Linux-Development-Sys Digest #707, Volume #7 Sun, 26 Mar 00 21:13:16 EST
Contents:
System.map location ("D. Stimits")
Re: seq num generation ("D. Stimits")
Re: kernelmessage ... and now? ("D. Stimits")
Re: System.map location (Paul Kimoto)
Re: Windows CE target for GCC? (Tuomas Airaksinen)
Loadable device drivers and debugging ("Luis F. Ortiz")
Re: help - byte aligned vs. word aligned structures using gcc... (Dan McGuirk)
Re: /usr/include messed up how to I get the files ? (Stefan Rieken)
Re: system hung when compiling kernel (Karsten Bickel)
call for papaers: We are very very excited about a book on Linux Device Drivers.
("Nicholas Jordan")
USB->ethernet (linksys) and 2.3.99-pre3 (Bryan)
Re: Linux Serial Port Interrupt Problem (Kaz Kylheku)
Re: System.map location ("D. Stimits")
Re: Loadable device drivers and debugging ("D. Stimits")
Reiserfs + Raid? ("Kai Leibrandt")
libpthread causes segfault (Michael Dipperstein)
----------------------------------------------------------------------------
Date: Sat, 25 Mar 2000 22:40:49 -0700
From: "D. Stimits" <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: System.map location
I'm trying to find out where in the kernel it decides that it must
read /boot/System.map, so I can create alternate locations and names.
I'd like to boot multiple kernels without having to repoint sym links
first.
I found /usr/src/linux/scripts/ksymoops/Makefile with a #define for
DEF_MAP, but this is apparently not relevant for the location that the
kernel reads System.map from (change attempts here were ignored). No
matter what I do, it seems that the kernel has made up its mind to use
/boot/System.map. Where is the location of boot time reading of
System.map located or defined? Apparently not lilo, apparently not
from /usr/src/linux/scripts/ksymoops/ (although it looks to be
involved in creation of System.map). I would like to create
/boot/2.2.14/, /boot/2.2.12/, and so on...and subdirectories in each
of those based on the date of creation or a managable numbering
scheme. Any ideas?
Thanks
------------------------------
Date: Sat, 25 Mar 2000 22:50:28 -0700
From: "D. Stimits" <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: seq num generation
mas wrote:
>
> Does anyone know where in the kernel source is the for the TCP sequence
> number generators?
>
> --
> Posted via CNET Help.com
> http://www.help.com/
A grep of 2.2.12 showed some promising start locations:
/usr/src/linux/ is assumed....
drivers/char/random.c, near lines 47, 1295, and 1553
include/net/slhc_vj.h, near line 44.
Your mileage may vary.
------------------------------
Date: Sat, 25 Mar 2000 23:04:24 -0700
From: "D. Stimits" <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: kernelmessage ... and now?
Daniel Schulze wrote:
>
> Ooops, seems to be an english list...
>
> Hi Folks
>
> following problem:
> I put a second 64MB bar into my pc and then...
> Early on booting the kernel (SuSE 2.2.13) says the following and stops
> working:
>
> Memory: 127680k/131072k (1240k kernel, 412k reserved, 1676k data, 64k
> init, 0k big)
> kmem-alloc: Bad slab magic (corrupt) name=kmem_cache
> canot create uid taskcount SLAB cache
>
> What wants he to tell me?!
> With mem-option <= 64MB everything works fine and the BIOS believes to
> have 128MB.
> What is the kernel doing at this time an why does he fail?
>
> (System: MoBo: NMC 5VMMC Rev1.0, K6-II-500, 64MB RAM (100MHz), 64MB
> (133MHz), ATI@Work AGP, SuSE6.3)
>
> Thanx
>
> By Daniel
Some motherboards are picky about mixing types of ram. Sometimes they
are picky about the order that it appears in the slots. Your ram could
be bad, or maybe you just have an incompatible mix. Are there any
specification differences beside one being PC100 and the other PC133?
------------------------------
From: [EMAIL PROTECTED] (Paul Kimoto)
Subject: Re: System.map location
Date: 26 Mar 2000 01:43:02 -0500
Reply-To: [EMAIL PROTECTED]
In article <[EMAIL PROTECTED]>, D. Stimits wrote:
> I'm trying to find out where in the kernel it decides that it must
> read /boot/System.map, so I can create alternate locations and names.
> I'd like to boot multiple kernels without having to repoint sym links
> first.
The kernel does not read System.map files. ps(1) and klogd(8) do; see
their man pages. You may also want to look at the (sys)klogd startup
script.
--
Paul Kimoto <[EMAIL PROTECTED]>
------------------------------
From: [EMAIL PROTECTED] (Tuomas Airaksinen)
Subject: Re: Windows CE target for GCC?
