Linux-Setup Digest #564, Volume #19 Wed, 6 Sep 00 13:13:12 EDT
Contents:
freebsd nfs and linux client (rw access?) (Matthew Willis)
Re: installation problem ("Nisi")
creating an extended partition from within Redhat Kickstart (Surinder S. Dio)
My Lan card can't determine IP Information (Michael)
Re: installation problem (Eric)
Re: set path (Ron Ross)
Re: how to use tar to restore backup file ([EMAIL PROTECTED])
Re: module loading failure at boot time. (depmod) (Harring Figueiredo)
Re: set path (Harring Figueiredo)
Re: set path (Harring Figueiredo)
strange ppp problem, i need some help... ("Harm Cronie")
Re: Promise ATA100 and Linux (Torsten Pinkert)
Re: NVIDIA RIVA TNT2 Video Card ("Pat Parsons")
Install with self-made drivers disk (shane)
Installing RH 6.2 screwed up Win98 (Fubar X.)
Scanner (Martin Racette)
timezone problem during boot ([EMAIL PROTECTED])
Re: Install Linux from floppy ("Gregory D. Horne")
Re: Syslog problem ("Sylvain")
Re: timezone problem during boot ([EMAIL PROTECTED])
Re: Promise ATA100 and Linux ("Frederic Vanneste")
----------------------------------------------------------------------------
From: Matthew Willis <[EMAIL PROTECTED]>
Subject: freebsd nfs and linux client (rw access?)
Date: Wed, 06 Sep 2000 14:15:18 GMT
I am trying to link together a freebsd nfs server and a linux client.
I have way more practice with linux than freebsd, but I have read the
freebsd handbook on nfs to no avail. I would like to get read-write
access
to NFS.
The freebsd server is on a lan at 10.0.0.2; The linux client is
rh6.2 based at 10.0.0.1. The freebsd /etc/exports only seems to export
as readonly mode; using -maproot=0 etc. means that the linux box can't
mount the directory.
# /etc/exports at 10.0.0.2 freebsd box
/usr/home 10.0.0.1
this mounts fine on my rh box but I get permission denied if I try some
-maproot=0 business. I suspect I need to authenticate but I don't know
where to start. I have much of /etc/inetd.conf commented out on the
linux box.
Also, when I make changes to the /etc/exports in the freebsd box I have
been using init 6 to test changes. Is there a better way? I tried
grepping
for nfs and using kill -HUP but there are about four nfs daemons
running.
------------------------------
From: "Nisi" <[EMAIL PROTECTED]>
Subject: Re: installation problem
Date: Wed, 6 Sep 2000 10:38:39 -0400
You need to partition your drive.
Check this out:
http://nuintari.net/linux/mistakes/part.shtml
Nisi
www.mindlessmayhem.com
Eric wrote in message <[EMAIL PROTECTED]>...
>[EMAIL PROTECTED] wrote:
>>
>> I have a Gateway 633c with a 15 gig hard drive. I partition my hard drive
>> into one 10 gig drive for windows and a empty 4 gig drive for linux.
>> i tried to install Red Hat Linux 6.1 but when i tried to assign a root
>> partition, it gave me an error message "Boot Partition too big." Is there
>> any way to get around this, other than buying a new hard drive?
>> Please help, thanx.
>>
>> --
>> Posted via CNET Help.com
>> http://www.help.com/
>
>Yes, but are you sure you want to?
>The installer is warning you that you will not be able to boot linux
>in the current setup, because the PC BIOS cannot access data beyond
>cylinder 1024 (I agree, the message isn't very clear). So the best
>solution would be to create a small (~15 MB) partition below this
>boundary (usually around 8 G), or be willing to boot from floppy
>(temporarily) and update the linux loader (LILO) or continue to boot
>from floppy. If you insist on creating the partitions and boot from
>floppy, you'll need fdisk to partition the disc, 'cause the partitioner
>UI from RH6.1 won't let you do this.
