• Advertisement
Sign in to follow this  

OpenGL Video RAM fragmentation and resource management for continuously running system?

This topic is 3405 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Basically I don't know how to handle this situation: Consider a continuously running system which loads and unloads modules based on a schedule very frequently (on average every minute) but the OpenGL rendering never leaves full-screen; the module switch is completely seamless (maybe a half-second fade in between one module stopping drawing to the OpenGL render thread and the next starting). Now, since I can't preload all the data simultaneously at system startup , and things like textures and VBOs will have to be dynamically loaded, how do I avoid video memory fragmentation from eventually killing performance (and possibly stability)? I need the system to be very stable and cannot do restarts more often than once in 24 hours (preferably once a week).

Share this post


Link to post
Share on other sites
Advertisement
First load in all the data that won't be changing between module loads/unloads, if any. After that, load in a module's texture data. When the module is done, unload it completely before loading the next module data. If you do it this way, there won't be any fragmentation.

Share this post


Link to post
Share on other sites
Quote:
Original post by Promit
Is this actually an observed problem, or just paranoia?

My boss wants a system that is near unconditionally stable. I better be paranoid if I want to keep my job.

Quote:
Original post by Numsgil
First load in all the data that won't be changing between module loads/unloads, if any. After that, load in a module's texture data. When the module is done, unload it completely before loading the next module data. If you do it this way, there won't be any fragmentation.

Thanks for the suggestion. I'm wondering what the best way is to load stuff so that I can do a quick switchover. I would have to preload stuff for the next module before the first is finished, but I'm wondering if a whole bunch of glTexImage, glBufferData, and glMapBuffer will not screw up framerate of whatever's currently running. I could have the next module queueing messages to the render thread to do these with some sparseness (in time) but that seems like not the simplest solution.

Share this post


Link to post
Share on other sites
In that case, you're going to have to do heavy stress testing anyway. Why not write a simpler first pass implementation, and then refine based on what you see?

Share this post


Link to post
Share on other sites
I guess just that in my experience major refinements end up being rewrites of half the stuff heh...

Numsgil's suggestion requires that I unload a module completely before loading the next one. That's a problem in terms of seamlessly going from one to the other, with a half-second black screen or logo being displayed between them acceptable. I'd really need to start loading the next one before unloading the current.

Now imagine doing such a switch around every minute for 24 hours. How likely that video memory would get fragmented and it'll end up with constant swapping with main memory?
At least with system RAM there there is a clear solution--preallocate memory pools. But no void* analogue for VRAM; to do memory management there would be a pain since there'd be preallocate 'free' lists of all different types of objects such as texture with x channels and y resolution (with glTexImage2D), and checking the list for a match to put the new texture in with glTexSubImage2D when loading actual textures, then sending back to free list (example at http://www.codeguru.com/cpp/g-m/opengl/texturemapping/article.php/c5573/ ), and then for VBOs there'd be something like this, etc. and it'll be horrible...

Share this post


Link to post
Share on other sites
The completely-unload-one-before-loading-the-next is the method we're using at work. In our experience unloading a level/module/whatever should be nearly instantaneous. Certainly not anywhere near as long as loading a level/module/whatever. This makes sense when you consider that the driver doesn't have to upload any data to the card when it unloads a texture, it just deletes a handle somewhere.

Share this post


Link to post
Share on other sites
Sure, the unloading is near instantaneous, but the loading isn't. That's the problem. Consider timeline:

<--0.5sFadeIn----1minModule1Draws-----0.5sFadeOut--><--0.5sSplash--><--0.5sFadeIn----1minModule2Draws-----0.5sFadeOut-->... [to module n then repeat]

After one module finishes running I can't go much over half a second till the second starts drawing. So completely unloading the first before loading the second doesn't seem viable.

Share this post


Link to post
Share on other sites
what you could do is to load the resource you need to load (textures, vbo, etc...) in system mem buffers only (you can do this asynchronously - in a loading thread).

