Linux-Development-Sys Digest #751, Volume #6 Fri, 28 May 99 09:14:30 EDT
Contents:
Re: rmdir() source code (Alexander Viro)
How to get eth_stats (Holger Voigt)
Re: rmdir() source code ("cjc86")
Re: Terabite Plus Filesystems (Marc SCHAEFER)
Re: Large CD-ROM file errors...? (brian moore)
Re: Large CD-ROM file errors...? (Peter Samuelson)
event variable ("cjc86")
sci cluster (Klaus Leopold)
Re: Large CD-ROM file errors...? (Mark Tranchant)
Re: SMP needlessly migrating processes between processors? (Robert Kaiser)
2.2.9 is stable? (lckun)
Re: Large CD-ROM file errors...? (Mogens Kjaer)
Re: Booting Diskless Client - Linux 2.2.6 (Andy Pershin)
[q] 2.2.9 is stable? (lckun)
Re: RedHat 6.0 SMP compile (Andrew Daviel)
Re: Large CD-ROM file errors...? (Mark Tranchant)
Re: C++ IDE ? (Ralf Gerwien)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Alexander Viro)
Subject: Re: rmdir() source code
Date: 28 May 1999 01:33:03 -0400
In article <7il938$26vi$[EMAIL PROTECTED]>,
cjc86 <[EMAIL PROTECTED]> wrote:
>
>>
>> Yes, it does. It's done by dput(dentry); after double_unlock()
>
> I think the dput(dentry) after double_unlock() should correspond to
> lookup_dentry(path) in the first line of the function. I don't think
> it's used to decrement dentry->d_count which is incremented by
> dentry->d_count++;
As you like it. Result is the same - both that place and dput()s in
double_unlock() decrease d_count. Notice that double_lock() doesn't
increase d_count on anything. Another dput() in double_unlock() corresponds
to dget(dentry->d_parent);
--
"You're one of those condescending Unix computer users!"
"Here's a nickel, kid. Get yourself a better computer" - Dilbert.
------------------------------
From: Holger Voigt <[EMAIL PROTECTED]>
Crossposted-To: comp.programming
Subject: How to get eth_stats
Date: Thu, 27 May 1999 17:43:14 +0200
Hi there,
can anybody told me how to get ethernet-device statistics (collisions,
etc) from linux without using /proc?
I think it must be available by using ioctl but i can't figure out how.
thanx
------------------------------
From: "cjc86" <[EMAIL PROTECTED]>
Subject: Re: rmdir() source code
Date: Fri, 28 May 1999 13:21:17 +0800
>
> Yes, it does. It's done by dput(dentry); after double_unlock()
>
>
I think the dput(dentry) after double_unlock() should correspond to
lookup_dentry(path) in the first line of the function. I don't think
it's used to decrement dentry->d_count which is incremented by
dentry->d_count++;
------------------------------
From: Marc SCHAEFER <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.ms-windows.nt.admin.misc,comp.sys.sun.admin,comp.sys.hp.misc,comp.os.linux.hardware
Subject: Re: Terabite Plus Filesystems
Date: 27 May 1999 14:31:47 GMT
In comp.os.linux.development.system Jake Maizel <[EMAIL PROTECTED]> wrote:
: We are building a system that needs to handle a huge number of files
: that are 500KB-1MB in size (1-2TB total). Our only constraint right now
To my knowledge, today's ext2fs on ix86 hardware can support file
systems upto 2 TB (2^31 blocks of 512 bytes). However, file sizes
are limited to 2 GB (which is not a problem in your case).
Now, ext2fs probably doesn't have a problem handling millions of files
in such large a fs.
Also, you might benefit of Linux's RAID5 features to create this huge
device. However, Linux, at this time, as no volume management: you cannot,
for example, extend a filesystem by giving it a few more
disks. However, if your files are on /big-disk/area1 and /big-disk/area2,
area1 and area2 can live on different filesystems, making expansion
possible, but painful.
