Heya everyone,
some friends and me decided to use OpenCL for a teamproject at university.
Well only 1 worked with OpenCL yet and built most of those parts.
Our Project works fine with NVIDIA yet but as I'm the only one using ATI they can't nor
want to help me.
.
I just started to read about OpenCL and try some samples.
I have general ideas of each methods used.
But in our project we use OpenCL C++ (I didn't find a sample of those 😕 ) and I nailed
down my problem to this command:
manager.getCurrentQueue().enqueueAcquireGLObjects(&vec, NULL, NULL);
Well we did glFinish() before that line and manager.getCurrentQueue() returns the 'current' cl::CommandQueue,
vec is a vector of cl::Memory:
std::vector<cl::Memory> vec;
vec.push_back(m_particlePosBuffer);
Compiling and starting the programm works fine, but as soon as the program reaches this command it just ignores it and everything beyond.
But it does NOT freeze, our glutwindow just keeps updating and simulating, but as that simulateScene never finishs (because of that problem)
it never reachs renderScene() so my window is just empty -.-
I'm sorry if I missed a Discussion about a similar problem.
Edith1:
Well there are OpenCL C++ samples, but non of those uses a shared Buffer with OpenGL 😕
And all i get from debugging info is:
Eine Ausnahme (erste Chance) bei 0x776e76fd in tp.exe: Microsoft C++-Ausnahme: cl::Error an Speicherposition 0x002decd0..
So it's some kind of memory problem -.-
CL-ERROR: clEnqueueAcquireGLObjects(-5) -> OUT_OF_RESOURCES - no idea why... it shouldn't -.-
I read through everything that could cause that problem (imho) and couldn't see anything weird...
AND it's working fine with NVIDIA >.<
Edit 2:
Nope, all OpenGLBuffers are created before OpenCL context creation,
in some forums i read similar weird stuff with some OpenCL C(++) methods
won't work on some GPUs... but all OpenCL C samples using interop with OpenGL
work fine here, same for all OpenCL C++ samples i tried yet.