>
>Eric
------------------------------
From: [EMAIL PROTECTED] (Surinder S. Dio)
Subject: creating an extended partition from within Redhat Kickstart
Date: 6 Sep 2000 14:22:18 GMT
Reply-To: [EMAIL PROTECTED]
Hi,
Just wondering if anyone had any pointers to the following:
1. I've Got a load of PC's with NT loaded on them, the intention is to
blast Redhat 6.2 onto them via Kickstart. I asked the NT guys to
leave an extended partition on the disk for my install - instead
they left a 2Gb unused partiton.
The layout currently is as follows:
2Gb NTFS (marked primary) - lets call it NTFS-a
2Gb Free (not labelled) " FREE
10Gb NTFS (ditto) " NTFS-b
5Gb NTS (ditto) " NTFS-c
I wanted to have the spare space as a 1800 / ext2 FS and a 200 swap -
however as I can only have 4 primary partitions I really need to
convert the 2Gb FREE partition into a an extended partion so that I
would have
/ on /dev/hda5 and swap on /dev/hda6
So really I need to convert the 2Gb FREE into a 2Gb EXTENDED
partition. I can do this manually via NT disk administrator or via
fdisk under linux (after booting from Redhat cdrom).
However there are over 200 PCs and I want to keep this as hands-off as
possible so I was looking for some way to create an extended partition
on these disks - prefabably as part of a Redhat Kickstart. I've been
thru the kickstart docs and havent hit upon the way to do this.
ANy ideas or alternatives welcome.
thanks in advance
Surinder
--
Please remove _NOSPAM_ in my reply address
------------------------------
From: Michael <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking,comp.os.linux.help
Subject: My Lan card can't determine IP Information
Date: Wed, 06 Sep 2000 22:28:52 +0800
When I use ifup to call my ethernet card, the error message is shown.
I am trying to use i-cable broadband and it uses dhcp.What can I do?
Thank you very much!
Michael
------------------------------
From: Eric <[EMAIL PROTECTED]>
Subject: Re: installation problem
Date: Wed, 06 Sep 2000 16:48:10 +0200
Reply-To: [EMAIL PROTECTED]
Nisi wrote:
>
> You need to partition your drive.
I need to do no such thing, thank you very much ;-)
> Check this out:
> http://nuintari.net/linux/mistakes/part.shtml
Not much use, after taking a short look at it (though it's a nice
tutorial on why and how to partition)
The problem the original poster had, was that he/she couldn't partition,
due to a very user-(un)friendly partitioning tool. There's a good reason
to warn you for not making the mistake of putting /boot beyond cyl.
1024, but the RH partitioner simply doesn't allow you to continue: Damn
thing!
For the rest, read my previous reply.
Eric
>
> Nisi
> www.mindlessmayhem.com
>
> Eric wrote in message <[EMAIL PROTECTED]>...
> >[EMAIL PROTECTED] wrote:
> >>
> >> I have a Gateway 633c with a 15 gig hard drive. I partition my hard drive
> >> into one 10 gig drive for windows and a empty 4 gig drive for linux.
> >> i tried to install Red Hat Linux 6.1 but when i tried to assign a root
> >> partition, it gave me an error message "Boot Partition too big." Is there
> >> any way to get around this, other than buying a new hard drive?
> >> Please help, thanx.
> >>
> >> --
> >> Posted via CNET Help.com
> >> http://www.help.com/
> >
> >Yes, but are you sure you want to?
> >The installer is warning you that you will not be able to boot linux
> >in the current setup, because the PC BIOS cannot access data beyond
> >cylinder 1024 (I agree, the message isn't very clear). So the best
> >solution would be to create a small (~15 MB) partition below this
> >boundary (usually around 8 G), or be willing to boot from floppy
> >(temporarily) and update the linux loader (LILO) or continue to boot
> >from floppy. If you insist on creating the partitions and boot from
> >floppy, you'll need fdisk to partition the disc, 'cause the partitioner
> >UI from RH6.1 won't let you do this.
> >
> >Eric
------------------------------
From: Ron Ross <[EMAIL PROTECTED]>
Subject: Re: set path
Date: 06 Sep 2000 11:23:12 -0500
"Brian" <[EMAIL PROTECTED]> writes:
> I have been unable to edit my $PATH or find any info on how to do it.