------------------------------
From: [EMAIL PROTECTED] (brian moore)
Crossposted-To: comp.os.linux.misc
Subject: Re: Large CD-ROM file errors...?
Date: 28 May 1999 06:20:52 GMT
On Wed, 26 May 1999 08:09:01 +0100,
Mark Tranchant <[EMAIL PROTECTED]> wrote:
> Matt Starnes wrote:
> >
> > It might have something to do with filesystems. The software we use in
> > Windoze 95 defaults to create a Joliet filesystem on the CD-ROM which is I
> > believe different from the standard ISO9660 format. You might want to
> > reburn it and check that. Or you can compile Joilet support into your
> > kernel.
>
> Well, thanks for the effort, but to quote myself: "I booted up Linux
> 2.2.9, with full CD support including Joliet compiled in)"
And precisely how does Joliet handle such things as owners, permissions,
symlinks and other features of a typical Unix filesystem? We won't even
get into how device files would work on a Joliet CD.
You don't want to burn it in Joliet: sure, Linux will read it, but it
will be missing important things like symlinks and (probably why it
doesn't work) device files.
--
Brian Moore | "The Zen nature of a spammer resembles
Sysadmin, C/Perl Hacker | a cockroach, except that the cockroach
Usenet Vandal | is higher up on the evolutionary chain."
Netscum, Bane of Elves. Peter Olson, Delphi Postmaster
------------------------------
From: [EMAIL PROTECTED] (Peter Samuelson)
Crossposted-To: comp.os.linux.misc
Subject: Re: Large CD-ROM file errors...?
Date: 28 May 1999 01:56:00 -0500
Reply-To: Peter Samuelson <[EMAIL PROTECTED]>
[Mark Tranchant <[EMAIL PROTECTED]>]
> > Well, thanks for the effort, but to quote myself: "I booted up
> > Linux 2.2.9, with full CD support including Joliet compiled in)"
[brian moore <[EMAIL PROTECTED]>]
> And precisely how does Joliet handle such things as owners,
> permissions, symlinks and other features of a typical Unix
> filesystem? We won't even get into how device files would work on a
> Joliet CD.
Of course. You wouldn't expect Microsoft to come up with something as
useful as Rock Ridge, now, would you? *They* don't have any use for
symlinks....
> You don't want to burn it in Joliet: sure, Linux will read it, but it
> will be missing important things like symlinks and (probably why it
> doesn't work) device files.
Bzzzt ... he burned a single 70MB tarfile, not a whole directory tree.
(Read the original post.) The issue is that Linux shows a truncated
(16MB) file instead of the whole thing. Needless to say, Windoze sees
the CD the way it wrote it.
--
Peter Samuelson
<sampo.creighton.edu!psamuels>
------------------------------
From: "cjc86" <[EMAIL PROTECTED]>
Subject: event variable
Date: Fri, 28 May 1999 14:28:16 +0800
Hi, All
I often suffer from a variable called "event" in linux source code.
It seems that it is used as following:
f_version = event++;
i_version = event++;
...
Could anybody tell me what the "event" variable is used for ?
Thanks ...
------------------------------
From: Klaus Leopold <[EMAIL PROTECTED]>
Subject: sci cluster
Date: Fri, 28 May 1999 09:17:57 +0200
hello!
i=B4ve 3 sci dolphin pci adapters and want to make a linux cluster with
them. now i=B4m searching for interesting and informative url=B4s how to =
do
this best.
thanks in advance, Klaus.....
-- =
Klaus Leopold
University of Klagenfurt
Institute of Information Technology Tel: ++43(0)463/2700-863
Research group: Systemintegration Fax: ++43(0)463/2700-867
mailto:[EMAIL PROTECTED]
http://www.ifi.uni-klu.ac.at/cgi-bin/staff_home?kleop
------------------------------
From: Mark Tranchant <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: Large CD-ROM file errors...?