Once all the resources have been loaded into RAM, unload your previous module (including GPUs objects), then simply creates you GPUs buffer and fill them from you RAM buffers. That should be pretty fast, at least this should be the fastest solution (appart from sharing resource between modules, i.e. re-using texture objects/vbo - but this is nearly impossible as they should all have the same size...)

Share this post


Link to post
Share on other sites
gjaegy, that's the approach I intend to start with. I thought about virtual texturing but the complexity and shader overhead are a turnoff...

Quote:
Original post by V-man
module?

Mostly independent media content packages, such as games, videos, etc. Since the functionality of each is different, i.e. game logic, animation, simulation, rendering and GI algorithms, etc., they're individual dynamically linked (at runtime) libraries and associated data, and only a small subset will fit in memory at a time, in general. The system is non-deterministic due to users interaction, but it needs to run without supervision and users are not intended to have any supervisory role whatsoever and can only interact with the content, not the system it runs on. That's about as descriptive as I can be without violating my NDA LOL

Share this post


Link to post
Share on other sites
Some will be small and all of those could be preloaded. But some will be using around a GB of RAM (I already have one that does that).
There will be around a dozen or so modules running one after another, and then the cycle repeats indefinitely.
I'm ignoring for now the problem of system memory fragmentation if I'm loading from HD (very likely given the memory limit on a 32-bit system), but might do a custom new/delete for that if it's a problem.

Share this post


Link to post
Share on other sites
Quote:
Original post by Prune
Some will be small and all of those could be preloaded. But some will be using around a GB of RAM (I already have one that does that).
And how much data will be shared between modules?

It sounds like you should be able to load all the new assets into RAM, and then flush out all the data in the GPU and submit the new data. This way you can preload from the HD, and you shouldn't have any fragmentation in the GPU.

