Linux-Misc Digest #376, Volume #27 Fri, 16 Mar 01 23:13:02 EST
Contents:
Re: Are mount, unmount, shutdown, sync standalone utilities (Juergen Heinzl)
expert on midi needed (preamp)
Re: 10 gig disk in a 500 meg BIOS (David Efflandt)
how to cp only dot files? (Cevat Ustun)
Re: checking email add (David Efflandt)
Re: KDE in Redhat 6.2 is missing the taskbar on bottom of screen (Robert Jones)
Re: how to cp only dot files? (Michael Heiming)
Re: *.po and *.mo (Stefan Braun)
81 day uptime, then console fonts corrupted? (David Efflandt)
Re: real player 8 (Vladimir Florinski)
Re: real player 8 ([EMAIL PROTECTED])
Re: Less space used by Reiserfs (Reinhard Karcher)
XF86Config files (charlie)
Re: how to cp only dot files? (Floyd Davidson)
Scrambled mouse cursor in X11? (Doug O'Leary)
Re: 10 gig disk in a 500 meg BIOS (Juergen Pfann)
Re: SOS No modem com4??? (Albert Schwartz)
Re: Extracting the bootimage from a bootable (El Torito) CD ? (Juergen Pfann)
lockd question (Christopher W. Aiken)
Re: real player 8 ("Jeffrey S. Kline")
Kernel oops with X (Chris Spiegel)
Re: pipe to rm? (Joseph Guerrant)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Subject: Re: Are mount, unmount, shutdown, sync standalone utilities
Date: Fri, 16 Mar 2001 23:21:32 GMT
In article <984733608.770365@sj-nntpcache-3>, Siddharth Vajirkar wrote:
>Hi,
>Are there any libraries that utilities like mount, unmount, shutdown and
>sync use?
>Or are they standalone utilities?
[-]
.file "sync.S"
.version "01.01"
.text
.align 4
.globl main
.type main,@function
main:
#include <sys/syscall.h>
.text
.globl _main
main:
movl $0x24,%eax
int $0x80
movl $0x01,%eax
movl $0,%ebx
int $0x80
.L1:
.size main,.L1-main
[-]
It depends 8)
Just for the fun of it,
Juergen
--
\ Real name : Juergen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
From: preamp <[EMAIL PROTECTED]>
Subject: expert on midi needed
Crossposted-To: comp.os.linux,comp.os.linux.questions
Reply-To: [EMAIL PROTECTED]
Date: Sat, 17 Mar 2001 09:45:11 +1000
I wish to be able to play (on an internal synth and via the midi port on my
SBlive) and record midi (from the midi port) under mandrake 7.2, KDE 2.1.
Both under KDE2 and now 2.1, the midi/karaoke player reports "Couldn't
open /dev/sequencer. Probably there is another program using it" (but as
far as I know there isn't)
In the control centre under sound/midi, it shows no midi devices at all.
Can anyone tell me for sure that the SBLive supports midi under linux, and
if so, how to enable it?
------------------------------
From: [EMAIL PROTECTED] (David Efflandt)
Crossposted-To: comp.os.linux.hardware
Subject: Re: 10 gig disk in a 500 meg BIOS
Date: Fri, 16 Mar 2001 23:49:50 +0000 (UTC)
Reply-To: [EMAIL PROTECTED]
On Fri, 16 Mar 2001, Rick Griffiths <[EMAIL PROTECTED]> wrote:
>I've read the hard drive howto, and I am still unclear. Here is my
>question:
>
>Does it matter that the BIOS in my Nec Versa E (circa 1992) can't
>handle a 10 gig drive if I am using Linux exclusively on the machine?
>The way I read the howto, all I have to do is make a 5meg first
>partition as the root. Is this correct? If I make the first partition 5
>meg, will the installation system find another partition and use it if
>I mount the second partition with /root as the mount point?
A primary boot partition of around 16 MB entirely below 1024 cyl should
do. This gives you room for compiling and installing alternate
kernels. Once the kernel loads it has its own software BIOS that can
handle larger drives.