> One article said my path was in the '.profile' file, but neither /root or
> /home/<user> directories contain this file.
> I magazine article described such utilities as : addpath, delpath, edpath,
> listpath, uniqpath, but none of these appear to be in the man or info pages.
> Where is the path variable kept and how do I edit it?
On my system, Slackware 7 (Bigslack), this resides in /etc/profile, and
that's where I've successfully changed $PATH and other global variables.
It's also possible to set environment variables per user through scripts
in the home directory (like .bashrc - depends on your shell... but I'm a
newbie, so you'd best look it up).
Ron.
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: how to use tar to restore backup file
Date: Wed, 06 Sep 2000 15:07:30 GMT
On Wed, 6 Sep 2000 10:58:30 +0800, "@" <[EMAIL PROTECTED]> wrote:
>I'm use tar command to restore xx.tar.gz file(root:tar xvfZ xxx.tar.gz), but
>the cursor move one line down and nothing happen. So is the parameters i use
>correct or...?
I'm not sure what tools you are using (doesn't look like a command
line with that root: business). But, the capital Z indicates that the
file was compressed with the compress program. A small z would
indicate it was compressed with the GNU gzip program. So,
tar -xvzf xxx.tar.gz
should work.
Gord
------------------------------
From: Harring Figueiredo <[EMAIL PROTECTED]>
Subject: Re: module loading failure at boot time. (depmod)
Date: 6 Sep 2000 11:19:13 -0400
Thanks for the hint, !!!
I have been trying to figure out the error on the file "modules.dep" for
about 45 minutes.. What a pain in the *&^. I completely ignored the
conf.modules and that was the one that was messed up. Somehow, a process
(maybe netconf ??) inserted a line "options 3c501 " on, surpise, line
6(for my 3COM net card.) but it was not passing any param, which is required
by depmod). Thus the error "cof:6: missing module argument."
I manually removed that line it it is now rolling!!
Thank for the help !
Harring
The Black Unicorn wrote:
> you can always try deleting (or moving to another place) the modules.dep
> file and then run depmod -a again. if there's a problem with your
> modules.dep file itself, this should fix it as it builds a new one.
>
> if that doesn't help, you may want to check you /etc/conf.modules or
> modules.conf for strange things (check around line 6, maybe ?). this is
> just a guess, though, but i won't hurt :-)
>
> TbU
>
> Harring Figueiredo wrote:
> >
> > Thank you in advance for reading this and helping me.
> >
> > I am running into the follwoing problem:
> >
> > When booting my machine (kernel 2.2.5.15 version - RedHat 6.0
> > build) it stops on the "Finding modules dependencies" , then it gives
> > me the following error:
> > " conf:6: missing module argument
> >
> > can't open "
> >
> > Then the boot process continues.
> >
> > I have noticed that one of the modules that I cannot use once it
> > is booted is the ppp.o (kppp fails). Right now, I have to MANUALLY
> > insmod slhr.o and then ismod ppp.o for it to work.
> >
> > I have had the machine running on and off for about 3 months
> > without a problem and now it is having this problem. Any suggestions
> > how to fix it ?
> >
> > I also have noticed that when I try to manually run "depmod -a"
> > on the command line, it gives me the same error. ( conf:6: module
> > missing agument) , Also I have tried using "depmod -a 2.2.5-15 " on the
> > command line and I get the error: "conf:6: module missing argument.
> > can't open /lib/modules/2.2.5-15/modules.dep").; however, when I do a
> > " l s-al /lib/modules/2.2.5-15/modules.dep" I can see that the file is
> > there. (I have even chmod to 777 on it to see if it was reading perm.
> > problem. Could not get results.
> > I have not upgraded or built a new kernel nor installed any
> > major packages (other than ymessenger -yahoo messenger).
> >
> > I have also noticed that the NFS fails with the error
> > "nfsvc function not implemented. "
> >
> > What have I done wrong ?
> >
> > Thanks for all the help.
