Linux-Development-Sys Digest #299, Volume #8 Sat, 25 Nov 00 05:13:09 EST
Contents:
Re: Can't find regcmp command (Nix)
Re: Packet ordering (Nix)
Re: What distro does Linus Torvalds use? (Nix)
Re: Interprocess Communication ... (Nix)
Re: Interprocess Communication ... (Nix)
Re: Runtime file size modifying (Nix)
Re: Research survey on cross platform development (Nix)
Re: linux software-interupt open file? (Kaz Kylheku)
Re: Runtime file size modifying (Alexander Viro)
Re: Modconf doesn't find any modules (kernel 2.4.0-test10) (Allin Cottrell)
Re: Database in Redhat Linux (Victor Wagner)
linux operating system installable files (nitin sharma)
Re: Linux on mainframe (Pete Zaitcev)
Re: [Q] How to get the Ethernet address ? ("Paul Pluzhnikov")
Re: Software RAID (John Nelson)
Re: rdtsc() timestamps synchronized between SMP CPU's? (Philip Armstrong)
Problem: sigaction with SI_SIGINFO flag (Petr Bavorovsky)
Re: Linux on mainframe ([EMAIL PROTECTED])
----------------------------------------------------------------------------
From: Nix <$}xinix{$@esperi.demon.co.uk>
Subject: Re: Can't find regcmp command
Date: 24 Nov 2000 23:16:41 +0000
Josef Moellers <[EMAIL PROTECTED]> writes:
> work on "regex_t"s, which are complex structures contianing more than
> just the precompiled pattern, e.g. the "Syntax setting with which the
> pattern was compiled".
Ah, Emacs. ;)
--
`The phrase `causes storage to be reserved', doesn't mean that it causes
storage to be reserved. This is a fundamental misunderstanding of
Standardeze.' --- Mike Stump on the GCC list
------------------------------
From: Nix <$}xinix{$@esperi.demon.co.uk>
Subject: Re: Packet ordering
Date: 24 Nov 2000 23:12:33 +0000
Ed Hudson <[EMAIL PROTECTED]> writes:
> We have noticed that when a fragmented ping is sent to a linux system,
> it is returned in reverse order (ie last fragment first). Is there a
> way to force linux to return the packets in the order received?
Why on earth do you care? This trick allows some fairly useful
optimizations.
--
`The phrase `causes storage to be reserved', doesn't mean that it causes
storage to be reserved. This is a fundamental misunderstanding of
Standardeze.' --- Mike Stump on the GCC list
------------------------------
From: Nix <$}xinix{$@esperi.demon.co.uk>
Subject: Re: What distro does Linus Torvalds use?
Date: 24 Nov 2000 23:21:57 +0000
[EMAIL PROTECTED] (bill davidsen) writes:
> Wish there was a Slackware like that, the upgrading is only so
> automatable.
There is. It's called `Debian'.
HTH. ;}
--
`The phrase `causes storage to be reserved', doesn't mean that it causes
storage to be reserved. This is a fundamental misunderstanding of
Standardeze.' --- Mike Stump on the GCC list
------------------------------
From: Nix <$}xinix{$@esperi.demon.co.uk>
Subject: Re: Interprocess Communication ...
Date: 24 Nov 2000 23:24:43 +0000
Kaelin Colclasure <[EMAIL PROTECTED]> writes:
> [EMAIL PROTECTED] writes:
>
> > 1. POSIX Message Queues
> > 2. POSIX Shared Memory
> > 3. POSIX Semaphores.
> > 4. Mutex that can be shared across PROCESSES.
> >
> > Any help in this direction will be greatly appreciated.
>
> All of the above. Collectively these are known as "System V IPC."
Wrong-o. sysvipc != POSIX IPC.
Linux does not as of the 2.2 kernel series support the mq_*(), sem_() or
shm_*() functions --- but this may well not be true of Linux-2.4 (I
can't remember whether it's been implemented there yet).
