Linux-Development-Sys Digest #151, Volume #8 Sat, 16 Sep 00 23:13:15 EDT
Contents:
Re: ext2 file size limit? (Gene Montgomery)
Re: printk output in X? (arne)
Re: Memory allocation Strangeness. (Update) (Nix)
Re: ext2 file size limit? (Mikko Rauhala)
Re: is a given IP adress local ("Vik Heyndrickx")
Re: ext2 file size limit? ("Anders Gulden Olstad")
Re: interrupt HD access, how ? (ZIP-Prob.) (=?iso-8859-1?Q?Stolm=E1r=20Tam=E1s?=)
Re: ext2 file size limit? (Christopher Browne)
Re: ext2 file size limit? (Christopher Browne)
Re: Swap Atomically? (Nix)
Re: Curious: compiling test8 series: "warning... token" (Nix)
Re: ext2 file size limit? (David M. Cook)
Re: new windowing system (Karl Heyes)
nVidia riva tnt2 and dri in kernel? (Jason Cox)
Re: ext2 file size limit? (Alexander Viro)
Re: printk output in X?
----------------------------------------------------------------------------
From: Gene Montgomery <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: ext2 file size limit?
Date: Sat, 16 Sep 2000 17:14:22 GMT
This is a multi-part message in MIME format.
==============E4B747D5DC6559F701F2AB08
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Paul Reilly wrote:
>
> Hi
>
> Can someone tell me what the max size for a single file is in linux?
2 Gig - lucky you, you found it! I tried to tar off a backup of more
than that amount and ran into it also. Best bet - break the
problem space up into digestable chunks. I just descended one level
via a forloop and did several tar files vice one.
> I'm trying to creat a 6GB loopback device, but using dd if=/dev/zero
> of=file
> crashes out after filling the file with 2GB. I presume this is happening
> as I've reached some file system limit? Is there any way around this or
> any plans on making ext2 handle larger files?
I think I read somewhere that there are plans afoot to increase the
file size limit, but can't remember where I saw it.
I don't know what a loopback device is, but maybe you can break up the
problem space into smaller chunks...
Gene
==============E4B747D5DC6559F701F2AB08
Content-Type: text/x-vcard; charset=us-ascii;
name="g.montgomery.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Gene Montgomery
Content-Disposition: attachment;
filename="g.montgomery.vcf"
begin:vcard
n:Montgomery;Gene
x-mozilla-html:FALSE
adr:;;;;;;
version:2.1
email;internet:[EMAIL PROTECTED]
note;quoted-printable:-- =0D=0APursuant to U.S. code,title 47, Chapter 5, Subchapter
II, Section 227,=0D=0Aand consistent with Oregon State Law, any and all nonsolicited
commercial=0D=0AE-mail sent to this address is subject to a consulting fee of $500.00
U.S.=0D=0AE-Mailing denotes acceptance of these terms.=0D=0AConsult
<http://www.law.cornell.edu/uscode/47/227.html> for details.
x-mozilla-cpt:;11168
fn:Gene Montgomery
end:vcard
==============E4B747D5DC6559F701F2AB08==
------------------------------
From: arne <[EMAIL PROTECTED]>
Subject: Re: printk output in X?
Date: Sat, 16 Sep 2000 12:44:26 -0600
I don't know if this is related to your problem, but I found that
printk's sometimes don't appear in /var/log/messages.
The reason for this was that I deleted my messages file to start with a
fresh one. Don't do that, syslogd opens the file at startup and uses the
same filehandle all through it's life. If you deleted the file, the
filehandle isn't valid anymore, but syslogd doesn't know that. You can
restart syslogd by sending it a TERM signal, and then it works again.
You might have to restart klogd too.
On some RedHat releases, there's a problem with rotatelogs which moves
messages to messages.1 and so on. I don't know which versions though. It
should send a TERM signal to syslogd and klogd after moving the messages
file.