> >
> > Harring
> > [EMAIL PROTECTED]
------------------------------
From: Harring Figueiredo <[EMAIL PROTECTED]>
Subject: Re: set path
Date: 6 Sep 2000 11:30:19 -0400
To edit the path for a single shell, do
export PATH=$PATH:<your/new/path/goes/here>
Example:
Let's try to execute the fortune program:
========== On the system shell =========
$ fortune
bash: fortune: command not found
$
=========
Now, if I do :
$export PATH=$PATH:/usr/games/fortune
and run the program again, I get:
$fortune
Truth has no special time of its own. Its hour is now -- always.
-- Albert Schweitzer
to see the value, type
$ echo $PATH
To edit a path to have it "fixed" everytime you log in or open a shell,
insert the export statemnet above at the end of the file $HOME/.bashrc . if you
don't have one, try copying one from either the root account or someone else.
I am assuming that you are using LINUX and bash as a shell.
If it is another shell, let me know.
Hope this helps.
Harring
Ron Ross wrote:
> "Brian" <[EMAIL PROTECTED]> writes:
>
> > I have been unable to edit my $PATH or find any info on how to do it.
> > One article said my path was in the '.profile' file, but neither /root or
> > /home/<user> directories contain this file.
> > I magazine article described such utilities as : addpath, delpath, edpath,
> > listpath, uniqpath, but none of these appear to be in the man or info pages.
> > Where is the path variable kept and how do I edit it?
>
> On my system, Slackware 7 (Bigslack), this resides in /etc/profile, and
> that's where I've successfully changed $PATH and other global variables.
> It's also possible to set environment variables per user through scripts
> in the home directory (like .bashrc - depends on your shell... but I'm a
> newbie, so you'd best look it up).
>
> Ron.
------------------------------
From: Harring Figueiredo <[EMAIL PROTECTED]>
Subject: Re: set path
Date: 6 Sep 2000 11:32:02 -0400
Harring Figueiredo wrote:
> To edit the path for a single shell, do
> export PATH=$PATH:<your/new/path/goes/here>
>
> Example:
> Let's try to execute the fortune program:
> ========== On the system shell =========
> $ fortune
> bash: fortune: command not found
> $
> =========
> Now, if I do :
>
> $export PATH=$PATH:/usr/games/fortune ***** SORRY , this is wrong, should have
> been only
export PATH=$PATH:/usr/games
>
> and run the program again, I get:
>
> $fortune
> Truth has no special time of its own. Its hour is now -- always.
> -- Albert Schweitzer
>
> to see the value, type
> $ echo $PATH
>
> To edit a path to have it "fixed" everytime you log in or open a shell,
> insert the export statemnet above at the end of the file $HOME/.bashrc . if you
> don't have one, try copying one from either the root account or someone else.
>
> I am assuming that you are using LINUX and bash as a shell.
>
> If it is another shell, let me know.
>
> Hope this helps.
>
> Harring
> Ron Ross wrote:
>
> > "Brian" <[EMAIL PROTECTED]> writes:
> >
> > > I have been unable to edit my $PATH or find any info on how to do it.
> > > One article said my path was in the '.profile' file, but neither /root or
> > > /home/<user> directories contain this file.
> > > I magazine article described such utilities as : addpath, delpath, edpath,
> > > listpath, uniqpath, but none of these appear to be in the man or info pages.
> > > Where is the path variable kept and how do I edit it?
> >
> > On my system, Slackware 7 (Bigslack), this resides in /etc/profile, and
> > that's where I've successfully changed $PATH and other global variables.
> > It's also possible to set environment variables per user through scripts
> > in the home directory (like .bashrc - depends on your shell... but I'm a
> > newbie, so you'd best look it up).
> >
> > Ron.
------------------------------
From: "Harm Cronie" <[EMAIL PROTECTED]>
Subject: strange ppp problem, i need some help...
