[Linux] Crashes got worse in B257

rkfg · September 2013

Is it just me or the game became really less stable? I can't complete a single game now, before it was 2-3 games in a row before the crash. First it crashed because of OOM, now it just segfaulted after less than 10 minutes of playing. Here's my dmesg part:

[623626.855274] ns2_linux32[2097]: segfault at 0 ip 00000000f6c57802 sp 00000000ffbc7e50 error 6 in libSpark_Core.so[f6aa0000+6c0000]
[625745.186117] ns2_linux32[3457]: segfault at 0 ip 00000000f17c8182 sp 00000000ffbbdc00 error 4 in libnvidia-glcore.so.319.49[f02cc000+1f5b000]
[626497.634770] ns2_linux32[6097]: segfault at 0 ip 00000000f17af1a6 sp 00000000ffdd9780 error 4 in libnvidia-glcore.so.319.49[f02b3000+1f5b000]

So it's something OpenGL-related it seems at least two times out of three. log.txt attached as well as stdout dump.

rkfg · September 2013

I also noticed that if you start the game with maxed up settings and then lower them midgame, the FPS becomes terrible (~20-25) with huge input lag and it crashes after that. Possibly it's because of memory fragmentation, the game loads high quality textures and doesn't unload it after settings change or unloads and leaves tiny free blocks which can't be reused. Anyway, the performance was horrible but then after the crash I entered another game and it was ok (the game was started with medium textures, streaming on, everything else off). Though it crashed very soon as I said, couldn't complete the round.

IronHorse · September 2013

Thanks, will pass this on!

Zaggy · September 2013

Same for me, where do I find stdout?
Log: http://pastebin.com/YCBcdudu

rkfg · September 2013

Just run Steam from the terminal and look there. It mostly duplicates the log.txt but also includes OOM/segfault messages so you could know why exactly your epic battle was interrupted and which library is to blame. Puts me at ease a bit.

Zaggy · September 2013

I only get:

[zaggynl@nexusi5 Natural Selection 2]$ stea[zaggynl@nexusi5 Natural Selection 2]$ steam steam://rungameid/4920
Running Steam on manjarolinux 0.8.7 64-bit
STEAM_RUNTIME is enabled automaticallym steam://rungameid/4920
[zaggynl@nexusi5 Natural Selection 2]$

rkfg · September 2013

Possibly you're running a wrapper instead of the real thing. I can't say for sure but Steam should be installed in ~/Steam or ~/.steam (I moved it to the SSD together with games). Search there for steam.sh and run it, should do the trick.

Gorguts · September 2013

257 is indeed less stable than the previous build 255. First off textures set to high has never worked. In 255 with textures set to high I could at least get into a game and play for a little bit before it crashed. Now the game crashes right when the map is done loading and sometimes even crashes in the main menu. In 255 with textures set to medium I could play for hours with no problem. Now with textures on medium Im lucky if the game goes for 5 minutes without crashing. Im glad you guys wanna support linux but if this is the quality youre gonna produce when working with linux, dont bother. As of now you guys have made the game un playable on linux...very sad as I had high hopes for this game.

OS = ElementaryOS Luna with kernel 3.11.1
Video = Nvidia gtx 660 (325.15 drivers)
Cpu = Intel i5 3570k
Ram = 8gigs

IronHorse · September 2013

@Gorguts its the first patch after a beta release (notice the words beta next to Opengl) for Linux support from a small indie studio.. I think you're being a bit premature with that judgement. Valve didn't even get it right in the first 4 weeks of release iirc

I'm too unfamiliar with the multitudes of Linux distros to give instructions on how, but giving debugging information, dumps and logs would be far more helpful in pointing the devs to address specific issues.

Rkfg has been very helpful with this already, ty

trinity.ns · September 2013

@Gorguts: Be patient, it'll get there. They are hard working guys & girls, but fixes like this take a while.

Gorguts · September 2013

Im being harsh on them because its obvious they did 0 testing of the new build on linux because if they did it wouldve never been released in the state that its in. I'll repeat what I said before in simpler terms...they took a game that was playable, and made it unplayable. What I have a problem with is them thinking that thats an okay thing to do...If they need testers Im volunteering right now. Theres no excuse for not testing it. Id be more than happy to help out with that.

rkfg · September 2013

I'm glad to help with reports! They could be more useful with debug symbols and full stack traces but this idea hasn't excited anyone, unfortunately.

ominide · September 2013

Confirming @rkfg symptoms
I've played a couple games for a bit and training seems to load. biodome like clockwork fails though.
Would a core dump be helpful?

