Geeks3D Forums

GeeXLab => GeeXLab - english forum => Topic started by: Stefan on January 31, 2016, 08:23:43 AM

Title: Lonely Sea Shell vs. NVIDIA compiler
Post by: Stefan on January 31, 2016, 08:23:43 AM

Lonely Sea Shell (https://www.shadertoy.com/view/ldy3WR) - ported to GeeXLab (https://drive.google.com/file/d/0BykQ4pHxfGQWWDJ2UUVTWEx1dUE/view)
runs fine with INTEL and ATI, but NVIDIA compiler fails with error C1103: too few parameters in function call

Work-around: replace  float r = length(target-source);  with
   

Code: [Select]

#ifdef GL_NV_command_list // check for NVIDIA, patch would fail with ATI and INTEL
float r = length(0.0, target-source); // ignore warning C7011: implicit cast from "float" to "vec3"
#else
float r = length(target-source); // original code compiles fine with ATI and Intel
#endif
I wonder why it runs at shadertoy.com with NVIDIA, but not in GeeXLab?

I added also vendor dependent overlay text

Code: [Select]

    local ret = gh_renderer.check_opengl_extension("GL_NV_command_list")
    if (ret == 1) then
    gfx.write_text(10, 160, 0, 1, 0, 1, "patched for NVIDIA")
    end
   
    local ret = gh_renderer.check_opengl_extension("GL_AMD_debug_output")
    if (ret == 1) then
    gfx.write_text(10, 160, 0, 1, 0, 1, "AMD works fine")
    end
   
    local ret = gh_renderer.check_opengl_extension("GL_INTEL_performance_query")
    if (ret == 1) then
    gfx.write_text(10, 160, 0, 1, 0, 1, "Intel works fine")
    end