code wiki / (root) / nx_aead_bench.nx

nx_aead_bench.nx

buildroot/runtime/nx_aead_bench.nx

2061 B54 linesdepth 5pulls 7 transitivereach 0 importersview sourcekind benchtopic aead
docsdependenciesstructsconstsfunctions

about

nx_aead_bench.nx -- reliable cipher cost split (run in WSL where the monotonic clock is real ns, not the Windows ~15ms-granular FILETIME). Times ChaCha20 and Poly1305 over one large buffer to find which half of the AEAD decrypt is the measured TLS-transfer bottleneck, and the raw MB/s of each (hardware-up target). expect_exit: 0

dependencies 3 imports · 0 importers

nx_syscalls.nx nx_chacha20.nx nx_poly1305.nx nx_aead_bench.nx

imports: nx_syscalls.nxnx_chacha20.nxnx_poly1305.nx

imported by: nobody (leaf or entry point)

call flow from main pre-order; caps 40 nodes / depth 6 declared; ↻ = already shown

main sys_mmap sys_now_ms sys_mmap ↻ sys_clock_gettime_mono chacha20_encrypt sys_mmap ↻ _chacha20_set_consts_key load_u32_le load_u32_le ↻ chacha20_perm_core store_u32_le poly1305_mac sys_mmap ↻ poly1305_clamp p_load_u32_le p_store_u32_le b_ps sys_write b_pn sys_mmap ↻ sys_write ↻ sys_exit

structs

none

consts

11const K_MAGIC_2970000: i64 = 2970000
12const K_MAGIC_1000000: i64 = 1000000

functions

14func b_ps(s: *u8, n: i64) -> i64 { sys_write(1, s, n); return 0 }
called by 1: main calls 1: sys_write
15func b_pn(v: i64) -> i64
called by 1: main calls 2: sys_mmapsys_write
26func main() -> i64