Linux-Development-Sys Digest #819, Volume #6 Fri, 11 Jun 99 17:15:26 EDT
Contents:
Re: /etc/inittab suggestion for all distributions (Alan Curry)
Re: TAO: the ultimate OS (Donovan Rebbechi)
Re: Any Journaling FS development? (Frank Sweetser)
Re: TAO: the ultimate OS (Alexander Viro)
Re: APM problems with RH6.0 Linux on Thinkpad 560 (Albert C. Lee)
Re: the ultimate OS (Christopher Browne)
Alpha 4-Sale (Anonymous)
Re: 64-bit arithmetic inside kernel (Juergen Heinzl)
Re: pThreads and STL in RedHat 6.0 (Juergen Heinzl)
Linux- Kernel Thread to user process copy question. (Cuneyt Akinlar)
Re: egcs bootstrap (Juergen Heinzl)
Re: 64-bit arithmetic inside kernel (Juergen Heinzl)
Re: pThreads and STL in RedHat 6.0 (Juergen Heinzl)
Re: Pinning a thread to a processor (bill davidsen)
Re: 64-bit arithmetic inside kernel (Phil Howard)
----------------------------------------------------------------------------
Subject: Re: /etc/inittab suggestion for all distributions
From: [EMAIL PROTECTED] (Alan Curry)
Date: Thu, 10 Jun 1999 19:39:25 GMT
In article <[EMAIL PROTECTED]>,
Abdullah Ramazanoglu <[EMAIL PROTECTED]> wrote:
>Alan Curry wrote:
>>
>> Here's a better suggestion. Don't run X in continuous-respawn mode.
>>
>> Use xdm without any servers in the Xservers file. Then start the server with
>> X -indirect localhost. If it goes bad you can ctrl-alt-backspace it and it
>> doesn't come back.
>>
>It may be trivial for you. But new users are using the system as it is
>shipped. If a solution doesn't come with the distribution, newcomers are
>most likely affected.
Actually I meant my suggestion to be taken BY the distribution makers, so
that newcomers will not suffer the respawning-xdm-from-hell problem.
--
Alan Curry |Declaration of | _../\. ./\.._ ____. ____.
[EMAIL PROTECTED]|bigotries (should| [ | | ] / _> / _>
==============+save some time): | \__/ \__/ \___: \___:
Linux,vim,trn,GPL,zsh,qmail,^H | "Screw you guys, I'm going home" -- Cartman
------------------------------
From: [EMAIL PROTECTED] (Donovan Rebbechi)
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: TAO: the ultimate OS
Date: 11 Jun 1999 01:54:32 GMT
On Fri, 11 Jun 1999 01:12:25 GMT, Jimen Ching wrote:
>Donal K. Fellows ([EMAIL PROTECTED]) wrote:
>The essay, ESR wrote, was about the development models; cathedral and
>bazaar. Are you suggesting that the cathedral model does not use
>peer-review?
It can't use peer review to the same extent that open source can.
I am free to review the linux kernel. Can I do the same with
the MS Office source ? Are cathedral projects audited by an external
review team ? if not, then you don't really have "peer review" ( even
if you have a group project, there are "groupthink" issues )
> Do you have proof that peer-review is not used in the
>cathedral model?
It's obvious that *more* peer review takes place with open source.
The only question is "how much more".
You mention that the essay does not attempt to substantiate its
points. Sure it doesn't. It doesn't have to. It's not an attempt
to "prove" that open source is better. It's a manifesto. You have
every right to question or disagree with it: manifestos are difficult
to swallow whole.
-- Donovan
------------------------------
From: Frank Sweetser <[EMAIL PROTECTED]>
Subject: Re: Any Journaling FS development?
Date: 10 Jun 1999 15:43:18 -0400
[EMAIL PROTECTED] writes:
> In article <[EMAIL PROTECTED]>,
> Frank Sweetser <[EMAIL PROTECTED]> wrote:
>
> >yup - stephen tweedie is working on journaling code for e2fs, and in
>
> Does he have a cvs repository where we can get our grubby little hands
> on the work-in-progress?
no idea... you'd have to ask him.
> >addition SGI is working in porting and releasing linux code for the XFS
> >filesystem.
>
> I wonder why they're waiting until "later this summer" to "begin
> releasing code"
they need to make sure that they don't release any bits of code they're not
allowed to.
--
Frank Sweetser rasmusin at wpi.edu fsweetser at blee.net | PGP key available
paramount.ind.wpi.edu RedHat 5.2 kernel 2.2.5 i586 | at public servers
Woody: How are you today, Mr. Peterson?
