Linux-Development-Sys Digest #676, Volume #6 Tue, 4 May 99 09:15:23 EDT
Contents:
Re: swapon, util-linux 2.9r (Allin Cottrell)
Re: PPP over Ethernet (Pete Zaitcev)
Re: Powering on terminal (Olav Woelfelschneider)
Re: 80x25 Video Memory & Color (Olav Woelfelschneider)
Re: /dev/hda1 has reached maximal mount count, check forced (Horst von Brand)
what is safe optimisation level for kernel? (Jacek Pop�awski)
Re: Threads >> PThreads or LinuxThreads?? (Marcus Sundberg)
Re: Segmentation fault (J.H.M. Dassen (Ray))
Re: Mac-emulation on Linux?
Help with source routing (Bernat Ginard)
mmap for a multi user device driver (fwd) ("[EMAIL PROTECTED]")
Re: redhat 6.0? (Craig Kelley)
Re: Linux Real-Time I/O (Gerard van der Sel)
Re: Linux Real-Time I/O ("G. Sumner Hayes")
Re: CVS (Re: Bill Gates, self made man, NOT!) (Christopher Browne)
Re: using interrupt handlers ("Hermann-J. Mathes")
Segmentation fault (Michele)
Re: Q: What can I do when SVGALIB did not support my video card(chipset)??? (Jacek
Pop�awski)
----------------------------------------------------------------------------
From: Allin Cottrell <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Re: swapon, util-linux 2.9r
Date: Mon, 03 May 1999 22:56:18 -0400
NF Stevens wrote:
> chmod 600 /dev/hda2
OK, silly me. Thanks also to the others who said the same.
Didn't realize you could do that for non-mounted devices.
Allin Cottrell.
------------------------------
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.networking
Subject: Re: PPP over Ethernet
From: [EMAIL PROTECTED] (Pete Zaitcev)
Date: Tue, 04 May 1999 00:09:23 GMT
Frank Sweetser <[EMAIL PROTECTED]> writes:
>Mattias Wildeman <[EMAIL PROTECTED]> writes:
>> Is there anyone out there who is working on a PPP over Ethernet client
>> for Linux?
>??? PPP requires a clear text channel to send data over. no such channel
>exists for ethernet, though you could it over an IP session. what could
>this be for?
How is Ethernet not a clear text? IMHO it is quite clear.
There is an RFC about PPP over Ethernet.
It may be useful if you want to use full-duplex Ethernet over
long haul between routers. So basically it turns Ethernet into
a high speed serial.
--Pete
------------------------------
From: Olav Woelfelschneider <[EMAIL PROTECTED]>
Subject: Re: Powering on terminal
Date: Mon, 3 May 1999 22:16:05 +0200
Arthur Rinkel <[EMAIL PROTECTED]> wrote:
AR> Hi, how do I configure Linux so that when I turn on my terminal (DEC
AR> VT320) I always get a login prompt on that terminal? In that case, Linux
Make sure the terminal's DTR output is wired to the computers DCD input.
A correct Nullmodem cable will do just this. Too bad there are too many
broken cables and broken wiring plans out there.
A correct, full nullmodem cable goes like this, check your favorite
hardware guide for pin numbers.
Terminal Computer
RXD TXD
TXD RXD
RTS CTS
CTS RTS
DTR DSR,DCD
DSR,DCD DTR
For a DEC VTxxx, RTS/CTS is usually not needed, and the terminal isn't
interested in DSR/DCD either, so what's left is:
Terminal Computer
RXD TXD
TXD RXD
DTR DSR,DCD
Four wires, including ground (Leaving RTS/CTS/RI/DTR of the computer free
for other geeky hacks, like a radio clock :-)
The DTR/DCD trick will make the computer think a modem has just called in
when you turn on the terminal. Also the computer will think the modem has
hung up if you turn the terminal off. So you will be automagically logged
out with a -click-.
You must make sure the getty you run on the serial line checks for DCD.
--
Olav "Mac" W�lfelschneider [EMAIL PROTECTED]
PGP fingerprint = 06 5F 66 B3 2A AD 7D 2D B7 19 67 3C 95 A7 9D AF
Mer mu� doch nur emol e bissje nochdenke. -- Mundstuhl
------------------------------
From: Olav Woelfelschneider <[EMAIL PROTECTED]>
Subject: Re: 80x25 Video Memory & Color
Date: Mon, 3 May 1999 22:06:24 +0200
[EMAIL PROTECTED] wrote:
dhc> Please dont!! A full screen refresh is what, about 4KB, with a modern
dhc> system (say newer then a 386) screen refresh time using a decent library
dhc> (curses or slang) is still tiny!
