Segmentation fault (core dumped) -- (with gdb backtrace)

Segmentation fault (core dumped) -- (with gdb backtrace)

Postby nana03435 » 22 Jan 2013, 23:22

Run it regularly, there is only one report without any other information:
Segmentation fault (core dumped)

So gdb backtrace below:
{l Code}: {l Select All Code}
(gdb) backtrace
#0  0x00007ffff2c54eb3 in ir_array_splitting_visitor::split_deref(ir_dereference**) () from /usr/lib64/dri/r600_dri.so
#1  0x00007ffff2c54fb9 in ir_array_splitting_visitor::handle_rvalue(ir_rvalue**) () from /usr/lib64/dri/r600_dri.so
#2  0x00007ffff2d5f23f in ir_rvalue_base_visitor::rvalue_visit(ir_dereference_array*) () from /usr/lib64/dri/r600_dri.so
#3  0x00007ffff2d5a69a in ir_expression::accept(ir_hierarchical_visitor*) ()
   from /usr/lib64/dri/r600_dri.so
#4  0x00007ffff2d5a69a in ir_expression::accept(ir_hierarchical_visitor*) ()
   from /usr/lib64/dri/r600_dri.so
#5  0x00007ffff2d5a69a in ir_expression::accept(ir_hierarchical_visitor*) ()
   from /usr/lib64/dri/r600_dri.so
#6  0x00007ffff2d5aa66 in ir_assignment::accept(ir_hierarchical_visitor*) ()
   from /usr/lib64/dri/r600_dri.so
#7  0x00007ffff2d5a37d in visit_list_elements(ir_hierarchical_visitor*, exec_list*, bool) () from /usr/lib64/dri/r600_dri.so
#8  0x00007ffff2d5a557 in ir_function_signature::accept(ir_hierarchical_visitor*) () from /usr/lib64/dri/r600_dri.so
#9  0x00007ffff2d5a37d in visit_list_elements(ir_hierarchical_visitor*, exec_list*, bool) () from /usr/lib64/dri/r600_dri.so
#10 0x00007ffff2d5a5ee in ir_function::accept(ir_hierarchical_visitor*) ()
   from /usr/lib64/dri/r600_dri.so
#11 0x00007ffff2d5a37d in visit_list_elements(ir_hierarchical_visitor*, exec_lis---Type <return> to continue, or q <return> to quit---
t*, bool) () from /usr/lib64/dri/r600_dri.so
#12 0x00007ffff2c55237 in optimize_split_arrays(exec_list*, bool) ()
   from /usr/lib64/dri/r600_dri.so
#13 0x00007ffff2c4e0fd in do_common_optimization(exec_list*, bool, bool, unsigned int) () from /usr/lib64/dri/r600_dri.so
#14 0x00007ffff2d245f2 in _mesa_glsl_compile_shader ()
   from /usr/lib64/dri/r600_dri.so
#15 0x00007ffff2cbbe88 in compile_shader () from /usr/lib64/dri/r600_dri.so
#16 0x00007ffff3783c35 in Ogre::GLSLProgram::compile(bool) ()
   from /usr/lib64/OGRE/RenderSystem_GL.so
#17 0x00007ffff378426d in Ogre::GLSLProgram::loadFromSource() ()
   from /usr/lib64/OGRE/RenderSystem_GL.so
#18 0x000000373f9fb392 in Ogre::HighLevelGpuProgram::loadHighLevelImpl() ()
   from /lib64/libOgreMain.so.1.7.4
#19 0x000000373f9fb9d1 in Ogre::HighLevelGpuProgram::loadHighLevel() ()
   from /lib64/libOgreMain.so.1.7.4
#20 0x000000373f9fb884 in Ogre::HighLevelGpuProgram::createParameters() ()
   from /lib64/libOgreMain.so.1.7.4
#21 0x000000373f9dc17c in Ogre::GpuProgram::getDefaultParameters() ()
   from /lib64/libOgreMain.so.1.7.4
#22 0x000000373fb5f5b8 in Ogre::GpuProgramTranslator::translateHighLevelGpuProgram(Ogre::ScriptCompiler*, Ogre::ObjectAbstractNode*) ()
   from /lib64/libOgreMain.so.1.7.4
---Type <return> to continue, or q <return> to quit---
#23 0x000000373fb612d5 in Ogre::GpuProgramTranslator::translate(Ogre::ScriptCompiler*, Ogre::SharedPtr<Ogre::AbstractNode> const&) ()
   from /lib64/libOgreMain.so.1.7.4