>Need I even do all this if I am only using Linux? Can I, in other
>words, make the first partition 5 gig and still expect the BIOS to find
>the master boot record in the first 1024 cylinders?
>
> Use reply-to address.
Unfortunately your reply-to address does not show up while composing a
reply.
Since you are only using Linux, you would put LILO in the MBR (which is at
the beginning of the disk and certainly below 1024 cyl, since I think the
MBR is only 512K). LILO and the kernel (in /boot) are what need to be
found using system BIOS. You might need to use the linear lilo option
(whatever the large hard disk HOWTO says).
When I had a 386 BIOS that did not support drive translation, I used a
Promise EideMAX add on card that worked with my existing drive controller
(or can add 3rd & 4th drive). It was $40 (1/2 the price of BIOS upgrade).
This allowed me to use large drives automatically without having to pay
attention to the large drive HOWTO (way back in Slackware 2.0) and have
never had to use linear mode.
--
David Efflandt [EMAIL PROTECTED] http://www.de-srv.com/
http://www.autox.chicago.il.us/ http://www.berniesfloral.net/
http://cgi-help.virtualave.net/ http://hammer.prohosting.com/~cgi-wiz/
------------------------------
From: Cevat Ustun <[EMAIL PROTECTED]>
Subject: how to cp only dot files?
Date: Fri, 16 Mar 2001 23:50:38 GMT
in other words,
cp -a .*
doesn't work...
Cev
------------------------------
From: [EMAIL PROTECTED] (David Efflandt)
Subject: Re: checking email add
Date: Sat, 17 Mar 2001 00:02:47 +0000 (UTC)
Reply-To: [EMAIL PROTECTED]
On 16 Mar 2001 15:36:29 GMT, Holland King <[EMAIL PROTECTED]> wrote:
>i am currently in the process of writing an admin tool that deletes accounts
>as needed, one of its jobs is to parse the .forward file and to set up
>aliases from what it finds in there. obviously there are some problems with
>just making each line an alias or even making each address it finds an
>alias. my boss had mentioned that there was some way to poll the address
>with some system calls and make sure it at least exist but was not sure
>about the implementation or details. does anyone know anything about this,
>or can anyone point me in the direction to do something like that. if not
>do you have other suggestions about how to get only the email addresses out
>of the file? thank you.
Some of the tools that you think might verify e-mail addresses (like
sendmail -bv) really only parse the address to see if it 'could' be valid.
The only sure way is to actually attempt to send e-mail using something
like sendmail -v [possibly using open3() in a Perl script] to monitor the
response. A crude example I used just to see if e-mail could get through
with sendmail -v and the normal method is
http://cgi-help.virtualave.net/pub/mailtest.txt
--
David Efflandt [EMAIL PROTECTED] http://www.de-srv.com/
http://www.autox.chicago.il.us/ http://www.berniesfloral.net/
http://cgi-help.virtualave.net/ http://hammer.prohosting.com/~cgi-wiz/
------------------------------
From: Robert Jones <[EMAIL PROTECTED]>
Subject: Re: KDE in Redhat 6.2 is missing the taskbar on bottom of screen
Date: Fri, 16 Mar 2001 18:08:36 -0600
mike wrote:
> Hi,
> a while after I installed Redhat 6.2 on my system the lower
> taskbar of KDE is missing. It is the one that has the
> button (K) for the main KDE menu, the shortcuts, as for example,
> the one to bring up an xterm, and the system time on the right hand
> of the screen.
> How can I restore KDE?
>
> Thanks
>
> Mike
>
> P.S. Is the taskbar a seperate program that can be reinstalled
> or is it majorly interwoven into the KDE desktop software?
You have described the *panel*, not the taskbar. The panel is hideable
by clicking a button at either end. The one on the left end is adjacent
to the "K" and you probably clicked it inadvertently, causing the panel
to hide. There should be a button at one side of the screen in the
vertical position normally occupied by the panel. Clicking this button
will restore the panel.
Also, both panel and taskbar can be configured to "auto-hide". I've
never tried it because I found the autohiding taskbar in Window$ to be
annoying as hell and I can't see that I'd like it any better here. I
can see where it might be of some use if you had a very small monitor
and needed all the space you could get to make you application windows
readable.