Confused · September 2013

Guys,

If you want to join the pt team there is an application thread in the general forum. Please apply. I am one of the few PT folks with a linux box to test with. So, any Linux bugs are pretty much directly my fault. Good times, right? More linux people means that more of you can be accused of not testing the game by random guys on the internet. That means less blame for me. Less blame for me means I can stop replying to every linux post in tech support so I can try to reproduce them all. That means I can start billing those hours to real paying clients again. Everyone wins. And I get paid!

Eh? Eh?

@rkfg if we ship the windows ones I will see if we can start shipping linux debug symbols as well.

@omninide,

Generally, I would say that more info is better. The out of memory issue is cross platform but my guess is an extra dump wouldn't kill anyone.

rkfg · September 2013

I'd be glad to help with playtesting but don't feel qualified enough. Also the time isn't well suited for me, it will be midnight here on Monday and Wednesday and I have a job to wake up at 8:00. Also don't want to kill my interest with obligatory playing. Sorry. Waiting for symbols and a manual on how to get traces with gdb.

tuxator · September 2013

Confused wrote: »

Guys,

If you want to join the pt team there is an application thread in the general forum. Please apply. I am one of the few PT folks with a linux box to test with. So, any Linux bugs are pretty much directly my fault.

I did that shortly after 255 appeared...never got a response.

Confused · September 2013

Good news,

I talked to Max he says they are currently shipping debug symbols on linux. So, you should already have them. With regard to what to do with the stack traces, you can send them directly to max@unknownworlds.com or tech-support@unknownworlds.com which goes to all of the programmers.

gdb basics

You are not going to get ns2 to run from the command line easily so you will likely need to do something like:

gdb attach $(pid of ns2_linux32) -ex cont

once ns2 is running. pid from ps -e|grep ns2_linux32

@tuxator, They generally add new PTs in groups. I will talk to @ironhorse and @obraxis about the whole thing see if we can try to get some linux pts in.

rkfg · September 2013

Wow, really good news! NS2 won't start from terminal because it needs lots of libs from Steam itself and I don't want to mess with LD_LIBRARY_PATH. It's safer to attach. Thank you, I'll start immediately.

Max · September 2013

You can run NS2 outside the Steam environment by using the Steam runtime. This is what we do for development. This readme describes more https://github.com/ValveSoftware/steam-runtime/blob/master/sdk/README.txt

rkfg · September 2013

Today was a good session. Played 3 or 4 rounds without a crash with GDB attached. I set windowed mode instead of Fullscreen Windowed and it seems the game now can survive Alt+Tab and also became more stable. Anyway, it crashed because of OOM and I've collected this backtrace:

Program received signal SIGABRT, Aborted.
0xf777a430 in __kernel_vsyscall ()
(gdb) backtrace
#0 0xf777a430 in __kernel_vsyscall ()
#1 0xf369380f in raise () from /lib/i386-linux-gnu/i686/cmov/libc.so.6
#2 0xf3696cc3 in abort () from /lib/i386-linux-gnu/i686/cmov/libc.so.6
#3 0xf4324d0d in base::debug::BreakDebugger() () from /opt/SteamLinux/SteamApps/common/Natural Selection 2/libawesomium-1-7.so.0
#4 0xf433f6ff in logging::LogMessage::~LogMessage() () from /opt/SteamLinux/SteamApps/common/Natural Selection 2/libawesomium-1-7.so.0
#5 0xf43542af in T.471 () from /opt/SteamLinux/SteamApps/common/Natural Selection 2/libawesomium-1-7.so.0
#6 0xf5a5e5da in tc_malloc () from /opt/SteamLinux/SteamApps/common/Natural Selection 2/libawesomium-1-7.so.0
#7 0xf3c0f609 in ?? () from /usr/lib32/libGL.so.1
#8 0xf15ceb3c in ?? () from /usr/lib32/libnvidia-glcore.so.319.49
#9 0xf1695ea9 in ?? () from /usr/lib32/libnvidia-glcore.so.319.49
#10 0xf1696c2c in ?? () from /usr/lib32/libnvidia-glcore.so.319.49
#11 0xf139bdcd in ?? () from /usr/lib32/libnvidia-glcore.so.319.49
#12 0xf13a1c06 in ?? () from /usr/lib32/libnvidia-glcore.so.319.49
#13 0xf13a22bd in ?? () from /usr/lib32/libnvidia-glcore.so.319.49
#14 0xf7272de2 in M4::OpenGLTexture::CreateInternal (this=0xc8d24be0, flags=3) at ../Source/Render/OpenGLTexture.cpp:535
#15 0xf7273038 in M4::OpenGLTexture::Create (this=0xc8d24be0, updateContext=0xe7b8ee08, format=M4::RenderFormat_R16G16B16A16F, xSize=1920, ySize=1200, numMipMaps=<optimized out>,
type=M4::RenderTextureType_2D, flags=3) at ../Source/Render/OpenGLTexture.cpp:258
#16 0xf726f9c2 in M4::OpenGLDevice::CreateTexture (this=0xe7b8ebe0, format=M4::RenderFormat_R16G16B16A16F, xSize=1920, ySize=1200, numMipMaps=1, type=M4::RenderTextureType_2D, flags=3)
at ../Source/Render/OpenGLDevice.cpp:779
#17 0xf72d071e in M4::RenderTargetManager::AcquireRenderTarget (this=0xdedacfc0, xSize=1920, ySize=1200, format=M4::RenderFormat_R16G16B16A16F) at ../Source/Render/RenderTargetManager.cpp:51
#18 0xf7299c0a in M4::RenderPipeline::CreateResourceRenderTargets (this=0xae2cadc8, resourceIndex=0) at ../Source/Render/RenderPipeline.cpp:1511
#19 0xf7299cb5 in M4::RenderPipeline::BeginLayerRendering (this=0xae2cadc8, layerIndex=0) at ../Source/Render/RenderPipeline.cpp:1459
#20 0xf72aafdf in M4::RenderRenderer::Render (this=0xdf3cb640, pipeline=0xae2cadc8, colorTarget=0xccc4ba70, depthTarget=0x0, faceIndex=0, sourceTexture=0xccc4bc00, sceneSettings=0xd8703382,
setupSettings=0xb2016c38, scene=0x0, time=1076.1129150390625, camera=..., viewport=..., visibleObjects=..., background=0x0) at ../Source/Render/RenderRenderer.cpp:1240
