papermc/Spigot-Server-Patches/0531-Improve-Chunk-Status-Transition-Speed.patch

124 lines
6.4 KiB
Diff
Raw Normal View History

Improve Chunk Status Transition Speed When a chunk is loaded from disk that has already been generated, the server has to promote the chunk through the system to reach it's current desired status level. This results in every single status transition going from the main thread to the world gen threads, only to discover it has no work it actually needs to do.... and then it returns back to main. This back and forth costs a lot of time and can really delay chunk loads when the server is under high TPS due to their being a lot of time in between chunk load times, as well as hogs up the chunk threads from doing actual generation and light work. Additionally, the whole task system uses a lot of CPU on the server threads anyways. So by optimizing status transitions for status's that are already complete, we can run them to the desired level while on main thread (where it has to happen anyways) instead of ever jumping to world gen thread. This will improve chunk loading effeciency to be reduced down to the following scenario / path: 1) MAIN: Chunk Requested, Load Request sent to ChunkTaskManager / IO Queue 2) IO: Once position in queue comes, submit read IO data and schedule to chunk task thread 3) CHUNK: Once IO is loaded and position in queue comes, deserialize the chunk data, process conversions, submit to main queue 4) MAIN: next Chunk Task process (Mid Tick or End Of Tick), load chunk data into world (POI, main thread tasks) 5) MAIN: process status transitions all the way to LIGHT, light schedules Threaded task 6) SERVER: Light tasks register light enablement for chunk and any lighting needing to be done 7) MAIN: Task returns to main, finish processing to FULL/TICKING status Previously would have hopped to SERVER around 12+ times there extra.
2020-05-30 05:12:18 +00:00
From 0000000000000000000000000000000000000000 Mon Sep 17 00:00:00 2001
From: Aikar <aikar@aikar.co>
Date: Fri, 29 May 2020 23:32:14 -0400
Subject: [PATCH] Improve Chunk Status Transition Speed
When a chunk is loaded from disk that has already been generated,
the server has to promote the chunk through the system to reach
it's current desired status level.
This results in every single status transition going from the main thread
to the world gen threads, only to discover it has no work it actually
needs to do.... and then it returns back to main.
This back and forth costs a lot of time and can really delay chunk loads
when the server is under high TPS due to their being a lot of time in
between chunk load times, as well as hogs up the chunk threads from doing
actual generation and light work.
Additionally, the whole task system uses a lot of CPU on the server threads anyways.
So by optimizing status transitions for status's that are already complete,
we can run them to the desired level while on main thread (where it has
to happen anyways) instead of ever jumping to world gen thread.
This will improve chunk loading effeciency to be reduced down to the following
scenario / path:
1) MAIN: Chunk Requested, Load Request sent to ChunkTaskManager / IO Queue
2) IO: Once position in queue comes, submit read IO data and schedule to chunk task thread
3) CHUNK: Once IO is loaded and position in queue comes, deserialize the chunk data, process conversions, submit to main queue
4) MAIN: next Chunk Task process (Mid Tick or End Of Tick), load chunk data into world (POI, main thread tasks)
5) MAIN: process status transitions all the way to LIGHT, light schedules Threaded task
6) SERVER: Light tasks register light enablement for chunk and any lighting needing to be done
7) MAIN: Task returns to main, finish processing to FULL/TICKING status
Previously would have hopped to SERVER around 12+ times there extra.
diff --git a/src/main/java/net/minecraft/server/PlayerChunk.java b/src/main/java/net/minecraft/server/PlayerChunk.java
index 52da9c78c0cc2b21533a1477a25a3dda492700e4..69899c100dd86c6c4795013364472336327ce036 100644
Improve Chunk Status Transition Speed When a chunk is loaded from disk that has already been generated, the server has to promote the chunk through the system to reach it's current desired status level. This results in every single status transition going from the main thread to the world gen threads, only to discover it has no work it actually needs to do.... and then it returns back to main. This back and forth costs a lot of time and can really delay chunk loads when the server is under high TPS due to their being a lot of time in between chunk load times, as well as hogs up the chunk threads from doing actual generation and light work. Additionally, the whole task system uses a lot of CPU on the server threads anyways. So by optimizing status transitions for status's that are already complete, we can run them to the desired level while on main thread (where it has to happen anyways) instead of ever jumping to world gen thread. This will improve chunk loading effeciency to be reduced down to the following scenario / path: 1) MAIN: Chunk Requested, Load Request sent to ChunkTaskManager / IO Queue 2) IO: Once position in queue comes, submit read IO data and schedule to chunk task thread 3) CHUNK: Once IO is loaded and position in queue comes, deserialize the chunk data, process conversions, submit to main queue 4) MAIN: next Chunk Task process (Mid Tick or End Of Tick), load chunk data into world (POI, main thread tasks) 5) MAIN: process status transitions all the way to LIGHT, light schedules Threaded task 6) SERVER: Light tasks register light enablement for chunk and any lighting needing to be done 7) MAIN: Task returns to main, finish processing to FULL/TICKING status Previously would have hopped to SERVER around 12+ times there extra.