HTH
--
It seems intuitively obvious to me, which means that it might be wrong.
-- Chris Torek
5:36pm up 18 days, 10:07, 3 users, load average: 0.07, 0.07, 0.02
------------------------------
Date: Sat, 17 Mar 2001 01:14:25 +0100
From: Michael Heiming <[EMAIL PROTECTED]>
Subject: Re: how to cp only dot files?
Cevat Ustun wrote:
>
> in other words,
> cp -a .*
be carefull with .* as it includes .. !
> doesn't work...
>
> Cev
Try something like this:
ls -a . | grep ^\\.[a-zA-Z0-9] | xargs --replace=cdat cp cdat
/path_to/copy_to/
Good luck
Michael Heiming
------------------------------
Date: Fri, 16 Mar 2001 13:31:53 +0100
From: Stefan Braun <[EMAIL PROTECTED]>
Subject: Re: *.po and *.mo
Brian Lere wrote:
>
> Hi,
>
> Where can I get docs about *.po and *.mo? I think the files are
> about localization but I don't know how to understand it and
> create it.
>
Take a look at gettext (info gettext).
Stefan
------------------------------
From: [EMAIL PROTECTED] (David Efflandt)
Subject: 81 day uptime, then console fonts corrupted?
Date: Sat, 17 Mar 2001 00:25:06 +0000 (UTC)
Reply-To: [EMAIL PROTECTED]
Running Mandrake 7.0 with generic kernel 2.2.17, boot to runlevel 3,
ancient S3 Trio64 video. After 81 days of uptime (load ave 1.00 due to
SETI@home), I noticed that the fonts in the console were scrambled when I
switched to the console from gnome (white background with lines trough
it). After exiting X, I got the blinking cursor, but the font was
invisible even with inverse background.
In either case 'reset' did not work (which usually clears up graphic
fonts). 'clear' or ^L did not help. 'setfont' or 'consolechars -d' would
clear it up, but only temporarily. The next time I startx, the console
font corrupted again. Everything was fine from xterm or remote login
(telnet or ssh).
I shutdown the box for a few minutes, rebooted, and everything was back to
normal. Any idea what could cause the console fonts to switch to
something other than normal?
I don't think it is due to overheating, because the only time that
happened (inlet fan packed with lint), I just had drive access problems
(no data loss except abreviated logs).
--
David Efflandt [EMAIL PROTECTED] http://www.de-srv.com/
http://www.autox.chicago.il.us/ http://www.berniesfloral.net/
http://cgi-help.virtualave.net/ http://hammer.prohosting.com/~cgi-wiz/
------------------------------
From: Vladimir Florinski <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.misc
Subject: Re: real player 8
Date: Fri, 16 Mar 2001 17:21:38 -0700
"Jeffrey S. Kline" wrote:
>
> No there isn't and probably won't be anytime soon unless someone reverse
> engineers the current stuff.
>
> I get absolutely no response from Real.com about any inquiries on anything
> not tied to Microsoft platforms. Not even Mac!!! You'd think you even get
> the infamous "automated response" mail but not even this. The only official
> response I got on a phone call was that they only support Windows and
> Macintosh, and that I'd have to use those platforms. The versions that were
> available, are rather dated, weren't apparently coded by them directly, and
> no plans exist to support foreign operating systems. They even dropped the
> BeOS version they put on the Professional Be 5.x CD's...
>
rpm -qi RealPlayer
Name : RealPlayer Relocations: (not relocateable)
Version : 8.0 Vendor: (none)
Release : 1 Build Date: Thu Dec 21 10:14:33 2000
Install date: Tue Dec 26 22:39:38 2000 Build Host: wellington
Group : Applications/Multimedia Source RPM: RealPlayer-8.0-1.src.rpm
Size : 11907641 License: commercial
Packager : [EMAIL PROTECTED]
Summary : Welcome to RealPlayer 8.0!
Description :
RealPlayer 8.0 (C) 2000 RealNetworks, Inc.
Streaming audio/video/flash/pix/text player.