--
`The phrase `causes storage to be reserved', doesn't mean that it causes
storage to be reserved. This is a fundamental misunderstanding of
Standardeze.' --- Mike Stump on the GCC list
------------------------------
From: Nix <$}xinix{$@esperi.demon.co.uk>
Subject: Re: Interprocess Communication ...
Date: 24 Nov 2000 23:25:45 +0000
[EMAIL PROTECTED] (bill davidsen) writes:
> AFAIK only the SvyV versions of these are available. Without a
> standard in fromt of me I can't remember how they differ. This question
> comes up from time to time, and the only time I looked into it, I had
> the impression that the functionality is the same, but the calling
> sequences are different.
The APIs of the POSIX variants are nicer, and they return fds, so you
can select on them and so forth.
i.e. they're SysVIPC as they should have been implemented all along.
--
`The phrase `causes storage to be reserved', doesn't mean that it causes
storage to be reserved. This is a fundamental misunderstanding of
Standardeze.' --- Mike Stump on the GCC list
------------------------------
From: Nix <$}xinix{$@esperi.demon.co.uk>
Subject: Re: Runtime file size modifying
Date: 24 Nov 2000 23:49:54 +0000
[EMAIL PROTECTED] (Alexander Viro) writes:
> In article <[EMAIL PROTECTED]>,
> Nix <$}xinix{$@esperi.demon.co.uk> wrote:
>
> >Unix needs a holepunch() function :(
>
> No, it doesn't. truncate() is evil enough. OTOH, if you want to implement
> it... Feel free - it's going to be interesting to watch.
I meant `it would fill a hole[1] in the APIs' in that at present there is
no way to make a hole in a file anywhere but at the end.
But yes, it does have many of the same uglinesses and horrors as
truncate() (although it is not as bad as rename()) --- but just because
these are monsters to implement does not mean that they should not exist.
[1] er. Possibly that was not the best form of words. You know what I mean.
Make a hole. Er.
--
`The phrase `causes storage to be reserved', doesn't mean that it causes
storage to be reserved. This is a fundamental misunderstanding of
Standardeze.' --- Mike Stump on the GCC list
------------------------------
From: Nix <$}xinix{$@esperi.demon.co.uk>
Subject: Re: Research survey on cross platform development
Date: 25 Nov 2000 00:04:08 +0000
Garrett Banuk <[EMAIL PROTECTED]> writes:
> The purpose of this survey is to find out why most software programs are
> not written as cross platform compatible.
This is probably the wrong group to ask that.
The reason, IMHO, is `most programmers are stupid and do not care about
portability'. Plus `most managers in commercial organizations do not
care about portability *or* reliability, and actively contest you when
you attempt to make programs portable and/or reliable'.
(Cynical? Me?)
> Most software is written to run on one specific platform. A platform is
`Most software is badly written garbage'.
(Equally, most software is not written using GNU autoconf or a similar
program.)
> Less used operating systems are now becoming popular and new operating
> systems are being developed, therefore, we believe that cross platform
> software design needs to become common practice in software development.
All the software on *my* system is cross-platform. (Well, except for
some limited things like IA32 boot loader code ;} )
(Well, most of it's cross-platform to anything POSIXish; some will only
work on Linux and *BSD platforms, but I try to keep that to a minimum.
Some stuff is inherently nonportable, but things like hard drive
tuning software and so on can't really be made terribly portable.)
--
`The phrase `causes storage to be reserved', doesn't mean that it causes
storage to be reserved. This is a fundamental misunderstanding of
Standardeze.' --- Mike Stump on the GCC list
------------------------------
From: [EMAIL PROTECTED] (Kaz Kylheku)
Crossposted-To: comp.lang.asm.x86
Subject: Re: linux software-interupt open file?
