Updated my (nvidia) driver, now RTT is broken?

Started by
8 comments, last by EIRIKMEDKOMPIS 13 years, 3 months ago
In my application I'm rendering my scene into a cubemap, so I create the cubemap texture and then render the scene 6 times into the six faces. I've had this working fine for some time, but I just recently updated my driver for my NVIDIA 8800.

Now when I try to run the program with the new driver, I get GL_FRAMEBUFFER_UNSUPPORTED when I try to create my framebuffer. Anyone have any experience with this error?

	glGenTextures(1,&_texHandle);	glBindTexture(GL_TEXTURE_CUBE_MAP,_texHandle);	glTexParameteri(GL_TEXTURE_CUBE_MAP, GL_TEXTURE_MAG_FILTER, GL_NEAREST);	glTexParameteri(GL_TEXTURE_CUBE_MAP, GL_TEXTURE_MIN_FILTER, GL_NEAREST);	glTexParameteri(GL_TEXTURE_CUBE_MAP, GL_TEXTURE_WRAP_S, GL_CLAMP_TO_EDGE);	glTexParameteri(GL_TEXTURE_CUBE_MAP, GL_TEXTURE_WRAP_T, GL_CLAMP_TO_EDGE);	for(uint side = 0;side<6;side++){		glTexImage2D(GL_TEXTURE_CUBE_MAP_POSITIVE_X+side, 0, GL_RGBA, resolution, resolution, 0, GL_RGBA, GL_UNSIGNED_BYTE, 0);	}	glGenerateMipmap(GL_TEXTURE_CUBE_MAP);//I don't need mipmaps, but I thought this might help my error.	glBindTexture(GL_TEXTURE_CUBE_MAP,0);	GLCHECKERROR();	GLuint framebuffer;	glGenFramebuffers(1,&framebuffer);	glBindFramebuffer(GL_DRAW_FRAMEBUFFER,framebuffer);	glm::mat4 proj = OpenGL::PerspectiveProject(90,1,1,0.1f,1000);		GLint viewport[4];	glGetIntegerv(GL_VIEWPORT,viewport);	for(uint side=0;side<6;side++){		glm::mat4 view;		switch(side){			case 0: view = OpenGL::LookAt(glm::vec3(0,0,0),glm::vec3(1,0,0),glm::vec3(0,-1,0));break;			case 1: view = OpenGL::LookAt(glm::vec3(0,0,0),glm::vec3(-1,0,0),glm::vec3(0,-1,0));break;			case 2: view = OpenGL::LookAt(glm::vec3(0,0,0),glm::vec3(0,1,0),glm::vec3(0,0,1));break;			case 3: view = OpenGL::LookAt(glm::vec3(0,0,0),glm::vec3(0,-1,0),glm::vec3(0,0,-1));break;			case 4: view = OpenGL::LookAt(glm::vec3(0,0,0),glm::vec3(0,0,1),glm::vec3(0,-1,0));break;			case 5: view = OpenGL::LookAt(glm::vec3(0,0,0),glm::vec3(0,0,-1),glm::vec3(0,-1,0));break;		}		//+x/-x/+y/-y/+z/-z		glFramebufferTexture2D(GL_DRAW_FRAMEBUFFER,GL_COLOR_ATTACHMENT0,GL_TEXTURE_CUBE_MAP_POSITIVE_X+side,_texHandle,0);		OpenGL::CheckFramebufferError(GL_DRAW_FRAMEBUFFER); //GL_UNSUPPORTED_FRAMEBUFFER}


I saw this on opengl.org, but I'm not sure if it applies to me. How can I tell if my texture is 'complete'? I've set the min/max filters and even called generate mipmaps.
Quote:
Warning: NVIDIA's OpenGL driver has a known issue with using incomplete textures. If the texture is not texture complete, the FBO itself will be considered GL_FRAMEBUFFER_UNSUPPORTED, or will have GL_FRAMEBUFFER_INCOMPLETE_ATTACHMENT. This is a driver bug, as the OpenGL specification does not allow implementations to return either of these values simply because a texture is not yet complete. Until this is resolved in NVIDIA's drivers, it is advised to make sure that all textures have mipmap levels, and that all glTexParameteri values are properly set up for the format of the texture. For example, integral textures are not complete if the mag and min filters have any LINEAR fields.