Norm: Never been better, Woody. ... Just once I'd like to be better.
-- Cheers, Chambers vs. Malone
------------------------------
From: [EMAIL PROTECTED] (Alexander Viro)
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: TAO: the ultimate OS
Date: 10 Jun 1999 16:11:10 -0400
In article <[EMAIL PROTECTED]>,
Vladimir Z. Nuri <[EMAIL PROTECTED]> wrote:
>Alexander Viro ([EMAIL PROTECTED]) wrote:
>: Or about cryptography. Or about Fermat. Or about Termination Problem.
>: Or about new protocols. Or about evil elitists and censors on xxx.lanl.gov.
>: Or about the future of net.journalistics. Or about Vision of Future(tm).
>: Or about FAQs.
>
>hahaha....... you ascribe me much more than I can humbly take
Really? Want to turn it into point-by-point list? Puh-lease.
cryptorgaphy: cypherpunks list archives.
Termination Problem: your own postings (from your current address) + stuff on
your website.
xxx.lanl.gov - search on dejanews for Vladimir & Nuri & xxx.lanl.gov and read
the thread.
net.journalistics? Puh-lease. L.D., forgot IWRG FAQ? Reread it. Compare with
the current postings. Language, style, you name it.
Vision of Future - WTF else you are wanking about in *this* *very* *thread*?
yodda, yodda...
>credit for !! you seem to have stumbled into a hall of mirrors
>in your self-admitted "hours" of search engine meanderings!!
>but now I am curious!! what is it I said, in particular, that
>caused you to spend so much time ?? hahaha whatever it was
Grasshopper, it's not you, it's boredom. It was not something
particular that you've said - general style was very familiar, but I
couldn't recall it. Well, boredom is a great thing - especially after a
16-18 hours of tube time when you have to wait for testsuite to run
before you can go away and sleep... To make the long story short - I
had a couple of hours when I had to sit and watch the system going
through tests and I was bored senseless. I was curious - I knew that
I've seen your style somewhere but couldn't recall where it was. Well,
archives are there more or less for such situations. You somehow
disappeared from the net.visibility in '95. Stumbling upon you 4 years
later made me wonder - what did this loon do all those years... The
rest is obvious.
--
"You're one of those condescending Unix computer users!"
"Here's a nickel, kid. Get yourself a better computer" - Dilbert.
------------------------------
From: [EMAIL PROTECTED] (Albert C. Lee)
Crossposted-To: comp.os.linux.setup,comp.sys.laptops
Subject: Re: APM problems with RH6.0 Linux on Thinkpad 560
Date: Thu, 10 Jun 1999 16:20:11 -0400
> I've configured APM with the following options:
>
> CONFIG_APM=y
> # CONFIG_APM_IGNORE_USER_SUSPEND is not set
> CONFIG_APM_DO_ENABLE=y
> CONFIG_APM_CPU_IDLE=y
> CONFIG_APM_DISPLAY_BLANK=y
> CONFIG_APM_POWER_OFF=y
> CONFIG_APM_IGNORE_MULTIPLE_SUSPEND=y
> CONFIG_APM_IGNORE_SUSPEND_BOUNCE=y
> CONFIG_APM_RTC_IS_GMT=y
> # CONFIG_APM_ALLOW_INTS is not set
>
> Does anyone have any insights or suggestions on this?
Try this (works on my TP600E running RH 6.0, and got the APM setup from a
560X user):
CONFIG_APM=y
CONFIG_APM_IGNORE_USER_SUSPEND=N
CONFIG_APM_DO_ENABLE=N
CONFIG_APM_CPU_IDLE=Y
CONFIG_APM_DISPLAY_BLANK=Y
CONFIG_APM_POWER_OFF=Y
CONFIG_APM_IGNORE_MULTIPLE_SUSPEND=Y
CONFIG_APM_IGNORE_SUSPEND_BOUNCE=N
CONFIG_APM_RTC_IS_GMT=N
CONFIG_APM_ALLOW_INTS=Y
This should enable proper suspend and resume as well as hibernate without
lockups. Make sure that before you try and suspend or hibernate to issue
the following command:
cardctl eject 1
To turn off PCMCIA -- it will redetect when you resume.