--
Vladimir
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: real player 8
Crossposted-To: linux.redhat.misc
Date: Sat, 17 Mar 2001 01:03:26 GMT
In comp.os.linux.misc Jeffrey S. Kline <[EMAIL PROTECTED]> wrote:
> No there isn't and probably won't be anytime soon unless someone reverse
> engineers the current stuff.
Hmm.. Don't know about your side.. I've got RP 8 working fine on my Linux
box here... Granted, you have to tell RPM to ignore the arch type on the
RPM, but that is well known...
Also note that all of their unix players (among others) are "community
supported software," meaning??? Hmm... Community support - ie. not Real Inc
supported directly.
Kris
------------------------------
Date: Fri, 16 Mar 2001 17:45:11 +0100
From: Reinhard Karcher <[EMAIL PROTECTED]>
Subject: Re: Less space used by Reiserfs
Rod Smith <[EMAIL PROTECTED]> wrote:
>In article <0asr6.903$[EMAIL PROTECTED]>,
> "Justin R. Smith" <[EMAIL PROTECTED]> writes:
>> I'm running the 2.4.2 kernel and
>> I converted my /usr/local partition to the reiserfs
>>
>> The odd thing is that the amount of disk space used has gone down by more
>> than half. Before the conversion, df reported 61% of the partition used,
>> and afterwards, it reports 25%! Yet nothing APEARS to be missing...
>>
>> Possible reasons:
>>
>> 1. df underreports reiserfs partitions
>>
>> 2. the reiserfs is more efficient in storing small files (this partition
>> contained thousands of very small files).
>ReiserFS is definitely more efficient at storing small files. Still,
>the drop from 61% filled to 25% seems pretty extreme. What are the
>exact figures on the number of files and total disk space used? I might
>accept this if the files were VERY tiny, on average, although I don't
>know precisely HOW tiny they'd have to be. You might try checking a
>bunch of the files to be sure they're intact.
If your inode-size was 4k and your mean filesize is 1k, you gain a lot of
space.
Reinhard
------------------------------
From: charlie <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.portable
Subject: XF86Config files
Date: Fri, 16 Mar 2001 21:02:03 -0500
Reply-To: [EMAIL PROTECTED]
What's the difference between XF86Config and XF86Config-4? Are both files
used if I have XFree86 4.0.1 installed? If I want to add a second input
device to get a touchpad and a mouse both work at the same time, do I have
to add new sections in both of these files?
------------------------------
From: Floyd Davidson <[EMAIL PROTECTED]>
Subject: Re: how to cp only dot files?
Date: 16 Mar 2001 15:51:50 -0900
Cevat Ustun <[EMAIL PROTECTED]> wrote:
>in other words,
>cp -a .*
>doesn't work...
Try something like
cp -a .[A-z]* /newdir
You might need to add to that regular expression if you have
some strange filenames.
--
Floyd L. Davidson <http://www.ptialaska.net/~floyd>
Ukpeagvik (Barrow, Alaska) [EMAIL PROTECTED]
------------------------------
From: Doug O'Leary <[EMAIL PROTECTED]>
Subject: Scrambled mouse cursor in X11?
Date: Fri, 16 Mar 2001 20:01:12 -0800
Hey, all;
I just got a new monitor, plugged it in, reran the Xconfigurator,
selected the right monitor, what not.
When I get into X11, everything looks great except the mouse cursor; it's
all scrambled and shifting back and forth. Only the mouse cursor.
Anyone know what's going on and better yet how to fix it?
Thanks.
Doug
--
===================
Douglas K. O'Leary
Senior System Admin
[EMAIL PROTECTED]
------------------------------
From: Juergen Pfann <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.hardware
Subject: Re: 10 gig disk in a 500 meg BIOS
Date: Sat, 17 Mar 2001 02:57:55 +0100
David Efflandt wrote:
>
> A primary boot partition of around 16 MB entirely below 1024 cyl should
> do. This gives you room for compiling and installing alternate
> kernels. Once the kernel loads it has its own software BIOS that can
> handle larger drives.