[size=2]My Projects:
[size=2]Portfolio Map for Android - Free Visual Portfolio Tracker
[size=2]Electron Flux for Android - Free Puzzle/Logic Game
Advertisement
Guess I can just put this down to a screwy driver install. Reinstalled it with driver cleaner and the problem went away. Thought something suspicious was going on when it was taking 100ms to render a simple untextured cube :)
[size=2]My Projects:
[size=2]Portfolio Map for Android - Free Visual Portfolio Tracker
[size=2]Electron Flux for Android - Free Puzzle/Logic Game
Hey, can you tell me what driver version is infected and what could work?
I suddenly experience similar stupid behaviour and cannot figure a working driver :( November 16, 2010 (GPU Caps show this, I guess it is 260.99, i.e. the newest one now) driver does not seem to work on both my GTS250 as 9600GSO.
Unfortunately I never found a resolution to this yet. I thought that upgrading my drivers fixed the issue, but a few days later it started creeping back up again.

What I constantly found happening was that after running my program a few times it started having all sorts of crazy problems: GL_OUT_OF_MEMORY errors everywhere, rendering SUPER slow (took 10 whole seconds to render a frame that used to render at 60fps), and these random framebuffer errors.

Its like there was some memory leak that was filling up my GPU, and once it got to the limit it stopped working, and the only solution I found was to reboot my computer, then everything would be fine for a day or two.

It seemed to work fine as long as I properly exited my program every time (manually deleted all the textures and framebuffers), but anytime my program exited uncleanly either through an exception or through abortion through the debugger it started happening again, it's like windows wasn't properly cleaning up the context when the program terminated.

Don't know if this is anything like your issue, but that's been my experience.
[size=2]My Projects:
[size=2]Portfolio Map for Android - Free Visual Portfolio Tracker
[size=2]Electron Flux for Android - Free Puzzle/Logic Game
You have to understand that OpenGL is pretty complex so there are few bugs.
Sig: http://glhlib.sourceforge.net
an open source GLU replacement library. Much more modern than GLU.
float matrix[16], inverse_matrix[16];
glhLoadIdentityf2(matrix);
glhTranslatef2(matrix, 0.0, 0.0, 5.0);
glhRotateAboutXf2(matrix, angleInRadians);
glhScalef2(matrix, 1.0, 1.0, -1.0);
glhQuickInvertMatrixf2(matrix, inverse_matrix);
glUniformMatrix4fv(uniformLocation1, 1, FALSE, matrix);
glUniformMatrix4fv(uniformLocation2, 1, FALSE, inverse_matrix);
yea but this has been going on now for 5 months or so... and this is a regression, something that should be easy to track down.

I have the EXACT same problem.

I have had this problem with my Quadro and posted a thread about it on here somewhere.

sucks that it still hasnt been resolved. i brought it up on nvidias website months ago as well and nobody at nvidia acknowledged the problem last time i checked, yet there are a lot of people encountering it... pretty frustrating.

------------------------------

redwoodpixel.com

Same here. I found that killing dwm.exe on Vista helped a bit when the problem occurs. But it comes back after some time and sooner or later I have to reboot. I managed to get GLexpert running to get more info but all I could find was that all occurring errors are related to some out of memory condition. For instance when binding a FBO the first time the driver tries to allocate memory for all attached textures/renderbuffers. When this fails it returns GL_FRAMEBUFFER_UNSUPPORTED.

I'm chasing this issue for month now and didn't find anything in our code that could cause this. gDebugger shows no resource leaks and even if there were some the driver should release all resources when the application exits.

