physics: Jolt's floor sized for what moves - pairs 4096, contacts 2048 and a 4 MB temp heap by default (phys_open_sized for more), and a body past max_bodies or a step past the pairs or contacts is Mem.over, never a quiet -1; world.ludic split from shapes.ludic

An empty world at the game's size held 51839 KB of Jolt's heap (200000 bodies, 65536 pairs, 20480
contacts, a 32 MB temp heap), the jolt= floor under every walk. Pairs and contacts come from what
moves - drops, boats, a few walkers - and a valley's still things make none. Now 11743 KB at 200000
bodies and 8461 KB at the 98304 the game opens for (jolt_floor_test holds it under 16 MB). The temp heap
still falls back to malloc for a step that wants more.

lib/macos-arm64 rebuilt; lib/windows-x64 needs native/build.sh on the PC (jph_world_new's new sizes).

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Orkun ÇAKILKAYA 2026-09-29 16:18:42 +03:00
parent 031fcad641
commit b4af12025d
11 changed files with 95 additions and 52 deletions

View file

@ -0,0 +1,25 @@
# jolt_floor_test.ludic - what an empty world costs in Jolt's own heap at the game's size (plan 25):
# the floor under every body, from the body table, the broad phase, the contact caches and the temp heap
import "ludic.physics"
import "ludic.base"
program JoltFloorTest {
numbers float
function floor_at(physics_st: mut PhysicsState, bodies: int) -> long {
expect(phys_open(physics_st, 64, 1))
phys_close(physics_st)
let before = phys_heap_bytes()
expect(phys_open(physics_st, bodies, 0))
let f = phys_heap_bytes() - before
phys_close(physics_st)
return f
}
# it was 51839 KB at 200000 bodies with 65536 pairs, 20480 contacts and a 32 MB temp heap
test "an empty world at the game's size costs under 16 MB" (physics_st: mut PhysicsState) {
let a = floor_at(physics_st, 200000)
let b = floor_at(physics_st, 98304)
print(`jolt floor: {a / long(1024)} KB at 200000 bodies, {b / long(1024)} KB at 98304`)
expect(b < long(16 * 1024 * 1024))
expect(a > b)
}
}