2020-05-30 05:12:18 +00:00
--- a/src/main/java/net/minecraft/server/PlayerChunk.java
+++ b/src/main/java/net/minecraft/server/PlayerChunk.java
@@ -56,6 +56,13 @@ public class PlayerChunk {
this.playersInChunkTickRange = this.chunkMap.playerChunkTickRangeMap.getObjectsInRange(key);
}
// Paper end - optimise isOutsideOfRange
+ // Paper start - optimize chunk status progression without jumping through thread pool
+ public boolean canAdvanceStatus() {
+ ChunkStatus status = getChunkHolderStatus();
+ IChunkAccess chunk = getAvailableChunkNow();
+ return chunk != null && (status == null || chunk.getChunkStatus().isAtLeastStatus(getNextStatus(status)));
+ }
+ // Paper end
// Paper start - no-tick view distance
public final Chunk getSendingChunk() {
diff --git a/src/main/java/net/minecraft/server/PlayerChunkMap.java b/src/main/java/net/minecraft/server/PlayerChunkMap.java
index 3139d55bb801b6c433de7274be4a1d0c0fd1eef7..98adcb6390105a183d66975ed9659906b609ce08 100644
Improve Chunk Status Transition Speed When a chunk is loaded from disk that has already been generated, the server has to promote the chunk through the system to reach it's current desired status level. This results in every single status transition going from the main thread to the world gen threads, only to discover it has no work it actually needs to do.... and then it returns back to main. This back and forth costs a lot of time and can really delay chunk loads when the server is under high TPS due to their being a lot of time in between chunk load times, as well as hogs up the chunk threads from doing actual generation and light work. Additionally, the whole task system uses a lot of CPU on the server threads anyways. So by optimizing status transitions for status's that are already complete, we can run them to the desired level while on main thread (where it has to happen anyways) instead of ever jumping to world gen thread. This will improve chunk loading effeciency to be reduced down to the following scenario / path: 1) MAIN: Chunk Requested, Load Request sent to ChunkTaskManager / IO Queue 2) IO: Once position in queue comes, submit read IO data and schedule to chunk task thread 3) CHUNK: Once IO is loaded and position in queue comes, deserialize the chunk data, process conversions, submit to main queue 4) MAIN: next Chunk Task process (Mid Tick or End Of Tick), load chunk data into world (POI, main thread tasks) 5) MAIN: process status transitions all the way to LIGHT, light schedules Threaded task 6) SERVER: Light tasks register light enablement for chunk and any lighting needing to be done 7) MAIN: Task returns to main, finish processing to FULL/TICKING status Previously would have hopped to SERVER around 12+ times there extra.
