Linux-Development-Sys Digest #618, Volume #8 Sat, 7 Apr 01 19:13:20 EDT
Contents:
Re: usbutils
Re: usbutils (Phil Ehrens)
Re: Kernel Panic in Network Interface
Glibc 2.2.2 Make Check Errors ("Fruitbat")
Re: [Help:] Kernel Upgrade (Florian =?iso-8859-1?Q?Gro=DFe=2DCoosmann?=)
Re: The never ending socket :-) ([EMAIL PROTECTED])
halt problem from Turbo Linux ("mike chan")
Re: Signal, Kernel and Co ("DongKook Park")
Re: [Help:] Kernel Upgrade ("Ian D. Stewart")
LinuxThreads and thread-safety, re-entrancy, async-safety! (Billy Bob Jameson)
Re: TCP/IP socket buffering (Nix)
Re: usleep() is unreliable when sleeping for less then 10000 micro (Pasztor Szilard)
Kernel 2.2.18 ("J. Liu")
Re: [Help:] Kernel Upgrade ("J. Liu")
Re: Kernel 2.2.18 (nordi)
----------------------------------------------------------------------------
From: <[EMAIL PROTECTED]>
Subject: Re: usbutils
Date: Fri, 6 Apr 2001 16:33:31 -0700
My apologies to you Phil. I did not read through the documentation correctly
before I went off half cocked with my reply. I hope you will not take
offense. The guide does tell one how to create the /proc/bus/usb stuff and
it works. Thank you very much.
"Phil Ehrens" <[EMAIL PROTECTED]> wrote in message
news:9ali6c$[EMAIL PROTECTED]...
> Geez. It tells you how to create the /proc/usb stuff. That was what
> your question was about, right???
>
> [EMAIL PROTECTED] wrote:
> >Dear Phil:
> > The link to the linuxusbguide is useful and I thank you. I have
enabled
> >USB support in the kernel and I can load some of the experimental
modules.
> >The guide is useful in providing additional information about some of
those
> >modules. Unfortunately, in this particular case, usbutils and not
mentioned
> >in the linuxusbguide. I suspect my problem is particular to usbutils and
not
> >...
>
------------------------------
From: [EMAIL PROTECTED] (Phil Ehrens)
Subject: Re: usbutils
Date: 6 Apr 2001 23:30:24 GMT
Reply-To: -@-
Apology accepted. Have fun with USB!
[EMAIL PROTECTED] wrote:
>My apologies to you Phil. I did not read through the documentation correctly
>before I went off half cocked with my reply. I hope you will not take
>offense. The guide does tell one how to create the /proc/bus/usb stuff and
>it works. Thank you very much.
------------------------------
From: [EMAIL PROTECTED] ()
Subject: Re: Kernel Panic in Network Interface
Date: Sat, 07 Apr 2001 04:26:16 -0000
In article <9airic$4nv8$[EMAIL PROTECTED]>,
bill davidsen <[EMAIL PROTECTED]> wrote:
>The 2.4 kernel supports "nosmp" as an option at boot time, but for the
>old kernel... who knows? I certainly don't remember that option, but I
>find new old features in linux all the time.
It's a compile time option in 2.2.
--
http://www.spinics.net/linux/
------------------------------
From: "Fruitbat" <[EMAIL PROTECTED]>
Subject: Glibc 2.2.2 Make Check Errors
Date: Sat, 07 Apr 2001 04:48:55 GMT
Hi, I have just built Glibc 2.2.2 on my Redhat 6.2 based pentium II system
using GCC 2.95.3, and a 2.4.2 kernel with the following options:
--prefix=/usr
--enable-add-ons
--enable-kernel=2.4.0
--disable-profile
Everything builds ok, but when I run "make check", check fails in the
linuxthreads directory with the following output:
make[2]: Entering directory `/usr/src/glibc-2.2.2/linuxthreads'
gcc
Examples/ex4.c -c -O2 -Wall -Winline -Wstrict-prototypes -Wwrite-strings -g
-I../include -I. -I/usr/src/biuld/linuxthreads -I.. -I../libio -I/usr/
src/biuld -I../sysdeps/i386/elf -I../linuxthreads/sysdeps/unix/sysv/linux/i3
86 -I../linuxthreads/sysdeps/unix/sysv/linux -I../linuxthreads/sysdeps/pthre