#24 0x000000373fb3c703 in Ogre::ScriptCompiler::compile(Ogre::SharedPtr<std::list<Ogre::SharedPtr<Ogre::ConcreteNode>, Ogre::STLAllocator<Ogre::SharedPtr<Ogre::ConcreteNode>, Ogre::CategorisedAllocPolicy<(Ogre::MemoryCategory)0> > > > const&, std::string const&) () from /lib64/libOgreMain.so.1.7.4
#25 0x000000373fb3c86a in Ogre::ScriptCompiler::compile(std::string const&, std::string const&, std::string const&) () from /lib64/libOgreMain.so.1.7.4
#26 0x000000373fb3c9a6 in Ogre::ScriptCompilerManager::parseScript(Ogre::SharedPtr<Ogre::DataStream>&, std::string const&) () from /lib64/libOgreMain.so.1.7.4
#27 0x000000373faeca92 in Ogre::ResourceGroupManager::parseResourceGroupScripts(Ogre::ResourceGroupManager::ResourceGroup*) () from /lib64/libOgreMain.so.1.7.4
#28 0x000000373faed27f in Ogre::ResourceGroupManager::initialiseAllResourceGroups() () from /lib64/libOgreMain.so.1.7.4
#29 0x00000000004c0c53 in liren_internal_init (self=0x15a15c0,
    mode=0x7fffffffdb40) at ../src/lipsofsuna/render/internal/render.cpp:240
#30 0x0000000000489874 in liren_render_new (paths=0x7c00b0,
    mode=0x7fffffffdb40) at ../src/lipsofsuna/render/render.c:85
#31 0x00000000004febfc in liext_graphics_new (program=0x7bfe80)
    at ../src/lipsofsuna/extension/graphics/ext-module.c:74
#32 0x000000000043ab5a in limai_program_insert_extension (self=0x7bfe80,
    name=0x159f0b8 "graphics") at ../src/lipsofsuna/main/main-program.c:483
---Type <return> to continue, or q <return> to quit---
#33 0x00000000004d0fc8 in Program_load_extension (args=0x7fffffffdc30)
    at ../src/lipsofsuna/script/script-program.c:78
#34 0x00000000004cbfbb in liscr_marshal_CLASS (lua=0x7c98c0)
    at ../src/lipsofsuna/script/script-args.c:1129
#35 0x0000003ac300c434 in luaD_precall () from /lib64/liblua-5.1.so
#36 0x0000003ac3016d38 in luaV_execute () from /lib64/liblua-5.1.so
#37 0x0000003ac300c85d in luaD_call () from /lib64/liblua-5.1.so
#38 0x0000003ac3008446 in lua_call () from /lib64/liblua-5.1.so
#39 0x0000003ac3020d4c in ll_require () from /lib64/liblua-5.1.so
#40 0x0000003ac300c434 in luaD_precall () from /lib64/liblua-5.1.so
#41 0x0000003ac3016d38 in luaV_execute () from /lib64/liblua-5.1.so
#42 0x0000003ac300c85d in luaD_call () from /lib64/liblua-5.1.so
#43 0x0000003ac300bada in luaD_rawrunprotected () from /lib64/liblua-5.1.so
#44 0x0000003ac300ca0f in luaD_pcall () from /lib64/liblua-5.1.so
#45 0x0000003ac30084e1 in lua_pcall () from /lib64/liblua-5.1.so
#46 0x00000000004d7b48 in private_exec_script (self=0x7c9850)
    at ../src/lipsofsuna/script/script.c:311
#47 0x00000000004d7914 in liscr_script_load_file (self=0x7c9850,
    path=0x7c4c40 "/home/bolan/lipsofsuna-0.6.0/data/lipsofsuna/main.lua",
    path_mod=0x7c0260 "/home/bolan/lipsofsuna-0.6.0/data/lipsofsuna",
    path_core=0x7c0230 "/home/bolan/lipsofsuna-0.6.0/data")
    at ../src/lipsofsuna/script/script.c:208
#48 0x000000000043a6b3 in limai_program_execute_script (self=0x7bfe80,
---Type <return> to continue, or q <return> to quit---
    file=0x54cdae "main.lua") at ../src/lipsofsuna/main/main-program.c:309
#49 0x000000000043c37c in main (argc=1, argv=0x7fffffffe278)
    at ../src/lipsofsuna/main/main.c:106
(gdb)
nana03435
 