-Al
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Crossposted-To: alt.os.linux,comp.os.linux.advocacy,comp.os.misc,comp.unix.advocacy
Subject: Re: the ultimate OS
Date: 11 Jun 1999 18:12:34 GMT
Reply-To: [EMAIL PROTECTED]
On 11 Jun 1999 12:47:19 -0400, Alexander Viro <[EMAIL PROTECTED]> wrote:
>In article <7jr2r9$v5k$[EMAIL PROTECTED]>,
>Stefaan A Eeckels <[EMAIL PROTECTED]> wrote:
>>This whole discussion about "opening objects and poking
>>around in their innards" is daft. It's about as useful
>>as having a cross between gdb and truss (strace for
>>Linuxers) as shell. Now if you'd been talking about
>Hey, ITS folks would probably beg to differ ;-) DDT *was* their shell.
>Seriously, at some moment I had to cruft up an ugly hack - line
>discipline that gave a simple set of commands allowing to walk through
>dcache, call inode and dentry methods, set/reset semaphores on inodes
>and tweak waitqueues. It was *not* pleasant, but it saved my ass
>several times - when you are hacking on VFS and it decides to hose
>itself... not too much can be done from userland. At least that way you
>can do accurate postmortem and make a relatively clean shutdown. From
>what I heard from -10 folks it's pretty close to their equivalent of
>shell <shudder>...
And I suspect that the Lisp Machine folk would have some somewhat
analagous stories, and this also sounds like it somewhat parallels the
Smalltalk environment, where the whole system can be regarded as a giant
debugging environment.
Given a suitable front end, there might be some merit to the idea. Not
that this can be mapped back onto the TAO ideas...
--
REALITY is a mescaline deficiency.
[EMAIL PROTECTED] <http://www.hex.net/~cbbrowne/lsf.html>
------------------------------
Subject: Alpha 4-Sale
Date: 11 Jun 1999 14:40:49 -0500
From: Anonymous <[EMAIL PROTECTED]>
I'm selling a Dec/Compaq Alpha Linux 533Mhz Workstation/Server.
I need the money, so I'm selling it very cheap.
Specs:
533Mhz Alpha-Processor 21164
AlphaPC 164LX Motherboard
RedHat Linux 6.0 Preinstalled With Updates + 2.2 Kernel
Windows NT 4.0 (optional)
2MB 3rd Level Synchronous SRAM Cache, 9ns
256 Megs SDRAM!
6.5 GiG HD
24x Cd-Rom Drive
Sony 1.44MB Floppy
SCSI Card
Sound Blaster 16
104-Key Keytronics Keyboard
Logitech Mouse
64-Bit 4meg Matrox Video Card
25 watt amp/eq with 4 "phono out".
Choice of 10mbs or 100mbs ethernet card (3com)
Axxion DL-17 Midtower Case w/Extra Cooling Fan and 300W Power Supply
Fastest and the best Linux Workstation/Server I've had... ever.
$1599
Links:
http://www.alphapowered.com
http://www.alphalinux.org
I'm in the San Jose, California area.
If your interested...
email: [EMAIL PROTECTED] or [EMAIL PROTECTED] (same)
Michael Asistido
3247 Almansa Ct.
San Jose CA 95127
--------== Posted Anonymously via Newsfeeds.Com ==-------
Featuring the worlds only Anonymous Usenet Server
-----------== http://www.newsfeeds.com ==----------
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Crossposted-To: comp.os.linux.questions,comp.os.linux.misc,comp.os.linux.help
Subject: Re: 64-bit arithmetic inside kernel
Date: Thu, 10 Jun 1999 21:07:42 GMT
In article <[EMAIL PROTECTED]>, Bill Woessner wrote:
>Hi. I'm trying to write a kernel module that uses 64 bit arithmetic and
>am having some problems with it. When I try to insert the module, I get
>the following errors:
>
>longlong.o: unresolved symbol __udivdi3
>longlong.o: unresolved symbol __umoddi3
Since the usual Intel CPU's are not 64 bit long long's are done in
software, part of libgcc.a. I cannot tell whether you can force gcc / egcs
to generate inline code here, sorry. If not you're probably up to some
asm code.
Cheers,
Juergen
--
\ Real name : J�rgen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Crossposted-To: comp.os.linux.development.app
Subject: Re: pThreads and STL in RedHat 6.0
Date: Thu, 10 Jun 1999 21:07:42 GMT
In article <[EMAIL PROTECTED]>, Allen Curtis wrote:
>I have glibc version 2.1.1-6 and do not seem to be able to link a
>program that uses both pThreads and STL. I know that has worked before.
>Can anyone tell me what the secret is to doing this?
g++ -pthread source.C does it fine here ... what is it complaining about (no
RH or other distribution but 2.1.1 too).
