Linux-Development-Sys Digest #713, Volume #7 Tue, 28 Mar 00 21:13:18 EST
Contents:
Re: Text segment and virtual memory (Nate Eldredge)
Re: BIOS support (Neil Koozer)
Re: BIOS support (H. Peter Anvin)
Re: Free BSD (Grant Edwards)
Kernel modules development ("Fabrizio")
Re: BIOS support (Michael Meding)
Re: Free BSD (Pjtg0707)
Re: BIOS support (Stefan Reinauer)
Re: Network Daemon App ("Boris")
SIGIO not delivered by kernel ("Ruppert R. Koch")
B_FREE on queue (bill davidsen)
Re: latest 2.3.99-pre3 won't boot on I-OPENER (bill davidsen)
Re: POSIX timer implementation available. (David Wragg)
Re: Zero padding in sprintf() doesn't work for strings - HELP ! (bill davidsen)
Re: Zero padding in sprintf() doesn't work for strings - HELP ! (bill davidsen)
Re: Can my server handle 500,000 page views per day ? ([EMAIL PROTECTED])
Re: Free BSD (Marc Britten)
Re: Zero padding in sprintf() doesn't work for strings - HELP ! (Juergen Heinzl)
software raid patch ("D. Stimits")
Re: Can my server handle 500,000 page views per day ? (H. Peter Anvin)
----------------------------------------------------------------------------
From: Nate Eldredge <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Text segment and virtual memory
Date: 28 Mar 2000 13:14:12 -0800
TJ Walls <[EMAIL PROTECTED]> writes:
> Hello,
>
> How does the the linux program loader allocate virtual memory for the text
> and data segments, and is there any user-land way to free them? I having tried
> unmapping the VM pages allocated for both regions from a shared library with
> the following results (reproduced on a Solaris 7 box)...
>
> I can unmap the page with the text segment (from a shared library routine),
> cannot mmap another page into the same VM region. The mmap call does not
> return an error, but core dumps. I _can_ however mmap a different physical
> page overtop of the already mapped VM page for the text segment!
>
> I cannot unmap the page containing the data segment or map other memory
> overtop of it. Again, the munmap calls core dumps.
>
> I've tried playing with various permission and locking settings to no
> avail... the only conclusion I can come up with is the kernel is not mmapping
> these pages into memory, but if it is not then how is it doing it?
You're sure this isn't a page that somehow needs to be used later?
Because if anything tries to access it again, you will core dump. But
AFAIK, munmap should work on arbitrary pages.
--
Nate Eldredge
[EMAIL PROTECTED]
------------------------------
From: Neil Koozer <[EMAIL PROTECTED]>
Subject: Re: BIOS support
Date: Tue, 28 Mar 2000 13:21:42 -0800
Mark Langsdorf wrote:
>
> Robert Redelmeier wrote in message <[EMAIL PROTECTED]>...
> >Sebastien Dessimoz wrote:
> >
> >> I'd like to know if Linux uses the BIOS (except loading the Linux
> kernel
> >> image of course)?
> >
> >The canonical answer is no, Linux does not use the BIOS for anything
> >beyond loading the kernel image. BIOS is 16 bit real-mode code, and
> >Linux is 32 bit pmode. The segregs do not line up at all.
>
> How much of the BIOS does Linux require to load the kernel
> image, then? Just enough to get the GP-BUS and SDRAM set up,
> or something more? If one were trying to to write a pared down
> BIOS to boot Linux, could the VGA or floppy drive controller be
> left off?
There is a project someplace where some people are working on a linux
bios, but I don't know the url :(
The bios must set up the programmable chips on the m/b, and this would
include the video controller unless those things come up in mode 3 or
something.
Neil.
------------------------------
From: [EMAIL PROTECTED] (H. Peter Anvin)
Subject: Re: BIOS support
Date: 28 Mar 2000 13:20:42 -0800
Reply-To: [EMAIL PROTECTED] (H. Peter Anvin)
Followup to: <8bqq9h$pj3$[EMAIL PROTECTED]>
By author: "Mark Langsdorf" <[EMAIL PROTECTED]>
In newsgroup: comp.os.linux.development.system
>
> How much of the BIOS does Linux require to load the kernel
> image, then? Just enough to get the GP-BUS and SDRAM set up,
> or something more? If one were trying to to write a pared down
> BIOS to boot Linux, could the VGA or floppy drive controller be
> left off?