Date: Wed, 6 Sep 2000 17:35:06 +0200
Hello,
I've set up a Redhat 6.2 Linux machine to serve as an internet gateway. I
configured IPCHAINS to masquarade from the inside network to the internet. I
succesfully configured pppd with dial on demand. I can surf from
machines(win 98 clients) on the network, download files (ftp, http) and send
email messages. When I try to send an email message with an attachment from
Outlook Express it starts sending but suddenly stops. From that moment my
Linux machine acts strange, I can't even PING its own IP adres. Neither any
other adress on the internet or internal network. If I kill the ppp0 link I
can ping its own IP adress. There aren't any messages in the log after the
normal messages that the ppp link went up. So does anybody know what is
causing this problem?
I'm using an PIII 650, 128 Mb, WD 10.2 Gb, RedHat Linux 6.2 as 'gateway'
It dials out with an elsa microlink external ISDN adapter.
Thanks Harm.
------------------------------
From: Torsten Pinkert <[EMAIL PROTECTED]>
Crossposted-To:
alt.comp.periphs.mainboard.asus,alt.os.linux.slackware,comp.os.linux.hardware
Subject: Re: Promise ATA100 and Linux
Date: Wed, 6 Sep 2000 17:42:11 +0200
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] says...
> yep, I have the same problem with hdparm. no dma. which leaves me with a
> fantastic 4.7 M/sec throughput. hurrah for ATA100 ?
As always with linux - patch your kernel and install tools ;-)
http://www.linux-ide.org/
------------------------------
From: "Pat Parsons" <[EMAIL PROTECTED]>
Subject: Re: NVIDIA RIVA TNT2 Video Card
Date: Wed, 6 Sep 2000 10:57:20 -0700
worked fine for me - don't remember what I did, try selecting the Generic
VGA compatible video card & Generic Monitor
"Ken White" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Has anyone sucessfully used this video card (with an LG 775N 17" colour
> monitor) with Linux (Redhat 6.0?)? If so can they please let me know what
> settings they used, as I am having a lot of difficulty in configuring
> Xfree86 to use it.
>
> TIA
>
> Ken White
>
>
>
------------------------------
From: shane <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.redhat
Subject: Install with self-made drivers disk
Date: Wed, 06 Sep 2000 12:09:30 -0400
I'm trying to install Redhat6.2 on a Micron MV5000. I found the
driver for the raid card at ami.com. I know I need to boot the
Redhat install floppy and then choose expert so that the installer
will read the raid driver off another floppy.
My question is this: how to create the driver floppy?
I tried formatting it with MSdos and copying the the driver files
to it.
The redhat installer says that the disk is not a driver's disk.
I tried formatting the disk with ext2 and copying the driver files
on it.
The redhat installer still says that the disk is not a driver's disk.
Anyone know what I'm doing wrong??
------------------------------
From: Fubar X. <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.install
Subject: Installing RH 6.2 screwed up Win98
Date: Wed, 06 Sep 2000 19:10:30 +0300
Hi,
On my primary 10 GB disk, I set aside 3 GB for Linux and installed
WIn98 on the other 7GB. I now installed RH 6.2 w/o any problems.
Installation was done for 'workstation', meaning the install program
automatically selected the partitioning. Linux runs w/o any problems.
However, when I start Win98, all of a sudden I have a new partition,
which, from Win98's perspective, is unformatted. Therefore, Win98
slows down terribly for almost any operation. And another bad
side-effect is that many programs don't run anymore, since they were
on drive E:, but now they're on F:, because the new drive was inserted
in the middle. This problem did not happen in RH 6.0.
I reinstalled RH 6.2 in custom mode, and selected the main Linux
partition as 'Linux native' (I assume it means ext2), but the problem
persists.
I searched the HOWTOs, but didn't find an answer.
Thanks.
------------------------------
From: Martin Racette <[EMAIL PROTECTED]>
Subject: Scanner
Crossposted-To: alt.os.linux.mandrake,comp.os.linux.hardware,comp.os.linux.misc
Date: Wed, 06 Sep 2000 16:16:32 GMT
Hi guys,
I know that SANE was installed when I did the installation of Mandrake=20
7.1 , since I selected everything that was available to install, but I=20
can't find out how to use it with my HP-5p, and I know for a fact that=20
this scanner is supported since its a SCSI, and is hook-up an dpowered=20
up
Thank you in advance
Merci a l'avance
Martin
------------------------------
From: [EMAIL PROTECTED]
Subject: timezone problem during boot
Date: Wed, 06 Sep 2000 16:34:26 GMT
I started building some modules today, which caused me to notice a boot
problem with timezones.
