spdis/voxy
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

_spent 3 hours on a single missing *32_

Browse Source
This commit is contained in:
mcrcortex
2025-05-02 22:58:55 +10:00
parent 8b04be72ee
commit e82d2a875c
1 changed files with 84 additions and 55 deletions
Download Patch File Download Diff File Expand all files Collapse all files

139
src/main/java/me/cortex/voxy/client/core/rendering/ChunkBoundRenderer.java
View File

@@ -1,23 +1,22 @@
package me.cortex.voxy.client.core.rendering;
import it.unimi.dsi.fastutil.ints.Int2IntOpenHashMap;
import it.unimi.dsi.fastutil.ints.Int2LongLinkedOpenHashMap;
import it.unimi.dsi.fastutil.ints.Int2LongOpenHashMap;
import it.unimi.dsi.fastutil.longs.Long2ByteOpenHashMap;
import it.unimi.dsi.fastutil.longs.Long2IntOpenHashMap;
import me.cortex.voxy.client.core.gl.GlBuffer;
import me.cortex.voxy.client.core.gl.GlFramebuffer;
import me.cortex.voxy.client.core.gl.GlTexture;
import me.cortex.voxy.client.core.gl.shader.AutoBindingShader;
import me.cortex.voxy.client.core.gl.shader.Shader;
import me.cortex.voxy.client.core.gl.shader.ShaderType;
import me.cortex.voxy.client.core.rendering.util.SharedIndexBuffer;
import me.cortex.voxy.client.core.rendering.util.UploadStream;
import me.cortex.voxy.common.Logger;
import net.minecraft.util.math.ChunkPos;
import net.minecraft.util.math.MathHelper;
import org.joml.Matrix4f;
import org.joml.Vector3i;
import org.lwjgl.system.MemoryUtil;
import static org.lwjgl.opengl.ARBDirectStateAccess.glCopyNamedBufferSubData;
import static org.lwjgl.opengl.GL11.GL_TRIANGLES;
import static org.lwjgl.opengl.GL11.GL_UNSIGNED_BYTE;
import static org.lwjgl.opengl.GL15.GL_ELEMENT_ARRAY_BUFFER;
@@ -32,68 +31,37 @@ import static org.lwjgl.opengl.GL45.glClearNamedFramebufferfv;
//This is a render subsystem, its very simple in what it does
// it renders an AABB around loaded chunks, thats it
public class ChunkBoundRenderer {
public static final int MAX_CHUNK_COUNT = 10_000;
private final GlBuffer chunkPosBuffer = new GlBuffer(MAX_CHUNK_COUNT*8);//Stored as ivec2
private static final int INIT_MAX_CHUNK_COUNT = 1<<12;
private GlBuffer chunkPosBuffer = new GlBuffer(INIT_MAX_CHUNK_COUNT*8);//Stored as ivec2
private final GlBuffer uniformBuffer = new GlBuffer(128);
private final Long2IntOpenHashMap chunk2idx = new Long2IntOpenHashMap(MAX_CHUNK_COUNT);
private final long[] idx2chunk = new long[MAX_CHUNK_COUNT];
private final Long2IntOpenHashMap chunk2idx = new Long2IntOpenHashMap(INIT_MAX_CHUNK_COUNT);
private long[] idx2chunk = new long[INIT_MAX_CHUNK_COUNT];
private final Shader rasterShader = Shader.makeAuto()
.add(ShaderType.VERTEX, "voxy:chunkoutline/outline.vsh")
.add(ShaderType.FRAGMENT, "voxy:chunkoutline/outline.fsh")
.compile()
.ssbo(0, this.uniformBuffer)
.ubo(0, this.uniformBuffer)
.ssbo(1, this.chunkPosBuffer);
private GlTexture depthBuffer = new GlTexture().store(GL_DEPTH_COMPONENT24, 1, 128, 128);
private final GlFramebuffer frameBuffer = new GlFramebuffer().bind(GL_DEPTH_ATTACHMENT, this.depthBuffer).verify();
private final Long2ByteOpenHashMap updates = new Long2ByteOpenHashMap();
public ChunkBoundRenderer() {
this.chunk2idx.defaultReturnValue(-1);
}
public void addChunk(long pos) {
if (this.chunk2idx.size() >= MAX_CHUNK_COUNT) {
throw new IllegalStateException("At capacity");
}
if (this.chunk2idx.containsKey(pos)) {
//Logger.warn("Chunk already in map: " + new ChunkPos(pos));
return;
}
int idx = this.chunk2idx.size();
this.chunk2idx.put(pos, idx);
this.idx2chunk[idx] = pos;
this.put(idx, pos);
public void addSection(long pos) {
this.updates.addTo(pos, (byte) 1);
}
public void removeChunk(long pos) {
int idx = this.chunk2idx.remove(pos);
if (idx == -1) {
//Logger.warn("Chunk not in map: " + new ChunkPos(pos));
return;
}
if (idx == this.chunk2idx.size()) {
//Dont need to do anything as heap is already compact
return;
}
if (this.idx2chunk[idx] != pos) {
throw new IllegalStateException();
}
//Move last entry on heap to this index