Tasos Kotaras wrote:
>
> Thank you all for your replies. Unfortuntaly nothing happens still. Here is
> what I did:
>
> 1. I changed the console_loglevel parameter. (I used a function I found in A.
> Rubini's book). I checked in text mode and it works fine for all priorities.
> 2. I tried xterm -C (as a root of course). Nothing appears to it.
> 3. I tried xterm -e tail -f /var/log/messages. Again nothing. The reason is
> that there is nothing writen
> in the /var/log/messages file.
>
> I got desperate, but eventually I found the solution: I used the following
> function that prints anything to the
> current console. Credits to Ori Pomerantz.
------------------------------
From: Nix <$}xinix{[email protected]>
Subject: Re: Memory allocation Strangeness. (Update)
Date: 16 Sep 2000 20:25:00 +0100
[EMAIL PROTECTED] (Kaz Kylheku) writes:
> I don't follow; that is only true if your program allocates more memory than it
> needs. In that case, the program is broken. Most programs immediately
> initialize an allocated object with meaningful data.
No, most programs immediately initialize *some* of every set of objects
that they initialize. In the case of collections, a fairly small amount
may be used. As I said, the time complexity requirements placed on the
C++ (and some of the Java) collection classes *require* virtual-address-
space-inefficient allocation schemes.
> A
> fork() could ensure that enough pages are reserved so that every page of the
> parent could be copied, or else fail. You don't have to actually copy pages to
> reserve them. You just have to record that they are reserved. The usual copy
> on write optimization still takes place.
Really?
So you can't run `ls' from your Emacs unless you have as much swap free
as is allocated by Emacs --- even though all you need is that needed by
`ls'.
You can't have 50 servers on your web server unless you have as much
swap as is needed by 50 complete Apaches, even though the children share
most of their pages. (The same is even more true of Oracle; picture
hundreds of children, each with 100Mb VSZs, changing a few hundred K of
it.)
The scheme you propose is wildly and horrendously inefficient and would
never be used in practice.
--
`no amount of Zen contemplation will ever make you at one
with a 3c905B ethernet card.' --- Alan Cox
------------------------------
From: [EMAIL PROTECTED] (Mikko Rauhala)
Crossposted-To: comp.os.linux.misc
Subject: Re: ext2 file size limit?
Date: 16 Sep 2000 19:58:47 GMT
Reply-To: [EMAIL PROTECTED]
On Sat, 16 Sep 2000 17:53:18 -0700, Paul Reilly <[EMAIL PROTECTED]> wrote:
>Can someone tell me what the max size for a single file is in linux?
On 32-bit platforms it's the 2 gigs you just hit. This limit is removed
in the 2.4-series kernels (currently in testing phase), though you of course
need a recent enough libc as well, and to compile the app with a 64-bit
size_t.
Support for 64-bit sizes has also been backported to the 2.2-series, and
I believe at least Red Hat's default kernel has the patch included.
Also, on 64-bit platforms 64-bit files are of course supported.
--
Mikko Rauhala - [EMAIL PROTECTED] - http://www.iki.fi/mjr/
------------------------------
From: "Vik Heyndrickx" <[EMAIL PROTECTED]>
Subject: Re: is a given IP adress local
Date: Sat, 16 Sep 2000 20:34:27 GMT
"Michael F Gordon" <[EMAIL PROTECTED]> wrote in message
news:8q06t6$9qt$[EMAIL PROTECTED]...
> if (ioctl(s,SIOCGIFCONF,&ifconf)==-1) {
Thanks a lot!
_Exactly_ what I was looking for :)
Kind regards,
--
Vik
------------------------------
From: "Anders Gulden Olstad" <[EMAIL PROTECTED]>
Subject: Re: ext2 file size limit?
Crossposted-To: comp.os.linux.misc
Date: Sat, 16 Sep 2000 22:54:25 +0200
Paul Reilly <[EMAIL PROTECTED]> wrote:
> Hi
>
> Can someone tell me what the max size for a single file is in linux?
