code wiki / (root) / nx_parexec_gate.nx

nx_parexec_gate.nx

buildroot/runtime/nx_parexec_gate.nx

3414 B77 linesdepth 5pulls 6 transitivereach 0 importersview sourcekind gate/proof
docsdependenciesstructsconstsfunctions

about

nx_parexec_gate.nx -- proves the PARALLEL executor is CORRECT: forking one worker per partition over shared memory yields byte-identical results to the sequential run, for several worker counts, including a worker count that does not divide the row count. Timing is nondeterministic and lives in the `bench` verb; this gate proves only correctness -- that real forked children over shared memory do not lose, double, or corrupt a single row. D001 verdict via nx_gate_verdict. expect_exit: 0

dependencies 2 imports · 0 importers

nx_gate_verdict.nx nx_parexec.nx nx_parexec_gate.nx

imports: nx_gate_verdict.nxnx_parexec.nx

imported by: nobody (leaf or entry point)

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

main gv_ctr sys_mmap gv_head gv_puts sys_write pg_frame1 cf_encoded_bytes cf_slen cf_align8 cf_encode cf_align8 ↻ de_run_agg cf_nrows sys_mmap ↻ de_bound de_agg_range cf_col cf_data_off de_merge pe_agg_parallel cf_nrows ↻ sys_mmap_shared sys_mmap ↻ de_bound ↻ sys_fork de_agg_range ↻ sys_exit sys_wait4 de_merge ↻ gv_check gv_puts ↻ pg_eq gv_verdict gv_puts ↻ gv_num sys_mmap ↻ sys_write ↻ sys_munmap gv_journal

structs

none

consts

none

functions

9func pg_eq(a: i64, b: i64) -> i64 { if a == b { return 1 } return 0 }
called by 1: main
10func pg_frame1(a: *i64, nm: *u8, n: i64) -> *u8
called by 1: main calls 2: cf_encoded_bytescf_encode
21func main() -> i64