code wiki / _hdl_build / nx_emu_bench.nx
nx_emu_bench.nx
buildroot/runtime/_hdl_build/nx_emu_bench.nx
about
nx_emu_bench.nx -- MEASURE the sovereign rv64 emulator's throughput (instructions/sec) so the "behind QEMU on perf"
claim is a NUMBER, not a hand-wave (measured-not-asserted). Assembles a compute loop with the sovereign assembler
(nx_rv64_asm), runs it IN-PROCESS on rv64im_min_sim with a high step cap, times the run with sys_now_us, and reports
the measured instr/sec + a bounded honest comparison to QEMU-TCG's ballpark. expect_exit:0
dependencies 11 imports · 0 importers
diagram shows first 10 each side; +1 more imports, +0 more importers in the complete lists below.
imports: nx_syscalls.nxnx_itoa_lib.nxnishi_hdl_primitives.nxrv64im_min_decoder.nxrv64im_min_alu.nxrv64im_min_regfile.nxrv64im_min_csr.nxrv64im_min_clint.nxrv64im_min_uart.nxrv64im_min_sim.nxnx_rv64_asm.nx
imported by: nobody (leaf or entry point)
call flow from main pre-order; caps 40 nodes / depth 6 declared; ↻ = already shown
structs
| none |
consts
| 16 | const BENCH_MAGIC_4096: i64 = 4096 |
| 17 | const BENCH_MAGIC_200000000: i64 = 200000000 |
| 18 | const BENCH_MAGIC_1000000: i64 = 1000000 |
| 19 | const BENCH_MAGIC_16000000: i64 = 16000000 |
| 29 | const BENCH_MEM_BASE: i64 = 0x80000000 |
| 30 | const BENCH_MEM_SIZE: i64 = 65536 |
| 31 | const BENCH_TX_CAP: i64 = 256 |
functions
| 21 | func b_puts(s: *u8) -> i64 { var n: i64=0; while s[n]!=(0 as u8){n=n+1} sys_write(1,s,n); return 0 } |
| 26 | func b_pn(v: i64) -> i64 { nxi_out(v); return 0 } |
| 27 | func ck(name: *u8, c: i64) -> i64 { if c==1 { b_puts(" PASS " as *u8) } else { b_puts(" FAIL " as *u8) } b_puts(name); b_puts("\n" as *u8); return c } |
| 33 | func main() -> i64 |