Date: Sat, 25 Nov 2000 00:14:00 -0000
Reply-To: [EMAIL PROTECTED]
On Fri, 24 Nov 2000 19:26:00 -0000, Daniel Schroeter <[EMAIL PROTECTED]> wrote:
>
>hi,
>does anybody know a software interupt where i can open a file under
>LINUX. is there somewhere a (sorted) list with interupts?
>THNX
Doh what? I think you are looking for the open() system call!
There is a corresponding interrupt, and there is a list of them. but you
should use the high level interface through the system library.
On Linux, these interrupts are an interface for the library developers,
not for application developers.
------------------------------
From: [EMAIL PROTECTED] (Alexander Viro)
Subject: Re: Runtime file size modifying
Date: 24 Nov 2000 19:48:53 -0500
In article <[EMAIL PROTECTED]>,
Nix <$}xinix{$@esperi.demon.co.uk> wrote:
>I meant `it would fill a hole[1] in the APIs' in that at present there is
>no way to make a hole in a file anywhere but at the end.
Sure, there is. If your filesystem allows holes and you care to do such
"optimization" - feel free to insert the check that data being written
consists of zeroes and let write() (->commit_write(), actually) punch
the hole for you. Transparent, doesn't require any new syscalls, doesn't
give POSIX wankers[2] any chance to invent new amazing mandatory
braindamage in that area...
>But yes, it does have many of the same uglinesses and horrors as
>truncate() (although it is not as bad as rename()) --- but just because
>these are monsters to implement does not mean that they should not exist.
>
>[1] er. Possibly that was not the best form of words. You know what I mean.
> Make a hole. Er.
[2] and that's a _very_ mild term, trust me. The only thing that stops me
from posting a long rant about choice pieces of POSIX/SuS idiocy to Other
Place is the lack of a thread that could be conveniently drifted in that
direction. I'm in the middle of getting the union-mounts[3] into the minimal
compliance with POSIX semantics and it's not a pretty sight. OTOH, it had
forced redesign of the original patch that made the thing work even when
we have remote filesystems as components. Currently the only thing that
is right off is autofs - everything else seems to be tolerable...
[3] Example of braindamage: chdir(2) requires exec permissions on new pwd.
That's nice, but what, in name of Cthulhu, does it mean for a union? And
it's not like we had a warranty that pwd is searchable - chmod 0 . does
the trick quite fine... Good for Plan 9 folks, they could send POSIX to
hell where it belongs. We are slightly less lucky...
--
"You're one of those condescending Unix computer users!"
"Here's a nickel, kid. Get yourself a better computer" - Dilbert.
------------------------------
From: Allin Cottrell <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: Modconf doesn't find any modules (kernel 2.4.0-test10)
Date: Fri, 24 Nov 2000 21:36:16 -0500
Otto Wyss wrote:
>
> I just installed kernel 2.4.0-test10 to see if my crashes of X v4 goes
> away but modconf doesn't find any modules, even if I have modutils
> 2.3.19...
What's "modconf"? Try using the standard tools /sbin/depmod and
/sbin/modprobe (reading their man pages first if necessary).
Note that /etc/conf.modules is no longer an acceptable name for
the modules configuration file: it must be modules.conf. This
is in the kernel's Changes documentation.
Alln Cottrell.
------------------------------
From: [EMAIL PROTECTED] (Victor Wagner)
Crossposted-To:
comp.os.linux.development.apps,comp.os.linux.apps,comp.os.linux.misc,comp.lang.java.databases,comp.databases.informix,comp.databases.oracle.server,comp.databases.sybase,linux.redhat.misc,linux.postgres,comp.databases.pick,comp.databases.ibm-db2,comp.databases.oracle.misc
Subject: Re: Database in Redhat Linux
Date: 24 Nov 2000 22:47:57 +0300
In comp.os.linux.misc Freelancer <[EMAIL PROTECTED]> wrote:
: I need decide which database going to run for Redhat Linux.
: I know MySQL is the most popular one in Linux world. I need
: you help me to fill out the blank and hole (?) in table below.