Ta'
Juergen
--
\ Real name : J�rgen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
From: [EMAIL PROTECTED] (Cuneyt Akinlar)
Crossposted-To: alt.os.linux,comp.os.linux
Subject: Linux- Kernel Thread to user process copy question.
Date: 11 Jun 1999 19:55:33 GMT
Hello,
I have a question about kernel to user space copy under Linux Kernel.
Basicly, I have a kernel thread which needs to copy data to a user buffer
of another process. That is, I am trying to copy data into a buffer in the
address space of a user process in the context of a kernel thread.
To make the question clear, I will give an example below:
Process X
{
read( fd, user_buffer, N );
}
|
|
\/ USER
=====================================================
KERNEL
interruptible_sleepon( &wait_queue ); /* Now Process X sleeps */
Kernel Thread wakes up, and
Kernel_thread()
{
char kernel_buffer[100];
memcpy_tofs( user_buffer, kernel_buffer, 100 ); /* copy data into user buf */
wake_up_interruptible( &wait_queue ); /* Wake-up process X */
}
As the above picture depicts, after I get a user buffer, I try to copy the
data into user_buffer of Process X within the context of Kernel_thread.
The kernel Thread crashes when it tries to copy the data.
Basicly, I get the following error:
"Cannot handle kernel page request" ( something like this )
And then kernel_thread crashes.
My question is, is there a solution to this problem. That is, is it
possible to copy data into user_buffer owned by process X in the context
of a kernel thread( Kernel_thread)?
Any help would be appriciated.
Regards
Cuneyt Akinlar
[EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Subject: Re: egcs bootstrap
Date: Fri, 11 Jun 1999 20:01:10 GMT
In article <7jr80k$am8$[EMAIL PROTECTED]>, gene wrote:
>I think I deleted something and now get this when I try to compile some
>programs:
>
>/tmp/ccPpUPmM.o: In function `main':
>/usr/src/gatos/yuvsum.c:64: undefined reference to `lstat'
>collect2: ld returned 1 exit status
>make: *** [yuvsum] Error 1
Since you still can run something it sounds like you might have
a glibc based system. If so make sure /usr/lib/libc.so is there
and looks like this ...
/* ----------------------------------- >8 ------------------------- */
/* GNU ld script
Use the shared library, but some functions are only in
the static library, so try that secondarily. */
GROUP ( /lib/libc.so.6 /usr/lib/libc_nonshared.a )
/* ----------------------------------- 8< ------------------------- */
... but do not blindly overwrite it of course.
Cheers,
Juergen
--
\ Real name : J�rgen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Crossposted-To: comp.os.linux.questions,comp.os.linux.misc,comp.os.linux.help
Subject: Re: 64-bit arithmetic inside kernel
Date: Fri, 11 Jun 1999 20:01:12 GMT
In article <[EMAIL PROTECTED]>, Peter Samuelson wrote:
>
>
> [Bill Woessner]
>> > Hi. I'm trying to write a kernel module that uses 64 bit
>> > arithmetic and am having some problems with it. When I try to
>> > insert the module, I get the following errors:
>> >
>> > longlong.o: unresolved symbol __udivdi3
>> > longlong.o: unresolved symbol __umoddi3
>
>[Juergen Heinzl <[EMAIL PROTECTED]>]
>> Since the usual Intel CPU's are not 64 bit long long's are done in
>> software, part of libgcc.a. I cannot tell whether you can force gcc
>> / egcs to generate inline code here, sorry. If not you're probably
>> up to some asm code.
>
>Inline code requires complete implementation in the header files,
>which AFAIK isn't there. But ... is there some reason you *couldn't*
>use libgcc.a in the kernel? I can't think of a good one, but I know
>how much I do not know about kernels, linking, etc. Of course, if
>__udivdi3 and __umoddi3 make use of the FPU, you're SOL.
Yes; I thought about a builtin (sorry for the confusion) but as you
said, __udivdi3 and __umoddi3 are probably self contained, a dissasemble
ought to show, and then linking to libgcc.a should do the job.
Ta'
Juergen
--
\ Real name : J�rgen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
From: [EMAIL PROTECTED] (Juergen Heinzl)
Crossposted-To: comp.os.linux.development.app
Subject: Re: pThreads and STL in RedHat 6.0
Date: Fri, 11 Jun 1999 20:01:11 GMT
In article <[EMAIL PROTECTED]>, Allen Curtis wrote:
>This is a multi-part message in MIME format.