>
Not if you want to use stock bootloaders and a stock setup.S. If
you're willing to get the kernel images into RAM some other way,
e.g. by your custom BIOS, and are willing to modify setup.S, then you
can get away with not having a BIOS at all.
-hpa
--
<[EMAIL PROTECTED]> at work, <[EMAIL PROTECTED]> in private!
"Unix gives you enough rope to shoot yourself in the foot."
------------------------------
From: grant@nowhere. (Grant Edwards)
Subject: Re: Free BSD
Date: Tue, 28 Mar 2000 21:41:17 GMT
In article <8br656$3l9$[EMAIL PROTECTED]>, Peter T. Breuer wrote:
>Does that answer your question? If not, please explain to me the
>advantages of using BeOS instead of Plan9.
BeOS I can buy in a shrink-wrapped box at Best Buy. Plan9, I
can't.
--
Grant Edwards grante Yow! ... bleakness...
at desolation... plastic
visi.com forks...
------------------------------
From: "Fabrizio" <[EMAIL PROTECTED]>
Subject: Kernel modules development
Date: Tue, 28 Mar 2000 21:36:54 GMT
Hello!
I'm having a (basic) problem while building a kernel module using Linux
2.2.x (RedHat and Mandrake).
Practically after the includes:
#include <linux/kernel.h>
#include <linux/modules.h>
when I add
#include <linux/fs.h>
the gcc reports me a long list of errors, starting with a missing definition
of ino_t types. I'm following the "Linux kernel module programming" of O.
Pomerantz.
Can somebdoy help me ?
Thanks!
Fabrizio
------------------------------
From: Michael Meding <[EMAIL PROTECTED]>
Subject: Re: BIOS support
Date: Tue, 28 Mar 2000 22:56:59 +0200
Reply-To: [EMAIL PROTECTED]
Hi there,
thats not correct. Linux uses the bios. For example the ACPI stuff in
the linux 2.3.99-pre-3. How do you think this is possible without the
bios ?
There are other functions: For example adressing disk space with bios in
newer kernels.
PCI access mode ? Also a switch do use the bios.
There are a couple more though.
Best regards
Michael Meding
------------------------------
From: [EMAIL PROTECTED] (Pjtg0707)
Subject: Re: Free BSD
Reply-To: [EMAIL PROTECTED]
Date: Tue, 28 Mar 2000 22:26:53 GMT
On Tue, 28 Mar 2000 21:02:08 GMT, Kaz Kylheku <[EMAIL PROTECTED]> wrote:
>On Tue, 28 Mar 2000 20:48:55 GMT, Pjtg0707 <[EMAIL PROTECTED]> wrote:
>>On Tue, 28 Mar 2000 11:06:46 -0800, Sebastien Dessimoz
>><[EMAIL PROTECTED]> wrote:
>>>What is the advantage to use Linux instead of FreeBSD or the opposite???
>>>
>>
>>One is a bagel, the other is a donut. Which do you prefer?
>
>The one without the big security hole in the middle.
>
>Flame away, brother. ;)
>
>
>--
>#exclude <windows.h>
In that case, give Bill Gates a call. He will sell you the donut hole
as extra to plug that hole! :-)
------------------------------
From: Stefan Reinauer <[EMAIL PROTECTED]>
Subject: Re: BIOS support
Date: Wed, 29 Mar 2000 01:26:45 +0200
Mark Langsdorf wrote:
> >The canonical answer is no, Linux does not use the BIOS for anything
> >beyond loading the kernel image. BIOS is 16 bit real-mode code, and
> >Linux is 32 bit pmode. The segregs do not line up at all.
>
> How much of the BIOS does Linux require to load the kernel
> image, then? Just enough to get the GP-BUS and SDRAM set up,
> or something more? If one were trying to to write a pared down
> BIOS to boot Linux, could the VGA or floppy drive controller be
> left off?
Yes. this is possible. And it is already being done. You might want to
have a look at the OpenBIOS project
http://www.freiburg.linux.de/OpenBIOS/
This project aims at implementing a free IEEE-1275-1994 compliant
Firmware. Any questions to this project are welcome.