#21 0xf72ab901 in M4::RenderRenderer::RenderScreenEffects (this=0xdf3cb640, context=..., sceneSettings=0xd8703382, scene=0xa9b82df0, camera=..., viewport=..., visibleObjects=...,
background=0x0) at ../Source/Render/RenderRenderer.cpp:3797
#22 0xf72ab239 in M4::RenderRenderer::Render (this=0xdf3cb640, pipeline=0xad5a3478, colorTarget=0xdd279e60, depthTarget=0xdbaece50, faceIndex=0, sourceTexture=0x0, sceneSettings=0xd8703382,
setupSettings=0xffb86bd8, scene=0xa9b82df0, time=1076.1129150390625, camera=..., viewport=..., visibleObjects=..., background=0x0) at ../Source/Render/RenderRenderer.cpp:1266
#23 0xf72bf55c in M4::RenderScene::InternalRender (this=0xa9b82df0, settings=..., colorTarget=0xdd279e60, depthTarget=0xdbaece50, faceIndex=0, pipeline=0xad5a3478, camera=...,
prevCamera=0x0, viewport=..., renderMask=4294967295, cullingMode=M4::RenderScene_CullingMode_Occlusion, background=0x0, deltaTime=0.0135498047) at ../Source/Render/RenderScene.cpp:816
#24 0xf72c0131 in M4::RenderScene::InternalRender (this=0xa9b82df0, settings=..., colorTarget=0xdd279e60, depthTarget=0xdbaece50, faceIndex=0, camera=0xb1a98de8, _viewport=0x0,
background=0x0, deltaTime=0.0135498047) at ../Source/Render/RenderScene.cpp:758
#25 0xf72c050c in M4::RenderScene::RenderMainCamera (this=0xa9b82df0, settings=..., colorTarget=0xdd279e60, depthTarget=0xdbaece50, faceIndex=0, camera=0xb1a98de8, viewport=0x0,
background=0x0, time=1076.1129150390625) at ../Source/Render/RenderScene.cpp:713
#26 0xf72c061e in M4::RenderScene::Render (this=0xa9b82df0, settings=..., colorTarget=0xdd279e60, depthTarget=0xdbaece50, camera=0xb1a98de8, background=0x0, time=1076.1129150390625)
at ../Source/Render/RenderScene.cpp:496
#27 0x08082344 in M4::ClientWorld::Render (this=0xd8701988, drawer=..., swapChain=0xe7b3ae88) at ../Source/Spark_Client/ClientWorld.cpp:2264
#28 0x0806caed in M4::ClientGame::Render (this=0xe7b38960, swapChain=0xe7b3ae88) at ../Source/Spark_Client/ClientGame.cpp:1640
#29 0x0807699f in M4::ClientGame::Update (this=0xe7b38960, deltaTime=0.027430015000390995) at ../Source/Spark_Client/ClientGame.cpp:891
#30 0x0806bfa6 in M4::ClientEngine::Update (this=0xf3ca2170, deltaTime=0.027430015000390995) at ../Source/Spark_Client/ClientEngine.cpp:148
#31 0x080bf044 in M4::BaseEngine::RunFrame (this=0xf3ca2170) at ../Source/Spark_Server/BaseEngine.cpp:188
#32 0x080bf070 in M4::BaseEngine::Run (this=0xf3ca2170) at ../Source/Spark_Server/BaseEngine.cpp:152
#33 0x0806a4c3 in main (argc=<error reading variable: Cannot access memory at address 0x104a>, argv=<error reading variable: Cannot access memory at address 0x104e>)
at ../Source/ShooterExe/Main.cpp:82

