Hello there.Y hope i will get the help i need here.
I have an HLDS hosted on unbuntu 10.10 server and it keeps crashing with these errors
----------------------------------------------
CRASH: Thu Dec 23 22:25:26 EET 2010
Start Line: ./hlds_amd -game cstrike -insecure +ip 195.2.197.62 +port 27015 +maxplayers 32 +map de_dust2
[New Thread 5823]
[New Thread 5828]
[New Thread 5826]
Core was generated by `./hlds_amd -game cstrike -insecure +ip 195.2.197.62 +port 27015 +maxplayers 32'.
Program terminated with signal 11, Segmentation fault.
#0 0x00ced492 in mysql_next_result () from cstrike/addons/amxmodx/modules/mysql_amxx_i386.so
#0 0x00ced492 in mysql_next_result () from cstrike/addons/amxmodx/modules/mysql_amxx_i386.so
#1 0x00ce908c in ?? () from cstrike/addons/amxmodx/modules/mysql_amxx_i386.so
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
No symbol table info available.
From To Syms Read Shared Object Library
0x004aea40 0x004af988 Yes (*) /lib/libdl.so.2
0x0045a610 0x00466ad8 Yes (*) /lib/libpthread.so.0
0x00b5dac0 0x00c67a64 Yes (*) /lib/libc.so.6
0x004e5830 0x004fc5cf Yes (*) /lib/ld-linux.so.2
0x00529d90 0x005accb4 Yes /usr/hlds/engine_amd.so
0x00f685f0 0x00f76ec4 Yes ./libsteam_api_c.so
0x001134b0 0x0012d7f8 Yes (*) /lib/libm.so.6
0x00fdba60 0x00fe9b4d Yes /usr/hlds/filesystem_stdio_i386.so
0x08f0809c 0x09031668 Yes ./cstrike/dlls/cs_i386.so
0x0013b184 0x0014e6e0 Yes ./cstrike/addons/metamod/dlls/metamod_i386.so
0x087602e0 0x08776ef8 Yes (*) /usr/hlds/cstrike/addons/dproto/dproto_i386.so
0x002542e0 0x002c5cd8 Yes (*) /usr/lib/libstdc++.so.6
0x001ddf50 0x001f4438 Yes /lib/libgcc_s.so.1
0x00384630 0x003de9e4 Yes (*) /usr/hlds/cstrike/addons/amxmodx/dlls/amxmodx_mm_i386.so
0x00ce12c0 0x00d63294 Yes (*) cstrike/addons/amxmodx/modules/mysql_amxx_i386.so
0x00ac3ca0 0x00ac9e34 Yes (*) cstrike/addons/amxmodx/modules/csx_amxx_i386.so
0x00202150 0x00207904 Yes (*) cstrike/addons/amxmodx/modules/cstrike_amxx_i386.so
0x002fcbe0 0x002ff6f4 Yes (*) cstrike/addons/amxmodx/modules/fun_amxx_i386.so
0x003120e0 0x0031c0c4 Yes (*) cstrike/addons/amxmodx/modules/engine_amxx_i386.so
0x0047d9f0 0x004a83b4 Yes (*) cstrike/addons/amxmodx/modules/hamsandwich_amxx_i386.so
0x00a55ac0 0x00a80b84 Yes (*) cstrike/addons/amxmodx/modules/fakemeta_amxx_i386.so
0x00334f30 0x00338264 Yes (*) cstrike/addons/amxmodx/modules/geoip_amxx_i386.so
0x06679ef0 0x0670c958 Yes (*) cstrike/addons/amxmodx/modules/orpheu_amxx_i386.so
0x072378d0 0x072a1fb4 Yes (*) cstrike/addons/amxmodx/modules/sqlite_amxx_i386.so
0x00351000 0x0035bd64 Yes (*) cstrike/addons/amxmodx/modules/csdm_amxx_i386.so
0x046765f0 0x04c0a5a4 Yes (*) ./steamclient.so
0x00ea33c0 0x00ec1a68 Yes (*) ./libtier0_s.so
0x00edd050 0x00f14fe0 Yes (*) ./libvstdlib_s.so
0x06f7a4b0 0x0705ba04 Yes ./libsteamvalidateuseridtickets.so
0x0033da20 0x00344e08 Yes (*) /lib/libnss_files.so.2
(*): Shared library is missing debugging information.