(We are also looking for kernel hackers and FORTH programmers interested
in joining the project)
The second one is LinuxBIOS. It is based on OpenBIOS' lowlevel code and
loads a kernel directly from flash. Via a new system call, it's then
possible to
start a new kernel from within Linux.
Have a look at: http://www.acl.lanl.gov/linuxbios/
Regards,
Stefan Reinauer, <[EMAIL PROTECTED]>
------------------------------
From: "Boris" <[EMAIL PROTECTED]>
Subject: Re: Network Daemon App
Date: Wed, 29 Mar 2000 00:31:25 +0200
Brian Patterson <[EMAIL PROTECTED]> schrieb in im Newsbeitrag:
FL%C4.1412$[EMAIL PROTECTED]
> Hi all,
> I came upon a need for a application - that when run, accepts up to 144
> client connections. These clients - will send a string of data once a
> second (or so). once the data is recieved - the program needs to either
> fork or create a new thread where I will process the line of data. I've
run
> into a problem that (*Sometimes*) when the data is sent from the client -
> the server doesn't see it until the client closes that connection.
> Currently the data is a fixed length but in my initial app - I made it
> accept everything up to a '\n'. Does anyone know of some good -simple-
> source of a daemon that will operate in this fashion? (i.e. - accepts
> network connections - listens for data and then forks or creates a new
> thread to handle it). I have all of Stevens books but I'm rather new to
> network programming so I'm afraid they don't help me much.
I didn't understand quite what the server and the client should do but for
sending and receiving some data terminated by \n use a Telnet client and
connect to your listening port. Thus you can check exactly what's going
wrong.
HTH,
Boris
------------------------------
From: "Ruppert R. Koch" <[EMAIL PROTECTED]>
Subject: SIGIO not delivered by kernel
Date: Tue, 28 Mar 2000 15:39:12 -0800
Hi all,
I wrote an application using an asynchonous socket. Incoming messages
trigger the kernel to send the SIGIO signal to the application, which
invokes a signal handler. Unfortunately Linux does not deliver the
signal when messages arrive. Generating the signal manually with
kill -29 creates the signal and my handler is invoked correctly.
Has this problem been discussed before? Any suggestions/solutions?
I tried PC Linuxlication 2.2.14. Guy Bormann has posted a solution for
MKLinux
a while ago, but his one is not directly portable.
Thanks,
Ruppert
------------------------------------------------------------------------
Ruppert Koch [EMAIL PROTECTED]
Dept of Electrical & Computer Engineering
4525 Hollister Av University of California, Santa Barbara
Santa Barbara, CA 93110 Phone: (805) 893-7788 Fax: (805) 893-3262
Phone: (805) 683-5606 http://alpha.ece.ucsb.edu/~ruppert
------------------------------------------------------------------------
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Subject: B_FREE on queue
Date: 29 Mar 2000 00:03:58 GMT
I have a system running software RAID 0+1, which has been running for
about 100 days. It got taken down hard (hardware), and when I tried to
boot I got a message including the title. I dropped back from 2.2.14 to
2.2.10 and get the system back up without a problem.
I had to hand apply the RAID patches from 2.2.10 to the 2.2.14 system,
but I have a number of systems running using that source, and the system
was totally stable for months, and has been through the hardware
shutdown many times before. When I got the md up, the fsck found no
problems with the filesystem, the error seems to come from the RAID.
Comments?
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
When taking small children to a carnival, always have them go potty
*before* you let them go on the rides, and let them eat all the junk
food and candy *after*.
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Crossposted-To: comp.os.linux.hardware
Subject: Re: latest 2.3.99-pre3 won't boot on I-OPENER
Date: 29 Mar 2000 00:09:49 GMT
In article <1wAD4.1159$[EMAIL PROTECTED]>,
Bryan <[EMAIL PROTECTED]> wrote:
| but when I use lilo to boot 2.3.99-pre3, it just hangs. lilo doesn't
| even get very far, it hangs within 1-2 seconds of printing '...'
| (dots). again, 2.2.13 works just fine: lilo loads to completion and
| the kernel/system works.
No answers, but something you can try, make your boot floppy as a
direct boot (bzdisk) instead of LILO, and don't forget to specify the
root device.