ad -I../linuxthreads/sysdeps/unix/sysv -I../linuxthreads/sysdeps/unix -I../l
inuxthreads/sysdeps/i386/i686 -I../linuxthreads/sysdeps/i386 -I../sysdeps/un
ix/sysv/linux/i386/i686 -I../sysdeps/unix/sysv/linux/i386 -I../sysdeps/unix/
sysv/linux -I../sysdeps/gnu -I../sysdeps/unix/common -I../sysdeps/unix/mman
-I../sysdeps/unix/inet -I../sysdeps/unix/sysv/i386 -I../sysdeps/unix/sysv -I
../sysdeps/unix/i386/i686 -I../sysdeps/unix/i386/i586 -I../sysdeps/unix/i386
-I../sysdeps/unix -I../sysdeps/posix -I../sysdeps/i386/i686/fpu -I../sysdep
s/i386/i686 -I../sysdeps/i386/i486 -I../sysdeps/i386/fpu -I../sysdeps/i386 -
I../sysdeps/wordsize-32 -I../sysdeps/ieee754/ldbl-96 -I../sysdeps/ieee754/db
l-64 -I../sysdeps/ieee754/flt-32 -I../sysdeps/ieee754 -I../sysdeps/generic/e
lf -I../sysdeps/generic -D_LIBC_REENTRANT -include
../include/libc-symbols.h -o /usr/src/biuld/linuxthreads/ex4.o
gcc -nostdlib -nostartfiles -o
usr/src/biuld/linuxthreads/ex4 -Wl,-dynamic-linker=/lib/ld-linux.so.2
/usr/src/biuld/csu/crt1.o /usr/src/biuld/csu/crti.o
`gcc --print-file-name=crtbegin.o` /usr/src/biuld/linuxthreads/ex4.o
/usr/src/biuld/linuxthreads/libpthread.so -Wl,-rpath-link=/usr/src/biuld:/u
sr/src/biuld/math:/usr/src/biuld/elf:/usr/src/biuld/dlfcn:/usr/src/biuld/nss
:/usr/src/biuld/nis:/usr/src/biuld/rt:/usr/src/biuld/resolv:/usr/src/biuld/c
rypt:/usr/src/biuld/linuxthreads /usr/src/biuld/libc.so.6
/usr/src/biuld/libc_nonshared.a -lgcc `gcc --print-file-name=crtend.o`
/usr/src/biuld/csu/crtn.o
GCONV_PATH=/usr/src/biuld/iconvdata LC_ALL=C
/usr/src/biuld/elf/ld-linux.so.2 --library-path
/usr/src/biuld:/usr/src/biuld/math:/usr/src/biuld/elf:/usr/src/biuld/dlfcn:/
usr/src/biuld/nss:/usr/src/biuld/nis:/usr/src/biuld/rt:/usr/src/biuld/resolv
:/usr/src/biuld/crypt:/usr/src/biuld/linuxthreads
/usr/src/biuld/linuxthreads/ex4 > /usr/src/biuld/linuxthreads/ex4.out
make[2]: *** [/usr/src/biuld/linuxthreads/ex4.out] Error 139
make[2]: Leaving directory `/usr/src/glibc-2.2.2/linuxthreads'
make[1]: *** [linuxthreads/tests] Error 2
make[1]: Leaving directory `/usr/src/glibc-2.2.2'
make: *** [check] Error 2
Can any one tell me if it is safe to install Glibc with this check failure,
and what Error 139 may possibly be?
Also has anyone else had this error and know the cause?
Much appreciated
--
Regards Geoff :)
==================
[EMAIL PROTECTED]
------------------------------
From: Florian =?iso-8859-1?Q?Gro=DFe=2DCoosmann?= <[EMAIL PROTECTED]>
Subject: Re: [Help:] Kernel Upgrade
Date: Sat, 07 Apr 2001 10:39:19 +0200
Hi Jun,
"J. Liu" wrote:
> I am trying to upgrade Kernel 2.2.14 to 2.2.18. I compiled the source
> code of 2.2.18
> and copied them to /boot.
>
> When I reboot the system, it failed. The error message is "Kernel panic:
> No init found".
You don't give enough informations for a better answer:
To upgrade, do the following (in case of a running system!):
1) copy the kernel sources to the proper place
2) do a "make menuconfig"
always select the drivers of your boot and root device and their
file systems normally. YOU SHOULDN'T MAKE MODULES OF THIS
DRIVERS. This eleminates the need of createing an initial
ram disk containing needed modules. This might be your
mistake.
3) do a "make dep; make clean; make bzImage" (NOT MORE)
4) do a "make modules"
5) do a "make modules_install"
6) copy the kernel, e.g. "cp arch/i386/boot/bzImage /boot/A_SEPARATE_NAME"
never delete or overwrite an existing running kernel, never, never, never!!!
7) edit "/etc/lilo.conf" and ADD a new entry. There are enough entries
to copy but don't use initrd.
8) "lilo"
9) boot and wait until your boot manager or boot selection came up.
Then load the new kernel.