Date: Fri, 28 May 1999 08:27:18 +0100
Reply-To: [EMAIL PROTECTED]
brian moore wrote:
>
> On Wed, 26 May 1999 08:09:01 +0100,
> Mark Tranchant <[EMAIL PROTECTED]> wrote:
> > Matt Starnes wrote:
> > >
> > > It might have something to do with filesystems. The software we use in
> > > Windoze 95 defaults to create a Joliet filesystem on the CD-ROM which is I
> > > believe different from the standard ISO9660 format. You might want to
> > > reburn it and check that. Or you can compile Joilet support into your
> > > kernel.
> >
> > Well, thanks for the effort, but to quote myself: "I booted up Linux
> > 2.2.9, with full CD support including Joliet compiled in)"
>
> And precisely how does Joliet handle such things as owners, permissions,
> symlinks and other features of a typical Unix filesystem? We won't even
> get into how device files would work on a Joliet CD.
>
> You don't want to burn it in Joliet: sure, Linux will read it, but it
> will be missing important things like symlinks and (probably why it
> doesn't work) device files.
>
But all I wanted was a single long-file-named 70MB file...
Bottom line: Linux couldn't read it, Windows 95 could.
Mark.
------------------------------
From: [EMAIL PROTECTED] (Robert Kaiser)
Subject: Re: SMP needlessly migrating processes between processors?
Date: 28 May 1999 09:09:08 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Arun Sharma) writes:
> On Thu, 27 May 1999 01:22:42 GMT, bryan <[EMAIL PROTECTED]> wrote:
>
> But if your goal is to bind a process to a particular processor, you
> should use one of the processor binding patches floating around.
>
Could you (or anyone else) please post a few related URLs ? The only
thing I was able to locate is Rik van Riel's "scheduler bigpatch"
at http://www.nl.linux.org/~riel/patches/schedule-bigpatch-2.2.5-2
which _appears_ to support processor binding (not really sure though).
Thanks
Rob
================================================================
Robert Kaiser email: rkaiser AT sysgo DOT de
SYSGO RTS GmbH
Mainz / Germany
------------------------------
From: lckun <[EMAIL PROTECTED]>
Subject: 2.2.9 is stable?
Date: Fri, 28 May 1999 18:21:16 +0900
Hi
I have 2.2.9 running on Pentium II machines (dual 450 MHz Pentium-MMX).
It seems for me to stable unless i don't make a debugging with printk in
sched.c.
But why hangs the system always, if i use some debugging in
sched.c??????
Anyone can tell me why???
Regards
lckun
------------------------------
From: Mogens Kjaer <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: Large CD-ROM file errors...?
Date: Fri, 28 May 1999 13:06:26 +0200
Mark Tranchant wrote:
>
> I got a friend to burn the StarOffice 5.1 download onto a CD-R for me
> rather than taking it home on 51 floppies. This CD-R was burned under
> Windows 95, with a file name of so51_lnx_01.tar (note: a long file
> name!).
>
> On getting home, I booted up Linux (2.2.9, with full CD support
> including Joliet compiled in) and tried to copy the 70.6MB file.
> However, Linux could only see the first 16MB or thereabouts. The copy
> succeeded and tar de-archived the file fine up to the truncation.
>
> So I tried DOS (7.0), which saw it as many files of about 650KB each,
> all with the same name. Argh!
>
> Windows 95 read it correctly and copied it fine, although it took *ages*
> (about 20 minutes on a 486DX4/120 with 24x EIDE drive).
Was this done on the same machine (i.e. the same CDROM drive)
as the one running Linux? The fact that it takes a long time to
read this cd in windows might suggest that it
is more or less defect. Maybe some drives are better for error
correction,
maybe windows allows for more retries...
I have seen many examples of problems with homeburned cd's, some won't
be read by some cdrom drives, and some will.
Mogens
--
Mogens Kjaer, Carlsberg Laboratory, Dept. of Chemistry
Gamle Carlsberg Vej 10, DK-2500 Valby, Denmark
Phone: +45 33 27 53 25, Fax: +45 33 27 47 08
Email: [EMAIL PROTECTED] Homepage: http://www.crc.dk
------------------------------
From: Andy Pershin <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking
Subject: Re: Booting Diskless Client - Linux 2.2.6
Date: 28 May 1999 06:37:54 GMT
Hello!