Textures and Awesomium... quite interesting. I hope it's not Awesomium who is the culprit though Chrome likes memory a lot. I have tabs consuming 1 Gb of RAM in it.
@Max oh, nice tools, thanks! May be of use for us definitely.

rkfg · September 2013

Aaaaaand here's a segfault. Happened just midgame while shooting, not related to out of memory crashes. This time it's inside NVIDIA's GL lib.

#0 0xf17ab1a6 in ?? () from /usr/lib32/libnvidia-glcore.so.319.49
#1 0xf17aff3d in ?? () from /usr/lib32/libnvidia-glcore.so.319.49
#2 0xf138b15d in ?? () from /usr/lib32/libnvidia-glcore.so.319.49
#3 0xf7229cc8 in M4::OpenGLDevice::DrawIndexedPrimitives (this=0xe7b49be0,
primitiveType=M4::RenderPrimitiveType_TriangleList, minVertexIndex=0, numVertices=65,
startIndex=0, numPrimitives=48, numInstances=1) at ../Source/Render/OpenGLDevice.cpp:998
#4 0xf72510e8 in M4::RenderModelProxy::RenderFaceSet (this=0x91da6e10, faceSetIndex=0,
numInstances=1) at ../Source/Render/RenderModelProxy.cpp:239
#5 0xf724dc37 in M4::RenderModelArray::RenderFaceSet (this=0x9c864f40, faceSetIndex=0)
at ../Source/Render/RenderModelArray.cpp:423
#6 0xf72678d1 in RenderFaceSet (this=<optimized out>, faceSet=...)
at ../Source/Render/RenderRenderer.cpp:324
#7 M4::RenderRenderer::RenderFaceSets<M4::RenderModelArray, M4::ObjectRenderer<M4::RenderModelArray> > (this=0xdd11f2c0, context=..., viewerParams=..., renderer=..., faceSets=0xdd39d010,
numFaceSets=87) at ../Source/Render/RenderRenderer.cpp:3415
#8 0xf726465f in M4::RenderRenderer::RenderModelArrays (this=0xdd11f2c0, context=...,
viewerParams=..., modelArray=0xff827df0, numModelArrays=87, techniqueIndex=1,
reverseCulling=false, shadows=false) at ../Source/Render/RenderRenderer.cpp:2617
#9 0xf72658ef in M4::RenderRenderer::RenderObjects (this=0xdd11f2c0, context=...,
viewerParams=..., objects=..., techniqueIndex=1, objectsMask=4294967039)
at ../Source/Render/RenderRenderer.cpp:2166
#10 0xf7265c68 in M4::RenderRenderer::RenderObjects (this=0xdd11f2c0, scene=0xdbf49c98,
context=..., objects=0xff827630, camera=..., viewport=..., techniqueIndex=1,
objectsMask=4294967039, zonesMask=4294967295) at ../Source/Render/RenderRenderer.cpp:2989
#11 0xf72661be in M4::RenderRenderer::Render (this=0xdd11f2c0, pipeline=0xdb4ef848,
colorTarget=0xe7af5eb8, depthTarget=0xe7af5f08, faceIndex=0, sourceTexture=0x0,
sceneSettings=0xdc0d9fc2, setupSettings=0xff8338c8, scene=0xdbf49c98,
time=2606.00537109375, camera=..., viewport=..., visibleObjects=..., background=0x0)
at ../Source/Render/RenderRenderer.cpp:1312
#12 0xf727a55c in M4::RenderScene::InternalRender (this=0xdbf49c98, settings=...,
colorTarget=0xe7af5eb8, depthTarget=0xe7af5f08, faceIndex=0, pipeline=0xdb4ef848,
camera=..., prevCamera=0x0, viewport=..., renderMask=4294967295,
cullingMode=M4::RenderScene_CullingMode_Occlusion, background=0x0, deltaTime=0.0288085938)
at ../Source/Render/RenderScene.cpp:816
#13 0xf727b131 in M4::RenderScene::InternalRender (this=0xdbf49c98, settings=...,
colorTarget=0xe7af5eb8, depthTarget=0xe7af5f08, faceIndex=0, camera=0xa8d18e50,
_viewport=0x0, background=0x0, deltaTime=0.0288085938)
at ../Source/Render/RenderScene.cpp:758
#14 0xf727b50c in M4::RenderScene::RenderMainCamera (this=0xdbf49c98, settings=...,
colorTarget=0xe7af5eb8, depthTarget=0xe7af5f08, faceIndex=0, camera=0xa8d18e50,
viewport=0x0, background=0x0, time=2606.00537109375)
at ../Source/Render/RenderScene.cpp:713
#15 0xf727b61e in M4::RenderScene::Render (this=0xdbf49c98, settings=...,
colorTarget=0xe7af5eb8, depthTarget=0xe7af5f08, camera=0xa8d18e50, background=0x0,
time=2606.00537109375) at ../Source/Render/RenderScene.cpp:496
#16 0x08082344 in M4::ClientWorld::Render (this=0xdc0d85c8, drawer=..., swapChain=0xe7af5e88)
at ../Source/Spark_Client/ClientWorld.cpp:2264
#17 0x0806caed in M4::ClientGame::Render (this=0xe7af3960, swapChain=0xe7af5e88)
at ../Source/Spark_Client/ClientGame.cpp:1640
#18 0x0807699f in M4::ClientGame::Update (this=0xe7af3960, deltaTime=0.02818026299974008)
at ../Source/Spark_Client/ClientGame.cpp:891
#19 0x0806bfa6 in M4::ClientEngine::Update (this=0xf3c5d170, deltaTime=0.02818026299974008)
at ../Source/Spark_Client/ClientEngine.cpp:148
#20 0x080bf044 in M4::BaseEngine::RunFrame (this=0xf3c5d170)
at ../Source/Spark_Server/BaseEngine.cpp:188
#21 0x080bf070 in M4::BaseEngine::Run (this=0xf3c5d170)
at ../Source/Spark_Server/BaseEngine.cpp:152
#22 0x0806a4c3 in main (argc=<error reading variable: Cannot access memory at address 0x3>,
argv=<error reading variable: Cannot access memory at address 0x7>)
at ../Source/ShooterExe/Main.cpp:82

This appeared in stdout right after detaching GDB:

Segmentation fault
Game removed: AppID 4920 "Natural Selection 2", ProcID 12822 
Inconsistency detected by ld.so: dl-close.c: 765: _dl_close: Assertion `map->l_init_called' failed!