Its a pity for Linux World, that most hype is done by people who don't
know what real database is. So they promote mySQL which is no more than
fast flat-file search engine with SQL-like syntax.
It cannot be considered real SQL just becouse SQL stands for
Structured Query language, and mySQL doesn't support structured, i.e.
nested queries.
But database is much more than just search engine. It also should ensure
integrity of data both by enforcing some conditions of them (i.e.
foreign keys and triggers) and by rolling failed transactio back to
consistent state.
So, only free database is PostgreSQL. But PostgreSQL start to
resemble real database only since 7.0 version, becouse before there was
no foreign keys. I would consider that it IS a database, not RESEMBLES
one only when it begin to support outer joins and binary large objects.
Both are scheduled for 7.1.
BTW, Inprise/Borland recently have open sources for Interbase, which IS
the database.
: I know Oracle, Sybase, Informix and DB2 support most or all of them in
: UNIX (Solaris, HP-UX, AIX, etc.) But I am
: not sure are they also support in Linux.
All ones you've mention are supported.
--
It's hard to tune heavily tuned code. :-)
-- Larry Wall in <[EMAIL PROTECTED]>
------------------------------
Date: 25 Nov 2000 05:30:29 +0000
From: nitin sharma <[EMAIL PROTECTED]>
Subject: linux operating system installable files
Crossposted-To:
comp.os.linux.powerpc,comp.os.linux.misc,de.comp.os.unix.linux.misc,no.it.os.unix.linux.diverse,uk.comp.os.linux
kindly specify the files which i should download for the installing linux on my system
from the corresponding sites.
linuxppc from http://download.sourceforge.net/mirrors/linuxppc/
yellowdog linux from ftp://ftp.yellowdoglinux.com/pub/yellowdog/
debian linux from
http://www.debian.org/distrib/ftplist
after downloading the files how do you install the operating system on the target
system? Do u need to write the files on a CD before installation or there is some
other method?
Chequemail.com - a free web based e-mail service that also pays!!!
http://www.chequemail.com
------------------------------
From: [EMAIL PROTECTED] (Pete Zaitcev)
Subject: Re: Linux on mainframe
Date: Sat, 25 Nov 2000 05:34:40 GMT
> | We have a Linux running on a Mainframe. It is 2.2.16.
> | When I try to do cat /proc/interrupts it returns with bad file
> | descriptor error.
> [...] Since the
> /proc/interrupts interface was really for these IRQ levels, and since
> that IRQ code is not in the 390 architecture specific parts, that might
> explain why it isn't working.
Why not to return EOF instead of EBADF? Sounds like a half-baked
kernel port to me.
--Pete
------------------------------
From: "Paul Pluzhnikov" <[EMAIL PROTECTED]>
Crossposted-To: comp.sys.sun.misc,comp.os.linux.setup
Subject: Re: [Q] How to get the Ethernet address ?
Date: Sat, 25 Nov 2000 05:42:34 GMT
"Pillonel Cédric, NWS-PD-71" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> How can I get the Ethernet address from my C++ application ?
>
Presumably you want IP address of the host your application is executing on.
man uname
man gethostbyname
------------------------------
From: John Nelson <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.hardware,alt.linux,comp.os.linux.misc
Subject: Re: Software RAID
Date: Fri, 24 Nov 2000 23:14:42 -0800
This one time, at band camp, <[EMAIL PROTECTED]>, [EMAIL PROTECTED]
said...
> Tut tut. No, it doesn't slow it appreciably, provided the disk writes
> are on two separate controllers (ide). Yes, two commands will be
This is the only "right" way to do RAID1 on IDE/ATA disk drives. Otherwise, if you put
both drives on the same channel and the master dies, the array dies. Rather defeats
the
purpose.
------------------------------
From: [EMAIL PROTECTED] (Philip Armstrong)
Subject: Re: rdtsc() timestamps synchronized between SMP CPU's?