Tom Daley <[EMAIL PROTECTED]> wrote:
>> Experiencing weirdness with Linux 2.2.6 kernel booting on a diskless
>> Alpha client. When mounting the local filesystems I see:
>>
>> RPC: sendmsg returned error 51
>> RPC: sendmsg returned error 51
>> RPC: sendmsg returned error 51
>> RPC: sendmsg returned error 51
>> portmap: server localhost not responding, timed out
>> lockd_up: makesock failed, error=-5
>>
>> This happens numerous times. However, if I wait long enough all my
>> filesystems get mounted properly. Anybody else see this? Is there a
>> patch?
>>
TD>
TD> I see this as well. I have not figured it out yet, so I still run
TD> 2.0.36 on the diskless client. The server is running 2.2.7.
Run portmap before mounting remote fs.
--
Andy Pershin [ Registered Linux user #71727 ]
[ mailto:[EMAIL PROTECTED] ][ http://www.penza.com.ru/~apa ]
------------------------------
From: lckun <[EMAIL PROTECTED]>
Subject: [q] 2.2.9 is stable?
Date: Fri, 28 May 1999 20:39:40 +0900
lckun wrote:
> Hi
>
> I have 2.2.9 running on Pentium II machines (dual 450 MHz Pentium-MMX).
> It seems for me to stable unless i don't make a debugging with printk in
> sched.c.
> But why hangs the system always, if i use some debugging in
> sched.c??????
> Anyone can tell me why???
>
> Regards
>
> lckun
After using for kernel output, kernel hangs almost at booting "Starting
cron daemon: "
lckun
------------------------------
From: [EMAIL PROTECTED] (Andrew Daviel)
Subject: Re: RedHat 6.0 SMP compile
Reply-To: [EMAIL PROTECTED]
Date: Fri, 28 May 1999 11:56:57 GMT
Rolf Welde Skeie ([EMAIL PROTECTED]) wrote:
: No, there is nothing wrong with the SMP support.
: The problem lies in the proper procedure (which of course is
: undocumented...).
Thanks for the tip re. EXTRAVERSION
I have been having very similar problems.
I have RedHat 6.0 on a dual Celeron system with one UIDE drive, one
SCSI on ASUS mb with Adaptec AIC-7890/1 ; I did an ftp install
and selected PPro/686 and SMP, so got (I think)
kernel-2.2.5-15.i686.rpm
kernel-BOOT-2.2.5-15.i386.rpm
kernel-headers-2.2.5-15.i386.rpm
kernel-smp-2.2.5-15.i686.rpm
Due to some glitch I couldn't boot LILO off the SCSI so
put it on the IDE with the boot images on SCSI. So far so good.
Then I wanted to build a kernel - not sure why really; wanted to try -march=686 -O3
with the installed egcs, and I wanted to build the sensor modules, and watchdog support
(though that is probably in the RedHat build).
Loaded kernel-source-2.2.5-15.i386.rpm ....
when I did make config; make dep; make bzImage ; make modules ; make modules_install ;
mkinitrd
then depmod -a I got a lot of errors about unresolved symbols, including some for the
adaptec
driver and my 3com card, so when I tried booting the system hung.
I just tried setting EXTRAVERSION to -15smp2 (in the RPM it is "-15"
and going through the procedure you outline, viz.
( /etc/conf.modules contains
alias eth0 3c59x
alias scsi_hostadapter aic7xxx
pre-install pcmcia_core /etc/rc.d/init.d/pcmcia start
)
make xconfig - set SMP and PPro/686
make dep
make clean
make bzImage
make install
make modules
make modules_install
insmod loop
mkinitrd -f /boot/initrd-2.2.5-15smp2.img 2.2.5-15smp2
cat <<! >>/etc/lilo.conf
image=/boot/vmlinuz-2.2.5-15smp2
label=linux225smp2
root=/dev/sda1
initrd=/boot/initrd-2.2.5-15smp2.img
read-only
!
lilo
When I try to boot this I get
lib/aic7xxx.o
unresolved symbol __global_sti
and then
kmod - failed to exec
/sbin/modprobe -s -k block major-8 error=2
I just downloaded a kernel tarball for 2.3.2
and did a similar exercise. Somehow the SCSI driver module
didn't get loaded. So I rebuilt it with the aic7xxx driver
resident. That boots, and loads the 3com driver, even.