>
> I'm trying to creat a 6GB loopback device, but using dd if=/dev/zero
> of=file
> crashes out after filling the file with 2GB. I presume this is happening
> as I've reached some file system limit? Is there any way around this or
> any plans on making ext2 handle larger files?
ext2fs has a filesize limit of 2GB, I'm sorry.
The Filesystems-HOWTO might be a good place to start seeking for
alternatives.
http://www.linux.no/biblioteket/HOWTO/Filesystems-HOWTO.html
--
Sing While You May!
------------------------------
From: =?iso-8859-1?Q?Stolm=E1r=20Tam=E1s?= <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.hardware
Subject: Re: interrupt HD access, how ? (ZIP-Prob.)
Date: Mon, 11 Sep 2000 23:17:38 +0200
Dirk Schenkewitz wrote:
> Hi All,
>
> Is there a way to make accesses to _locally_mounted_ discs interruptible ?
>
> The problem is: when I acces a ZIP drive, it sometimes happens that the drive
> goes into a loop of rhythmic stepping and such, over and over, until I pull
> and replug its power (It's a parallel ZIP, accessed via ppa). After that,
> the ZIP-drive calms down, but now the linux system hangs completely - no
> console switching, no shitdown er shutdown via alt-ctrl-del, only hard reset
> works.
> Even if could not access the (replugged) ZIP-drive, it would be nice to (at
> least) get back control of the rest of the system, so I could do a shutdown
> - and don't have to wait for 'fsck' to get finished.
>
> I know that nfs-mounts can be made interruptibe - is there a solution for
> local HDs ?
fuser -mv /mnt/zip will show you which pids are using the mount point.
So you can kill them.
After that do a sync if you want, so the system will write out all puffered data.
Somehow it's possible to disable caching on specified devices.
Then umount, and then
echo "remove-scsi-singledevice 0 0 0 0" > /proc/scsi
then you can discoonect the device safely.
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Crossposted-To: comp.os.linux.misc
Subject: Re: ext2 file size limit?
Reply-To: [EMAIL PROTECTED]
Date: Sat, 16 Sep 2000 21:34:20 GMT
Centuries ago, Nostradamus foresaw a time when Paul Reilly would say:
>Can someone tell me what the max size for a single file is in linux?
There is no single answer to this.
>I'm trying to creat a 6GB loopback device, but using dd if=/dev/zero
>of=file
>crashes out after filling the file with 2GB. I presume this is happening
>as I've reached some file system limit? Is there any way around this or
>any plans on making ext2 handle larger files?
No, you have _not_ reached a filesystem limit. ext2 supports files as
large as 2TB in size.
Apparently your version of VFS and the interface from there to LIBC and
on to the dd application does not support this.
--
[EMAIL PROTECTED] - <http://www.ntlug.org/~cbbrowne/linuxkernel.html>
All extremists should be taken out and shot.
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Crossposted-To: comp.os.linux.misc
Subject: Re: ext2 file size limit?
Reply-To: [EMAIL PROTECTED]
Date: Sat, 16 Sep 2000 21:34:22 GMT
Centuries ago, Nostradamus foresaw a time when Anders Gulden Olstad would say:
>Paul Reilly <[EMAIL PROTECTED]> wrote:
>> Hi
>>
>> Can someone tell me what the max size for a single file is in linux?
>>
>> I'm trying to creat a 6GB loopback device, but using dd if=/dev/zero
>> of=file
>> crashes out after filling the file with 2GB. I presume this is happening
>> as I've reached some file system limit? Is there any way around this or
>> any plans on making ext2 handle larger files?
>
>ext2fs has a filesize limit of 2GB, I'm sorry.
ext2fs has a file size limit of _2TB,_ I'm sorry, but the limit that
is being hit is _not_ a filesystem limit, but rather a limitation in
the interface between VFS and GLIBC on 32 bit platforms.