>
I'd like to add some considerations espec. on a separate /boot partition
:
Most CHS-LBA translation schemes I've seen recently use a mapping of
255 heads and 63 sectors, thus 1 cylinder in this mapping is
16065 sectors or nearly 8 MB. Normally, Linux fdisk prints these
numbers at the beginning.
Linux fdisk sets partition sizes to multiples of cylinders, as this
is the recommended, safe partitioning scheme across operating systems.
Thus, I recommend to stick to this and divide your HD(s) in (c)fdisk
in cylinder units anyway.
I still regard 1 cylinder / 8 MB for the /boot partition to be plenty
enough for most of us - as you usually don't compile your kernels *in*
there (remember this would require >100 MB for 2.4.x), you just _store_
your kernel, the System.map, LILO' secondary loaders (boot.b, chain.b
etc.) and - probably most important - LILO's "map" file there.
Even with a "bloated" 2.4 kernel and its System.map, you're unlikely
to exceed 2 MB - thus, 16 MB is overkill IMHO.
OTOH, with nowadays' HD sizes, it doesn't matter too much, though.
Juergen
------------------------------
From: Albert Schwartz <[EMAIL PROTECTED]>
Subject: Re: SOS No modem com4???
Date: Fri, 16 Mar 2001 21:35:49 -0500
On Fri, 16 Mar 2001 00:40:45 -0500, Glitch <[EMAIL PROTECTED]> wrote:
>Albert Schwartz wrote:
>
>> My modem is set on com4. Comp. is Pent111, Win98 + Linux Caldero
>> USR Sportster, 56k (no Winmodem).
>> Installed Caldero couple days ago. Had Mandrake 7.1 before.
>> Linux will not find Modem. Msg, Modem busy, or sorry cannot open.
>> I have tried KDE, dev/modem, and tty 0 to 3. No good.on any. Tho I
>> think the setting for com 4 should be tty3.
>>
>> Win98 works fine, linux also works fine except for modem.
>> Used OS\2 but want to change to Linux.
>> I need to keep com port 4. Others in use.
>> Linux finds modem on a Pent1 with the same modem, but com port 2 is
>> used.
>> What do I need to get Linux to recognize Com4???
>>
>> Help much appreciated.
>> It took me 84 years to get this dumb.
>>
>> Albert, Wa3fib.
>
>are u using tty3? You should be using ttyS3, note the capital "S".
>
>ln -s /dev/modem /dev/ttyS3
>
>will link 'modem' to ttyS3
>
>try that.
IT WORKS!!! and we thank you very much.
I needed this.
Albert.
------------------------------
From: Juergen Pfann <[EMAIL PROTECTED]>
Subject: Re: Extracting the bootimage from a bootable (El Torito) CD ?
Date: Sat, 17 Mar 2001 03:33:19 +0100
Rainer Krienke wrote:
>
> In article <[EMAIL PROTECTED]>,
> David <[EMAIL PROTECTED]> writes:
> >
> > dd if=/path/to/boot.img of=/dev/fd0 bs=1440k
> >
>
> This is roughly the image but not exactly. It seems that there is still a header
> in front of the image because the first 1440K contain for example the options of
> mkisofs, the command the ISO filesystem on the CD was created with.
> Thats why I asked for a tool that exactly extracts only the image leaving all
> other iso9660 stuff alone).
>
If the image's size is 1,474,560 bytes, then it *is* an image of each
and every single byte of a "1.44" MB floppy disk, period.
What doesn't matter here is the name of that file, of course.
And maybe you mix it up with a file normally called "boot.catalog"
(2048 bytes IIRC) that for instance mkisofs creates on the CD image.
To my understanding that's similar to LILO's "map" file: a map where
to find the bootable sectors, or machine code to load the floppy
image via BIOS function calls - not sure about that, I admit.
Anyway, you don't need the "boot.catalog" when creating the bootable
floppy - only the image itself, I repeat that.
BTW : the "bs" argument to dd isn't really necessary, it just speeds
up the transfer.