Posts: 9
Joined: 09 Jun 2012, 22:47

Re: Segmentation fault (core dumped) -- (with gdb backtrace)

Postby amuzen » 23 Jan 2013, 00:15

Looks like the crash occurs inside the shader compiler of your graphics driver. Since the shader compiler should never crash with any input, this is likely a bug in the driver, though it is hard to say for sure. If you have an old graphics card, that might also be a factor. Nevertheless, I doubt we can do anything about it since it's highly unlikely that the problem would be in LoS itself.
User avatar
amuzen
LoS Moderator
 
Posts: 327
Joined: 05 Dec 2009, 02:49

Re: Segmentation fault (core dumped) -- (with gdb backtrace)

Postby KroArtem » 23 Jan 2013, 20:21

I had something similar long time ago (~ when LoS switched to ogre) and reported it to amuzen. However it really seems to be a driver bug so nothing was done with it.

P.S. I had this crash with mesa from git and ATi Radeon HD3850
KroArtem
 
Posts: 375
Joined: 26 Aug 2010, 19:04

Re: Segmentation fault (core dumped) -- (with gdb backtrace)

Postby nana03435 » 24 Jan 2013, 19:30

Hi, amuzen and KroArtem,

After testing the mesa driver and AMD official driver, I guess it is a bug of open source driver since the game can work on official driver. However, at this time, AMD stop support graphic cards which older than HD4XXX. But I don't know whether this gdb backtrace information is good for report it to the upstream of mesa or xorg-radeon.
nana03435
 
Posts: 9
Joined: 09 Jun 2012, 22:47

Re: Segmentation fault (core dumped) -- (with gdb backtrace)

Postby amuzen » 24 Jan 2013, 20:22

It would probably be quite hard to locate and fix it with the backtrace alone since there aren't even line numbers in it. Reproducing the issue would also be a pain for them since LoS is far from a nice, minimal test case that developers generally prefer. Furthermore, the issue is likely to be at least somewhat GPU specific since the game works fine with the open source driver on the HD 5xxx series of cards. Well, nevertheless, if they'll pay attention to the bug and you'll be able to provide them with the extra information they'll demand, it should get fixed eventually.
User avatar
amuzen
LoS Moderator
 
Posts: 327
Joined: 05 Dec 2009, 02:49

Re: Segmentation fault (core dumped) -- (with gdb backtrace)

Postby amuzen » 16 Jul 2013, 17:50

Just in case any of the Intel users who had this issue are still around, it would be nice if someone could try the latest revision from Git and report back whether it's working. We came up with a solution that fixed it on newer versions of the open source Radeon driver, so it might now work on Intel too.
User avatar
amuzen
LoS Moderator
 
Posts: 327
Joined: 05 Dec 2009, 02:49

Who is online

Users browsing this forum: No registered users and 1 guest