Linux-Development-Sys Digest #216, Volume #8 Sun, 15 Oct 00 14:13:14 EDT
Contents:
Re: Question About Kernel 2.4.0 Beta's, When Do You Think It will be a Stable
Release? (Rick Ellis)
Re: A new directory hierarchy standard - need opinions (The Ghost In The Machine)
Re: executing applications without using filesystem (Roger Leigh)
queue_task in 2.2.14 ("Brian VanLeeuwen")
Re: Time problem (Karl Heyes)
Re: host names resolved from /etc/networks!!! (Karl Heyes)
Re: Question About Kernel 2.4.0 Beta's, When Do You Think It will be a Stable
Release? ("Emu")
Re: Question About Kernel 2.4.0 Beta's, When Do You Think It will be a Stable
Release? (Kaz Kylheku)
Re: host names resolved from /etc/networks!!! (Tor Slettnes)
Re: Printer status (Lew Pitcher)
Re: How does a computer boot? (Brian Juergensmeyer)
Re: How does a computer boot? (Brian Juergensmeyer)
Re: Problem with ATHLON processor (Daniel Migowski)
to sync or not to sync ([EMAIL PROTECTED])
What is /boot/System.map ? ("J.Smith")
init ramdisk vs. root ramdisk ([EMAIL PROTECTED])
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Rick Ellis)
Subject: Re: Question About Kernel 2.4.0 Beta's, When Do You Think It will be a Stable
Release?
Date: 14 Oct 2000 19:33:13 GMT
In article <8s7vh9$lqc$[EMAIL PROTECTED]>,
Emu <[EMAIL PROTECTED]> wrote:
>I see you point about my Post. Keep in mind that myself and others are
>volunteering their time to make the kernel a successfully stable one.
Then you should be aware of the "schedule".
>All I asked was a simple question.
I answered it. Ask Linus and you'll get a similar answer.
>Its weird that I asked this same question in
>#linuxhelp in Undernet and they responded with out such a remark :)
So?
------------------------------
From: [EMAIL PROTECTED] (The Ghost In The Machine)
Crossposted-To: comp.os.linux.admin,comp.os.linux.networking,comp.os.linux.setup
Subject: Re: A new directory hierarchy standard - need opinions
Date: Sat, 14 Oct 2000 19:37:31 GMT
In comp.os.linux.development.system, David Wragg
<[EMAIL PROTECTED]>
wrote
on 10 Oct 2000 14:18:00 +0000
<[EMAIL PROTECTED]>:
>[EMAIL PROTECTED] (The Ghost In The Machine) writes:
>> 2. NFS is easily sniffable, as it does not have cryptographic capability
>> as far as I know. This may be addressed sometime in the future,
>> and there may be good replacements for NFS that do have crypto.
>> I certainly hope so, but I haven't looked.
>
>It was addressed sometime in the past. RFC2203, which specifies the
>RPCSEC_GSS security flavour for ONC RPC (which NFS sits on top of), is
>dated September 1997. RFC2623 goes into more detail about applying
>RPCSEC_GSS to NFS.
>
>Solaris has implemented this since 2.6 I think, with Diffie Hellman
>public key encryption as the security mechanism. 2.8 (maybe 2.7 too?)
>can also use Kerberos v5 as the security mechanism.
>
>The group (at the University of Michigan?) doing the Sun-sponsored
>NFSv4 implementation for Linux were implementing RPCSEC_GSS as a part
>of that.
>
>
>David Wragg
Interesting; I was not aware of that.
Is there a RFC2203/2623-compatible implementation of NFS available
for Linux yet, then? (Will there be part of one as of Linux 2.4? :-) )
--
[EMAIL PROTECTED] -- insert random misquote here
------------------------------
From: [EMAIL PROTECTED] (Roger Leigh)
Crossposted-To: comp.os.linux.development.apps
Subject: Re: executing applications without using filesystem
Date: 14 Oct 2000 20:56:16 GMT
Reply-To: [EMAIL PROTECTED]
On Sat, 14 Oct 2000 17:56:18 GMT, The Ghost In The Machine wrote:
>I'm curious as to that, too, if only because the Amiga's ramdisk
>device didn't have a fixed size and I kinda liked that. :-)
A bit like tmpfs on Solaris then? That uses part of the VM as a
filesystem, and it grows and shrinks as needed (there is a hard limit on
the maximum size IIRC). As it's VM, and not just real memory, it can be in
swap space too. Has anyone tried porting this to Linux as it's really
neat?