Reply-To: [EMAIL PROTECTED]
Date: Sun, 26 Mar 2000 09:12:49 GMT
Thu, 16 Mar 2000 08:08:25 GMT, Tom & Ada Campbell kirjoitti:
>Is there version of GCC that can output Windows CE executables?
Can you put Linux on Windows CE-machine???
I suppose you can't...
--
kirjoitti: Tuomas Airaksinen linux
kotisivu: http://tumasites.cjb.net is
s�hk�posti: [EMAIL PROTECTED] all we
icq: 11870110, ircnet: tuma need
------------------------------
From: "Luis F. Ortiz" <[EMAIL PROTECTED]>
Subject: Loadable device drivers and debugging
Date: Sun, 26 Mar 2000 10:58:28 GMT
Ok. I admit it. I'm stumped.
I'm using GDB 4.18 remotely to try to debug a loadable module from a
remote system running RedHat 6.1 . I use the add-symbol-file command
with the load address of the loaded module and then try to look at the
addresses of the symbols loaded and try to set breakpoints.
Well, they are all wrong. Many of the symbols of functions are all
mapped to the same address. When the debugger stops, all the wrong
source line are indicated. This makes debugging awkward.
Anyone out there doing this kind of debugging and can offer any
pointers/hints?
I hate to think I have to debug the debugger. I'd prefer to think I
had done something wrong.
------------------------------
From: Dan McGuirk <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Re: help - byte aligned vs. word aligned structures using gcc...
Date: Sun, 26 Mar 2000 03:14:52 -0800
Robert Nickel wrote:
> Can anyone tell me how to make a structure byte-aligned instead of word
> aligned?
>
> I'm using a structure as the layout for an mmaped region of /dev/mem and
> need to make sure that my structure is kept in order.
>
> I'm certain this is trivially simple, but I'm (again) without the know how
> on this one.
With gcc you use the 'packed' keyword. It is not portable, but being
that you're reading from /dev/mem, I don't think you're exactly going
for maximum portability.
>From the gcc manual (amazing what you can learn from these things):
`packed'
The `packed' attribute specifies that a variable or structure field
should have the smallest possible alignment--one byte for a
variable, and one bit for a field, unless you specify a larger
value with the `aligned' attribute.
Here is a structure in which the field `x' is packed, so that it
immediately follows `a':
struct foo
{
char a;
int x[2] __attribute__ ((packed));
};
--
Dan McGuirk
[EMAIL PROTECTED]
De omnibus dubitandum est.
------------------------------
From: Stefan Rieken <[EMAIL PROTECTED]>
Subject: Re: /usr/include messed up how to I get the files ?
Date: Sun, 26 Mar 2000 15:47:46 +0200
Michael Meding wrote:
>
> Hi everybody,
>
> I messed up my directory /usr/include and now I am trying to dig a place
> up from where thos files originate ;-))
>
> Anybody knows anything about this and is willing to help a dumb user ?
>
> Thanks in advance
>
> With best regards
>
> M.E.Meding
/usr/include is the *general* place for include files. In other words,
the files come from everywhere. I assume you have a Linux with a
reasonable package system. On such a system, the following is likely:
- standard C includes are part of your compiler package (GCC/ EGCS) or
maybe part of your libc package (I think the last one) which should be
either libc or glibc.
- Linux kernel includes are part of your kernel package.
- Any library includes are either of your libraryname package or your
libraryname-dev package.
Just re-install the packages of which you screwed up (but hey, look out
when installing kernels/ clib!!).
Greets,
Stefan
- includes for
--
NOTICE:
Due to the new advertisement policies of softhome.net, I am
looking for a new email address. If I change, you will find
my new address in this signature.
If you want to be informed ASAP, mail me at my old address.
------------------------------
From: Karsten Bickel <[EMAIL PROTECTED]>
Subject: Re: system hung when compiling kernel
Date: Sun, 26 Mar 2000 15:30:38 GMT
Hallo Anders Larsen,
thanks for your reply.
> > Now I tried to build my first own kernel based on the kernel-sources
> > 2.0.36.
> >
> > All works fine till the point gcc tries to compile 'ide.c'. It starts
> > compiling and then the system(!) hung. Nothings works (e.g. switch to
> > another virtual console, numlock key, ...) - the only exit is a hard
> > reset
> > or power off/on.
> >
> > I tried it several times - exactly the same.
> > I reinstalled the sources, compiler, libraries - exactly the same.
> >
> > Any suggestions?
> > Do you need more info?
>
> That's common behaviour when the system runs out of memory.
> If you can't add RAM, you must increase the available swap space.