Stack level 0, frame at 0xbff71370:
eip = 0xced492 in mysql_next_result; saved eip 0xce908c
called by frame at 0x21a0
Arglist at 0xbff71368, args:
Locals at 0xbff71368, Previous frame's sp is 0xbff71370
Saved registers:
ebx at 0xbff71360, ebp at 0xbff71368, esi at 0xbff71364, eip at 0xbff7136c
End of crash report
----------------------------------------------
I requested help to discover the problem on a lot of forums and thei sey it`s because of conflicts bettwen the ubnuntu glibc version and the glibc version wich metamod-p was compiled and i should recompile metamod again on mi server.
Mi server has
GNU C Library (Ubuntu EGLIBC 2.12.1-0ubuntu10) stable release version 2.12.1, by Roland McGrath et al.
i downloaded the source of metamod-p from the site and tried to compile it with make command but i get this
Not that I have stopped developing, it's 'mature'. No bugs have been reported (that is bugs that are actually in metamod-p) and it's old for game. So nothing ground breaking changes happening there.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello,
Would it be possible Post a new Linux i586/i686 binary? With set the compiler flag to "intrinsic" SSE2
I have tried to compile the source code in Debian 6.0, with gcc (Debian 4.4.5-8) 4.4.5
many errors.
Thank You,
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Since this Adam person has forked the metamod-p (not submitted his fix upstream), I recommend that you ask him.
I see he is posting binary, then by GPL license (of metamod-p), he's obligated to give you the source-code (metamod-p32 + the fix) he used to create that binary.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I went through compiler generated assembler code of metamod-p32 (originally compiled with gcc-3.3.6), and recompile of metamod-p using ubuntu gcc-4.6.3.
I just can't see what enabling SSE2 is supposed to archive. I think what is seen here is effect of compiling metamod-p with newer compiler, tuned for current processor architechtures. And before actual performance numbers are shown, 'SSE2+friends make things better' is just placebo effect.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
You're still compiling in the root directory. Change directory to "~/metamod-1.19p33/metamod" then use "make OPT=opt". Other directories are example plugins for metamod and I didn't fix those for gcc-4.x (yet).
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
edman747@sc:~/metamod-p/metamod-p-1.19p32/metamod$ make OPT=opt
gcc -O2 -fomit-frame-pointer -ffast-math -march=i586 -mtune=pentium4 -fno-exceptions -fno-rtti -Wall -Wno-unknown-pragmas -DOPT_TYPE="\"optimized\"" -D__METAMOD_BUILD__ -D__INTERNALS_USE_REGPARAMS__ -fPIC -I. -I../metamod -I../hlsdk/engine -I../hlsdk/common -I../hlsdk/pm_shared -I../hlsdk/dlls -I../hlsdk -o opt.linux_i386/api_hook.o -c api_hook.cpp
api_hook.cpp:1: error: CPU you selected does not support x86-64 instruction set
api_hook.cpp:1: error: CPU you selected does not support x86-64 instruction set
make: *** Error 1
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Core was generated by `./hlds_i686 -num_edicts 3072 -game svencoop -pingboost 2 +ip 178.33.47.48 +port'.
Program terminated with signal 11, Segmentation fault.
#0 0xb7668873 in _IO_vfprintf_internal (s=0xbf83793c, format=0xb36eb138 "Player '%s' isn't in the transition volume %s, aborting\n",
ap=0xbf838a98 "\264\177\210\374\070\b\210\262p\215\203\277") at vfprintf.c:1614
1614 vfprintf.c: No such file or directory.
in vfprintf.c
(gdb) bt
#0 0xb7668873 in _IO_vfprintf_internal (s=0xbf83793c, format=0xb36eb138 "Player '%s' isn't in the transition volume %s, aborting\n",
ap=0xbf838a98 "\264\177\210\374\070\b\210\262p\215\203\277") at vfprintf.c:1614
#1 0xb7688eb0 in _IO_vsnprintf (string=0xbf837a74 "Player ' \277\002", maxlen=4096, format=0xb36eb138 "Player '%s' isn't in the transition volume %s, aborting\n",
args=0xbf838a98 "\264\177\210\374\070\b\210\262p\215\203\277") at vsnprintf.c:120
#2 0xb380f074 in safevoid_vsnprintf(char*, unsigned int, char const*, char*) () from ./svencoop/addons/metamod/dlls/metamod_i386.so
#3 0xbf837a74 in ?? ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
(gdb)
stack overflow?