You are right. Too bad that programming curses is a pain in the ass.
And to the original poster: Please try to stick to something like curses
or slang, so your app can run through a network connection or serial
link.
However, if you insist on restricting your app to the console, try to learn
about the /dev/vcsX devices. These give you direct access to the screen
memory.
There are a few special cases, where this makes sense. The vche
(VirtualConsoleHexEditor) is such a case.
--
Olav "Mac" W�lfelschneider [EMAIL PROTECTED]
PGP fingerprint = 06 5F 66 B3 2A AD 7D 2D B7 19 67 3C 95 A7 9D AF
Mer mu� doch nur emol e bissje nochdenke. -- Mundstuhl
------------------------------
From: [EMAIL PROTECTED] (Horst von Brand)
Crossposted-To: comp.os.linux.hardware
Subject: Re: /dev/hda1 has reached maximal mount count, check forced
Date: 2 May 1999 18:08:34 GMT
On Fri, 30 Apr 1999 16:59:13 -0700, David Peavey <[EMAIL PROTECTED]> wrote:
>/dev/hda1 has reached maximal mount count, check forced
>I get this message sometimes when I boot up. Can someone tell me what this
>means (besides the obvious)? How do fix it? We are doing some system
>integration work here and this message occurs occasionally.
Just what it says: The filessystems are checked on each boot, after a few
boots a full check is made. Nothing to fix here. Just boot less if it
bothers you ;-)
tunefs(8)
--
Horst von Brand [EMAIL PROTECTED]
Casilla 9G, Vi�a del Mar, Chile +56 32 672616
------------------------------
From: [EMAIL PROTECTED] (Jacek Pop�awski)
Subject: what is safe optimisation level for kernel?
Reply-To: [EMAIL PROTECTED]
Date: 2 May 1999 10:11:46 GMT
I read that people compile 2.2.x with egcs, and I heard that 100% safe
is only gcc. What is the true? Can I safe compile kernel with -march=pentium?
What about pgcc? I read that XFree-3.3.3 hangs when compiled with pgcc,
so pgcc is bad and egcs is safe? Is it bug in pgcc and will it be corrected?
------------------------------
From: Marcus Sundberg <[EMAIL PROTECTED]>
Subject: Re: Threads >> PThreads or LinuxThreads??
Date: Wed, 28 Apr 1999 10:29:48 +0200
Martin Recktenwald wrote:
>
> "G. Sumner Hayes" <[EMAIL PROTECTED]> writes:
>
> > Not with current glibc and kernel if I understand correctly. (glibc
> > 2.1 and linux 2.2) The LinuxThreads package in glibc2.1 uses its own
> > signals is you are running a kernel that supports them.
> >
> > At least that's what I've heard. I haven't actually tested it -- please
> > correct me if I'm wrong.
> >
> I heard the same but couldn't verify; on the linuxthreads-homepage it
> still says SIGUSR1/2 is not usable and glibc2.1 is currently not
> available.
This is from the linuxthreads FAQ in glibc2.1:
On ``old'' kernels (2.0 and early 2.1 kernels), there are only 32
signals available and the kernel reserves all of them but two:
SIGUSR1 and SIGUSR2. So, LinuxThreads has no choice but use those
two signals.
On recent kernels (late 2.1 kernels and the forthcoming 2.2 kernels),
more than 32 signals are provided in the form of realtime signals.
When run on one of those kernels, LinuxThreads uses two reserved
realtime signals for its internal operation, thus leaving
SIGUSR1 and SIGUSR2 free for user code.
//Marcus
--
===============================+====================================
Marcus Sundberg | http://www.stacken.kth.se/~mackan/
Royal Institute of Technology | Phone: +46 707 295404
Stockholm, Sweden | E-Mail: [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (J.H.M. Dassen (Ray))
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Segmentation fault
Date: 4 May 1999 09:25:59 GMT
Michele <[EMAIL PROTECTED]> wrote:
>How can I debug the core file left from a process that has operated a
>segmentation fault?Is there a tool that let you see what has happened over
>that process??
A core file represents a "snapshot" of the process at the moment it crashed.
You can examine it using a debugger like "gdb".
Read http://www.wi.leidenuniv.nl/home/kuyper/stuva2/debug.ps.gz for an
introduction to debugging code in a Unix environment.