--
Roger Leigh ** Registration Number: 151826, http://counter.li.org **
Need Epson Stylus Utilities? http://www-users.york.ac.uk/~rl117/linux/
For GPG Public Key: finger [EMAIL PROTECTED] or see
http://pgp5.ai.mit.edu/pks-commands.html#extract
------------------------------
From: "Brian VanLeeuwen" <[EMAIL PROTECTED]>
Subject: queue_task in 2.2.14
Date: Sat, 14 Oct 2000 21:10:20 GMT
Hello,
I written a small module, that simply prints "I am in kernel" -quite
useless I know. However, I tried to insert this module in the kernel task
list so that this message gets printed at every timer interrupt.
I tried using queue_task() to insert the module in the task list.
The problem is that every time I try to insert the module I get the
following error message:
"unresolved symbol"
Can someone help me on this please.
Thanks,
Brian
------------------------------
From: Karl Heyes <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Time problem
Date: Sun, 15 Oct 2000 03:06:01 +0000
In article <[EMAIL PROTECTED]>, David Yeung <[EMAIL PROTECTED]> wrote:
> I've installed the RedHat 6.2 in a Intel PIII dual CPU machine, but I found
> very strange system clock problem. Here are the problem:
>
> % uname -a
> Linux hqlx13.ust.hk 2.2.16-3smp #1 SMP Mon Jun 19 19:00:35 EDT 2000 i686
> unknown
When the kernel was built.
> % date
> Sat Oct 14 08:01:22 HKT 2000
> % touch file1
> % ls -l file1
> -rw-r--r-- 1 3987 1563 0 Oct 14 00:01 file1
> % date -u
> Sat Oct 14 00:01:36 UTC 2000
>
> When I type a 'date' command, it show the correct clock value in HKT
> timezone. However, when I 'touch' a file, the file creation time 8 hours
> before the local time. It seems that it is the UTC time.
>
It's probably your /etc/localtime file. use timeconfig,linuxconf or copy
straight from /usr/share/zoneinfo (rpm -ql glibc)
Don't forget there is also you CMOS clock as well.
karl.
------------------------------
From: Karl Heyes <[EMAIL PROTECTED]>
Subject: Re: host names resolved from /etc/networks!!!
Date: Sun, 15 Oct 2000 03:17:22 +0000
In article <[EMAIL PROTECTED]>, Thomas Drescher
<[EMAIL PROTECTED]> wrote:
....
> Iface isdn-gw-2 * 255.255.255.255 UH 0 0 0=20
> ippp1 hst1.thomas.net * 255.255.255.255 UH 0 0 0=20
> plip0 hst2.thomas.net * 255.255.255.255 UH 0 0 0=20
> plip1 isdn-gw-1 * 255.255.255.255 UH 0 0 0=20
> ippp0 loopback * 255.0.0.0 U 0 0 0 lo
> default isdn-gw-1 0.0.0.0 UG 0 0 0=20 ippp0
>
check you net-tools/route version, they may be old. Route to hosts entries
should resolv from hosts, but linux use to not have these entries normally.
My system is fairly hacked up from RH, but my route is 1.96 and net-tools
1.54
karl.
------------------------------
From: "Emu" <[EMAIL PROTECTED]>
Subject: Re: Question About Kernel 2.4.0 Beta's, When Do You Think It will be a Stable
Release?
Date: Sat, 14 Oct 2000 22:40:01 -0500
You know what, after what I saw today at a Bio-Medical Convention, I am
just not going to participate in any more Test Kernels. What I saw will
just plainly replace Linux totally, completely.
And I am not going to say what it is since I was told not to say a word in
the research department. I do not need sarcasism from what it seems all
of a sudden from a couple of ppl who reply that way.
I just actually waisted my time asking the question here in the first place.
I can't understand why this one question returned unforvorable answers. From
now on I will ask my questions in a more informative group, #linuxHelp in
Undernet.