Date: 24 Nov 2000 22:47:25 -0000
In article <[EMAIL PROTECTED]>,
Kaelin Colclasure <[EMAIL PROTECTED]> wrote:
>Are the 64-bit timestamps (well, clock cycle counts) that are accessed
>via rdtsc() synchronized between CPU's on an SMP Linux system? That
>is, supposing it were somehow possible to execute a rdtsc instruction
>on each CPU in the system at the exact same instant, would they all
>have the same value?
>
>If not, would they be close enough to allow reasonably meaningful
>comparisons between values from two different CPU's?
I believe that its perfectly possible on intel SMP systems to have two
processors with different clock rates both installed at the same
time. If true, then this would definately imply that the clock cycle
counters would differ between the processors.
However, I'm not sure as to the absolute veracity of the above...
Phil
--
http://www.kantaka.co.uk/ .oOo. public key: http://www.kantaka.co.uk/gpg.txt
------------------------------
From: Petr Bavorovsky <[EMAIL PROTECTED]>
Subject: Problem: sigaction with SI_SIGINFO flag
Date: Sat, 25 Nov 2000 15:37:59 +0600
Hi, all !
There is a strange problem.
here a code fragment:
===================================================
void on_SIGIO(int s, siginfo_t *si, void *v)
{
fprintf(stderr,"%d\n",si->si_fd);
}
int main()
{
struct sigaction sa;
bzero(&sa,sizeof(sa));
sa.sa_flags=SA_SIGINFO;
sa.sa_sigaction=on_SIGIO;
sigaction(SIGIO,&sa,0);
=======================================================
When signal SIGIO arrives, it calls on_SIGIO function.
In main I set the SA_SIGINFO flag and sa_sigaction instead of
sa_sighandler,
so I want to be called a function receives three arguments, the second
has type 'siginfo_t *'.
Then I want to get a file descriptor (in ASYNC mode), which causes a
signal SIGIO on data receive.
But there are always zero in si->si_fd.
Man page of sigaction says, that SIGIO (synonyms SIGPOLL, SIGURG) fills
si_fd field.
It seems difficult to find soft which uses sigaction call with
SI_SIGINFO flag, so help me to
solve this problem.
If it's impossible to use SI_SIGINFO in linux (at any reasons) here what
I want to do:
my process creates some ASYNC fullduplex sockets (PF_UNIX, SOCK_STREAM),
each of them may receive something at anytime. I should be able to
immediately read the data and process them (so I need to know event of
which socket raises the signal).
Petr Bavorovsky.
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Linux on mainframe
Date: Sat, 25 Nov 2000 10:03:14 -0000
On Sat, 25 Nov 2000 05:34:40 GMT Pete Zaitcev <[EMAIL PROTECTED]> wrote:
|> | We have a Linux running on a Mainframe. It is 2.2.16.
|> | When I try to do cat /proc/interrupts it returns with bad file
|> | descriptor error.
|
|> [...] Since the
|> /proc/interrupts interface was really for these IRQ levels, and since
|> that IRQ code is not in the 390 architecture specific parts, that might
|> explain why it isn't working.
|
| Why not to return EOF instead of EBADF? Sounds like a half-baked
| kernel port to me.
The code that implemented /proc/interrupts simply wasn't ported.
That means /proc/interrupts is treated like /proc/xyzzy123abc,
that is, it's unknown.
It would be nice to have /proc/interrupts implemented to see a
breakdown of I/O interrupt totals by channel and by device.
It would be a radically different format and probably would call
for a very different /proc pathname.
For a lot of people, the 360/370/390 architecture is radical.
Imagine trying to comprehend a machine in which there cannot
even be an IRQ conflict. I just wish I could afford one.
--
=================================================================
| Phil Howard - KA9WGN | Dallas | http://linuxhomepage.com/ |
| [EMAIL PROTECTED] | Texas, USA | http://phil.ipal.org/ |
=================================================================
------------------------------
** 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
******************************