Juergen
------------------------------
From: [EMAIL PROTECTED] (Christopher W. Aiken)
Subject: lockd question
Date: Sat, 17 Mar 2001 03:17:30 -0000
Reply-To: [EMAIL PROTECTED]
On one of our office machines (running Linux) we have 6 nfs
mounted file servers. the /var/log/messages file (and the system
console) is full of the following errors:
Mar 14 05:46:37 linux64 kernel: lockd: failed to monitor 10.3.11.10
Mar 14 05:46:37 linux64 kernel: lockd: cannot monitor 10.3.11.10
Any clue as to what that error is? Why only the server "10.3.11.10"
and not the rest on the mounted servers? We can get/use/put files
to that server w/o any problems. The /var/log/messages file is MB's in
size.
BTW... What is "lockd"? Can it be removed?
-=[cwa]=-
--
Christopher W. Aiken
Scenery Hill, Pa, USA
chris at cwaiken dot com
www.cwaiken.com
SuSE 7.1 Professional Linux
------------------------------
From: "Jeffrey S. Kline" <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.misc
Subject: Re: real player 8
Date: Fri, 16 Mar 2001 21:21:40 -0600
Yea, I know that. Duh... <grin> Apparently though, it's mostly the
"codec's" that are at issue and with those codec's installed, it should be
theoretically possible to have KDE's media player show stuff but they don't
release anything. I have followed all the instructions and doco's I can
find on RP8 under Linux and have not been able to get it going with anything
but local files. It doesn't stream like the Windoz side does. The version
they put out with BeOS also has the same problem. I can read local files but
it doesn't take streams either from across net... (unless it's downloaded as
a file then opened)... Our local TV station has a web site with numerous
Realvideo feeds of stories and such, and non of them work at all except
under Winblows...
Don't get me wrong on it as I like it but since it doesn't support all the
"claimed" feature sets, it's pretty much useless to me. I like the effort of
Quicktime going on much better personally. The codec's there appear to
function much better and more efficiently than does realaudio/realvideo
codecs. Plus, there already is quite a bit of supported Quicktime stuff out
there that does work! <grin>
Cheers;
Jeff
<[EMAIL PROTECTED]> wrote in message
news:yJys6.3261$[EMAIL PROTECTED]...
> In comp.os.linux.misc Jeffrey S. Kline <[EMAIL PROTECTED]>
wrote:
> > No there isn't and probably won't be anytime soon unless someone reverse
> > engineers the current stuff.
>
> Hmm.. Don't know about your side.. I've got RP 8 working fine on my
Linux
> box here... Granted, you have to tell RPM to ignore the arch type on the
> RPM, but that is well known...
>
> Also note that all of their unix players (among others) are "community
> supported software," meaning??? Hmm... Community support - ie. not Real
Inc
> supported directly.
>
> Kris
------------------------------
From: [EMAIL PROTECTED] (Chris Spiegel)
Subject: Kernel oops with X
Date: 17 Mar 2001 03:40:31 GMT
I get fairly regular kernel oopses generated by X. At least 2 a week,
sometimes a couple times daily. Running the output through ksymoops
gives:
>>EIP; c012096d <kmem_cache_free+14d/174> <=====
Trace; c021d0ea <uhci_unlink_urb_sync+122/150>
Trace; c021d3c5 <uhci_unlink_urb+4d/68>
Trace; c0211c9a <usb_unlink_urb+1e/2c>
Trace; c02181ce <hid_close+26/2c>
Trace; c0210ef9 <input_close_device+15/20>
Trace; c021adbe <mousedev_release+7e/108>
Trace; c012494b <__fput+1f/48>
Trace; c0125a0f <fput+17/48>
Trace; c01249c6 <filp_close+52/5c>
Trace; c0124a2d <sys_close+5d/68>
Trace; c0109030 <system_call+34/38>
Code; c012096d <kmem_cache_free+14d/174>
00000000 <_EIP>:
Code; c012096d <kmem_cache_free+14d/174> <=====
0: c7 05 00 00 00 00 00 movl $0x0,0x0 <=====
Code; c0120974 <kmem_cache_free+154/174>
7: 00 00 00
Code; c0120977 <kmem_cache_free+157/174>
a: eb 12 jmp 1e <_EIP+0x1e> c012098b
<kmem_cache_free+16b/174>
Code; c0120979 <kmem_cache_free+159/174>
c: 8d 76 00 lea 0x0(%esi),%esi
Code; c012097c <kmem_cache_free+15c/174>
f: 56 push %esi
Code; c012097d <kmem_cache_free+15d/174>
10: 53 push %ebx
Code; c012097e <kmem_cache_free+15e/174>
11: 68 7e 60 00 00 push $0x607e
Since this has begun, I've replaced my motherboard and mouse (for
different reasons), so it doesn't appear to be a problem with either of
those. This is kernel 2.2.18 on a PII/300 (no overclocking) with 128MB
of RAM. It only happens when I'm trying to switch from X to the
console, but I can't intentionally cause it to happen.