HTH,
Ray
--
ART A friend of mine in Tulsa, Okla., when I was about eleven years old.
I'd be interested to hear from him. There are so many pseudos around taking
his name in vain.
- The Hipcrime Vocab by Chad C. Mulligan
------------------------------
Crossposted-To: comp.os.linux.advocacy,comp.os.linux.misc,comp.os.linux.powerpc
Date: Mon, 3 May 1999 20:53:00 -0400
From: <[EMAIL PROTECTED]>
Subject: Re: Mac-emulation on Linux?
> > Where can you get the LinuxDisks extension?
> > Its website seems to be down.
> > http://w3.teaser.fr/~mpollet/LinuxDisks/
> http://www.penguinpowered.com/~louisk
> check the macutils section
This file doesn't seem to uncompress with Stuffit Expander.
Do you know how to get this to work, or know of another
site with the file?
Greg
------------------------------
From: Bernat Ginard <[EMAIL PROTECTED]>
Subject: Help with source routing
Date: Tue, 04 May 1999 12:06:00 +0200
I'm trying to configure the source routing that comes
with 2.2.x kernels but I have the problem that none of
the utilities I have tried seems not support it.
They are net-tools 1.52 and iproute2 2.2.4. The documen-
tation that comes with them has no reference to it.
Could somebody tell me what I need to configure it.
Thanks in advance
-- =
_____________________________________________________
Bernat Ginard Llad=F3
Mailto:[EMAIL PROTECTED] http://www.kaos.es
------------------------------
From: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
Subject: mmap for a multi user device driver (fwd)
Date: Tue, 4 May 1999 12:16:40 +0200
Reply-To: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
Hi everybody,
Dirk asked me to forward this message to the appropiate newsgroup.
I hope here is the right place. Any help for the problem would be
appreciated.
Thanks
Hermann-Josef
***************************************************************************
* *
* Dr. Hermann-Josef Mathes *
* KASCADE Collaboration *
* Forschungszentrum Karlsruhe Phone: +49 7247 822429 *
* Institut fuer Kernphysik 3 FAX: +49 7247 824075 *
* POB 3640 *
* D-76021 Karlsruhe/Germany Mail: [EMAIL PROTECTED] *
* *
***************************************************************************
* Feel free to visit my homepage: *
* http://www-ik3.fzk.de/~mathes/Welcome.html *
* *
* Visit the institute and project homepages: *
* http://www-ik3.fzk.de/ *
* http://www-ik1.fzk.de/KASCADE_home.html *
***************************************************************************
========== Forwarded message ==========
Date: Mon, 3 May 1999 14:32:23 +0200 (MEST)
From: [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Subject: mmap for a multi user device driver
Hi,
I have a 2.0.36 kernel running on a PII. I use
this machine to develop a device driver for a
PCI-/VMEbus adaptor.
In the case of multiple process access to the
driver, it is vital to switch the vme descriptor
(contains parameters like address-, access- and
data-type besides a base address on the vme bus)
on each scheduling of a process using the mmapped
feature.
To achieve this goal, i use a nopage handler, which
does the switching of the vme descriptor and queues
a task (in tq_scheduler), which destroys the mapping
of this process window to the hardware window
of the VMEbus when the process will deschedule to kick
the nopage handler on next schedule.
the drivers mmap look like this:
static int foo_mmap(struct inode *inode, struct file *file,
struct vm_area_struct *vma)
{
.
.
vma->vm_ops = &foo_vmops;
vma_set_inode(vma,inode);
inode->i_count++;
// no remap done here to give the
// nopage handler a chance to work
foo_vma_open(vma);
return 0;
}
and the nopage handler:
static unsigned long foo_vma_nopage(struct vm_area_struct *vma,
unsigned long address, int no_share)
{
.
.
foo_vme_switch_context(...);
queue_task(&psn, &tq_scheduler);
return vme_base;
}
the struct psn contains the process_switch routine:
static void foo_process_switch_notifier(void *data)
{
.
.
do_mmap(file, user_page, vme_range,
PROT_READ|PROT_WRITE, MAP_FIXED|MAP_SHARED,
vme_base);
// the unmap is done in do_mmap to kick the
// nopage handler on next schedule
}
The mmapped area is released in foo_release() via
munmap(user_page, vme_range);
This mechanism works fine, the vme descriptor switching is
done each schedule, but after a test run the kernel is
UNSTABLE resulting in file system errors and many SIGSEGVs,
sometimes the /var/log/messages file contains logs from
free_one_pmd(): bad directory entry ...