(had to comment out some line in st.c pnet->waiting or somesuch..)
Doing "depmod -e -a" I get (only)
/lib/modules/2.3.2smp/block/floppy.o: unresolved symbol(s)
blk_ioctl
perhaps due to my stupidity in making the floppy driver a module instead of
resident.....
now .. on to building sensors .. make .. install .. ldconfig . depmod .. sensors .
it works; amazing; says my m/b's at 30C .
So .. I still don't understand why I can't rebuild a RedHat 6.0 kernel from the RPMs.
If the headers don't match the kernel, won't that mean that any extra modules (like
sensors)
that anyone tries to build will fail, even without trying to build a new kernel ?
--
Andrew Daviel
http://vancouver-webpages.com/andrew
Deniable unless digitally signed.
------------------------------
From: Mark Tranchant <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: Large CD-ROM file errors...?
Date: Fri, 28 May 1999 12:47:37 +0100
Reply-To: [EMAIL PROTECTED]
Mogens Kjaer wrote:
> Was this done on the same machine (i.e. the same CDROM drive)
> as the one running Linux? The fact that it takes a long time to
> read this cd in windows might suggest that it
> is more or less defect. Maybe some drives are better for error
> correction,
> maybe windows allows for more retries...
>
> I have seen many examples of problems with homeburned cd's, some won't
> be read by some cdrom drives, and some will.
>
Yes, all read attempts were on the same machine. Linux read the
truncated first 16MB very quickly (for my machine) to disk, with no
problems and no apparent retries. tar unarchived it quite happily up to
the cutoff point, so there was no data corruption.
It didn't sound defective. I have had disks that force retries, and you
can hear the lens tracking backwards and forwards. This disk just read a
bit, waited for a bit, read a bit more, etc. I monitored the file system
read/write with System Monitor - it was "pulsing" at about 500KB/s
briefly about once every two seconds, and eventually (20 minutes later)
completed successfully - StarOffice installed perfectly. CPU utilization
was at 100% throughout the operation. My Windows CD-ROM performance is
usually better than that, holding a steady 800KB/s or so, rather than
the 60KB/s of this transfer.
Mark.
------------------------------
From: Ralf Gerwien <[EMAIL PROTECTED]>
Subject: Re: C++ IDE ?
Date: Fri, 28 May 1999 14:03:45 +0200
This is a multi-part message in MIME format.
==============D0F8DA858AB56499BFF1B75D
Content-Type: text/plain; charset=iso-8859-1; x-mac-type="54455854";
x-mac-creator="4D4F5353"
Content-Transfer-Encoding: 8bit
Manohar Singh wrote:
> hello there,
> does anyone amongst u know of an IDE (C++) for Linux that i can download
> from the net ?
Take a look at the following sites:
www.xshare.com
freshmeat.net
bonn.linuxberg.com
www.fokus.gmd.de/linux
I am sure you'll find there what you need.
(sorry for my !"§$%&/ english).
==============D0F8DA858AB56499BFF1B75D
Content-Type: text/x-vcard; charset=us-ascii;
name="gerwien.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Ralf Gerwien
Content-Disposition: attachment;
filename="gerwien.vcf"
begin:vcard
n:Gerwien;Ralf
x-mozilla-html:FALSE
adr:;;;;;;
version:2.1
email;internet:[EMAIL PROTECTED]
x-mozilla-cpt:;1
fn:Ralf Gerwien
end:vcard
==============D0F8DA858AB56499BFF1B75D==
------------------------------
** 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
******************************