Skvate · September 2013

I got one on mineshaft, after going through a phasegate. It seems to happen often when loading a lot of stuff on screen at the same time.

ns2_linux32[5959]: segfault at 0 ip 00000000f1a9c4a6 sp 00000000ffba2980 error 4 in libnvidia-glcore.so.313.30[f05f8000+1cd3000]

rkfg · September 2013

I can only guess of course but it's unlikely a bug in NVIDIA's driver. Probably (looking at my backtrace) there's a null pointer in the triangle list that's being passed to the GL library. Or just an invalid pointer which points to the already free memory. Now let's guess what it is with 65 vertices and 48 primitives (triangles?), then patch the code in a hex editor and send the fix to the devs.

Skvate · September 2013

I havent gotten around to use the gdb thing yet, but will try later if its still needed.

rkfg · September 2013

You can do that just out of curiosity, maybe there will be a pattern in backtraces.

Factoid · September 2013

I've had a few crashes. Ubuntu 13.04 running the primus/bumblebee driver PPA. MSi GE-60 gaming laptop. Currently I log std out/err and bundle them up with log.txt. I've been trying to get core dumps to show up but no luck. I'll see about getting gdb attached to the process. Has anyone had any luck getting core dumps? I've tried ulimit -c unlimited but they're perhaps not showing up where I expect.

I'll put my crash logs up on dropbox.
https://www.dropbox.com/sh/zbgwti60z6i6miu/uH3KyGUZQo/ns2_crash

257 does seem less stable than 255

Also, I have a weird mic bug. Every time I start the game I have to go to the root options screen before my mic will respond. Audio otherwise works properly on startup.

OscarMayer · September 2013

I tried attaching gdb as well and here's a backtrace:

Program received signal SIGSEGV, Segmentation fault.
0xf7248c64 in M4::RenderObject::GetType (this=0x0) at ../Source/Render/RenderObject.cpp:33
33 ../Source/Render/RenderObject.cpp: No such file or directory.
(gdb)
(gdb) backtrace
#0 0xf7248c64 in M4::RenderObject::GetType (this=0x0) at ../Source/Render/RenderObject.cpp:33
#1 0xf7251d10 in M4::RenderRenderer::CollectObjects (this=0x8a6cdd40, object=0xffd83748, numObjects=1, objects=...) at ../Source/Render/RenderRenderer.cpp:2087
#2 0xf725ce55 in M4::RenderRenderer::Render (this=0x8a6cdd40, pipeline=0xd8840b18, colorTarget=0xe79e7470, depthTarget=0xe79e74c0, faceIndex=0, sourceTexture=0x0,
sceneSettings=0xbdc920e2, setupSettings=0xffd83748, scene=0xa570c6d0, time=524.3658447265625, camera=..., viewport=..., visibleObjects=..., background=0x0)
at ../Source/Render/RenderRenderer.cpp:1216
#3 0xf727155c in M4::RenderScene::InternalRender (this=0xa570c6d0, settings=..., colorTarget=0xe79e7470, depthTarget=0xe79e74c0, faceIndex=0, pipeline=0xd8840b18, camera=...,
prevCamera=0x0, viewport=..., renderMask=4294967295, cullingMode=M4::RenderScene_CullingMode_Occlusion, background=0x0, deltaTime=0.0259399414)
at ../Source/Render/RenderScene.cpp:816
#4 0xf7272131 in M4::RenderScene::InternalRender (this=0xa570c6d0, settings=..., colorTarget=0xe79e7470, depthTarget=0xe79e74c0, faceIndex=0, camera=0xbcb8b668, _viewport=0x0,
background=0x0, deltaTime=0.0259399414) at ../Source/Render/RenderScene.cpp:758
#5 0xf727250c in M4::RenderScene::RenderMainCamera (this=0xa570c6d0, settings=..., colorTarget=0xe79e7470, depthTarget=0xe79e74c0, faceIndex=0, camera=0xbcb8b668, viewport=0x0,
background=0x0, time=524.3658447265625) at ../Source/Render/RenderScene.cpp:713
#6 0xf727261e in M4::RenderScene::Render (this=0xa570c6d0, settings=..., colorTarget=0xe79e7470, depthTarget=0xe79e74c0, camera=0xbcb8b668, background=0x0,
time=524.3658447265625) at ../Source/Render/RenderScene.cpp:496
#7 0x08082344 in M4::ClientWorld::Render (this=0xbdc906e8, drawer=..., swapChain=0xe79e7440) at ../Source/Spark_Client/ClientWorld.cpp:2264
#8 0x0806caed in M4::ClientGame::Render (this=0xe79ef960, swapChain=0xe79e7440) at ../Source/Spark_Client/ClientGame.cpp:1640
#9 0x0807699f in M4::ClientGame::Update (this=0xe79ef960, deltaTime=0.014721259999532776) at ../Source/Spark_Client/ClientGame.cpp:891
#10 0x0806bfa6 in M4::ClientEngine::Update (this=0xf3c54170, deltaTime=0.014721259999532776) at ../Source/Spark_Client/ClientEngine.cpp:148
#11 0x080bf044 in M4::BaseEngine::RunFrame (this=0xf3c54170) at ../Source/Spark_Server/BaseEngine.cpp:188
#12 0x080bf070 in M4::BaseEngine::Run (this=0xf3c54170) at ../Source/Spark_Server/BaseEngine.cpp:152
#13 0x0806a4c3 in main (argc=<error reading variable: Cannot access memory at address 0x0>, argv=<error reading variable: Cannot access memory at address 0x4>)
at ../Source/ShooterExe/Main.cpp:82

