Path: gmd.de!xlink.net!sol.ctr.columbia.edu!spool.mu.edu!
darwin.sura.net!udel!news.intercon.com!panix!not-for-mail
From: ma...@panix.com (Phiber Optik)
Newsgroups: comp.unix.sys5.r4
Subject: syslogd
Date: 10 Aug 1993 09:25:42 -0400
Organization: PANIX Public Access Internet and Unix, NYC
Lines: 13
Message-ID: <2487km$63j@panix.com>
NNTP-Posting-Host: panix.com
Keywords: syslogd

I'm running Consensys, and having a problem with syslogd...
It starts up fine, and works fine locally, but it never
sets itself up on the net by binding to its port (514).
No error, it just never even tries.
A truss of syslogd shows that it is in fact checking 
/etc/netconfig and /etc/net/ticlts for connection-less
transports, but apparently isn't finding what it wants.
What do I have to do to accomplish this?
Any help appreciated, e-mail reply preferred.

Mark Abene
ma...@panix.com

Path: gmd.de!newsserver.jvnc.net!udel!news.intercon.com!panix!not-for-mail
From: ma...@panix.com (Phiber Optik)
Newsgroups: comp.unix.sys5.r4
Subject: kernel panic woes...
Date: 4 Sep 1993 10:04:11 -0400
Organization: PANIX Public Access Internet and Unix, NYC
Lines: 15
Message-ID: <26a78r$gag@panix.com>
NNTP-Posting-Host: panix.com
Keywords: kernel, panic, help

This is a dire emergency...  I'm currently getting continuous kernel
panics on my 486DX2-66 running Consensys 4.2.  I have 32 megs of memory.
The panics are consistently stopping at the same exact address in the 
kernel:  kmem_alloc+1dd.  Apparently some sort of problem with kernel
memory allocation.  Has anyone else been experiencing this?
I recall hearing talk about paging/swapping problems, and a buggy
mmap function.  Can anyone shed some light on this?
Any help would be appreciated, e-mail is preferred.
Again, this is an emergency, as I maintain a large multi-user conferencing
system on the net, and constant kernel panics every few hours is a hellish
nightmare.

Thanks,
Mark Abene

Path: gmd.de!newsserver.jvnc.net!udel!news.intercon.com!panix!not-for-mail
From: ma...@panix.com (Phiber Optik)
Newsgroups: comp.unix.sys5.r4
Subject: kernel panics revisited...
Date: 12 Sep 1993 15:45:35 -0400
Organization: PANIX Public Access Internet and Unix, NYC
Lines: 24
Message-ID: <26vu8v$e3@panix.com>
NNTP-Posting-Host: panix.com
Keywords: kernel, panic, consensys

Not long ago, I posted about my kernel panic woes with crashes
at kmem_alloc().  It turned out that IRC (version 2.2.1.2) was the culprit.
It appears that SVR4.2's malloc and mmap/brk's are seriously fucked.
One would run IRC, /quit, and the system would either hang immediately
or panic in a minute.  Malloc/free calls seemed to be in order, and I even
tried linking in GNU's malloc, which changed nothing.
It seriously bothers me that anyone at all can compile something that
should execute fine, yet can cause system crashes.

Now, continuing the story, an old kernel panic that went away has recently
resurfaced for some reason, and it's driving me batty.
I'm panicking repeatedly with the trap at crget().  Has anyone else
gotten this panic?  If anyone knows the function of crget, or what the
"cr" is referring to, I would be greatly indebted.
Someone, anyone, HELP!!!
Once again, I'm running Consensys SVR4.2 with 32MB.  And no, pulling out
16MB only slows down the system, but the panics remain, for those who
suggested it.  It changed nothing.
E-mail responses are preferred.
Thanks.

Mark Abene  SysAdmin of Echo BBS, NYC.