By the way, I had one of the Ops there tell me that sometime between late
3rd quater earlier 4th quarter for 2.4.0 stable. Its that simple. Only
wanted to know what this group thought.
"Rick Ellis" <[EMAIL PROTECTED]> wrote in message
news:8sacdp$ek8$[EMAIL PROTECTED]...
> In article <8s7vh9$lqc$[EMAIL PROTECTED]>,
> Emu <[EMAIL PROTECTED]> wrote:
>
> >I see you point about my Post. Keep in mind that myself and others are
> >volunteering their time to make the kernel a successfully stable one.
>
> Then you should be aware of the "schedule".
>
> >All I asked was a simple question.
>
> I answered it. Ask Linus and you'll get a similar answer.
>
> >Its weird that I asked this same question in
> >#linuxhelp in Undernet and they responded with out such a remark :)
>
> So?
>
------------------------------
From: [EMAIL PROTECTED] (Kaz Kylheku)
Subject: Re: Question About Kernel 2.4.0 Beta's, When Do You Think It will be a Stable
Release?
Reply-To: [EMAIL PROTECTED]
Date: Sun, 15 Oct 2000 02:57:21 GMT
On Sat, 14 Oct 2000 22:40:01 -0500, Emu <[EMAIL PROTECTED]> wrote:
>You know what, after what I saw today at a Bio-Medical Convention, I am
>just not going to participate in any more Test Kernels. What I saw will
>just plainly replace Linux totally, completely.
This convention wasn't held in a channel on Undernet by any chance?
------------------------------
Subject: Re: host names resolved from /etc/networks!!!
From: Tor Slettnes <[EMAIL PROTECTED]>
Date: Sun, 15 Oct 2000 06:09:48 GMT
Of course, you checked /etc/nsswitch.conf?
hosts: files nis dns
Or somesuch.
>>>>> "Thomas" == Thomas Drescher <[EMAIL PROTECTED]> writes:
Thomas> Hi folks, i have a very strange problem in my SuSE 6.2
Thomas> system. :-(everyone must also have that, thou...):
Thomas> Hosts are not resolved from /etc/hosts!
------------------------------
From: Lew Pitcher <[EMAIL PROTECTED]>
Subject: Re: Printer status
Date: Sat, 14 Oct 2000 13:59:33 -0400
Aulne wrote:
>
> Hello all,
>
> I am printing locally. I would like to get the printer's status so that I
> can report errors such as "no paper" or "paper jam", etc... I'm sending
> files through lpr. I looked at lpc, lpq and lprm, which works fine, but only
> seems to report errors about print spooling. In order words, I think I would
> have to maintain a timeout value and if the file's still in the spooler after
> such time, then there's an error.
It's part of the spooler's job to hide the mechanics of data printing
from the application that is generating print data. There _are_ ways
around this (have your appl popen() the 'lpc' program and query the
status of the printer), but these are non-portable and not easily
developed. In any case, this technique fails when tracking the status of
a networked printer. If you want direct control over the printer, it
would be better if you wrote your output to (and read the status of) the
appropriate /dev/lp? device yourself.
> Is there a way to get the printer err msgs?
>
> Alain
--
Lew Pitcher
Master Codewright and JOAT-in-training
------------------------------
From: Brian Juergensmeyer <[EMAIL PROTECTED]>
Date: Sun, 15 Oct 2000 14:47:39 GMT
Subject: Re: How does a computer boot?
Hi, Wolfram,
The book title and author are "Linux Device Drivers" by Alessandro=20
Rubini. The section you are talking about is Chapter 16.
HTH,
Brian
>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<
On 10/13/00, 3:06:55 AM, Wolfram Faul <[EMAIL PROTECTED]> wrote =
regarding Re: How does a computer boot?:
> Hi,
> > Please post me solutions, advices, book titles, links for sites...
> > Everything could help me.
> >
> In the O'Reilly Book "Device Drivers", sorry I forgot the authors name=
,
> is small chapter about booting the linux kernel.
> Wolfram
------------------------------
From: Brian Juergensmeyer <[EMAIL PROTECTED]>
Date: Sun, 15 Oct 2000 14:58:32 GMT
Subject: Re: How does a computer boot?
