constantly!
Date: Thu, 29 Jul 1999 02:31:46 -0600
Shane Owenby wrote:
>
> In comp.os.linux.development.system, Brian Gilman
> <[EMAIL PROTECTED]> wrote:
> >>Hello all!
> >> Well, I go my abit dual celeron board today and have had nothing but
> >>problems with different kernels.....I wanted to use this board to learn
> >>about smp and programming threads with smp but, it's just not stable
> >>enough.....Sigh......Does anyone know what kernel version is considered
> >>the most *stable* for smp? Thanks in advance!
>
> Well I read all the responses to your post and just like everone else
> I have been running 2.2.* on mymachine with little to no trouble from them.
On this note, I have been having the Stuck in TLB Wait Queue problem,
though it has lessened with 2.2.10.
It'd sure be nice to not have that problem.
Bill
------------------------------
From: [EMAIL PROTECTED] (Stefaan A Eeckels)
Crossposted-To:
linux.dev.c-programming,linux.dev.kernel,linux.dev.scsi,comp.os.linux.hardware
Subject: Re: Linux SCSI Performance Issues
Date: 29 Jul 1999 09:59:41 GMT
In article <[EMAIL PROTECTED]>,
Dimi Shahbaz <[EMAIL PROTECTED]> writes:
> Hello,
> My question is regarding linux (2.0.36) SCSI performance degredation
> when we add disks to the SCSI bus. We are getting some strange results
> with regards to performance degredation as the number of scsi disks on
> the same SCSI bus goes from 1 to 4 disks. Sorry for the attachments,
> but it is the best way to explain whats happening. Our test program
> seeks and reads random sectors all across the disk(s) and prints out
> the histogram of the number of calls that took X microseconds (which is
> the x axis). The test was run for 30 minutes. Multiple disks are
> accessed at the same time through the use of concurrently running
> threads, each handling IO on one disk. We tried the test on kernel
^^^^^^^^
> version 2.2.10 but got similar results.
I think that what you're coming up against is not caused by the
SCSI subsystem, but by the fact you're using threads. Try to run
the test with multiple processes, instead of multiple threads.
--
Stefaan
--
PGP key available from PGP key servers (http://www.pgp.net/pgpnet/)
___________________________________________________________________
Perfection is reached, not when there is no longer anything to add,
but when there is no longer anything to take away. -- Saint-Exup�ry
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup
Subject: Re: drives mount in linux and dos-like OS's
Date: 29 Jul 1999 09:34:58 GMT
YamYam <[EMAIL PROTECTED]> wrote:
> I want to ask about the difference in mount/umount the drivers in
> linux 'is it necessary, if so why?' and the easy one in Dos-like
> OS's 'e.g., Win95, Win98, ...'.
necessary? well, others posted about the automounter.
The IMHO most sensible way (=the Mac way) is not possible due to technical
limitations of PC-hardware. It does not generate an interrupt when a floppy
is inserted into the drive. The only way I could think of to work around
this is to poll the drive (=frequently query its state). But this would be
useless most of the time and create unnecessary load on the system which is
less tolerable on a multiuser system. (BTW: If a (DOS-) floppy is mounted
automatically, who will be its owner, i.e. which permission will be set? How
does the automounter handle this?)
The second limitation is: The floppy can be removed manually. To keep the
filesystem consistent, DOS (don't know about Windows 9x/NT) does not cache
writes. Linux and other unices do. Again, the Apple sollution (eject FDs
only through software) would be great: write caching could be enabled and
still the floppy could not be removed until all caches are flushed.
Peter
--
Peter Gritsch <[EMAIL PROTECTED]>
`... so I'd rather you didn't try any last-minute stuff.'
I *AM* LAST-MINUTE STUFF, said Death, standing up.
[Terry Pratchett, Hogfather]
------------------------------
From: [EMAIL PROTECTED] (Ivo)
Subject: Device Drivers Programming problems in linux
Date: 29 Jul 1999 09:39:44 GMT
Dear reader,
At school we need to make a character device driver for linux.
I already know how to programm the programm itself (the driver), but can't seem
to run it with insmod. I talked with my teacher, who is unfortunately on
vacation now. And he said something about recompiling the kernel with the new
code added to it in some way (mem.c ?).....
Anyway, I have been searching the internet for awhile now in search of people,
who ever build drivers for linux or know how to do is....all help is welcome.
Oh I use redhat 5.2 (we need to for school).
