| Bug ID | 92301 |
|---|---|
| Summary | Unigine Sanctuary freeze-crashes R390X |
| Product | Mesa |
| Version | 11.0 |
| Hardware | x86-64 (AMD64) |
| OS | Linux (All) |
| Status | NEW |
| Severity | normal |
| Priority | medium |
| Component | Drivers/Gallium/radeonsi |
| Assignee | dri-devel@lists.freedesktop.org |
| Reporter | mc.return@gmx.net |
| QA Contact | dri-devel@lists.freedesktop.org |
I understand that this benchmark might not use the latest version of the Unigine engine, but starting it should not make my machine lock up with actual high-end hardware, especially since the benchmark works flawless on pre-GCN AMD hardware.
Please attach the corresponding Xorg.0.log file and output of = dmesg and glxinfo. Are the 32-bit Mesa binaries up to date?
Created attachment 118697 [details]
console output of glxinfo
Hi Michel, thanks a lot for the swift response. I've built Mesa by checking out origin/11.0 and configuring the build with ./autogen.sh --prefix=/usr --enable-texture-float --libdir=/usr/lib/x86_64-linux-gnu/ --with-gallium-drivers=nouveau,radeonsi,r600 --with-egl-platforms=drm,x11 --enable-glx-tls --enable-shared-glapi --enable-glx --enable-driglx-direct --enable-gles1 --enable-gles2 --enable-experimental-targets=R600 How can I find out if my 32-bit Mesa binaries are up to date ?
(In reply to MC Return from comment #4) > How can I find out if my 32-bit Mesa binaries ar= e up to date ? It sounds like you only built 64-bit binaries of Mesa yourself, so you're probably still using the 32-bit binaries from Ubuntu. What does apt-cache policy libgl1-mesa-dri:i386 say?
(In reply to Michel D=C3=A4nzer from comment #5) >=20 > It sounds like you only built 64-bit binaries of Mesa yourself, so you= 're > probably still using the 32-bit binaries from Ubuntu. What does >=20 > apt-cache policy libgl1-mesa-dri:i386 >=20 > say? libgl1-mesa-dri:i386: Installed: 11.1~git1509280730.e413d2~gd~v Candidate: 11.1~git1509280730.e413d2~gd~v Version table: *** 11.1~git1509280730.e413d2~gd~v 0 100 /var/lib/dpkg/status 11.0.1~git20150928+11.0.7f1a77ae-0ubuntu0ricotz~trusty 0 500 ht= tp://ppa.launchpad.net/xorg-edgers/ppa/ubuntu/ trusty/main i386 Packages 10.1.3-0ubuntu0.4 0 500 http://archive.ub= untu.com/ubuntu/ trusty-updates/main i386 Packages 10.1.0-4ubuntu5 0 500 http://archive.ub= untu.com/ubuntu/ trusty/main i386 Packages
The crash seems to be so severe, that the last messages are not saved in .xsession-errors anymore, but once I managed to see some related errors printed out. I will try to reproduce this situation and report the errors.