long ePos = this.idx2chunk[this.chunk2idx.size()];// since is already removed size is correct end idx
if (this.chunk2idx.put(ePos, idx) == -1) {
throw new IllegalStateException();
}
this.idx2chunk[idx] = ePos;
//Put the end pos into the new idx
this.put(idx, ePos);
public void removeSection(long pos) {
this.updates.addTo(pos, (byte) -1);
}
//Bind and render, changing as little gl state as possible so that the caller may configure how it wants to render
public void render(Viewport<?> viewport) {
if (this.chunk2idx.isEmpty()) return;
if (this.chunk2idx.isEmpty() && this.updates.isEmpty()) return;
if (this.depthBuffer.getWidth() != viewport.width || this.depthBuffer.getHeight() != viewport.height) {
this.depthBuffer.free();
@@ -118,8 +86,9 @@ public class ChunkBoundRenderer {
UploadStream.INSTANCE.commit();
//TODO: NOTE: need to reverse the winding order since we want the back faces of the AABB, not the front
{
//need to reverse the winding order since we want the back faces of the AABB, not the front
glFrontFace(GL_CW);//Reverse winding order
//"reverse depth buffer" it goes from 0->1 where 1 is far away
@@ -132,19 +101,15 @@ public class ChunkBoundRenderer {
glBindVertexArray(RenderService.STATIC_VAO);
glBindFramebuffer(GL_FRAMEBUFFER, this.frameBuffer.id);
this.rasterShader.bind();
glBindBufferBase(GL_UNIFORM_BUFFER, 0, this.uniformBuffer.id);
glBindBuffer(GL_ELEMENT_ARRAY_BUFFER, SharedIndexBuffer.INSTANCE_BB_BYTE.id());
//TODO: BATCH with multiple cubes per instance, this helps fill the pipe and should greatly improve performance of this
//Batch the draws into groups of size 32
int count = this.chunk2idx.size();
int i = 0;
if (count/32 > 0) {
if (count > 32) {
glDrawElementsInstanced(GL_TRIANGLES, 6 * 2 * 3 * 32, GL_UNSIGNED_BYTE, 0, count/32);
i = (count/32)*32;
count -= i;
}
if (count > 0) {
glDrawElementsInstancedBaseInstance(GL_TRIANGLES, 6 * 2 * 3 * count, GL_UNSIGNED_BYTE, 0, 1, i);
if (count%32 != 0) {
glDrawElementsInstancedBaseInstance(GL_TRIANGLES, 6 * 2 * 3 * (count%32), GL_UNSIGNED_BYTE, 0, 1, (count/32)*32);
}
{
@@ -156,6 +121,70 @@ public class ChunkBoundRenderer {
glEnable(GL_CULL_FACE);
glEnable(GL_DEPTH_TEST);
}
this.processUpdates();
}
private void processUpdates() {
if (this.updates.isEmpty()) return;
var iter = this.updates.long2ByteEntrySet().fastIterator();
while (iter.hasNext()) {
var entry = iter.next();
if (entry.getByteValue()==0) continue;
long pos = entry.getLongKey();
if (0<entry.getByteValue()) {
if (this.chunk2idx.containsKey(pos)) {
//Logger.warn("Chunk already in map: " + new ChunkPos(pos));
continue;
}
this.ensureSize1();//Resize if needed
int idx = this.chunk2idx.size();
this.chunk2idx.put(pos, idx);
this.idx2chunk[idx] = pos;
this.put(idx, pos);
} else {
int idx = this.chunk2idx.remove(pos);
if (idx == -1) {
//Logger.warn("Chunk not in map: " + new ChunkPos(pos));
continue;
}
if (idx == this.chunk2idx.size()) {
//Dont need to do anything as heap is already compact
continue;
}
if (this.idx2chunk[idx] != pos) {
throw new IllegalStateException();
}
//Move last entry on heap to this index
long ePos = this.idx2chunk[this.chunk2idx.size()];// since is already removed size is correct end idx
if (this.chunk2idx.put(ePos, idx) == -1) {
throw new IllegalStateException();
}
this.idx2chunk[idx] = ePos;
//Put the end pos into the new idx
this.put(idx, ePos);
}
}
this.updates.clear();
}
private void ensureSize1() {
if (this.chunk2idx.size() < this.idx2chunk.length) return;
int size = (int) (this.idx2chunk.length*1.5);
Logger.info("Resizing chunk position buffer to: " + size);
//Need to resize
var old = this.chunkPosBuffer;
this.chunkPosBuffer = new GlBuffer(size * 8L);
glCopyNamedBufferSubData(old.id, this.chunkPosBuffer.id, 0, 0, old.size());
old.free();
var old2 = this.idx2chunk;
this.idx2chunk = new long[size];
System.arraycopy(old2, 0, this.idx2chunk, 0, old2.length);
//Replace the old buffer with the new one
((AutoBindingShader)this.rasterShader).ssbo(1, this.chunkPosBuffer);
}
private void put(int idx, long pos) {