1,106 Commits over 274 Days - 0.17cph!
Tests: perf test to measure ServerProfiler overhead
Going to try a different internal storage approach to make future changes easier (and potentially faster)
Tests: ran the new tests
Merge: from parallel_validatemove
- Removes PlayerCache.ValidPlayers allocs
Tests: took a snapshot on Craggy in editor
Optim: PlayerCache.ValidPlayers no longer allocates garbage
Tests: took snapshot on Craggy in editor to confirm
Merge: from main
Tests: none, no conflicts
Merge: from profiling_improvements
- Reduces overhead of serializing to/from ProtoBuf by not tracking BufferStream calls
Tests: snapshot on Craggy in editor
Update: further filtering of methods
- Dropping BaseEntity.Is* methods that are just HasFlag wrappers
- Dropping new Rust.Data BufferStream and RangeHandle
Should reduce overhead on serialization
Tests: snapshot on Craggy in editor
Merge: from main
Tests: none, no conflicts
Merge: from relationshipmanager_leaks
- Server and Client-side bugfixes for pooling around RelationshipManager types
Tests: local 2 player session with flicking open-closed the contacts screen and with pooling tracking
Merge: from main
Tests: none, no conflicts
Bugfix: return ProtoBuf.PlayerRelationships back to pool after client rpc is executed
Think it was the only PlayerRelationships leak on client
Tests: local session on craggy, opened&closed contacts - saw pool metrics stay stable
Update: using new RelationshipManager.ClearRelations to avoid potential pooling leaks
Tests: none
Bugfix: clear PlayerRelationshipInfo when returning PlayerRelationship to pool
Tests: none, since it turns out that we don't have any code that stresses this path - one more thing to fix then
Bugfix: return PlayerRelationshipInfo back to pool when forgetting a player
2 more places to fix
Tests: none
Bugfix: return to pool ProtoBuf.RelationshipManager.PlayerRelationships + nested types after client rpcs
There's still one more server leak and a client leak
Tests: local 2-player session. Made sure changed code is being stepped through.
Merge: from main
Tests: none, no conflicts
Merge: from reduce_appmarkersellorder_allocs
- Reduces pool spillage and misses of ProtoBuf.AppMarker.SellOrder
Tests: started server on procgen map
Optim: Increase ProtoBuf.AppMarker.SellOrder pool capacity to 2k
- Should reduce creation and spillage of SellOrders
Tests: booted server on procgen.
Merge: from profiling_improvements
Tests: snapshot on craggy in editor
Merge: from main
Tests: none, no conflicts
Update: Further removal of about 10% methods from perf snapshot
- Built using e70c083b
Tests: took a snapshot on craggy
Merge: from listcompare_optim
- ListHashSet can now be pooled and supports Compare
- Optim Networkable.UpdateSubscriptions via ListHashSet
Tests: unit tests
Merge: from main
Tests: none, no conflicts
Clean: simplify code via ?. notation
Tests: unit tests
Optim: Constain List<T>.Compare to only works on List<T>
This removes remaining 4 boxing allocations
Tests: unit tests
Optim: Use ListHashSet<T>.Compare in UpdateSubscriptions and UpdateHighPrioritySubscriptions
Tests: none, trivial change
Update: Network.Visibility.Provider works on ListHashSet
Tests: none, simple changes
Update: Facepunch.Pool now supports ListHashSet
- gave ListHashSet a default ctor
Tests: unit tests
Update: ListHashSet has it's own specialized static Compare method
- Replciated tests from ListExtensionTests for ListHashSet
Tests: ran unit tests
Optim: List.Compare now uses pooled hashsets instead of hashing-and-sorting
2x faster. Now, just need to see if I can get rid of these remaining allocs (just 4 per stable run)
Tests: unit tests
Merge: from profiling_improvements
- Reduce capture size by ~19% by filtering out more methods
Tests: snapshot on craggy in editor
Update: further reduce capture scope by ~19%
- Built from c4679e41
Tests: snapshot in editor on Craggy
Merge: from parallel_validatemove
- Bugfix: GamePhysics.OverlapCapsules no longer skips 0-length-capsule queries
- Additional unit tests
Tests: unit tests + staging demo playback
Tests: add GamePhysics.CheckShhere, CheckCapsule and related consistency tests
Contains divergence cases that are constantly warned about - this is to document current behavior
Tests: ran unit tests
Update: Adding GamePhysics.CheckSpheres
Tests: unit tests (next submit)
Merge: from main
Tests: none, no conflicts
Update: Made GamePhysics.CheckCapsules and related consistent with CheckCapsule around sphere queries
- Added a defaulted param(true) that controls whether we should scan for sphere-like capsule queries or not
- Added a couple comments to clarify things
This adds ~1ms on 10k player perf test scenario, but should silence Unity's warnings
Tests: unit tests + staging demo playback
Update: GamePhysics.OverlapCapsules and OverlapSpheres now check for invalid commands and report errros
Tests: unit tests
Clean: added a comment expanding the reason for previous test
Tests: none, trivial change
▉▅▄█▆: ▊▌▋▉█▊▇▇▄▌▋▅▄▄▇▍▆▍▍▉█▊▆▆▅▇▆▉▉▋▆▅ - ▆▌▄▍ ▅▉▌▅ ▇▄▋▅▆▄▉▇▊ ▇-▇▆▅▌▌▄ ▉█▇▅▇▌▌ ▇▉▉▌▅
▍▇▌▊▄ ▋▅█▍ ▄▄▅▌▍▌▅█▄ ▄█ ▍▄█▊ ▅ ▋▌▉▆█▊▊▅▇ ▇▆▉ ▊▆█▋ █▅▉▇ ▆▆█▍ ▊▌▉ ▉█▅▍ ▍▅ █▆▍ ▊▅
▉▇▇▍█: ▉▊▉ ▋▌▊ ▋▉▌▌▇
Bugfix: ServerDemoPlayer - skip trying to spawn invalid prefabs from the server demo stream
Tests: played back server demo
Update: make ServerDemoPlayer complient with recent Protobuf changes
Tests: played back a staging server demo
Merge: from main
Tests: none, no conflicts
Merge: from loading_entity_leak
- Fix zombie ProtoBuf.Entity objects never being returned to the pool after loading a game save
Tests: loaded craggy save, editor default procgen save and staging server save
Bugfix: don't leak ProtoBuf.Entity during loading of a save
Loading a 4.3k staging server save drops ~129k objects (almost all get spilled, but that's okay after a load)
Tests: loaded craggy save, loaded default editor procgen save and stagign server procgen save
Merge: from fix_treetoolrenderer
- Brings back tree rendering on editor start with a scene with trees without triggering database refresh
Tests: variosu editor open scenarios, various scene switching scenarios with domain reload
Update: TreeToolRenderer is now a static class instead of an EditorTool
State survives domain reload (somehow).
Tests: started editor on swamp_a, craggy, swamp_b - rendered okay. Triggered a couple domain reloads via code changes - still rendered with active scene + scene switching
Bugfix: bring back tree rendering on first scene opening
- Defer tool initialization to first scene load to avoid triggering asset refresh in InitializeOnLoad
Tests: closed and opened editor with various starting scenes. Switched scenes.
Merge: from ioentity_slacklevels_pooling
- Fixes IO entity spilling List<float> on save
Tests: build a couple water tanks and checked pool.print_memory
Bugfix: use pooled List<float> during IOEntity's slacklevels saving
Avoids inflating List<float> in the pool, and eventually spilling them, leading to GC pressure.
Tests: on craggy build a couple water tanks - checked that "List`1[System.Single" pooled active objects doesn't go negative