Thanks in advance
Ivo Klerkx
e-mail: [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] ()
Subject: Re: Server linux config
Date: Thu, 29 Jul 1999 10:41:56 GMT
try /proc/sys/fs/file-max and /proc/sys/fs/inode-max
you can do something like:
echo 32768 > /proc/sys/fs/file-max
On 22 Jul 1999 22:01:00 GMT, bill davidsen <[EMAIL PROTECTED]> wrote:
>In article <[EMAIL PROTECTED]>,
>Warren Young <[EMAIL PROTECTED]> wrote:
>
>Any more ideas gratefully taken, I hate to think that I have to go to
>some other o/s just because of this. The notes and patches I found all
>seem to be on 2.0 kernels and haven't shed light on the 2.2 problem. Are
>all the web servers running hacked 2.0 kernels?
>
>--
>bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
> The Internet is not the fountain of youth, but some days it feels like
>the fountain of immaturity.
>
------------------------------
From: [EMAIL PROTECTED] (Allen Ashley)
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup,linux.redhat.misc
Subject: Re: Unresolved symbols in module...
Date: 29 Jul 1999 10:38:26 GMT
Make clean does not erase the modules in /lib/modules/<whatever>. You have to
do that manually. If you are wary of this step, then rename the directory
before you do make modules_install. The error messages at boot are harmless,
but I don't like them either. I can't explain why the >>defiles came up
empty, I do it all the time.
------------------------------
From: "Alex Abreu" <simonet at bhnet dot com dot br>
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup,linux.redhat.misc
Subject: Re: Unresolved symbols in module...
Date: Thu, 29 Jul 1999 08:09:32 -0500
First of all, thank you and the other guys for the help.
Cameron L. Spitzer wrote in message ...
>Oops, if you're making a set of kernel + modules all at once, put the
>version info in them.
the reason I started making everything w/o version info is because with
version info I was getting a lot of messages about version mismatch
(something like "Your 2.0.36 modules do not match your kernel version
2.0.36", which sounds quite weird for me).
Should I rebuild everything now using version info? What about the version
mismatch error?
Again, thanks. This thread's been of great help.
Yours,
Alex
------------------------------
From: [EMAIL PROTECTED] (James Stevenson)
Subject: Re: HELP: how to measure hard disk access performance on Linux?
Reply-To: [EMAIL PROTECTED]
Date: Thu, 29 Jul 1999 09:23:00 +0100
Hi
it would really make a difference when other programs
are running
On 28 Jul 1999 18:45:02 -0700, Errin Watusikac <[EMAIL PROTECTED]> wrote:
>[EMAIL PROTECTED] (James Stevenson) writes:
>
>> Hi
>>
>> try droping down into single user mode
>> then you should get a better answer
>>
>> X uses a lot of ram
>
>Yeh, well I got a lot more RAM than X needs. In any case, do I care
>how fast the disk is in single user mode or in X mode? Is there
>something special about this test that needs tens of megabytes of
>memory or something? And if I run it twice or more, shouldn't it clear
>out unused memory so it has room to run? Looks like the sign of some
>poor kernel design or code or tuning. Does it look that way to anyone
>who knows something about it? I'm no kernel guru.
--
=============================================
Check Out: http://www.users.zetnet.co.uk/james/
E-Mail: [EMAIL PROTECTED]
9:20am up 38 days, 10:23, 2 users, load average: 1.37, 1.52, 1.31
------------------------------
From: Heeeeeeeez back! <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup,linux.redhat.misc
Subject: Re: Unresolved symbols in module...
Date: Thu, 29 Jul 1999 14:33:52 +0100
In comp.os.linux.misc Allen Ashley <[EMAIL PROTECTED]> wrote:
> "Alex Abreu" <simonet@(spam? no thanx)bhnet.com.br> writes:
>>How can I solve the "Unresolved symbols ...." problem?
> Delete all the modules that are giving you error messages,
> you don't need them with the kernel you compiled. You
> can do:
> depmod -a >>delfiles
> then edit delfiles to change the lines to a series of rm statements,
> then execute the file to delete all the unused modules.
Not really a good idea if, like me, you have two kernels...
One's optimised for pentium, but I'm not moving fully to it yet because it's
ppp is broken. The other is a fully modular 386 compiled kernel with a
working ppp...
--
______________________________________________________________________________
| [EMAIL PROTECTED] | |
| Andrew Halliwell BSc |"The day Microsoft makes something that doesn't |
| in | suck is probably the day they start making |
| Computer Science | vacuum cleaners" - Ernst Jan Plugge |
==============================================================================
|GCv3.12 GCS>$ d-(dpu) s+/- a C++ US++ P L/L+ E-- W+ N++ o+ K PS+ w-- M+/++ |
|PS+++ PE- Y t+ 5++ X+/X++ R+ tv+ b+ DI+ D+ G e++ h/h+ !r!| Space for hire |
==============================================================================
------------------------------
** 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
******************************
