I’d say you need a bit of Cake, he claims to be very Uber with Linux see if you can PM him 😎 😎

Map: bremen_b2
Signal: Segmentation fault (11)
Siginfo: 0xbffcfa9c
Code: 1
Faulting Memory Ref/Instruction: 0x37fb0b60
DSO Information:
0x00e2b000 (unknown)
0x00d1c000 /lib/tls/i686/cmov/libdl.so.2
0x00113000 /usr/lib/libX11.so.6
0x00b4e000 /usr/lib/libXext.so.6
0x008ed000 /lib/tls/i686/cmov/libm.so.6
0x00359000 /lib/tls/i686/cmov/libc.so.6
0x00262000 /lib/ld-linux.so.2
0x00b0e000 /usr/lib/libxcb.so.1
0x00242000 /usr/lib/libXau.so.6
0x009fc000 /usr/lib/libXdmcp.so.6
0x0028e000 /lib/tls/i686/cmov/libnss_compat.so.2
0x00296000 /lib/tls/i686/cmov/libnsl.so.1
0x0027f000 /lib/tls/i686/cmov/libnss_nis.so.2
0x002ad000 /lib/tls/i686/cmov/libnss_files.so.2
0x004bd000 /usr/lib/tls/libnvidia-tls.so.1
0x004a2000 /usr/lib/libXcursor.so.1
0x00603000 /usr/lib/libXrender.so.1
0x004ad000 /usr/lib/libXfixes.so.3
0x0773d000 /home/mess/.etwolf/pb/pbcl.so
0x004b3000 /lib/libnss_mdns4_minimal.so.2
0x00b67000 /lib/tls/i686/cmov/libnss_dns.so.2
0x00806000 /lib/tls/i686/cmov/libresolv.so.2
0x00508000 /home/mess/.etwolf/pb/pbag.so
0x07910000 /home/mess/.etwolf/pb/pbsv.so
0x006dc000 /usr/lib/libGL.so.1
0x03e2c000 /usr/lib/libGLcore.so.1
0x069bb000 /home/mess/.etwolf/etpub/ui.mp.i386.so
0xa3db8000 /home/mess/.etwolf/etpub/cgame.mp.i386.so
Stack frames: 1 entries
Backtrace:
(1) /home/mess/.etwolf/etpub/cgame.mp.i386.so(CG_Respawn+0x194) [0xa3e478a4]
-88-

Attempting to clean up.
Received signal 11, exiting…

Strange that it is coming from two different functions.
Something is definitely screwy there, however I have no idea what it is 🙁

Cheers

When do these crashes occur or are they random?
And what is this a paste of exactly (maybe post a file of something)

Looks like a broken installation mess 🙂
Format and use the winblows version!!!

i tried to post in this topic and instead my oz forum crashed… LOL :S
hahah but as i said, its cause you told my secret

EVERYONE STAND BACK. sUPER CAEK TO THE RESCUE!!!

I arent seen ET segment fault before. When does it crash? Bremen or every map? ET will crash if your com_hunkmegs is too low thats for sure. Some big maps will kill ET.

Feel free to paste in lots o logs and stuff. Those logs appear to be from ET itself and not teh kernel which is making me suspect cvars, Teh iD software engines are usually solid as the gear stick in your grannys car on a cold morning.

The crashes are really quite random and never on the same map.

Its happened about 5 times so far over the space of a week. Ive looked around the logs and thats the only information Ive found on the crash. Its only been happening since the latest etpub client (oz version) Is it possible they’re related?

Looks like a broken installation mess 🙂
Format and use the winblows version!!!

NEVER

i tried to post in this topic and instead my oz forum crashed… LOL :S
hahah but as i said, its cause you told my secret

Sorry! :<

The crashes are really quite random and never on the same map.

Its happened about 5 times so far over the space of a week. Ive looked around the logs and thats the only information Ive found on the crash. Its only been happening since the latest etpub client (oz version) Is it possible they’re related?

Segmentation fault can also occur under following circumstances:

a) A buggy program / command, which can be only fixed by applying patch.

seg faults are cause by programs/apps that try to load themselves into reserved memory,Windows user’s get this as Access Violation Error.(usually followed by BSOD) So mess yup it is probably the updated version of pub.

The crashes are really quite random and never on the same map.

Its happened about 5 times so far over the space of a week. Ive looked around the logs and thats the only information Ive found on the crash. Its only been happening since the latest etpub client (oz version) Is it possible they’re related?

Nah – I’ve been getting the same crashes since last official etpub client 20090416 and also the last unofficial etpub client 20090912.

I probably get crash once every 40 hours or so of playing time….very random, random maps, random times. It can happen on same map.

Maybe a coincidence but usually when ET crashes I have been getting & sending more then normal amount of PM and adminchat messages……

Do you regularly play on any other ETpub servers?? Do you have many other etpub client pk3 files in your ET etpub folder?

Cheers.

This might be a little bit too much information but I think an strace might be helpful. Only problem is that strace will generate far too much traffic. Like 30mb every 5 seconds with ET.
I might write some kinda script to discard all but the last 10 seconds of logs so you can play until it crashes.

Hrm..

*scratches head*
hrmm..

Cos I have ET running successfully on Ubuntu 9.04 and 9.10 so it cant be anything wrong with ET or Ubuntu. Its got to be something unique to your system. What drivers are you running dood? Nvidia or ATi? Or Intel? Do you know how to tell which driver u are running?
Try running this in a command prompt:
lspci | grep VGA

Cos I have ET running successfully on Ubuntu 9.04 and 9.10 so it cant be anything wrong with ET or Ubuntu. Its got to be something unique to your system.

Interesting….so you’ve never had this crash, Cake?? I always thought it was a problem that affected all Linux clients.
I have been getting it for years – even with old official etpub client 20070801 from etpub.org (before they died…)

So you’ve given me renewed optimism that it’s just something wrong with my system… 😉

Cheers.

PS: Cake – the ET segfault actually calls a function in ET client that writes information to etpub_crash.log file before dying.
So if you have any specific ideas, I can add code to that function in attempt to gather more information upon crash…

Hai

Correct I have never had this crash.
Normally strace (system call tracing) makes a HUGE amount of data for a big program/game like ET. And all your memory goes down the toilet after about 2 minutes. But strace has the information we want.

I have taken the liberty of writing a script for you. Its pretty rough and makes your ET rather slow. But what it does is ensures you have the latest system calls logged whenever ET quits.
It run strace, but saves all system calls to a dump file in RAM, and switches between two dump logs every 2000 lines. That way when ET crashes, youll have the latest information in one of those dump logs.

A word of warning: This will severely degrade ET’s performance to the point of taking roughly 30 seconds or more to load a map. So best to only use it if you think that youre likely to catch it crashing.

Instructions:
* Extract tracelatest.sh from the archive and save it somewhere (anywhere). It should already be marked as executable.
* Jump into a command console, and switch to the directory you saved tracelatest.sh
* Set everything in motion by typing out the below line exactly. What you’re doing here is running ET under strace, but piped to the script.
strace et 2>&1 | ./tracelatest.sh

* When you’ve finished with ET or ET crashes, one either these two text files will exist:
/dev/shm/tracelatest_dump_a
or
/dev/shm/tracelatest_dump_b

Which ever one of those exists, zip it up and submit it to the forums.
Meanwhile we can continue in all the other exciting ways to get to the bottom of the cause.

Cheers cake, I’ll have a fiddle tonight.