I set up an 32 MB swap partition. At one of my tries I switched to another
virtual console and startet 'top'. At the moment the system hung up I had
<300kb free memory and the swap was used with <3 MB -> 29 MB free(!).
Another suggestion?
> have fun
> Anders Larsen
so long
Karsten
--
Posted via CNET Help.com
http://www.help.com/
------------------------------
From: "Nicholas Jordan" <[EMAIL PROTECTED]>
Subject: call for papaers: We are very very excited about a book on Linux Device
Drivers.
Date: Sun, 26 Mar 2000 09:00:33 -0600
[full text at http://www.balista.com/njp/linux-device-drivers.eml ]
[I will put up the noted submission document in html as soon as I
can get a workable document]
>Hi Nicholas,
>
>We are very very excited about a book on Linux Device Drivers. I think
this
>would be great as a Black Book, with part of it devoted to writing the
>device drivers, but the vast majority devoted to installing,
configuring and
>troubleshooting. For specifics, see the email to me from Jeff Duntemann
>below.
>
>The next step would be to do a formal proposal, for which I am
attaching a
>blank proposal form (it's in Word, as I don't have access to a Linux
>machine, so I hope that you can use it). Along with the proposal, an
outline
>needs to be worked up, with an estimate page count. If you need help
with
>any of this, give me a call, and I can walk you through some of it.
>
>Regards,
>Charlotte Carpentier, MCP, A+ Certified
>Acquisitions Editor
>The Coriolis Group <http://www.coriolis.com/>
>Over 2 million certification titles in print
>480/483-0192, ext. 342
>-----------------------------------------------------------------------
----
>The present is the future you envisioned for yourself long ago.
>
>
>-----Original Message-----
>From: Jeff Duntemann [mailto:[EMAIL PROTECTED]]
>Sent: Monday, March 13, 2000 9:44 AM
>To: Tracy Schofield; Steve Sayre; Charlotte Carpentier
>Subject: Re: answer from slackwareFAQ maintainer
>(http://www.balista.com/njp/linux.htm)
>
>
>A book on Linux device drivers would be killer, and I've asked around in
the
>past, to no avoid. There's really two books in this space:
>
>1. Writing Linux Device Drivers
>2. Installing and Troubleshooting Linux Device Drivers
>
>Linux has a BIG weakness: There is no explicit hardware abstraction
layer as there is in Windows, so creating new device drivers is tricky, and
>installing them correctly is worse.
>
>Fewer people write device drivers than attempt to use them, so Book 2 would
>be the one to go after first. It must cover topics like recompiling the
>kernel and using gdb to spot problems. It must discuss ugly stuff like
>protection levels that are not exposed at the user level under Windows,
>which has that abstraction layer.
>
>But boy, this would be a good book to have. Charlotte, definitely go after
the guy.
>
>--73--
>
>--JD--
------------------------------
From: Bryan <[EMAIL PROTECTED]>
Subject: USB->ethernet (linksys) and 2.3.99-pre3
Date: Sun, 26 Mar 2000 16:18:54 GMT
just tried the 2.3 kernel for the first time (I needed usb/ethernet
for my new i-opener) ;-)
it does work! very cool..
but I'm not sure why the linksys adapter doesn't show LINK on its led.
the switch I connect the cat5 cable to shows link and I can ssh (etc)
to and from the box just fine. is the link light at all software
controlled? sounds unlikely but I'm not sure why a brand new usb
adapter would have a broken link led.
(as an aside, I was very surprised how easy it was to get usb/ethernet
working with this very new [to me] device; I know the 2.2 kernel and
its net.devices pretty well; and I was able to get the usb/ether
device up and working without reading ANY docs - just running 'make
menuconfig', taking good guesses at things, then make bzdisk and
reboot. wow. that was friggin EASY <g> )
--
Bryan, http://Grateful.Net (ANTISPAM: email is my name at my web's domain)
(c) 2000. Publishing and/or relaying of this material on all forums other than
USENET implies agreeing to a consultancy fee of US$150 per posting. You must
obtain a written permit before you publish. Violators are subject to civil
prosecution for Copyright Infringement as applicable. Publication by C|NET
and Microsoft Networks expressly prohibited.
------------------------------
From: [EMAIL PROTECTED] (Kaz Kylheku)
Subject: Re: Linux Serial Port Interrupt Problem
Reply-To: [EMAIL PROTECTED]
Date: Sun, 26 Mar 2000 16:55:17 GMT
On Sun, 26 Mar 2000 04:09:11 GMT, [EMAIL PROTECTED] <[EMAIL PROTECTED]>
wrote:
>I have been working on writing a device driver for
>linux that operates as an ethernet device but uses
>the serial port (COM1).