My question is, if it's possible to perform this task in this
way or have i used the do_mmap()/do_munmap() kernel routines
incorrectly?
Any help would be very appreciated.
Many Thanks
------------------------------
Crossposted-To: comp.os.linux.advocacy
Subject: Re: redhat 6.0?
From: Craig Kelley <[EMAIL PROTECTED]>
Date: 28 Apr 1999 10:48:07 -0600
[EMAIL PROTECTED] (Des Herriott) writes:
> My major gripe with GNOME is its incredible slowness as soon as you
> enable any kind of pretty theming, in either Gtk or Enlightenment. Am
> I missing something? I haven't seen anyone else complain, but I've
> tried it on two fairly high-end machines, and glacial is about the best
> way I can describe it.
I get decent performance when imlib's cache is pumped up. The machine
has 128MB of RAM and imlib caches all the widget-looks (I can't think
of a better term) so that no disk activity is needed. It is a P5/200
machine (and gnome is compiled with the Pentium compiler, pgcc).
Other than that solution, you should probably avoid "pixmap" themes if
you want decent performance. Unfortunately, pixmap themes are the
easiest to "write".
> The other thing I dislike about the default Red Hat (5.9) distribution
> is the fact that E is not as nicely integrated as one might like - E
> places a row of icons over the panel, in the bottom right of the
> screen. That's a pretty minor cosmetic issue, to be fair, but first
> impressions count, and that's not a good first impression.
That is the stupid ShineyMetal theme -- just edit the buttons out of
the configuration in /usr/local//Enlightenment/themes/ShineyMetal (or
wherever the RPM puts them) or use a different theme.
--
The wheel is turning but the hamster is dead.
Craig Kelley -- [EMAIL PROTECTED]
http://www.isu.edu/~kellcrai finger [EMAIL PROTECTED] for PGP block
------------------------------
From: Gerard van der Sel <[EMAIL PROTECTED]>
Subject: Re: Linux Real-Time I/O
Date: Tue, 04 May 1999 09:17:32 +0200
Reply-To: [EMAIL PROTECTED]
Alex Maranda wrote:
>
> Jim & Lisa Meils wrote:
> >
> > I am a newbie to Linux, and have a somewhat simple question. I have
> > written an increadibly simple I/O program for DOS using QBASIC, and
> > would like to write a similar program for Linux and possibly use the
> > timing and multitasking capabilities to dim lights with it. Currently I
> > use the
> > parallel port for the output and control up to 8 light circuits. What
> > is the best route to take with Linux? Can Linux control in real-time
> > i.e. run a routine EXACTLY 120 times/sec ?
> 8.3 ms...not with the standard kernel; check out RT-Linux (Real Time
> Linux) at http://luz.cs.nmt.edu/~rtlinux/
A rather small task (20 lines of c code) is running on 120 usec (8192
Hz) interval on my system (kernel module, irq 8 driven).
It is a standard RedHat 5.2 (kernel version 2.0.36) distro.
CPU: pentium 200 MHz MMX
So he doesn't need RT-linux.
--
Met vriendelijke groet,
Gerard van der Sel
Mailto:[EMAIL PROTECTED]
"De dinosaurussen hadden hun komeet, wij hebben de Windows computer" -
me
"The box said: 'install on Windows 95, NT 4.0 or better'.
So I installed it on Linux."
------------------------------
From: "G. Sumner Hayes" <[EMAIL PROTECTED]>
Subject: Re: Linux Real-Time I/O
Date: Tue, 04 May 1999 04:03:14 -0400
Gerard van der Sel wrote:
>
> Alex Maranda wrote:
> >
> > 8.3 ms...not with the standard kernel; check out RT-Linux (Real Time
> > Linux) at http://luz.cs.nmt.edu/~rtlinux/
> A rather small task (20 lines of c code) is running on 120 usec (8192
> Hz) interval on my system (kernel module, irq 8 driven).
> It is a standard RedHat 5.2 (kernel version 2.0.36) distro.
> CPU: pentium 200 MHz MMX
> So he doesn't need RT-linux.
Point of clarification:
In stock Linux, user-level processes only get scheduled at 100Hz. So
they can't normally be guaranteed 120th of a second responses.
Kernel modules don't suffer this limitation.