2020-05-30 05:12:18 +00:00
--- a/src/main/java/net/minecraft/server/PlayerChunkMap.java
+++ b/src/main/java/net/minecraft/server/PlayerChunkMap.java
@@ -88,6 +88,7 @@ public class PlayerChunkMap extends IChunkLoader implements PlayerChunk.d {
public final WorldServer world;
private final LightEngineThreaded lightEngine;
private final IAsyncTaskHandler<Runnable> executor;
+ final java.util.concurrent.Executor mainInvokingExecutor; // Paper
public final ChunkGenerator<?> chunkGenerator;
private final Supplier<WorldPersistentData> l; public final Supplier<WorldPersistentData> getWorldPersistentDataSupplier() { return this.l; } // Paper - OBFHELPER
private final VillagePlace m;
@@ -285,6 +286,15 @@ public class PlayerChunkMap extends IChunkLoader implements PlayerChunk.d {
this.world = worldserver;
this.chunkGenerator = chunkgenerator;
this.executor = iasynctaskhandler;
+ // Paper start - optimize chunk status progression without jumping through thread pool
+ this.mainInvokingExecutor = (run) -> {
+ if (MCUtil.isMainThread()) {
+ run.run();
+ } else {
+ iasynctaskhandler.execute(run);
+ }
+ };
+ // Paper end
ThreadedMailbox<Runnable> threadedmailbox = ThreadedMailbox.a(executor, "worldgen");
iasynctaskhandler.getClass();
@@ -724,7 +734,7 @@ public class PlayerChunkMap extends IChunkLoader implements PlayerChunk.d {
Improve Chunk Status Transition Speed When a chunk is loaded from disk that has already been generated, the server has to promote the chunk through the system to reach it's current desired status level. This results in every single status transition going from the main thread to the world gen threads, only to discover it has no work it actually needs to do.... and then it returns back to main. This back and forth costs a lot of time and can really delay chunk loads when the server is under high TPS due to their being a lot of time in between chunk load times, as well as hogs up the chunk threads from doing actual generation and light work. Additionally, the whole task system uses a lot of CPU on the server threads anyways. So by optimizing status transitions for status's that are already complete, we can run them to the desired level while on main thread (where it has to happen anyways) instead of ever jumping to world gen thread. This will improve chunk loading effeciency to be reduced down to the following scenario / path: 1) MAIN: Chunk Requested, Load Request sent to ChunkTaskManager / IO Queue 2) IO: Once position in queue comes, submit read IO data and schedule to chunk task thread 3) CHUNK: Once IO is loaded and position in queue comes, deserialize the chunk data, process conversions, submit to main queue 4) MAIN: next Chunk Task process (Mid Tick or End Of Tick), load chunk data into world (POI, main thread tasks) 5) MAIN: process status transitions all the way to LIGHT, light schedules Threaded task 6) SERVER: Light tasks register light enablement for chunk and any lighting needing to be done 7) MAIN: Task returns to main, finish processing to FULL/TICKING status Previously would have hopped to SERVER around 12+ times there extra.
2020-05-30 05:12:18 +00:00
return either.mapLeft((list) -> {
return (Chunk) list.get(list.size() / 2);
});
- }, this.executor);
+ }, this.mainInvokingExecutor); // Paper
}
@Nullable
@@ -1074,7 +1084,7 @@ public class PlayerChunkMap extends IChunkLoader implements PlayerChunk.d {
IChunkAccess ichunkaccess = (IChunkAccess) optional.get();
if (ichunkaccess.getChunkStatus().b(chunkstatus)) {
- CompletableFuture completablefuture1;
+ CompletableFuture<Either<IChunkAccess, PlayerChunk.Failure>> completablefuture1; // Paper
if (chunkstatus == ChunkStatus.LIGHT) {
completablefuture1 = this.b(playerchunk, chunkstatus);
@@ -1090,7 +1100,7 @@ public class PlayerChunkMap extends IChunkLoader implements PlayerChunk.d {
Improve Chunk Status Transition Speed When a chunk is loaded from disk that has already been generated, the server has to promote the chunk through the system to reach it's current desired status level. This results in every single status transition going from the main thread to the world gen threads, only to discover it has no work it actually needs to do.... and then it returns back to main. This back and forth costs a lot of time and can really delay chunk loads when the server is under high TPS due to their being a lot of time in between chunk load times, as well as hogs up the chunk threads from doing actual generation and light work. Additionally, the whole task system uses a lot of CPU on the server threads anyways. So by optimizing status transitions for status's that are already complete, we can run them to the desired level while on main thread (where it has to happen anyways) instead of ever jumping to world gen thread. This will improve chunk loading effeciency to be reduced down to the following scenario / path: 1) MAIN: Chunk Requested, Load Request sent to ChunkTaskManager / IO Queue 2) IO: Once position in queue comes, submit read IO data and schedule to chunk task thread 3) CHUNK: Once IO is loaded and position in queue comes, deserialize the chunk data, process conversions, submit to main queue 4) MAIN: next Chunk Task process (Mid Tick or End Of Tick), load chunk data into world (POI, main thread tasks) 5) MAIN: process status transitions all the way to LIGHT, light schedules Threaded task 6) SERVER: Light tasks register light enablement for chunk and any lighting needing to be done 7) MAIN: Task returns to main, finish processing to FULL/TICKING status Previously would have hopped to SERVER around 12+ times there extra.