Such a driver should be implemented as a serial line discipline, much like the
PPP or SLIP code.
>I've tested the driver on
>an AMD Athlon with SD11 motherboard that uses VIA
>686 south bridge, AMD 750 north bridge, and 16550A
>UART. The driver works fine and the UART
>interrupts work well, and the linux kernel calls
>my interrupt routine.
You should not need to install any interrupt service routines to do
this task, only implement the tty line discipline callbacks.
Back to the drawing board!
--
#exclude <windows.h>
------------------------------
Date: Sun, 26 Mar 2000 11:32:39 -0700
From: "D. Stimits" <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: System.map location
Paul Kimoto wrote:
>
> In article <[EMAIL PROTECTED]>, D. Stimits wrote:
> > I'm trying to find out where in the kernel it decides that it must
> > read /boot/System.map, so I can create alternate locations and names.
> > I'd like to boot multiple kernels without having to repoint sym links
> > first.
>
> The kernel does not read System.map files. ps(1) and klogd(8) do; see
> their man pages. You may also want to look at the (sys)klogd startup
> script.
>
> --
> Paul Kimoto <[EMAIL PROTECTED]>
Thanks! :)
------------------------------
Date: Sun, 26 Mar 2000 16:20:49 -0700
From: "D. Stimits" <[EMAIL PROTECTED]>
Subject: Re: Loadable device drivers and debugging
"Luis F. Ortiz" wrote:
>
> Ok. I admit it. I'm stumped.
>
> I'm using GDB 4.18 remotely to try to debug a loadable module from a
> remote system running RedHat 6.1 . I use the add-symbol-file command
> with the load address of the loaded module and then try to look at the
> addresses of the symbols loaded and try to set breakpoints.
>
> Well, they are all wrong. Many of the symbols of functions are all
> mapped to the same address. When the debugger stops, all the wrong
> source line are indicated. This makes debugging awkward.
>
> Anyone out there doing this kind of debugging and can offer any
> pointers/hints?
>
> I hate to think I have to debug the debugger. I'd prefer to think I
> had done something wrong.
A question I'd asked just above this message was on the location of
System.map. I got a surprising answer, that it wasn't the kernel that
reads System.map, rather ps, and klogd. When reading the man pages, I
found some interesting notes on kernel symbols and looking up
information from debugging dumps. I can't answer your question, but I'd
bet that the man page for klogd would provide some information. As it
turns out, your list of symbols and addresses for modules is not always
up to date. System.map does not contain information on module symbols,
and klogd can be out of synch. Possibly...making a big guess...sending a
kill for SIGUSR1 to klogd just might improve your address information.
Quite possibly klogd has nothing to do with your source of address info,
but then I would bet that something similar is going on elsewhere with
regards to your symbol addresses.
------------------------------
From: "Kai Leibrandt" <[EMAIL PROTECTED]>
Crossposted-To: redhat.config
Subject: Reiserfs + Raid?
Date: Mon, 27 Mar 2000 01:56:12 +0200
Hi all,
I am eagerly awaiting SGI's XFS filesystem (due this summer I think), but
until then I'm experimeting with Reiserfs. Does anyone have any information
on combining ReiserFS with a soft- or hardware RAID 1 or 5 array?
On a similar note, does anyone know of a site with benchmark results
comparing different CPU speeds but same hardware otherwise (i.e controllers,
disks, memory) on a software RAID 5?
Many thanks,
Kai.
------------------------------
From: Michael Dipperstein <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.development.apps
Subject: libpthread causes segfault
Date: Sun, 26 Mar 2000 17:59:13 -0800
Hi,
I'm hoping someone that reads this group might be able to point me in
the
right direction. I'm porting a multi-level thread package from IRIX to
SMP Linux. The user level portion of the package seems to have been
ported correctly. I'm now in the process of adding the Kernel level
control.
I had hoped that by using Linux Threads I could quickly get the kernel
level portion of the thread package running, unfortunately I'm having
some
trouble with the Linux Thread library, and I've boiled the trouble down
to
just compiling with the libpthread.
If I take my user level test code and the '-lpthread' directive to the
gcc
command line, the code fails with a segment fault that results happened
in
function errno while trying to handle a printf.
The user level test code does not make any calls to pthread_* functions
and runs fine if the '-lpthread' directive is removed.
I'm guessing that libpthread replaces the code for one of the standard
library functions with its own code, and that code is incompatible with
my
code.
My user level thread code swaps stacks when threads are switched out.
Are
then any restrictions on stack manipulation when the thread library is
linked in, but none of its functions are called?
-Mike
------------------------------
** 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
******************************