My hardware clock is local time. /etc/timezone contains "US/Central"
(no quotes). /etc/default/rcS contains "UTC=no" (no quotes).
Once the system is booted, the time is correct.
The problem is that in rcS.d, S20modutils runs before S40hwclock.sh. It
is S40hwclock.sh that sets the time correctly, but S20modutils run
depmod, which regenerates /lib/modules/2.2.13/modules.dep with the
wrong time (5 hours too early). The files time is not my local time
(GMT-5),
nor is it GMT. It is being set to GMT-10. It is as if the HW clock was
GMT, and 5 hours are subtracted from it for local time.
This problem then causes modprobe to complain about /etc/modules.conf
being newer than modules.dep, which causes S20modutils to execute
the last if statement (modprobe ... -t boot \*) when it shouldn't
causing more error messages.
If I edit S20 and add "touch /testtime" I get the same thing. I need
the HW clock as local since the system is dual boot with Win98.
libc6 2.1.3-10, Debian potato.
Any suggestions?
Thanks
-Michael
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: "Gregory D. Horne" <[EMAIL PROTECTED]>
Subject: Re: Install Linux from floppy
Date: Wed, 06 Sep 2000 11:41:46 -0500
No, you only need a few diskettes to get Debian GNU/Linux installed and use dselect or
apt-get via ftp to get the rest of the packages you want.
Igor Mercier wrote:
> G'day Somerton
> From floppy ??? I am like you new to linux but I would never install from floppy
> Why!
> if disk takes 1.4m per disc and the instalation I did took 540 meg and that was not
> full instalation how may floppys would you need
>
> If you have a cd rom why not use it I used red hat 6.0 this version of linux lets
> you install from cdrom if your pc via bios can be made to autoboot the cdrom , or it
> lets you create a insalation boot disk to use the cdrom
>
> I hope that this may be of help
>
> Igor
> from land down under
>
> SOMERTON KENNEDY wrote:
>
> > try woven goods for linux
> >
> > "David .." wrote:
> > >
> > > Daniel wrote:
> > > >
> > > > Hi there. Anyone had tried to install Linux distribution totally from floppy
> > > > disk to hardisk?
> > > > Any Linux distribution to recommend?
> > > > I 'm new to Linux and want to use my old laptop (with cdrom) to surf net
> > > > using linux os.
> > > > Thanks.
> > >
> > > Slackware is one of the few that you can install from floppy
> > >
> > > http://www.slackware.com/
> > >
> > > --
> > > Confucius say: He who play in root, eventually kill tree.
> > > Registered with the Linux Counter. http://counter.li.org
> > > ID # 123538
--
Gregory D. Horne L I N U X .~.
Systems Analyst The Choice /V\
of a GNU /( )\
Generation ^^-^^
------------------------------
From: "Sylvain" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: Syslog problem
Date: Wed, 6 Sep 2000 18:58:06 +0200
Peter T. Breuer <[EMAIL PROTECTED]> a �crit dans le message :
8p3d1j$ug1$[EMAIL PROTECTED]
> In comp.os.linux.misc Kart <[EMAIL PROTECTED]> wrote:
>
> : "Peter T. Breuer" <[EMAIL PROTECTED]> wrote in message
> : news:8p2don$m1q$[EMAIL PROTECTED]...
> :> In comp.os.linux.misc Sylvain <[EMAIL PROTECTED]>
wrote:
> :> : Note : I boot the system with a read-only root fs and a ramdisk over
/var.
> :>
> :> A read only root fs is a neat trick. How do you make /dev/console
> :> writable?
> :>
> :> : Kernel is 2.2.16 patched with devfs. When rebooting with a read-write
root
> :> : fs, I have no problem at all !!