ty,
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
ran gdb with a debug build of metamod-p33
(gdb) bt
#0 0xb7668873 in _IO_vfprintf_internal (s=0xbf83793c, format=0xb36eb138 "Player '%s' isn't in the transition volume %s, aborting\n",
ap=0xbf838a98 "\264\177\210\374\070\b\210\262p\215\203\277") at vfprintf.c:1614
#1 0xb7688eb0 in _IO_vsnprintf (string=0xbf837a74 "Player ' \277\002", maxlen=4096, format=0xb36eb138 "Player '%s' isn't in the transition volume %s, aborting\n",
args=0xbf838a98 "\264\177\210\374\070\b\210\262p\215\203\277") at vsnprintf.c:120
#2 0xb380f074 in MPluginList::ini_refresh (this=0xb3800690) at mlist.cpp:458
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
currently trying to compile an amxx debug build, not having much luck
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hello there.Y hope i will get the help i need here.
I have an HLDS hosted on unbuntu 10.10 server and it keeps crashing with these errors
I requested help to discover the problem on a lot of forums and thei sey it`s because of conflicts bettwen the ubnuntu glibc version and the glibc version wich metamod-p was compiled and i should recompile metamod again on mi server.
Mi server has
i downloaded the source of metamod-p from the site and tried to compile it with make command but i get this
Can you please tell me wath am i doing rong is how can i fix this.
PS why did you stop developing this great mod?
Beste regards and i hope you can help me.
Popa Andrei Florin
is there someone waching this forum?
Sorry, I read your message and then forgot.
I tested metamod with Ubuntu 10.10 server, loads/runs without problems.
Your problem seems to be in AMXModX, mysql plugin crashing. Maybe it has to be compiled for Ubuntu?
Try contact http://www.amxmodx.org/
Not that I have stopped developing, it's 'mature'. No bugs have been reported (that is bugs that are actually in metamod-p) and it's old for game. So nothing ground breaking changes happening there.
Hello,
Would it be possible Post a new Linux i586/i686 binary? With set the compiler flag to "intrinsic" SSE2
I have tried to compile the source code in Debian 6.0, with gcc (Debian 4.4.5-8) 4.4.5
many errors.
Thank You,
I cannot see any reason, why one would want to recompile metamod-p with SSE2.
Hey,
I understand it will be a little faster and Adam also says he "made a tiny memory fix"
http://forums.svencoop.com/showthread.php?p=484868#post484868
Thank You,
Since this Adam person has forked the metamod-p (not submitted his fix upstream), I recommend that you ask him.
I see he is posting binary, then by GPL license (of metamod-p), he's obligated to give you the source-code (metamod-p32 + the fix) he used to create that binary.
I went through compiler generated assembler code of metamod-p32 (originally compiled with gcc-3.3.6), and recompile of metamod-p using ubuntu gcc-4.6.3.
I just can't see what enabling SSE2 is supposed to archive. I think what is seen here is effect of compiling metamod-p with newer compiler, tuned for current processor architechtures. And before actual performance numbers are shown, 'SSE2+friends make things better' is just placebo effect.
CVS now contain updated version of metamod-p, that can be compiled with gcc-4.x. (however untested, might not work)
Hi,
Can you link me? I will download the source files and compile metamod-p in Debian 6.0
I will talk to Adam about the GPL.
Thank You,
http://www.student.oulu.fi/~jukivili/metamod-1.19p33.tar.bz2
still seeing errors with gcc-4.4.5
make: *** Error 1
make: Leaving directory `/home/edman747/metamod-1.19p33/trace_plugin'
make: *** Error 2
edman747@sc:~/metamod-1.19p33$
Go to "metamod-1.19p33/metamod/" directory. Then compile with "make OPT=opt".
OPT=opt is needed to build optimized binary instead of debug.
… after that new metamod should be at "opt.linux32/metamod_i386.so"
Ok,
I used make "OPT=opt"
still have errors:
engine_api.cpp:1050: error: invalid conversion from 'void (*)(void*, char*, …)' to 'void (*)(void*, const char*, …)'
make: *** Error 1
make: Leaving directory `/home/edman747/metamod-1.19p33/trace_plugin'
make: *** Error 2
edman747@sc:~/metamod-1.19p33$
You're still compiling in the root directory. Change directory to "~/metamod-1.19p33/metamod" then use "make OPT=opt". Other directories are example plugins for metamod and I didn't fix those for gcc-4.x (yet).