I managed to capture the sanctuary log: Loading "/home/mcr2010/.Unigine Sanctuary/unigine.cfg"... Loading "libGL.so.1"... Loading "libopenal.so.1"... Set 1024x768 windowed video mode Set 1.00 gamma value Unigine engine http://unigine.com/ Binary: Linux 32bit GCC 4.3.2 Release May 20 2010 App path: /home/mcr2010/Downloads/sanctuary/bin/ Data path: /home/mcr2010/Downloads/sanctuary/data/ Save path: /home/mcr2010/.Unigine Sanctuary/ ---- System ---- System: Linux 4.3.0-040300rc4-generic x86_64 CPU: Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz 3491MHz MMX SSE SSE2 SSE3 SSSE3 SSE41 SSE42 HTT GPU: Gallium 0.4 on AMD HAWAII (DRM 2.43.0, LLVM 3.6.2) 3.0 Mesa 11.1.0-devel (git-e413d2f 2015-09-28 vivid-oibaf-ppa) System memory: 7912 Mb Video memory: 256 Mb ---- MathLib ---- Set SSE3 simd processor ---- Sound ---- Renderer: OpenAL Soft OpenAL vendor: OpenAL Community OpenAL renderer: OpenAL Soft OpenAL version: 1.1 ALSOFT 1.14 Found AL_EXT_LINEAR_DISTANCE Found AL_EXT_OFFSET Found ALC_EXT_EFX Found EFX Filter Found EFX Reverb Found EAX Reverb Found QUAD16 format Found 51CHN16 format Found 61CHN16 format Found 71CHN16 format Maximum sources: 256 Maximum effect slots: 4 Maximum auxiliary sends: 2 ---- Render ---- GLRender::GLRender(): Unknown GPU OpenGL vendor: X.Org OpenGL renderer: Gallium 0.4 on AMD HAWAII (DRM 2.43.0, LLVM 3.6.2) OpenGL version: 3.0 Mesa 11.1.0-devel (git-e413d2f 2015-09-28 vivid-oibaf-ppa) Found required GL_ARB_map_buffer_range Found required GL_ARB_vertex_array_object Found required GL_ARB_vertex_buffer_object Found required GL_ARB_half_float_vertex Found required GL_ARB_half_float_pixel Found required GL_ARB_occlusion_query Found required GL_EXT_texture3D Found required GL_EXT_texture_cube_map Found required GL_EXT_texture_sRGB Found required GL_EXT_texture_swizzle Found required GL_ARB_shader_object Found required GL_ARB_vertex_shader Found required GL_ARB_fragment_shader Found required GL_ARB_draw_buffers Found required GL_ARB_framebuffer_object Found required GL_EXT_framebuffer_blit Found required GL_EXT_framebuffer_multisample Found optional GL_ARB_draw_instanced Found optional GL_ARB_transform_feedback Found optional GL_ARB_draw_elements_base_vertex Found optional GL_ARB_uniform_buffer_object Found optional GL_ARB_sample_shading Found optional GL_ARB_gpu_shader4 Found optional GL_ARB_texture_rg Found optional GL_ARB_texture_array Found optional GL_ARB_texture_multisample Found optional GL_ARB_texture_compression Found optional GL_ARB_texture_compression_rgtc Found optional GL_ARB_seamless_cube_map Shading language: 1.30 Maximum texture size: 16384 Maximum texture units: 80 Maximum draw buffers: 8 ---- Physics ---- Physics: Multi-threaded ---- Interpreter ---- Version: 2.31 Loading "sanctuary/unigine.cpp" 47ms Loading "demos/sanctuary/locale/unigine.en" dictionary Loading "core/materials/unigine_post.mat" 12 materials 12 shaders 0ms Loading "core/materials/unigine_render.mat" 34 materials 77 shaders 4ms Loading "core/materials/unigine_meshes.mat" 16 materials 9350 shaders 34ms Loading "core/materials/unigine_terrains.mat" 1 material 293 shaders 1ms Loading "core/materials/unigine_grass.mat" 1 material 81 shaders 3ms Loading "core/materials/unigine_particles.mat" 1 material 43 shaders 2ms Loading "core/materials/unigine_billboards.mat" 1 material 33 shaders 2ms Loading "core/materials/unigine_volumes.mat" 5 materials 29 shaders 5ms Loading "core/materials/unigine_guis.mat" 1 material 24 shaders 0ms Loading "core/materials/unigine_water.mat" 1 material 