> :>
> :> Yeah, well, somehow I'm not entirely surprised.
> :>
> :> The few occasions I've tried to make / readonly, I've regretted it.
> :> Needs mods to mount and umount for one thing. It probably is feasible
> :> if you start off rw and ro and some later stage, after replacing
> :> /etc/mtab, /etc/motd, /etc/issue and friends with symlinks to /var.
> :> The real stinker is /dev.
>
> : This is not really helpful ... Perhaps could give ideas as which files
or
> : services could be involved in this problem ??
>
> ls -lrt /etc | tail
The only files which are modified (in /etc) when booting with a "normal"
RedHat 6.2 station :
/etc/issue
/etc/issue.net
/etc/mtab
/etc/HOSTNAME
/etc/ioctl.save
/etc/sysconfig/hwconf
The first two are created/modified by rc.sysinit or rc.local (don't
remember) and easy to handle.
mtab has been handled with a symlink to /proc/mounts.
HOSTNAME : the easier !!
ioctl.save : This file is created by init (man 8 init) and could be deleted.
hwconf : hardware config updated by kudzu. Not important.
The files in /etc (and subdirs) which are modified when restarting the
syslog after boot : none.
>
> Somone already noted that using devfs allevaites most symptoms!
Details on the subject ?
/dev is writable (even with a ro root fs) and syslogd creates its socket
(/dev/log). System calls does not return any error !
>
> Peter
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: timezone problem during boot
Date: Wed, 06 Sep 2000 16:45:19 GMT
once i had the same think then i made a resetup said "do not change
my hrdware time "
In article <8p5rmh$rqa$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> I started building some modules today, which caused me to notice a
boot
> problem with timezones.
>
> My hardware clock is local time. /etc/timezone contains "US/Central"
> (no quotes). /etc/default/rcS contains "UTC=no" (no quotes).
>
> Once the system is booted, the time is correct.
>
> The problem is that in rcS.d, S20modutils runs before S40hwclock.sh.
It
> is S40hwclock.sh that sets the time correctly, but S20modutils run
> depmod, which regenerates /lib/modules/2.2.13/modules.dep with the
> wrong time (5 hours too early). The files time is not my local time
> (GMT-5),
> nor is it GMT. It is being set to GMT-10. It is as if the HW clock
was
> GMT, and 5 hours are subtracted from it for local time.
>
> This problem then causes modprobe to complain about /etc/modules.conf
> being newer than modules.dep, which causes S20modutils to execute
> the last if statement (modprobe ... -t boot \*) when it shouldn't
> causing more error messages.
>
> If I edit S20 and add "touch /testtime" I get the same thing. I need
> the HW clock as local since the system is dual boot with Win98.
>
> libc6 2.1.3-10, Debian potato.
>
> Any suggestions?
> Thanks
> -Michael
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: "Frederic Vanneste" <[EMAIL PROTECTED]>
Subject: Re: Promise ATA100 and Linux
Crossposted-To:
alt.comp.periphs.mainboard.asus,alt.os.linux.slackware,comp.os.linux.hardware
Date: Wed, 06 Sep 2000 17:08:06 GMT
In article <[EMAIL PROTECTED]>, The Black Unicorn
<[EMAIL PROTECTED]> wrote:
>
> yep, I have the same problem with hdparm. no dma. which leaves me with a
> fantastic 4.7 M/sec throughput. hurrah for ATA100 ?
My prob is solved, I had used my old .config to compile my kernel - just forgot
to rename it!
Now it's not perfect, but I get about 20M/sec.
Just get the ide-patch from
ftp://ftp.be.kernel.org/pub/linux/kernel/people/hedrick/
(or some other mirror)
And compile it with support for dma and your card,
eg. the Promise Ultra66 is the PDC20262.
See the kernel docs for your card...
Good luck.
Frederic
"The charm of knowledge would be small,
were it not that so much shame has to be
overcome on the way to it."
'Friedrich Nietzsche'
------------------------------
** 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.setup) 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-Setup Digest
******************************
