lune/tests/ffi/README.md

tests/ffi

Requirements

gcc for library compiling (for external-*)

Test Results

External tests

• external_math
• external_pointer
• external_print
• external_struct
• external_closure

Luau-side

• isInteger

• cast

• pretty_print

  need box, ref test

• write_boundary

  Failed, need fix

• read_boundary

  Failed, need fix

Benchmark Results

Note: LuaJit's os.clock function returns process CPU time (used) which much smaller then Luau's os.clock output. In this benchmark, luau uses 'time.h' instead of os.clock. See utility/proc_clock

benchmark/external_call

Target external c function

int add(int a, int b) {
    return a + b;
}

bench_scale = 1000000

Lune ffi

Command: cargo run run tests/ffi/benchmark/external_call Command: cargo run --profile=release run tests/ffi/benchmark/external_call

• Device1-Linux-PVE
  Lune release target: 0.205127 (sec)
  Lune dev target: 1.556489 (sec)

  Commit: ddf0c4c

• Device2-Windows-11 Lune release target: 0.1875 (sec)
  Lune dev target: ? SEGFUALT (sec)

  Commit: ddf0c4c

LuaJit ffi

Command: luajit tests/ffi/benchmark/external_call/luajit.lua

• Device1-Linux-PVE: 0.001682 (sec)

  LuaJIT 2.1.1727870382
  (flags = JIT ON SSE3 SSE4.1 BMI2 fold cse dce fwd dse narrow loop abc sink fuse)

Deno ffi

Command: deno run --unstable-ffi --allow-ffi ./tests/ffi/benchmark/external_call/deno.ts

• Device1-Linux-PVE: 0.006384 (sec)

  Deno 1.46.3 (v8 = 12.9.202.5-rusty)

Sysinformation

• Device1-Linux-PVE

  CPU: AMD Ryzen 5 7600 (12) @ 5.1
  MEM: 61898MiB 5600 MT/s
  KERNEL: 6.8.12-2-pve (Proxmox VE 8.2.7 x86_64)

• Device2-Windows-11

  CPU: AMD Ryzen 5 7600 (4) @ 3.800GHz
  MEM: 12250MiB 5600 MT/s
  KERNEL: 10.0.22631 (Windows 11 x86_64)
  HOST: QEMU Standard PC (Q35 + ICH9, 2009)