I really would like to report this to nVidia but it's nearly impossible to make a simple repro case. But since it's nearly impossible to become a registered nvidia developer nowadays I wouldn't even know how to submit it anyway. :-/

If any of you has found any solution or workaround please post it here. This is driving me crazy.
I'm sure it must be an unfixed bug in their driver. I'd suggest trying older ones unless you need the latest ones for some reason. I went back to 197.13 and haven't had any problems (yet), thought I haven't stressed it very much yet.
[size=2]My Projects:
[size=2]Portfolio Map for Android - Free Visual Portfolio Tracker
[size=2]Electron Flux for Android - Free Puzzle/Logic Game
I stumbled across this extension which is able to give some information about the memory as seen by the driver:
http://developer.download.nvidia.com/opengl/specs/GL_NVX_gpu_memory_info.txt

Under Vista or Win7 these values are global ones for all processes. I notices that for me these two numberse increase of time and over several restarts of the application:

GPU_MEMORY_INFO_EVICTION_COUNT_NVX
GPU_MEMORY_INFO_EVICTED_MEMORY_NVX

I discovered that these two numbers increase when my app resizes an opengl window while it is still hidden. For instance this happened in our editor that shows the opengl window after doing the hole opengl setup and preparing the layout for all windows. This involves resizing the render window. After that the main editor window is shown. I can reproduce this with a simple example app. I have not tested yet if the is really our original problem but it's the only clue I have right now.

Edit:
Here is a repro case for this in case you are still interested:
[source lang=cpp]#pragma comment(lib, "glut32.lib")#include "glut.h"#include <stdio.h>#define GL_GPU_MEMORY_INFO_DEDICATED_VIDMEM_NVX 0x9047#define GL_GPU_MEMORY_INFO_TOTAL_AVAILABLE_MEMORY_NVX 0x9048#define GL_GPU_MEMORY_INFO_CURRENT_AVAILABLE_VIDMEM_NVX 0x9049#define GL_GPU_MEMORY_INFO_EVICTION_COUNT_NVX 0x904A#define GL_GPU_MEMORY_INFO_EVICTED_MEMORY_NVX 0x904Bvoid PrintVideoMemory(){		static GLint evicted=0;	GLint vidmem=0, mem_available=0, vidmem_available=0, evicted_count=0, evicted_size=0; 	glGetIntegerv(GL_GPU_MEMORY_INFO_DEDICATED_VIDMEM_NVX, &vidmem);	glGetIntegerv(GL_GPU_MEMORY_INFO_TOTAL_AVAILABLE_MEMORY_NVX, &mem_available);	glGetIntegerv(GL_GPU_MEMORY_INFO_CURRENT_AVAILABLE_VIDMEM_NVX, &vidmem_available);	glGetIntegerv(GL_GPU_MEMORY_INFO_EVICTION_COUNT_NVX, &evicted_count);	glGetIntegerv(GL_GPU_MEMORY_INFO_EVICTED_MEMORY_NVX, &evicted_size);	if (evicted!=evicted_count)	{		printf("evicted_count=%d, evicted_size=%dkb, vidmem_available=%dkb\n",			evicted_count, evicted_size, vidmem_available);		evicted=evicted_count;	}}void renderScene(void) {	static bool flip=true;	flip=!flip;	glClear(GL_COLOR_BUFFER_BIT);	glBegin(GL_TRIANGLES);		glVertex3f(-0.5,-0.5,0.0);	glVertex3f(0.5,0.0,0.0);	glVertex3f(0.0,0.5,0.0);	glEnd();	glutSwapBuffers();	PrintVideoMemory();	if (flip)		glutReshapeWindow(300,200);		else		glutReshapeWindow(300,300);	}int main(int argc, char* argv[]){	glutInit(&argc, argv);	glutInitDisplayMode(GLUT_RGBA | GLUT_DOUBLE );	glutInitWindowPosition(100,100);	glutInitWindowSize(100,100);	glutCreateWindow("resoource leak provoker");	glutHideWindow();	//comment this to make the leak disappear		glutDisplayFunc(&renderScene);	glutIdleFunc(&renderScene);	glutMainLoop();	return 0;}
i have this exact same problem

geforce 9800 gt

This topic is closed to new replies.

Advertisement