63 shaders 8ms Loading "core/materials/unigine_skies.mat" 1 material 19 shaders 11ms Loading "core/materials/unigine_decals.mat" 1 material 69 shaders 3ms Loading "core/properties/unigine.prop" 2 properties 0ms Unigine~# world_load sanctuary Loading "sanctuary.cpp" 41ms Loading "sanctuary/sanctuary.mat" 54 materials 253ms Loading "sanctuary.world" 5618ms Unigine~# render_hdr 0 && render_srgb 0 && render_restart GLShader::loadFragment(): error in "core/shaders/render/fragment_occlusion_blur.shader" file defines: UNKNOWN,QUALITY_LOW,QUALITY_MEDIUM,QUALITY_HIGH,MULTISAMPLE_0,USE_INSTANCING,USE_TEXTURE_ARRAY,USE_DEFERRED,USE_TRANSLUCENT,USE_PARALLAX,USE_OCCLUSION,USE_REFLECTION,OPENGL,USE_PSEUDO_INSTANCING,USE_PSEUDO_TRANSFORM,USE_ARB_SAMPLE_SHADING,USE_ARB_TEXTURE_MULTISAMPLE,HAS_ARB_DRAW_INSTANCED 0:511(15): error: no function with name 'texelFetch2DOffset' 0:512(15): error: no function with name 'texelFetch2DOffset' 0:513(15): error: no function with name 'texelFetch2DOffset' 0:514(15): error: no function with name 'texelFetch2DOffset' 0:515(15): error: no function with name 'texelFetch2DOffset' 0:516(15): error: no function with name 'texelFetch2DOffset' 0:517(15): error: no function with name 'texelFetch2DOffset' 0:518(15): error: no function with name 'texelFetch2DOffset' 0:519(15): error: no function with name 'texelFetch2DOffset' 0:520(15): error: no function with name 'texelFetch2DOffset' 0:521(16): error: no function with name 'texelFetch2DOffset' 0:522(16): error: no function with name 'texelFetch2DOffset' 0:523(16): error: no function with name 'texelFetch2DOffset' 0:524(16): error: no function with name 'texelFetch2DOffset' 0:525(16): error: no function with name 'texelFetch2DOffset' 0:526(16): error: no function with name 'texelFetch2DOffset' I noticed it reports GPU: Gallium 0.4 on AMD HAWAII (DRM 2.43.0, LLVM 3.6.2) 3.0 Mesa 11.1.0-devel (git-e413d2f 2015-09-28 vivid-oibaf-ppa) with 256 Mb video-memory (should be 8Gb)...
(In reply to MC Return from comment #8) > I noticed it reports GPU: Gallium 0.4 on AMD HAW= AII (DRM 2.43.0, LLVM 3.6.2) > 3.0 Mesa 11.1.0-devel (git-e413d2f 2015-09-28 vivid-oibaf-ppa) [...] It's using the system 32-bit binaries (apparently left over from the oibaf = PPA, about one week old snapshot of Git master), because you haven't built any 32-bit binaries yourself.
(In reply to MC Return from comment #8) > GLShader::loadFragment(): error in > "core/shaders/render/fragment_occlusion_blur.shader" file > defines: > UNKNOWN,QUALITY_LOW,QUALITY_MEDIUM,QUALITY_HIGH,MULTISAMPLE_0,USE_INST= ANCING, > USE_TEXTURE_ARRAY,USE_DEFERRED,USE_TRANSLUCENT,USE_PARALLAX,USE_OCCLUS= ION, > USE_REFLECTION,OPENGL,USE_PSEUDO_INSTANCING,USE_PSEUDO_TRANSFORM, > USE_ARB_SAMPLE_SHADING,USE_ARB_TEXTURE_MULTISAMPLE,HAS_ARB_DRAW_INSTAN= CED > 0:511(15): error: no function with name 'texelFetch2DOffset' These errors may indicate that it's not picking up the correct drirc configuration. Does it work if you set the following environment variables = for running it: LIBGL_DEBUG=3Dverbose force_glsl_extensions_warn=3Dtrue disable_blend_func_extended=3Dtrue If not, please provide the corresponding terminal output and log file again= .
>
> It's using the system 32-bit binaries (apparently left over from the oibaf
> PPA, about one week old snapshot of Git master), because you haven't built
> any 32-bit binaries yourself.
I want to ensure that all binaries, including the 32bit ones, are self-built, I
am rebuilding with the --enable-32-bit flag added right now, hope this works...