10)You might want to delete old kernels after some weeks of testing.
Cheers, Florian
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: The never ending socket :-)
Date: Sat, 07 Apr 2001 08:48:30 -0000
On Fri, 06 Apr 2001 22:00:18 +0200 Arnaud Westenberg <[EMAIL PROTECTED]> wrote:
| I have a (newbe!) question about Unix domain sockets.
|
| Suppose I want to write a daemon wich supports multiple connections for
| multiple processes. However I don't want to mess with all the socket
| names at the filesystem level.
|
| You'll probably say: No problem, just use 'socketpair' to create an
| unnamed pair of connected sockets. Ok, so I use 'socketpair' from within
| the daemon code, but how do I tell the user process wich (socket) file
| descriptor to use for his end of the connection? Do I have to use the
| 'ancillary data' for this? My first guess would be no, because this data
| still has to be transmitted through the socket (or am I missing
| something here), but what should one use instead?
You would use socketpair if the other process is being forked from
the one creating the sockets. What you get from socketpair is two
file descriptors already connected. Typically you would then form
a new process. The new process inherits the same variables so it
would have the file descriptor numbers in the same place as the
parent process. The code for the parent would use one and close
the other, and the code for the child would do the reverse. The
child will most likely need to close all the other file descriptors
the parent had open, otherwise those connections can get hung even
after the parent closes them because other children have them open.
It can make an ugly mess.
--
=================================================================
| Phil Howard - KA9WGN | Dallas | http://linuxhomepage.com/ |
| [EMAIL PROTECTED] | Texas, USA | http://phil.ipal.org/ |
=================================================================
------------------------------
From: "mike chan" <[EMAIL PROTECTED]>
Subject: halt problem from Turbo Linux
Date: Sat, 7 Apr 2001 16:08:32 +0800
when i type commands "halt" or "init 0" both exiting a message:
mkdir: cannot make directroy '/var/lib/pgsql/base'; Read-only file system
Starting post gresql service: /usr/bin/post master does no find the database
system. Expected to find it in the PGDATA directory "/vab/lib/pgsql", but
unable open file with pathname "/var/lib/pgsql/base/template1/pg-class".
No data directory-can't proceed.
failed.
how can i fix the problem and why it is exiting the problem
?
Thank you all first!
------------------------------
From: "DongKook Park" <[EMAIL PROTECTED]>
Subject: Re: Signal, Kernel and Co
Date: Sat, 7 Apr 2001 20:06:08 +0900
"The kernel checks whether a signal for any process has arrived when
switching from Kernel Mode to User mode... (roughly every 10ms)"
above statement is from 'Understanding the Linux Kernel p.252)
Any signal sent is maintained by p->signal...
*^^*;;
"Mikael Chambon" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Hi,
>
> I am trying to understand how a signal can received by a process.
> I think I've mostly understood everything about the sending
> step, but after send_sig_info() did a sigaddset and put
> p->sigpending = 1 (p is a pointer to the target task_struct)
> how the kernel now that a signal has arrived on the target process ?
>
> What is the link with the ret_from_intr() and do_signal functions ??
>
>
> Thanks if you can help me.
> --
> Mikael Chambon | Paris France | ICQ 10249913
------------------------------
From: "Ian D. Stewart" <[EMAIL PROTECTED]>
Subject: Re: [Help:] Kernel Upgrade
Date: Sat, 07 Apr 2001 10:01:51 -0400
In article <[EMAIL PROTECTED]>, "J. Liu" <[EMAIL PROTECTED]>
wrote:
> Hi,
>
> I am trying to upgrade Kernel 2.2.14 to 2.2.18. I compiled the source
> code of 2.2.18
> and copied them to /boot.
>
> When I reboot the system, it failed. The error message is "Kernel panic:
> No init found".
> I heard I need to upgrade the whole system.
>
> Advice needed to help me upgrade the kernel.
A few things you can look into.
In the source distribution, there should be a file 'Changes' under the
Documentation directory (on my system it's
/usr/src/linux/Documentation/Changes). That file lists (amoung other
things) the minimum level required for different programs. Make sure you
have that level or higher installed.
Also, if you configured any of the options as modules, make sure to run
'make modules' followed by 'make modules_install'.
Finally, make sure /etc/lilo.conf points to the location of your new
kernel. And make sure to reinstall LILO (run /sbin/lilo as root). This
needs to be done even if you didn't change /etc/lilo.conf.
HTH,
Ian
>
> Jun
------------------------------
From: Billy Bob Jameson <[EMAIL PROTECTED]>
Subject: LinuxThreads and thread-safety, re-entrancy, async-safety!
Date: Sat, 07 Apr 2001 16:17:49 GMT
[Repost from other newsgroup]
Hi all.