Example:
make bzdisk ROOT_DEV=/dev/sdb3
I have been having problems with getting Slackware on a Netfinity
(IBM) box, which *may* be related to an odd floppy drive config. IBM
likes to play, that's why there's a "floppy=thinkpad" option.
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
When taking small children to a carnival, always have them go potty
*before* you let them go on the rides, and let them eat all the junk
food and candy *after*.
------------------------------
From: David Wragg <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Re: POSIX timer implementation available.
Date: 27 Mar 2000 22:30:47 +0000
"Robert H. de Vries" <[EMAIL PROTECTED]> writes:
> Complete in the sense of the RT-POSIX specification. AFAIK that
> specification does not mention SIGEV_THREAD.
The RT-POSIX spec was folded into the 1996 edition of POSIX.1, and
since then there is no distinct RT-POSIX spec, just optional parts of
POSIX.1.
That standard says (section 3.3.1.2) that if _POSIX_THREADS is defined
(i.e. POSIX threads are present), then SIGEV_THREAD is available.
David Wragg
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Zero padding in sprintf() doesn't work for strings - HELP !
Date: 29 Mar 2000 00:18:48 GMT
In article <[EMAIL PROTECTED]>,
Stephen Richards <[EMAIL PROTECTED]> wrote:
| I'm afraid that the %06s doesn't work as you expect. The 0 is a flag in
| the formatting rules and indicate a padding of 0's from the left.
| However this only applies to numeric field types %d, %f et al.
Note that the man page clearly says that it does work the way he
expects. I'm not sure that the ANSI standard says the same thing, when I
left the committee the sense was that it wouldn't be required to act
that way, but the man page is totally clear.
The old libc does it with blanks as well.
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
When taking small children to a carnival, always have them go potty
*before* you let them go on the rides, and let them eat all the junk
food and candy *after*.
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Zero padding in sprintf() doesn't work for strings - HELP !
Date: 29 Mar 2000 00:25:48 GMT
In article <[EMAIL PROTECTED]>,
Wolfram Gloger <[EMAIL PROTECTED]> wrote:
| Second, if such a case was changed _now_ in glibc, the surprise would
| come with other systems were the flag leads to some other behaviour.
| (And those systems most probably can't be changed.) I've seen far too
| many complaints about `Linux-specific' programming, and introducing
| this kind of special features would proliferate that sort of
| programming unnecessarily.
Another reason why it should be changed. Both AIX and hp-ux pad with
zeros as expected, so it would be at minimum acceptable to do it that
way. Following Plauger's "law of least astonishment."
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
When taking small children to a carnival, always have them go potty
*before* you let them go on the rides, and let them eat all the junk
food and candy *after*.
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.networking,comp.os.linux.security,comp.software.testing
Subject: Re: Can my server handle 500,000 page views per day ?
Date: Wed, 29 Mar 2000 00:32:56 GMT
James,
Try WebSpray - (s/w) Stress tests your web servers to detect errors,
see how much load they can handle, and predict performance. WebSpray
is not only capable of simulating up to 1,000 clients from a single IP
address, but also supporting multiple IP addresses with or without
aliases. It will test all pages - not just the home page. WebSpray will
report any broken links found. Available from our site for download
Free for 15 days.
http://www.redhillnetworks.com
In article <Pine.D-G.3.91.1000321102918.3202A-100000@unix1>,
Alan Corey <[EMAIL PROTECTED]> wrote:
> On Tue, 21 Mar 2000, Chiltown Worldwide wrote:
>
> > We have recently switched ISPs for our hosting needs. On April 1st,
our
> > online store (www.chiltown.com) will be live and based upon our
advertising
> > budget, we expect 100,000 page views per day for the first week.
> >
> > I would like to know how others actually test load and if someone
could help
> > us either prove or disprove that our site (www.chiltown.com) can
handle the
> > load. We are interested in testing only the performance / load on
the first
> > page.
> >
> > We would also like to know how others test their sites for load as
well as
> > the best ISPs that can take some serious pounding. Any information
you can
> > send to me is greatly appreciated. Thank you in advance.
