I am new to this forum, hoping someone can help me with my problem.
I am looking for fastest method to copy application generated stencil data to stencil buffer? My OpenGL app uses ATI Radeon 5970 video card and when I used glDrawPixels to copy the data as below the performance was terribly slow about 3 FPS.
glDrawPixels(768, 768, GL_STENCIL_INDEX, GL_UNSIGNED_BYTE, (const GLvoid* )data);
What is interesting is the same app runs 25 FPS on ATI Radeon 3800 video card.
I even tried using the pixel buffer objects in conjunction with glDrawPixels, the performance seems to be fine (note this is only when the width/height is a multiple of 32) but the stencil buffer is corrupted.
glBindBuffer(GL_PIXEL_UNPACK_BUFFER, m_pbo);
glBufferSubData(GL_PIXEL_UNPACK_BUFFER, 0, size, (const GLvoid *) plane_data);
glDrawPixels(768, 768, GL_STENCIL_INDEX, GL_UNSIGNED_BYTE, 0);
Note, I do not use shaders.
Please let me know if anyone needs a sample code.
thanks in advance
Sounds very silent in here...seems am I the only one using glDrawPixels on ATI 5970? Really need help with this issue. Any one any ideas?
It got fixed. You have to wait for two or three months to try the new driver.
Originally posted by: frali It got fixed. You have to wait for two or three months to try the new driver.
hey Frank,
Could it be possible to get a beta release with the fix? We are developing a simulation based on these vidoe cards and we want to make sure this is not a risk to our program.
Also let me know your e-mail ID for further communication on this.
thanks
You could contact me by frank.li@amd.com