edman747@sc:~/metamod-p/metamod-p-1.19p32/metamod$ make OPT=opt
gcc -O2 -fomit-frame-pointer -ffast-math -march=i586 -mtune=pentium4 -fno-exceptions -fno-rtti -Wall -Wno-unknown-pragmas -DOPT_TYPE="\"optimized\"" -D__METAMOD_BUILD__ -D__INTERNALS_USE_REGPARAMS__ -fPIC -I. -I../metamod -I../hlsdk/engine -I../hlsdk/common -I../hlsdk/pm_shared -I../hlsdk/dlls -I../hlsdk -o opt.linux_i386/api_hook.o -c api_hook.cpp
api_hook.cpp:1: error: CPU you selected does not support x86-64 instruction set
api_hook.cpp:1: error: CPU you selected does not support x86-64 instruction set
make: *** Error 1
oops,
edman747@sc:~/metamod-1.19p33/metamod$ make OPT=opt
gcc -m32 -MM -I. -I../metamod -I../hlsdk/engine -I../hlsdk/common -I../hlsdk/pm_shared -I../hlsdk/dlls -I../hlsdk api_hook.cpp api_info.cpp commands_meta.cpp conf_meta.cpp dllapi.cpp engine_api.cpp engineinfo.cpp game_support.cpp game_autodetect.cpp h_export.cpp linkgame.cpp linkplug.cpp log_meta.cpp meta_eiface.cpp metamod.cpp mlist.cpp mplayer.cpp mplugin.cpp mqueue.cpp mreg.cpp mutil.cpp osdep.cpp osdep_p.cpp reg_support.cpp sdk_util.cpp studioapi.cpp support_meta.cpp thread_logparse.cpp vdate.cpp osdep_linkent_linux.cpp osdep_detect_gamedll_linux.cpp | sed "s;\(^*\):\(.*\);opt.linux_i386/\1: Makefile Config.mak \2;" > opt.linux_i386/Rules.depend
make: Nothing to be done for `default'.
Thank You,
Hmm, interesting. On Debian Squeeze, I can compile without problem using 'make OPT=opt'.
Does 'make OPT=opt linux' work for you?
I did make clean, then make OPT=opt
Should I do 'make OPT=opt linux'?
Best Wishes,
Yes, try "make OPT=opt linux".
That works, ty
24779c61f4ea4ad16307c5018a67b014 metamod_i386.so
Hello,
segfault and core file.
Core was generated by `./hlds_i686 -num_edicts 3072 -game svencoop -pingboost 2 +ip 178.33.47.48 +port'.
Program terminated with signal 11, Segmentation fault.
#0 0xb7668873 in _IO_vfprintf_internal (s=0xbf83793c, format=0xb36eb138 "Player '%s' isn't in the transition volume %s, aborting\n",
ap=0xbf838a98 "\264\177\210\374\070\b\210\262p\215\203\277") at vfprintf.c:1614
1614 vfprintf.c: No such file or directory.
in vfprintf.c
(gdb) bt
#0 0xb7668873 in _IO_vfprintf_internal (s=0xbf83793c, format=0xb36eb138 "Player '%s' isn't in the transition volume %s, aborting\n",
ap=0xbf838a98 "\264\177\210\374\070\b\210\262p\215\203\277") at vfprintf.c:1614
#1 0xb7688eb0 in _IO_vsnprintf (string=0xbf837a74 "Player ' \277\002", maxlen=4096, format=0xb36eb138 "Player '%s' isn't in the transition volume %s, aborting\n",
args=0xbf838a98 "\264\177\210\374\070\b\210\262p\215\203\277") at vsnprintf.c:120
#2 0xb380f074 in safevoid_vsnprintf(char*, unsigned int, char const*, char*) () from ./svencoop/addons/metamod/dlls/metamod_i386.so
#3 0xbf837a74 in ?? ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
(gdb)
stack overflow?
ty,
ran gdb with a debug build of metamod-p33
(gdb) bt
#0 0xb7668873 in _IO_vfprintf_internal (s=0xbf83793c, format=0xb36eb138 "Player '%s' isn't in the transition volume %s, aborting\n",
ap=0xbf838a98 "\264\177\210\374\070\b\210\262p\215\203\277") at vfprintf.c:1614
#1 0xb7688eb0 in _IO_vsnprintf (string=0xbf837a74 "Player ' \277\002", maxlen=4096, format=0xb36eb138 "Player '%s' isn't in the transition volume %s, aborting\n",
args=0xbf838a98 "\264\177\210\374\070\b\210\262p\215\203\277") at vsnprintf.c:120
#2 0xb380f074 in MPluginList::ini_refresh (this=0xb3800690) at mlist.cpp:458
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
currently trying to compile an amxx debug build, not having much luck