> >
> > James McGovern
> > Webmaster
> > Chiltown Worldwide
> >
> >
>
> There are some Perl scripts around for stress testing a server that
will
> fire off thousands of requests and record response times. I saw
source to
> one in a magazine printed over a year ago but I don't remember the
> magazine, sorry. I didn't have much use for it personally.
>
> If you poke around www.perl.com, specifically CPAN, you may find
something.
>
> Alan Corey
>
> ----------------------------------------------------------------------
> It's time to FDISK Microsoft and reformat.
>
>
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: Marc Britten <[EMAIL PROTECTED]>
Subject: Re: Free BSD
Date: Wed, 29 Mar 2000 00:14:34 -0500
Reply-To: [EMAIL PROTECTED]
"Peter T. Breuer" wrote:
>
> Sebastien Dessimoz <[EMAIL PROTECTED]> wrote:
> : What is the advantage to use Linux instead of FreeBSD or the opposite???
>
> FreeBSD is spelled differently and is written by different people
> and has a different licence and different objectives and a different
> development model.
>
> Does that answer your question? If not, please explain to me the
> advantages of using BeOS instead of Plan9.
>
> Peter
BeOS was designed for the desktop, Plan9 was built for a distributed
computing environment
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Zero padding in sprintf() doesn't work for strings - HELP !
Date: Wed, 29 Mar 2000 02:00:03 GMT
In article <8briic$ofu$[EMAIL PROTECTED]>, bill davidsen wrote:
>
>In article <[EMAIL PROTECTED]>,
>Wolfram Gloger <[EMAIL PROTECTED]> wrote:
>
>| Second, if such a case was changed _now_ in glibc, the surprise would
>| come with other systems were the flag leads to some other behaviour.
>| (And those systems most probably can't be changed.) I've seen far too
>| many complaints about `Linux-specific' programming, and introducing
>| this kind of special features would proliferate that sort of
>| programming unnecessarily.
>
> Another reason why it should be changed. Both AIX and hp-ux pad with
>zeros as expected, so it would be at minimum acceptable to do it that
>way. Following Plauger's "law of least astonishment."
[...]
The latest draft standard I've got states zero padding is allowed
for ...
d,i,o,u,x,X,a,A,e,E,f,F,g and G
... conversions and the same as far as the Unix98 specification
is concerned, so your installations of AIX and HP-UX are doing it
wrong and you might see whether there is an appropriate patch.
As far as the "law of least astonishment" is concerned I am afraid
following that route would result in quite a lot of astonishment in
the future.
Since there are standards, I do not see and room left for discussions.
Cheers,
Juergen
--
\ Real name : Jürgen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
Date: Tue, 28 Mar 2000 19:03:41 -0700
From: "D. Stimits" <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: software raid patch
It appears that the software raid 0 patch (e.g., raidtools with
raid0145-19990824-2.2.11) is still not part of the kernel at 2.2.14.
Anyone know if software raid for striping will be built into the
kernel soon, since it is a time-consuming to manually adjust these
patches to work on newer kernels?
------------------------------
From: [EMAIL PROTECTED] (H. Peter Anvin)
Crossposted-To: comp.os.linux.networking,comp.os.linux.security,comp.software.testing
Subject: Re: Can my server handle 500,000 page views per day ?
Date: 28 Mar 2000 18:08:17 -0800
Reply-To: [EMAIL PROTECTED] (H. Peter Anvin)
Followup to: <8brivo$8hq$[EMAIL PROTECTED]>
By author: [EMAIL PROTECTED]
In newsgroup: comp.os.linux.development.system
>
> James,
>
> Try WebSpray - (s/w) Stress tests your web servers to detect errors,
> see how much load they can handle, and predict performance. WebSpray
> is not only capable of simulating up to 1,000 clients from a single IP
> address, but also supporting multiple IP addresses with or without
> aliases. It will test all pages - not just the home page. WebSpray will
> report any broken links found. Available from our site for download
> Free for 15 days.
>
Well, I can tell you that for the last 48-or-so hours,
{ftp,www}.kernel.org has been averaging 80 Mbit/s, with an estimated
600 connections per second (large files.)
All on one dual-CPU Linux box.
-hpa
--
<[EMAIL PROTECTED]> at work, <[EMAIL PROTECTED]> in private!
"Unix gives you enough rope to shoot yourself in the foot."
------------------------------
** 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
******************************