Path: gmd.de!newsserver.jvnc.net!udel!news.intercon.com!panix!not-for-mail
From: ma...@panix.com (Phiber Optik)
Newsgroups: comp.unix.sys5.r4
Subject: Consensys kdb...
Date: 22 Sep 1993 04:45:26 -0400
Organization: PANIX Public Access Internet and Unix, NYC
Lines: 11
Message-ID: <27p3b6$ocf@panix.com>
NNTP-Posting-Host: panix.com

Question:  is there a way to have kdb "timeout" if there's a panic and
it remains idle for some amount of time?  I'd like to leave kdb installed,
but if the system crashes and I'm not at the console, I'd like it to come
back up as quickly as possible, instead of sitting there till kingdom come
at the kdb prompt.
This is on Consensys SVR4.2 on an Intel 486/66.

Thanks, e-mail replies preferred.

Mark Abene, SysAdmin of Echo BBS, NYC.

Xref: gmd.de news.software.nn:2685 news.software.nntp:3976 
comp.unix.sys5.r4:5593
Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!
sol.ctr.columbia.edu!news.kei.com!echonyc.com!echonyc.com!not-for-mail
From: phi...@echonyc.com (Mark Abene)
Newsgroups: news.software.nn,news.software.nntp,comp.unix.sys5.r4
Subject: Big problems with NN
Date: 8 Oct 1993 02:39:54 -0400
Organization: ECHO BBS & Public Access Internet Site, NYC
Lines: 34
Distribution: inet
Message-ID: <2931vq$kd0@echonyc.com>
NNTP-Posting-Host: echonyc.com
X-Newsreader: TIN [version 1.1 PL9]

I'm running Consensys SVR4.2 (Unixware-like).
I've spent a great deal of time with NN and am still having
problems, so I'm making the following post:

While building the MASTER database seems to work without a hitch,
attempting to start nnmaster as a daemon fails with the log message
"Incomplete MASTER file".  So I tried having nnmaster invoked from
cron using the "-LM" flags, as suggested in the docs, and this bombs
soon after and mails me the fact that it failed with a segmentation
fault.  Is anyone else getting this?  It's driving me up a wall!!
Basically, for whatever reason, it doesn't like the MASTER file
it creates.  I'm hoping for relief from someone.  A good number of
the users here use NN frequently.

Oh, and before I forget, a little bug I am seeing in both NN and RN:
the main help screens have everything squished together.  I've seen
this on other systems as well.  The tabs are there, where they should
be, only they aren't expanded.  I believe someone forgot to leave
CBREAK mode before displaying help.  Why don't people notice these
little things?  Is this the "Emperor's New Help Screen"?

And last but not least, what makes the newsreaders ask the user if they
want to re-enter the last newsgroup they were in, the next time they
go to read news?  None of my readers are doing this.
Just incase anyone is wondering, I'm running NNTP with INN, but letting
my newsreaders get the articles straight out of /usr/spool/news, so as
not to waste net bandwidth.


-- 
Mark Abene - SysAdmin of Echo BBS and Public Access Internet Site, NYC
phi...@echonyc.com
phi...@panix.com
phi...@eff.org

Path: gmd.de!newsserver.jvnc.net!howland.reston.ans.net!sol.ctr.columbia.edu!
news.kei.com!echonyc.com!echonyc.com!not-for-mail
From: phi...@echonyc.com (Mark Abene)
Newsgroups: comp.unix.sys5.r4
Subject: TIOCNOTTY & TIOCGWINSZ ioctls broken...
Date: 18 Oct 1993 05:42:04 -0400
Organization: ECHO BBS & Public Access Internet Site, NYC
Lines: 20
Message-ID: <29todc$m3i@echonyc.com>
NNTP-Posting-Host: echonyc.com
X-Newsreader: TIN [version 1.1 PL9]

The title says it all.  I posted about this a while ago, but since I
got no response, I'm posting again.  TIOCNOTTY is supposed to disassociate
a process from its controlling tty.  I know this is SYSV and you'd commonly
use setpgrp(2) (or setsid(2)), but the ioctl call exists, and we're supposed
to be compatible, right?  The ioctl call fails everytime.  Go ahead and write
a 5 line program and see for yourself.

TIOCGWINSZ/TIOCSWINSZ are used to get/set the size of a (usually) text
window, and work with a structure called "winsize".  The ioctl fails.

Both TIOCNOTTY and TIOCGWINSZ are defined in sys/termios.h.