(In reply to Michel D=C3=A4nzer from comment #10) > (In reply to MC Return from comment #8) > > GLShader::loadFragment(): error in > > "core/shaders/render/fragment_occlusion_blur.shader" fi= le > > defines: > > UNKNOWN,QUALITY_LOW,QUALITY_MEDIUM,QUALITY_HIGH,MULTISAMPLE_0,USE= _INSTANCING, > > USE_TEXTURE_ARRAY,USE_DEFERRED,USE_TRANSLUCENT,USE_PARALLAX,USE_O= CCLUSION, > > USE_REFLECTION,OPENGL,USE_PSEUDO_INSTANCING,USE_PSEUDO_TRANSFORM, > > USE_ARB_SAMPLE_SHADING,USE_ARB_TEXTURE_MULTISAMPLE,HAS_ARB_DRAW_I= NSTANCED > > 0:511(15): error: no function with name 'texelFetch2DOffset' >=20 > These errors may indicate that it's not picking up the correct drirc > configuration. Does it work if you set the following environment varia= bles > for running it: >=20 > LIBGL_DEBUG=3Dverbose force_glsl_extensions_warn=3Dtrue > disable_blend_func_extended=3Dtrue >=20 > If not, please provide the corresponding terminal output and log file = again. I set those env vars, but the result was exactly the same - sanctuary crash= ed the machine after a few seconds running, but it showed these errors after changing screens to console mode: [drm: ci_dpm_enable [radeon]] *ERROR* ci_start_dpm failed [drm: radeon_pm_resume [radeon]] *ERROR* radeon: dpm resume failed [drm: radeon_pm_resume [radeon]] *ERROR* radeon: dpm resume failed
(In reply to MC Return from comment #11) > > > > It's using the system 32-bit binaries (apparently left over from the oibaf > > PPA, about one week old snapshot of Git master), because you haven't built > > any 32-bit binaries yourself. > > I want to ensure that all binaries, including the 32bit ones, are > self-built, I am rebuilding with the --enable-32-bit flag added right now, > hope this works... Nope --enable-32-bit is no longer recognized and I could not find any instructions on how to ensure a 32bit build... What flags do I need to add to ./autogen.sh or what needs to be done to to get that accomplished ?
(In reply to MC Return from comment #12) > I set those env vars, but the result was exactly= the same - sanctuary > crashed the machine after a few seconds running, but it showed these e= rrors > after changing screens to console mode: Thanks, but as I said, we need to see the corresponding terminal / log outp= ut from Sanctuary. (In reply to MC Return from comment= #13) > Nope --enable-32-bit is no longer recognized and= I could not find any > instructions on how to ensure a 32bit build... > What flags do I need to add to ./autogen.sh or what needs to be done t= o to > get that accomplished ? I use: --host=3Di386-linux-gnu --with-llvm-prefix=3D[path to 32-bit llvm directory] 'CC=3Dgcc -m32' 'CXX=3Dg++ -m32'
(In reply to Michel D=C3=A4nzer from comment #14) > Thanks, but as I said, we need to see the corres= ponding terminal / log > output from Sanctuary. > After upgrade to X.Org X Server 1.17.1 the freeze-crashing of a default Sanctuary run is fixed for me, but after setting and checking via printenv = the mentioned environment variables the Sanctuary log still shows the same erro= rs, here is the full log: $ ./1024x768_windowed.sh=20 libGL: pci id for fd 4: 1002:67b0, driver radeonsi libGL: OpenDriver: trying /usr/lib/i386-linux-gnu/dri/tls/radeonsi_dri.so libGL: OpenDriver: trying /usr/lib/i386-linux-gnu/dri/radeonsi_dri.so ATTENTION: default value of option force_glsl_extensions_warn overridden by environment. ATTENTION: default value of option