Share this post


Link to post
Share on other sites
Sign in to follow this  

  • Advertisement
  • Advertisement
  • Popular Tags

  • Advertisement
  • Popular Now

  • Similar Content

    • By LifeArtist
      Good Evening,
      I want to make a 2D game which involves displaying some debug information. Especially for collision, enemy sights and so on ...
      First of I was thinking about all those shapes which I need will need for debugging purposes: circles, rectangles, lines, polygons.
      I am really stucked right now because of the fundamental question:
      Where do I store my vertices positions for each line (object)? Currently I am not using a model matrix because I am using orthographic projection and set the final position within the VBO. That means that if I add a new line I would have to expand the "points" array and re-upload (recall glBufferData) it every time. The other method would be to use a model matrix and a fixed vbo for a line but it would be also messy to exactly create a line from (0,0) to (100,20) calculating the rotation and scale to make it fit.
      If I proceed with option 1 "updating the array each frame" I was thinking of having 4 draw calls every frame for the lines vao, polygons vao and so on. 
      In addition to that I am planning to use some sort of ECS based architecture. So the other question would be:
      Should I treat those debug objects as entities/components?
      For me it would make sense to treat them as entities but that's creates a new issue with the previous array approach because it would have for example a transform and render component. A special render component for debug objects (no texture etc) ... For me the transform component is also just a matrix but how would I then define a line?
      Treating them as components would'nt be a good idea in my eyes because then I would always need an entity. Well entity is just an id !? So maybe its a component?
      Regards,
      LifeArtist
    • By QQemka
      Hello. I am coding a small thingy in my spare time. All i want to achieve is to load a heightmap (as the lowest possible walking terrain), some static meshes (elements of the environment) and a dynamic character (meaning i can move, collide with heightmap/static meshes and hold a varying item in a hand ). Got a bunch of questions, or rather problems i can't find solution to myself. Nearly all are deal with graphics/gpu, not the coding part. My c++ is on high enough level.
      Let's go:
      Heightmap - i obviously want it to be textured, size is hardcoded to 256x256 squares. I can't have one huge texture stretched over entire terrain cause every pixel would be enormous. Thats why i decided to use 2 specified textures. First will be a tileset consisting of 16 square tiles (u v range from 0 to 0.25 for first tile and so on) and second a 256x256 buffer with 0-15 value representing index of the tile from tileset for every heigtmap square. Problem is, how do i blend the edges nicely and make some computationally cheap changes so its not obvious there are only 16 tiles? Is it possible to generate such terrain with some existing program?
      Collisions - i want to use bounding sphere and aabb. But should i store them for a model or entity instance? Meaning i have 20 same trees spawned using the same tree model, but every entity got its own transformation (position, scale etc). Storing collision component per instance grats faster access + is precalculated and transformed (takes additional memory, but who cares?), so i stick with this, right? What should i do if object is dynamically rotated? The aabb is no longer aligned and calculating per vertex min/max everytime object rotates/scales is pretty expensive, right?
      Drawing aabb - problem similar to above (storing aabb data per instance or model). This time in my opinion per model is enough since every instance also does not have own vertex buffer but uses the shared one (so 20 trees share reference to one tree model). So rendering aabb is about taking the model's aabb, transforming with instance matrix and voila. What about aabb vertex buffer (this is more of a cosmetic question, just curious, bumped onto it in time of writing this). Is it better to make it as 8 points and index buffer (12 lines), or only 2 vertices with min/max x/y/z and having the shaders dynamically generate 6 other vertices and draw the box? Or maybe there should be just ONE 1x1x1 cube box template moved/scaled per entity?
      What if one model got a diffuse texture and a normal map, and other has only diffuse? Should i pass some bool flag to shader with that info, or just assume that my game supports only diffuse maps without fancy stuff?
      There were several more but i forgot/solved them at time of writing
      Thanks in advance
    • By RenanRR
      Hi All,
      I'm reading the tutorials from learnOpengl site (nice site) and I'm having a question on the camera (https://learnopengl.com/Getting-started/Camera).
      I always saw the camera being manipulated with the lookat, but in tutorial I saw the camera being changed through the MVP arrays, which do not seem to be camera, but rather the scene that changes:
      Vertex Shader:
      #version 330 core layout (location = 0) in vec3 aPos; layout (location = 1) in vec2 aTexCoord; out vec2 TexCoord; uniform mat4 model; uniform mat4 view; uniform mat4 projection; void main() { gl_Position = projection * view * model * vec4(aPos, 1.0f); TexCoord = vec2(aTexCoord.x, aTexCoord.y); } then, the matrix manipulated:
      ..... glm::mat4 projection = glm::perspective(glm::radians(fov), (float)SCR_WIDTH / (float)SCR_HEIGHT, 0.1f, 100.0f); ourShader.setMat4("projection", projection); .... glm::mat4 view = glm::lookAt(cameraPos, cameraPos + cameraFront, cameraUp); ourShader.setMat4("view", view); .... model = glm::rotate(model, glm::radians(angle), glm::vec3(1.0f, 0.3f, 0.5f)); ourShader.setMat4("model", model);  
      So, some doubts:
      - Why use it like that?
      - Is it okay to manipulate the camera that way?
      -in this way, are not the vertex's positions that changes instead of the camera?
      - I need to pass MVP to all shaders of object in my scenes ?
       
      What it seems, is that the camera stands still and the scenery that changes...
      it's right?
       
       
      Thank you
       
    • By dpadam450
      Sampling a floating point texture where the alpha channel holds 4-bytes of packed data into the float. I don't know how to cast the raw memory to treat it as an integer so I can perform bit-shifting operations.

      int rgbValue = int(textureSample.w);//4 bytes of data packed as color
      // algorithm might not be correct and endianness might need switching.
      vec3 extractedData = vec3(  rgbValue & 0xFF000000,  (rgbValue << 8) & 0xFF000000, (rgbValue << 16) & 0xFF000000);
      extractedData /= 255.0f;
    • By Devashish Khandelwal
      While writing a simple renderer using OpenGL, I faced an issue with the glGetUniformLocation function. For some reason, the location is coming to be -1.
      Anyone has any idea .. what should I do?
  • Advertisement