2020-05-30 05:12:18 +00:00
return this.b(playerchunk, chunkstatus);
}
}
- }, this.executor);
+ }, this.mainInvokingExecutor).thenComposeAsync(CompletableFuture::completedFuture, this.mainInvokingExecutor); // Paper - optimize chunk status progression without jumping through thread pool - ensure main
Improve Chunk Status Transition Speed When a chunk is loaded from disk that has already been generated, the server has to promote the chunk through the system to reach it's current desired status level. This results in every single status transition going from the main thread to the world gen threads, only to discover it has no work it actually needs to do.... and then it returns back to main. This back and forth costs a lot of time and can really delay chunk loads when the server is under high TPS due to their being a lot of time in between chunk load times, as well as hogs up the chunk threads from doing actual generation and light work. Additionally, the whole task system uses a lot of CPU on the server threads anyways. So by optimizing status transitions for status's that are already complete, we can run them to the desired level while on main thread (where it has to happen anyways) instead of ever jumping to world gen thread. This will improve chunk loading effeciency to be reduced down to the following scenario / path: 1) MAIN: Chunk Requested, Load Request sent to ChunkTaskManager / IO Queue 2) IO: Once position in queue comes, submit read IO data and schedule to chunk task thread 3) CHUNK: Once IO is loaded and position in queue comes, deserialize the chunk data, process conversions, submit to main queue 4) MAIN: next Chunk Task process (Mid Tick or End Of Tick), load chunk data into world (POI, main thread tasks) 5) MAIN: process status transitions all the way to LIGHT, light schedules Threaded task 6) SERVER: Light tasks register light enablement for chunk and any lighting needing to be done 7) MAIN: Task returns to main, finish processing to FULL/TICKING status Previously would have hopped to SERVER around 12+ times there extra.
2020-05-30 05:12:18 +00:00
}
}
@@ -1201,6 +1211,12 @@ public class PlayerChunkMap extends IChunkLoader implements PlayerChunk.d {
Improve Chunk Status Transition Speed When a chunk is loaded from disk that has already been generated, the server has to promote the chunk through the system to reach it's current desired status level. This results in every single status transition going from the main thread to the world gen threads, only to discover it has no work it actually needs to do.... and then it returns back to main. This back and forth costs a lot of time and can really delay chunk loads when the server is under high TPS due to their being a lot of time in between chunk load times, as well as hogs up the chunk threads from doing actual generation and light work. Additionally, the whole task system uses a lot of CPU on the server threads anyways. So by optimizing status transitions for status's that are already complete, we can run them to the desired level while on main thread (where it has to happen anyways) instead of ever jumping to world gen thread. This will improve chunk loading effeciency to be reduced down to the following scenario / path: 1) MAIN: Chunk Requested, Load Request sent to ChunkTaskManager / IO Queue 2) IO: Once position in queue comes, submit read IO data and schedule to chunk task thread 3) CHUNK: Once IO is loaded and position in queue comes, deserialize the chunk data, process conversions, submit to main queue 4) MAIN: next Chunk Task process (Mid Tick or End Of Tick), load chunk data into world (POI, main thread tasks) 5) MAIN: process status transitions all the way to LIGHT, light schedules Threaded task 6) SERVER: Light tasks register light enablement for chunk and any lighting needing to be done 7) MAIN: Task returns to main, finish processing to FULL/TICKING status Previously would have hopped to SERVER around 12+ times there extra.
2020-05-30 05:12:18 +00:00
return CompletableFuture.completedFuture(Either.right(playerchunk_failure));
});
}, (runnable) -> {
+ // Paper start - optimize chunk status progression without jumping through thread pool
+ if (playerchunk.canAdvanceStatus()) {
+ this.mainInvokingExecutor.execute(runnable);
+ return;
+ }
+ // Paper end
this.mailboxWorldGen.a(ChunkTaskQueueSorter.a(playerchunk, runnable)); // CraftBukkit - decompile error
});
}