--
[EMAIL PROTECTED] - <http://www.hex.net/~cbbrowne/>
"It goes against the grain of modern education to teach children to
program. What fun is there in making plans, acquiring discipline in
organizing thoughts, devoting attention to detail and learning to be
self-critical?" -- Alan Perlis
------------------------------
From: Nix <$}xinix{[email protected]>
Subject: Re: Swap Atomically?
Date: 16 Sep 2000 20:28:28 +0100
Larry Ebbitt <[EMAIL PROTECTED]> writes:
> How about "mutaphim?"
I don't think there's a kind of angel called a `seraphex'.
--
`no amount of Zen contemplation will ever make you at one
with a 3c905B ethernet card.' --- Alan Cox
------------------------------
From: Nix <$}xinix{[email protected]>
Subject: Re: Curious: compiling test8 series: "warning... token"
Date: 16 Sep 2000 20:56:00 +0100
Kasper Dupont <[EMAIL PROTECTED]> writes:
[from gcc.info]
> It might be better to discard the last preprocessor token instead of
> the last preceding sequence of non-whitespace characters; in fact, we
> may someday change this feature to do so.
The integrated preprocessor has made this change; but you can revert to
the old behaviour, too.
--
`no amount of Zen contemplation will ever make you at one
with a 3c905B ethernet card.' --- Alan Cox
------------------------------
From: [EMAIL PROTECTED] (David M. Cook)
Crossposted-To: comp.os.linux.misc
Subject: Re: ext2 file size limit?
Date: 16 Sep 2000 22:04:27 GMT
On Sat, 16 Sep 2000 17:53:18 -0700, Paul Reilly <[EMAIL PROTECTED]> wrote:
>crashes out after filling the file with 2GB. I presume this is happening
>as I've reached some file system limit?
Yes, you've found the limit.
>Is there any way around this
You can move to a 64-bit platform.
>or any plans on making ext2 handle larger files?
You can get a kernel patched for Large File Support, which allows you to get
around the limit on 32-bit systems. Red Hat, for instance, has such a
kernel at
http://www.redhat.com/support/errata/enterprise/RHSA-2.2.16-4.lfs.html
Dave Cook
------------------------------
From: Karl Heyes <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.x,comp.windows.x
Subject: Re: new windowing system
Date: Sat, 16 Sep 2000 23:23:32 +0000
In article <[EMAIL PROTECTED]>, Aurel Balmosan <[EMAIL PROTECTED]> wrote:
....
>
> Hi all,
>
> I just want to give my 2 cents to this discussion:
>
> Where can the so called performance problem be found in X11?
> Personnaly I haven't found any yet. If one compares the
> 'performance' of X and other so called faster windowing
> systems with fair benchmarks, or even better, real applications
> no difference can be found.
>
> If one of you know a specific action that is supported
> by X11 and by another windowing system that its noticable
> slow in X11 please e-mail me. I would really like to know it.
>
> One point I have to make: Why is everyone thinking that a socket
> (tcp/ip, unix, ...) connection is slow? Especially when it comes
> to X the Xlib communication layer knows how to use it efficiently.
>
sockets add extra wrapping around the data thats coming from the
applications, things like tcp/ip. It's not much in itself but every
piece of information has its own, and each has to processed through
the networkl stack. They are useful when dealing over a network but
run locally it adds an overhead, which can add up.
A more popular thing happening these days is 3d work, which is going
though DRI which is not network based anyway. Linux is nice
and fast on TCP/IP but IP has a disadvantage when dealing locally.
karl
------------------------------
From: Jason Cox <[EMAIL PROTECTED]>
Subject: nVidia riva tnt2 and dri in kernel?
Date: Sun, 17 Sep 2000 01:56:44 GMT
Hi=20
has anyone looked at inserting the tnt2 into the dri of the kernel=20
2.4.0-test8 ?
Also think i'vwe got the M1621 apg going ans will supply a patch if=20
its requested.