>--------------51B1098564E65B4B7BAF80F5
>Content-Type: text/plain; charset=iso-8859-1
>Content-Transfer-Encoding: 8bit
>
>I get the following message when I try to build the program. I have also
>attached the files for your viewing pleasure.
>
>cc -D_REENTRANT qtest.cc -o qtest -lpthread
>/tmp/cchIKQsj.o: In function `__malloc_alloc_template<0>::oom_malloc(unsigned
>int)':
>/tmp/cchIKQsj.o(.__malloc_alloc_template<0>::gnu.linkonce.t.oom_malloc(unsigned
>int)+0x17): undefined reference to `endl(ostream &)'
>/tmp/cchIKQsj.o(.__malloc_alloc_template<0>::gnu.linkonce.t.oom_malloc(unsigned
>int)+0x21): undefined reference to `cerr'
>/tmp/cchIKQsj.o(.__malloc_alloc_template<0>::gnu.linkonce.t.oom_malloc(unsigned
>int)+0x26): undefined reference to `ostream::operator<<(char const *)'
>/tmp/cchIKQsj.o(.__malloc_alloc_template<0>::gnu.linkonce.t.oom_malloc(unsigned
>int)+0x31): undefined reference to `ostream::operator<<(ostream &(*)(ostream
>&))
>collect2: ld returned 1 exit status
>make: *** [qtest] Error 1
cc is the C compiler, use g++ -pthread or c++ -pthread qtest.cc -o qtest. No,
no -lpthread required (assuming egcs) and with using -pthread all what is
required will be defined right. I'd nor rely on -D_REENTRANT doing the job,
although this might be enough on Linux. On other systems there are similar
compiler options; it is safer to use them.
[...]
Cheers,
Juergen
--
\ Real name : J�rgen Heinzl \ no flames /
\ EMail Private : [EMAIL PROTECTED] \ send money instead /
------------------------------
From: [EMAIL PROTECTED] (bill davidsen)
Crossposted-To: comp.unix.solaris
Subject: Re: Pinning a thread to a processor
Date: 10 Jun 1999 22:16:52 GMT
In article <7jmjk5$ih9$[EMAIL PROTECTED]>,
Karsten Scholtyssik <[EMAIL PROTECTED]> wrote:
| I'm looking for a method to set the CPU affinity of a process/thread. Under
| Windows NT, I use the function SetProcessAffinityMask() - what is the Linux
| and/or Solaris equivalent? (I hope there *is* a way at all)
|
| I searched the man pages (pthread etc.) to no avail.
Out of curiousity, what do you think this binding with do for you? In
any case, Linux preferentially uses the previous processor for a
process, but doesn't bind exclusively (nor do you want it to, unless you
have one CPU which is different hardware than another).
Also, does this NT call bind the process to the processor or the
processor to the process?
--
bill davidsen <[EMAIL PROTECTED]> CTO, TMR Associates, Inc
The Internet is not the fountain of youth, but some days it feels like
the fountain of immaturity.
------------------------------
From: [EMAIL PROTECTED] (Phil Howard)
Subject: Re: 64-bit arithmetic inside kernel
Crossposted-To: comp.os.linux.questions,comp.os.linux.misc,comp.os.linux.help
Date: Thu, 10 Jun 1999 22:14:45 GMT
On Thu, 10 Jun 1999 21:07:42 GMT Juergen Heinzl ([EMAIL PROTECTED]) wrote:
| In article <[EMAIL PROTECTED]>, Bill Woessner wrote:
| >Hi. I'm trying to write a kernel module that uses 64 bit arithmetic and
| >am having some problems with it. When I try to insert the module, I get
| >the following errors:
| >
| >longlong.o: unresolved symbol __udivdi3
| >longlong.o: unresolved symbol __umoddi3
|
| Since the usual Intel CPU's are not 64 bit long long's are done in
| software, part of libgcc.a. I cannot tell whether you can force gcc / egcs
| to generate inline code here, sorry. If not you're probably up to some
| asm code.
Grab the code for these functions from libgcc source and compile them into
your module. That way the symbols will be there. Be sure to check for any
that those also call. You may also need to hack those functions to deal
with any assumptions they might make (e.g. other libraries) that do not
apply in the kernel. For example, serious error messages might need to
go to printk(). The quick hack would be to comment or ifdef out anything
you don't actually need.
--
Phil Howard KA9WGN
[EMAIL PROTECTED] [EMAIL PROTECTED]
------------------------------
** 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
******************************