Hi, Pozzana,
I'm going off memory on this, but what I'm about to say should be grossl=
y=20
accurate. When the power switch is first tripped, the BIOS performs a=20
Power On Self Test (or POST). At the (successful) completion of this=20
POST, the CPU automatically jumps to a known, standardized address in th=
e=20
system ROM and begins execution. This address contains code for loading=
=20
the boot sectors of the known bootable drives. The first bootable drive=
=20
is identified and the boot sector is loaded into memory. Generally, thi=
s=20
code is a secondary bootstrap loader, as (IIRC) the code for the primary=
=20
bootstrap can't be more than one sector big on the hard drive. Once the=
=20
secondary loader loads itself, the system will jump to the address to=20
which the secondary bootstrap loader was loaded, and will begin executio=
n=20
there. At this point, the boot process starts splitting off depending o=
n=20
which OS you're loading. Generally, however, the OS loads the base=20
kernel and initializes it (at this point the system switches into=20
protected mode), and the required device drivers are identified and=20
loaded.
The book I had for Operating Systems was Andrew Tanenbaum's "Operating=20
Systems: Design and Implementation" First edition.
Hope this helps,
Brian
>>>>>>>>>>>>>>>>>> Original Message <<<<<<<<<<<<<<<<<<
On 10/13/00, 6:48:24 AM, "GIULIO POZZANA" <[EMAIL PROTECTED]> wrote regardin=
g=20
How does a computer boot?:
> Thanks for opening my mail :D.
> I'm studying Informatics in Venice University, at the second year.
> Actually I'm studying for an exam about Operating Systems.
> Following the documentation given to me from the Institute, I've studi=
ed=20
a
> lot of theory about designing an OS, but nothing about the basics for
> programming an operating system.
> Fundamentally, I don't know how does a computer boot -and the basic
> operations that have to be executed by the computer in this phase- (Is=
it=20
a
> miracle?).
> Please post me solutions, advices, book titles, links for sites...
> Everything could help me.
> THANXX
> Pozzana Mauro
> Universit=E0 C=E0 Foscari di Venezia, CdL Informatica.
------------------------------
From: Daniel Migowski <[EMAIL PROTECTED]>
Subject: Re: Problem with ATHLON processor
Date: Sun, 15 Oct 2000 18:12:06 +0200
St�phane CLEMENCEAU wrote:
> I can't install linux on a athlon processor. How i proceed to pass this
> problem ?
Tried 2.2.16 and 2.4.0test6 andn had no probs. Maybe, u didn't set the
prozessortype correct? I used at least in the 2.4er kernel the
Athlon-Option.
Daniel Migowski
------------------------------
From: [EMAIL PROTECTED]
Subject: to sync or not to sync
Date: Sun, 15 Oct 2000 16:51:17 -0000
Asyncronous I/O is in some cases unsatisfactory. Yet I also find the
alternative, syncronous, to be equally unsatisfactory in many of the
same cases. While looking over kernel code for something unrelated,
I notice indications that the compromise solution really isn't being
considered.
What I run into is with asyncronous I/O, a process doing a lot of
writing is able to write much faster than the physical device, and
as a result, buffers can fill up RAM. Done syncronously, optimal
performance is not met because the process is blocked until the
write is completed. It is pretty close to optimal most of the time
but it could be a bit better.
The compromise I'd like to see is a "soft syncronous". What this would
do is when a write is requested, the process (or fd for non-blocking I/O)
would block until the write is STARTED. This would mean as in asyncronous
that the write status cannot reflect the conclusion status of the write
operation to the device. But like syncronous, it would not flood RAM with
buffers since the 2nd write would block until the first is completed, or
in the case of devices like SCSI that can schedule multiple writes, the
Nth write would block, where N is presumeably finite.
I brought the related issue of buffer flooding up a little over a year
ago, and tried several buffer tuning suggestions, none of which made any
significant impact. I am occaisionally doing very large scale writes
to some devices or an archive filesystem where I have gigabyte size tar
files (or splits thereof until ext2 can get past the 2G size limit).