The actual generated oops is:
kmem_free: Bad obj addr (objp=c5ff5040, name=urb_priv)
Unable to handle kernel NULL pointer dereference at virtual address
00000000
current->tss.cr3 = 0507b000, %cr3 = 0507b000
*pde = 00000000
Oops: 0002
CPU: 0
EIP: 0010:[<c012096d>]
EFLAGS: 00010292
eax: 0000003a ebx: c5ff5040 ecx: 00000009 edx: 00000032
esi: c7fefb00 edi: 00000292 ebp: 00000004 esp: c507df00
ds: 0018 es: 0018 ss: 0018
Process X (pid: 206, process nr: 39, stackpage=c507d000)
Stack: c5ff5040 c7f89900 c5ff507c 00000002 c021d0ea c7fefb00 c5ff5040 \
c7f91994
c7f91994 c3facb60 bffff620 c021d3c5 c7f89900 c7dd4c40 c7f91994 \
c0211c9a
c7dd4c40 c02181ce c7dd4c40 c0210ef9 c7dd4e44 c7f91980 c021adbe \
c7f91994
Call Trace: [<c021d0ea>] [<c021d3c5>] [<c0211c9a>] [<c02181ce>] \
[<c0210ef9>] [<c021adbe>] [<c012494b>] [<c0125a0f>] [<c01249c6>] \
[<c0124a2d>] [<c0109030>]
Code: c7 05 00 00 00 00 00 00 00 00 eb 12 8d 76 00 56 53 68 7e 60
Thanks for any insight you can give me,
Chris Spiegel
------------------------------
From: [EMAIL PROTECTED] (Joseph Guerrant)
Subject: Re: pipe to rm?
Date: 17 Mar 2001 02:57:22 GMT
On Wed, 14 Mar 2001 12:53:31 -0800, Steve Bui <[EMAIL PROTECTED]> wrote:
In comp.os.linux.misc, you wrote:
>Is there an easy way to remove a list of files generated by a command
>such as locate?
I use this command, kilf, to do it. Only removes if it's a file,
and echoes the filename. I wrote it myself. (An original patentable
idea? I suspect not... ;) You'd use it like this:
find-locate-echo-etc-allmyhopefullynonvaluablefiles | kilf
etc.
You could also put the list into a file, edit the file, then:
kilf < listfilename
Use at your own risk.
For safety, some people make commands like this that'd also put
the files in some trash dir, etc. - perhaps with serial#'s added
to names so as to avoid overwriting dupenames, and maybe a cron
script to delete regularly. (The overhead is just fiddling with
the filesystem database, not really moving all the data, but
maybe still considerable if lots of files.)
You could also/or log the names of all deleted files.
All pretty easy to do in the Bourne shell. Just add a few lines
of script. And, the general idea can be extended to all kinds of
new commands to enhance Bourne shell scripting speed and ease.
You can also use this in the Vi (or similar) editor. If lines
5,10 had names of files to delete:
:5,10!kilf
The actual files would be deleted, replaced with the std output
of this command, which was what you started with - the names. If
some of the names weren't regular files, they'd not get echoed
back in the editor.
I think you could do the same in maybe Pico, or Joe, or some of
that category of editors anyway.
#! /bin/bash
###############
#kilf - rm file list from standard input, outputing names.
while read line
do
if [ -f "$line" ]
then
echo "$line"
rm "$line"
fi
done
------------------------------
** 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 by posting to comp.os.linux.misc.
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-Misc Digest
******************************