Thanks Jason Cox
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Alexander Viro)
Crossposted-To: comp.os.linux.misc
Subject: Re: ext2 file size limit?
Date: 16 Sep 2000 21:04:11 -0400
In article <[EMAIL PROTECTED]>,
Christopher Browne <[EMAIL PROTECTED]> wrote:
>ext2fs has a file size limit of _2TB,_ I'm sorry, but the limit that
>is being hit is _not_ a filesystem limit, but rather a limitation in
>the interface between VFS and GLIBC on 32 bit platforms.
Yaaaaargh... OK, it seems to become a FAQ
Q: is it true that ext2 has 2Gb limit on file size?
A: BS
Q: so how comes that I can't create files larger than that?
A: because VM in Linux 2.2 and earlier can't cope with files larger than
2.2 on 32-bit architectures. Regardless of filesystem.
Q: will reiserfs help?
A: what part of "regardless of filesystem" is too hard to understand?
Q: OK, so what can I do, I'm stuck with 32-bit box?
A: use 2.4 _or_ 2.2 with LFS patches _or_ FreeBSD. All of them will handle
more than 2Gb on ext2.
Q: I've done that, and half of utilities doesn't work
A: That was a question?
Q: OK, _why_?
A: because if libc thinks that offsets are 32 bit it's not going to pass
anything larger to the kernel
Q: what should I do?
A: get sufficiently recent libc. And learn to use search engines, already -
all that stuff had been beaten to death _many_ times.
Q: why...
A: excuse me, what was your username, again?
Q: ... are you so... Hey, what's up with this NIC? It's sparAAAAAASSSHHH<thud>
--
Live phase 1 <--> RJ45 pin 3 Neutral <--> RJ45 pin 1
Live phase 2 <--> RJ45 pin 6 GND <--> RJ45 pin 8
Live phase 3 <--> RJ45 pin 2
------------------------------
From: [EMAIL PROTECTED] ()
Subject: Re: printk output in X?
Date: 17 Sep 2000 02:23:42 GMT
Reply-To: [EMAIL PROTECTED]
On my system, the output of the hello module appears in /var/adm/syslog
(rather than /var/adm/messages). I'm on Slackware so YMMV.
chris
In article <[EMAIL PROTECTED]>,
arne <[EMAIL PROTECTED]> writes:
|> I don't know if this is related to your problem, but I found that
|> printk's sometimes don't appear in /var/log/messages.
|> The reason for this was that I deleted my messages file to start with a
|> fresh one. Don't do that, syslogd opens the file at startup and uses the
|> same filehandle all through it's life. If you deleted the file, the
|> filehandle isn't valid anymore, but syslogd doesn't know that. You can
|> restart syslogd by sending it a TERM signal, and then it works again.
|> You might have to restart klogd too.
|>
|> On some RedHat releases, there's a problem with rotatelogs which moves
|> messages to messages.1 and so on. I don't know which versions though. It
|> should send a TERM signal to syslogd and klogd after moving the messages
|> file.
|>
|>
|> Tasos Kotaras wrote:
|> >
|> > Thank you all for your replies. Unfortuntaly nothing happens still. Here is
|> > what I did:
|> >
|> > 1. I changed the console_loglevel parameter. (I used a function I found in A.
|> > Rubini's book). I checked in text mode and it works fine for all priorities.
|> > 2. I tried xterm -C (as a root of course). Nothing appears to it.
|> > 3. I tried xterm -e tail -f /var/log/messages. Again nothing. The reason is
|> > that there is nothing writen
|> > in the /var/log/messages file.
|> >
|> > I got desperate, but eventually I found the solution: I used the following
|> > function that prints anything to the
|> > current console. Credits to Ori Pomerantz.
--
Chris Willing Ph: (61-2) 9351 3005
Vislab, A28 Fax: (61-2) 9351 7726
University of Sydney
NSW 2006 Australia http://www.vislab.usyd.edu.au/staff/chris/
------------------------------
** 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
******************************