I'm running Consensys SVR4.2.  If anyone has had these similar problems,
or has solutions, I'd like to hear about it.

-- 
Mark Abene - SysAdmin of Echo BBS and Public Access Internet Site, NYC
phi...@echonyc.com
phi...@panix.com
phi...@eff.org

Path: gmd.de!xlink.net!howland.reston.ans.net!usenet.ins.cwru.edu!eff!
news.kei.com!news.echonyc.com!echonyc!phiber
From: phi...@echonyc.echonyc.com (Mark Abene)
Newsgroups: alt.test
Subject: test, test.
Date: 8 Nov 1993 00:58:49 GMT
Organization: ECHO BBS & Public Access Internet Site, NYC
Lines: 7
Message-ID: <2bk5k9$cr@subway.echonyc.com>
NNTP-Posting-Host: echonyc.com
X-Newsreader: TIN [version 1.2 PL2]

This is a test posting with the new tin.

--
Mark Abene - SysAdmin of Echo BBS and Public Access Internet Site, NYC
phi...@echonyc.com
phi...@panix.com
phi...@eff.org

Xref: gmd.de comp.unix.sys5.r4:6411 comp.unix.unixware:359
Path: gmd.de!newsserver.jvnc.net!darwin.sura.net!haven.umd.edu!
news.umbc.edu!eff!news.kei.com!news.echonyc.com!echonyc!phiber
From: phi...@echonyc.com (Mark Abene)
Newsgroups: comp.unix.sys5.r4,comp.unix.unixware
Subject: problem w/ latest sendmail
Date: 30 Nov 1993 12:31:43 GMT
Organization: ECHO BBS & Public Access Internet Site, NYC
Lines: 12
Message-ID: <2dfeff$30b@subway.echonyc.com>
NNTP-Posting-Host: echonyc.com
X-Newsreader: TIN [version 1.2 PL2]

I'm having one major problem compiling the latest sendmail, version 8.6.4:
Unixware (Consensys) apparently has no flock(2) system call to do file
locking.  I did a little work-around that basically calls lockf, which does
exist, only the flock fails with a "bad file number" error while trying to
lock the aliases database.  Has anyone successfully compiled this latest
sendmail?  And why doesn't Unixware have flock(2)???

--
Mark Abene - SysAdmin of Echo BBS and Public Access Internet Site, NYC
phi...@echonyc.com
phi...@panix.com
phi...@eff.org

Path: gmd.de!xlink.net!sol.ctr.columbia.edu!news.kei.com!news.echonyc.com!
echonyc!phiber
From: phi...@echonyc.com (Mark Abene)
Newsgroups: comp.unix.sys5.r4
Subject: SoundBlaster Driver...
Date: 25 Dec 1993 07:02:34 GMT
Organization: ECHO BBS & Public Access Internet Site, NYC
Lines: 20
Message-ID: <2fgoia$ma@subway.echonyc.com>
NNTP-Posting-Host: echonyc.com
X-Newsreader: TIN [version 1.2 PL2]

Previously, Jan Brittenson (bs...@moomin.ai.mit.edu) posted about a driver
for the SoundBlaster.  I ftped the archive, and looked it over.  I had
idinstall convert the old format Master, System, Node, etc. files to 4.2
format, which it had no trouble doing, and changed my interrupt appropriately
in the sb System file.  I rebuilt the kernel, everything went fine.
I ran the tst_fm_note program, and a nice chord blurted out of the speakers.
Yay.  Now for the problem.  Whenever I try to play a .au file, either manually
or through Mosaic, sometimes it sounds fine, yet other times play seems not
to read the header properly and the speed is wrong.  Most of the time, there
is loud, random noise interspersed with the audio, other times, the entire
sample is just loud noise.  I ftped some .snd files, and they simply don't
play at all.  Just noise.  What's the deal with this driver?  Needs a little
more work?  Are there others?  It's really great with Mosaic, only it's very
sporadic and unreliable.  By the way, I'm using Consensys SVR4.2 (UnixWare).

--
Mark Abene - SysAdmin of Echo BBS and Public Access Internet Site, NYC
phi...@echonyc.com
phi...@panix.com
phi...@eff.org