In these cases, RAM (I have as much as 512 MB on some machines) can get
totally flooded with these buffers, causing other data to flushed or
swapped out needlessly, and delaying response times when other programs
need to run again. I would like to restrict any ONE file descriptor to
a fixed limit on buffer space in RAM when writing. A restriction to one
buffer would effect the "soft syncronous" idea suggested above (although
I would like to have a mount-wide "soft syncronous" option, too).
Any new ideas? Any specifically known improvements in 2.4 (I will ignore
suggestions to "try it and see if ..." because I already will be doing
that, but would consider expediting that only if I discover that someone
really does know that my specific need has finally been addressed)?
--
| Phil Howard - KA9WGN | My current websites: linuxhomepage.com, ham.org
| phil (at) ipal.net +----------------------------------------------------
| Dallas - Texas - USA | [EMAIL PROTECTED]
------------------------------
From: "J.Smith" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: What is /boot/System.map ?
Date: Sun, 15 Oct 2000 19:47:48 +0200
Hi there.
I just did a kernel compile on my linux system, and noticed, besides the
bzImage of course, a file called 'System.map'. I know I have to copy it to
my /boot dir, just like the kernel image, but I was just wondering what this
file exactly *is* and what it is needed for. Does lilo need it? Is it needed
for loadable modules? Do other applications use this file? And if I have two
kernels installed, I will also get me two System.map files, where each is
specific for the kernel it was compiled with. So how does the kernel or
other programs know which System.map to use then, if for example we have 2
kernels in /boot, like:
A) "vmlinuz.2.2.13 & System.map.2.2.13"
B) "vmlinuz.2.2.16 & System.map.2.2.16"
Maybe I am posting this to the wrong newsgroup here, but I guess that the
only people who can
really clarify what system.map is used for and how you should handle two
kernels&map files are the guys who actually know something about the
source-code :)
If anyone can explain all this to me, or point me to an URL, it would be
greatly appreciated.
Thanks.
------------------------------
From: [EMAIL PROTECTED]
Subject: init ramdisk vs. root ramdisk
Date: Sun, 15 Oct 2000 17:55:57 -0000
To run a system from CDROM w/o using any HD, ramdisk is essential.
A limitation that exists in loading the initial ramdisk is how
much data can be comrpessed into the boot image, which on Intel
PC devices with El Torito is limited to a 2.88 MB floppy image.
To get past this limitation, my first thought is to have the
initial ramdisk bring in a program run as /linuxrc which then
reads a larger image from a file on the CDROM into yet another
ramdisk which will then be remounted as root by the kernel.
An issue I am running into is which ramdisk to use. I cannot
use /dev/ram0 (1,0) because initrd uses that one, and it is
already mounted. Overwriting it would have nasty consequences.
So I will use a 2nd ramdisk. Now the issue is which one. Should
I use /dev/ram1? Maybe programs on a running system expect to
take the next ramdisk past one that might be used by initrd. Or
should I use /dev/ram15? Similarly, programs might grab ramdisks
from the top down. I'd really prefer to let whoever uses this
make that decision. But the obvious place, where one puts a
root=/dev/ram1 on the kernel argument string won't work because
the hard coded device names in init/main.c don't have any more
ramdisks besides just /dev/ram (I notice a similar limitation in
choice of SCSI CDROM devices).
The only thing I can see is to require users to rebuild the initrd
image that all this starts from and include a special file pointing
specifically to the same device that the kernel will mount (which
would have to know it strictly by major,minor to get to a ramdisk
other than the first one).
Any comments on making it easier for users? I don't want to require
them to recompile everything to change something like which ramdisk.
Another thought.
Is it possible for /linuxrc to unmount / (after 2 ramdisks are open)
and mount / again as the 2nd ramdisk? If that is possible, then I
could copy the initrd to the 2nd ramdisk and rebuild the 1st one to
be what will become the real device to mount root. But alas, the
mapping of /linuxrc for the running process would block unmounting
the / filesystem. OK, maybe switching executeables. But even then
something is mapped from a filesystem mounted over / anyway. How to
have no filesystems mounted and have a process running ... doubtful.
--
| Phil Howard - KA9WGN | My current websites: linuxhomepage.com, ham.org
| phil (at) ipal.net +----------------------------------------------------
| Dallas - Texas - USA | [EMAIL PROTECTED]
------------------------------
** 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
******************************