disable_blend_func_extended overridden by environment. ATTENTION: option value of option force_glsl_extensions_warn ignored. ATTENTION: option value of option disable_blend_func_extended ignored. libGL: Using DRI3 for screen 0 Loading "/home/mcr2010/.Unigine Sanctuary/unigine.cfg"... Loading "libGL.so.1"... Loading "libopenal.so.1"... Set 1024x768 windowed video mode libGL: pci id for fd 10: 1002:67b0, driver radeonsi libGL: OpenDriver: trying /usr/lib/i386-linux-gnu/dri/tls/radeonsi_dri.so libGL: OpenDriver: trying /usr/lib/i386-linux-gnu/dri/radeonsi_dri.so ATTENTION: default value of option force_glsl_extensions_warn overridden by environment. ATTENTION: default value of option disable_blend_func_extended overridden by environment. ATTENTION: option value of option force_glsl_extensions_warn ignored. ATTENTION: option value of option disable_blend_func_extended ignored. libGL: Using DRI3 for screen 0 Set 1.00 gamma value Unigine engine http://unigine.com/ Binary: Linux 32bit GCC 4.3.2 Release May 20 2010 App path: /home/mcr2010/Downloads/sanctuary/bin/ Data path: /home/mcr2010/Downloads/sanctuary/data/ Save path: /home/mcr2010/.Unigine Sanctuary/ ---- System ---- System: Linux 4.3.0-040300rc4-generic x86_64 CPU: Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz 3491MHz MMX SSE SSE2 SSE3= SSSE3 SSE41 SSE42 HTT GPU: Gallium 0.4 on AMD HAWAII (DRM 2.43.0, LLVM 3.6.2) 3.0 Mesa 11.1.0-dev= el (git-e413d2f 2015-09-28 vivid-oibaf-ppa) System memory: 7912 Mb Video memory: 256 Mb ---- MathLib ---- Set SSE3 simd processor ---- Sound ---- Renderer: OpenAL Soft OpenAL vendor: OpenAL Community OpenAL renderer: OpenAL Soft OpenAL version: 1.1 ALSOFT 1.15.1 Found AL_EXT_LINEAR_DISTANCE Found AL_EXT_OFFSET Found ALC_EXT_EFX Found EFX Filter Found EFX Reverb Found EAX Reverb Found QUAD16 format Found 51CHN16 format Found 61CHN16 format Found 71CHN16 format Maximum sources: 256 Maximum effect slots: 4 Maximum auxiliary sends: 2 ---- Render ---- GLRender::GLRender(): Unknown GPU OpenGL vendor: X.Org OpenGL renderer: Gallium 0.4 on AMD HAWAII (DRM 2.43.0, LLVM 3.6.2) OpenGL version: 3.0 Mesa 11.1.0-devel (git-e413d2f 2015-09-28 vivid-oibaf-= ppa) Found required GL_ARB_map_buffer_range Found required GL_ARB_vertex_array_object Found required GL_ARB_vertex_buffer_object Found required GL_ARB_half_float_vertex Found required GL_ARB_half_float_pixel Found required GL_ARB_occlusion_query Found required GL_EXT_texture3D Found required GL_EXT_texture_cube_map Found required GL_EXT_texture_sRGB Found required GL_EXT_texture_swizzle Found required GL_ARB_shader_object Found required GL_ARB_vertex_shader Found required GL_ARB_fragment_shader Found required GL_ARB_draw_buffers Found required GL_ARB_framebuffer_object Found required GL_EXT_framebuffer_blit Found required GL_EXT_framebuffer_multisample Found optional GL_ARB_draw_instanced Found optional GL_ARB_transform_feedback Found optional GL_ARB_draw_elements_base_vertex Found optional GL_ARB_uniform_buffer_object Found optional GL_ARB_sample_shading Found optional GL_ARB_gpu_shader4 Found optional GL_ARB_texture_rg Found optional GL_ARB_texture_array Found optional GL_ARB_texture_multisample Found optional GL_ARB_texture_compression Found optional GL_ARB_texture_compression_rgtc Found optional GL_ARB_seamless_cube_map Shading language: 1.30 Maximum texture size: 16384 Maximum texture units: 80 Maximum draw buffers: 8 ---- Physics ---- Physics: Multi-threaded ---- Interpreter ---- Version: 2.31 Loading "sanctuary/unigine.cpp" 47ms Loading "demos/sanctuary/locale/unigine.en" dictionary Loading "core/materials/unigine_post.mat" 12 materials 12 shaders= 0ms Loading "core/materials/unigine_render.mat" 34 materials 77 shade= rs 4ms Loading "core/materials/unigine_meshes.mat" 16 materials 9350 sha= ders 33ms Loading "core/materials/unigine_terrains.mat" 1 material 293 shad= ers 1ms Loading "core/materials/unigine_grass.mat" 1 material 81 shaders = 3ms Loading "core/materials/unigine_particles.mat" 1 material 43 shad= ers 2ms Loading "core/materials/unigine_billboards.mat" 1 material 33 sha= ders 2ms Loading "core/materials/unigine_volumes.mat" 5 materials 29 shade= rs 5ms Loading "core/materials/unigine_guis.mat" 1 material 24 shaders 0= ms Loading "core/materials/unigine_water.mat" 1 material 63 shaders = 8ms Loading "core/materials/unigine_skies.mat" 1 material 19 shaders = 10ms Loading "core/materials/unigine_decals.mat" 1 material 69 shaders= 3ms Loading "core/properties/unigine.prop" 2 properties 0ms Unigine~# world_load sanctuary Loading "sanctuary.cpp" 42ms Loading "sanctuary/sanctuary.mat" 54 materials 228ms Loading "sanctuary.world" 5446ms Unigine~# render_hdr 0 && render_srgb 0 && render_restart GLShader::loadFragment(): error in "core/shaders/render/fragment_occlusion_blur.shader" file defines: UNKNOWN,QUALITY_LOW,QUALITY_MEDIUM,QUALITY_HIGH,MULTISAMPLE_0,USE_INSTANCIN= G,USE_TEXTURE_ARRAY,USE_DEFERRED,USE_TRANSLUCENT,USE_PARALLAX,USE_OCCLUSION= ,USE_REFLECTION,OPENGL,USE_PSEUDO_INSTANCING,USE_PSEUDO_TRANSFORM,USE_ARB_S= AMPLE_SHADING,USE_ARB_TEXTURE_MULTISAMPLE,HAS_ARB_DRAW_INSTANCED 0:511(15): error: no function with name 'texelFetch2DOffset' 0:512(15): error: no function with name 'texelFetch2DOffset' 0:513(15): error: no function with name 'texelFetch2DOffset' 0:514(15): error: no function with name 'texelFetch2DOffset' 0:515(15): error: no function with name 'texelFetch2DOffset' 0:516(15): error: no function with name 'texelFetch2DOffset' 0:517(15): error: no function with name 'texelFetch2DOffset' 0:518(15): error: no function with name 'texelFetch2DOffset' 0:519(15): error: no function with name 'texelFetch2DOffset' 0:520(15): error: no function with name 'texelFetch2DOffset' 0:521(16): error: no function with name 'texelFetch2DOffset' 0:522(16): error: no function with name 'texelFetch2DOffset' 0:523(16): error: no function with name 'texelFetch2DOffset' 0:524(16): error: no function with name 'texelFetch2DOffset' 0:525(16): error: no function with name 'texelFetch2DOffset' 0:526(16): error: no function with name 'texelFetch2DOffset' Unigine~# quit Close "libopenal.so.1" Close "libGL.so.1" Memory usage: 64b Allocations: 1 Shutdown
| What | Removed | Added |
|---|---|---|
| Status | NEW | RESOLVED |
| Resolution | --- | WORKSFORME |
(In reply to Michel D=C3=A4nzer from comment #14) >=20 > I use: >=20 > --host=3Di386-linux-gnu --with-llvm-prefix=3D[path to 32-bit llvm dire= ctory] > 'CC=3Dgcc -m32' 'CXX=3Dg++ -m32' I have no idea how to find out the "path to 32-bit llvm directory"= ; and searching for it did not bring up any results I could use. Could you please tell me where this path is supposed to be and how it's nam= e is on a Ubuntu amd64 based system ? I think this report can be closed for now - or should it stay open and be renamed because of the log problems ?