And here's a dmesg:
[427040.365661] ns2_linux32[5505]: segfault at 0 ip 00000000f1738d37 sp 00000000ff97a660 error 4 in libnvidia-glcore.so.325.15[f021e000+2070000]

Is posting these here helpful for the NS2 devs?

Also, I'm having the same kind of mic issues.

Smeat · September 2013

I've also had a lot of crashes while playing. Most of them are segfaults in libnvidia-glcore.so.

Sep 25 00:43:46 gentoo kernel: [78431.125222] ns2_linux32[6308]: segfault at 0 ip 00000000f176e816 sp 00000000ffab8e00 error 4 in libnvidia-glcore.so.319.49[f02fe000+1f5b000]

But Today this error occurred while loading ns2_veil. I think it crashed during "Loading Map", but I am not sure.

(gdb) bt
#0  0xf397143d in pthread_cond_signal () from /lib32/libpthread.so.0
#1  0xf6ce8690 in M4::Event::SetSignaled (this=0xffcb1578)
    at ../Source/Engine/linux/Event.cpp:49
#2  0xf6cebbde in M4::Job::SignalComplete (this=0xffcb1574)
    at ../Source/Engine/linux/Job.cpp:46
#3  0xf6cec154 in M4::WorkerThread::Run (this=0x87eca00)
    at ../Source/Engine/linux/Job.cpp:94
#4  0xf6cecf4f in M4::RunThread (data=0x87eca00)
    at ../Source/Engine/linux/Thread.cpp:32
#5  0xf396cdef in start_thread () from /lib32/libpthread.so.0
#6  0xf37719de in clone () from /lib32/libc.so.6
(gdb) p $_siginfo
$3 = {si_signo = 11, si_errno = 0, si_code = 1, _sifields = {_pad = {12, 
      0 <repeats 28 times>}, _kill = {si_pid = 12, si_uid = 0}, _timer = {
      si_tid = 12, si_overrun = 0, si_sigval = {sival_int = 0, 
        sival_ptr = 0x0}}, _rt = {si_pid = 12, si_uid = 0, si_sigval = {
        sival_int = 0, sival_ptr = 0x0}}, _sigchld = {si_pid = 12, si_uid = 0, 
      si_status = 0, si_utime = 0, si_stime = 0}, _sigfault = {si_addr = 0xc}, 
    _sigpoll = {si_band = 12, si_fd = 0}}}

Gentoo 64Bit
Nvidia 319.49

synergysymphony · October 2013

im on ubuntu 13.04 64X, amd 13.4 on a hd5800, i5. first: textures dont show if particle quality is on low. second: fps 120+ everthing on low except PQ in an empty game, fps seems to drop in proportion to players in the game.

Factoid · October 2013

Figured out how to get dumps working properly under ubuntu. They were being rerouted to an apport application by way of the /proc/sys/kernel/core_pattern setup. Just working out some of the details on what exactly does an doesn't work, and I'll be able to write a script/guide to make it easier to handle crashes from ns2 on linux.

Problem I had with GDB was if I'm not in windowed mode, then the crash will lockup my desktop. I suppose I could attach it via one of the VTs...

synergysymphony · October 2013

turning catalyst "A.I." off made a huge difference...

[Linux] Crashes got worse in B257

Comments