You might be able to work around this by changing the value of HZ in
the kernel, but I'm not certain how effective or stable that would be --
I know some people claim to run with increased HZ without problems,
but I wouldn't do it in a production environment without serious
testing in your configuration. I don't know that it would be sufficient
to ensure the behavior you want, though it's probably necessary.
--Sumner
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Subject: Re: CVS (Re: Bill Gates, self made man, NOT!)
Reply-To: [EMAIL PROTECTED]
Date: Tue, 04 May 1999 02:28:26 GMT
On 03 May 1999 19:47:38 +0100, Nix <[EMAIL PROTECTED]> wrote:
>[EMAIL PROTECTED] (H. Peter Anvin) writes:
>> From what I've seen, several of these problems are structural.
>> Attempting another graft on top of CVS is just going futher down the
>> hole. Sometimes one just have to fix everything from scratch...
>
>That was meant to read `another graft on top of RCS', yes?
Hmmm... The answer probably doesn't matter. In the "old days," they'd
graft things onto RCS, which is where CVS came from.
Now, things like Aegis "graft onto" CVS...
I don't think there's a straight answer to that question...
On the other hand, with respect to building the "better Version Control
System," it sounds to me like there is pretty substantial agreement that
we don't quite have it right yet.
Hopefully once the "right one" comes along, there will correspondingly
be some equivalent to sccs2rcs constructed to transform CVS archives
into the new form.
[I've found RCS to be more-or-less good enough for my purposes, although
I'm starting to notice inadequacy for managing the situation where I
change my web page "name space..."]
--
"The idea that Bill Gates has appeared like a knight in shining armour to
lead all customers out of a mire of technological chaos neatly ignores the
fact that it was he who, by peddling second-rate technology, led them into
it in the first place." - Douglas Adams in Guardian, 25-Aug-95
[EMAIL PROTECTED] <http://www.hex.net/~cbbrowne/lsf.html>
------------------------------
From: "Hermann-J. Mathes" <[EMAIL PROTECTED]>
Subject: Re: using interrupt handlers
Date: Tue, 04 May 1999 12:43:08 +0200
TimL wrote:
>
> I was wondering if anyone has had experience with interrupt handlers. I
> have a PCI device that can generate an IRQ. I have written a simple handler
> for this within my driver/module, but for my application, I need the user
> application to be notified when an interrupt has taken place.
>
Hi Tim,
we developped a driver which uses to schemes for interrupt passing to
the users
program select() (suited for X applications) and signal.
Follow http://www-ik3.fzk.de/~mathes/software/software.html
and download atvme.0.29.tar.gz. File driver/atvme.c contains the code.
> What is the best method to have the driver/module inform the user
> application that an interrupt has taken place? Can I write something like a
> daemon, or would using threads be better? Or is there something easier?
>
IMHO this depends of the type of application and the speed requirement.
> In my application, I have a function that handles the interrupt request,
> however, I need to somehow call that function from my driver/module when an
> interrupt occurs.
>
> Thanks for any assistance,
> TL
In your case signal() seems to be appropiate. I don't know if all
functions of
user level interrupt handler are reentrant. Possible yes, if they are
used also
for multi-threading.
Hermann-Josef
--
***************************************************************************
*
*
* Dr. Hermann-Josef
Mathes *
* KASCADE
Collaboration *
* Forschungszentrum Karlsruhe Phone: +49 7247
822429 *
* Institut fuer Kernphysik 3 FAX: +49 7247
824075 *
* POB
3640 *
* D-76021 Karlsruhe/Germany Mail:
[EMAIL PROTECTED] *
*
*
***************************************************************************
------------------------------
From: Michele <[EMAIL PROTECTED]>
Subject: Segmentation fault
Date: Mon, 03 May 1999 17:03:49 +0200
Hello there.
How can I debug the core file left from a process that has operated a
segmentation fault?Is there a tool that let you see what has happened
over that process??
Ty in advance
Michele
------------------------------
From: [EMAIL PROTECTED] (Jacek Pop�awski)
Subject: Re: Q: What can I do when SVGALIB did not support my video card(chipset)???
Reply-To: [EMAIL PROTECTED]
Date: 2 May 1999 15:48:05 GMT
ei wrote:
>hi~~~
>could anyone please tell me what i can do when SVGALIB not support my
>Millennium II & G200 ?
>Is there any solution to use graphics for console mode on liunx or
>unix?
>Thanks for your reply...very very much...
...and my - similiar - question is:
when svgalib will support S3-Virge card?
------------------------------
** 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
******************************