Need advice!
My company takes on to porting a very sensitive library to Linux (RedHat
7.0, glibc 2.2.9). The library uses POSIX threads heavily.
On Linux, at this time, LinuxThreads seems to be a very mature package
and looks like being the right choice (and the only good one). However,
we have come upon some documentation (not very up-to-date but
anyways...) on the net saying that glibc is not fully thread-safe
(LinuxThreads package README says that too). Also, the documentation
warns about the non re-entrancy of other libraries in the system and
about possible problems when mixing with libraries like svgalib that
uses SIGUSR signals that LinuxThreads use too.
So: what is your opinion? Stick with LinuxThreads or look after a
user-level threads package? We were thinking that a user-level thread
package would alleviate some of the problems.
What is the latest news in kernel re-entrancy (2.2.x kernels)? Any
links, points, etc.?
Thanks++
------------------------------
From: Nix <$}xinix{[email protected]>
Crossposted-To: comp.os.linux.networking
Subject: Re: TCP/IP socket buffering
Date: 07 Apr 2001 18:15:48 +0100
On Tue, 03 Apr 2001, David Schwartz stated:
> If you have no information about where the boundaries are, how can you
> separate them?
In the application I fixed that tried to treat TCP as a datagram
protocol, the boundaries were separated by wishful thinking.
(i.e. `one read() will get one command back'. When asked why, the
original author said `because the other end told it to do that'. How,
exactly? `I dunno.')
--
`I tried to get rid of it... and oh my god the dragons.' --- Zack Weinberg
on -iprefix
------------------------------
From: Pasztor Szilard <[EMAIL PROTECTED]>
Subject: Re: usleep() is unreliable when sleeping for less then 10000 micro
Date: 7 Apr 2001 18:33:46 GMT
Norm:
> Investigate the Real-Time Linux system at
> http://www.rtlinux.org
>
> You'll find the answers to all of your questions.
The applications often must run on any standard linux distribution, and in
this case, rtlinux is not a viable alternative.
By the way, there is one other method: use the real time clock and IRQ8, you
can get the cpu at the irq occurrence.
----------------------------------------------------------
| Two billion men shave twenty-thousand years a day. |
----------------------------------------------------------
------------------------------
From: "J. Liu" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Kernel 2.2.18
Date: Sat, 07 Apr 2001 17:39:54 -0400
Hello,
I'd like to know which distribution comes with Kernel 2.2.18 ?
Thanks.
Jun
------------------------------
From: "J. Liu" <[EMAIL PROTECTED]>
Subject: Re: [Help:] Kernel Upgrade
Date: Sat, 07 Apr 2001 17:44:20 -0400
Thank for your information.
When I reboot the machine, it stooped before file system mounting.
So I wonder if there is any problem with "mount". How do I upgrade it
too ?
Florian Gro=DFe-Coosmann wrote:
> =
> Hi Jun,
> =
> "J. Liu" wrote:
> > I am trying to upgrade Kernel 2.2.14 to 2.2.18. I compiled the source=
> > code of 2.2.18
> > and copied them to /boot.
> >
> > When I reboot the system, it failed. The error message is "Kernel pan=
ic:
> > No init found".
> You don't give enough informations for a better answer:
> =
> To upgrade, do the following (in case of a running system!):
> 1) copy the kernel sources to the proper place
> 2) do a "make menuconfig"
> always select the drivers of your boot and root device and their
> file systems normally. YOU SHOULDN'T MAKE MODULES OF THIS
> DRIVERS. This eleminates the need of createing an initial
> ram disk containing needed modules. This might be your
> mistake.
> 3) do a "make dep; make clean; make bzImage" (NOT MORE)
> 4) do a "make modules"
> 5) do a "make modules_install"
> 6) copy the kernel, e.g. "cp arch/i386/boot/bzImage /boot/A_SEPARATE_NA=
ME"
> never delete or overwrite an existing running kernel, never, never, =
never!!!
> 7) edit "/etc/lilo.conf" and ADD a new entry. There are enough entries
> to copy but don't use initrd.
> 8) "lilo"
> 9) boot and wait until your boot manager or boot selection came up.
> Then load the new kernel.
> 10)You might want to delete old kernels after some weeks of testing.
> =
> Cheers, Florian
------------------------------
From: nordi <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Kernel 2.2.18
Date: Sun, 08 Apr 2001 00:37:00 +0200
"J. Liu" wrote:
> I'd like to know which distribution comes with Kernel 2.2.18 ?
SuSE 7.1 does, although the kernel is slightly
customized by the SuSE developers.
nordi
------------------------------
** 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 the
comp.os.linux.development.system newsgroup.
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
******************************
