Log once per switch when rt_knob reads a value other than 0, 1 or ab as on.

perf.c rt_knob reads "0" as off, "ab" as the A/B blocks and everything else as on, without a word: "cache_die = off",
"crash_filter_guard = false" and "one_per_core = 0 ; x" all read as 1 (the audit's probe, Fv/knob.exe, cut rt_knob out
with cfg.c: three 1s and no log line). A player who wrote such a value meant the opposite and had no way to learn it
from the log.

rt_knob now logs "cfg: NX2NAVI_<name> = '<value>' is not 0, 1 or ab: read as 1" the first time it reads such a value
for a switch. The value still reads as 1: reading words as 0 changes what a written setting does and waits for
approval with F-8. Most of the 109 call sites keep the result in a static but some read it per call, so the warned
names are kept in a table of 32 pointers (the names are string literals) filled by compare-and-swap; 0, 1 and ab pay
only one more string compare.

New tests/rt_knob_check.py cuts rt_knob and its table out of perf.c and builds them with stand-ins for the settings and
the log: 0, 1, ab, empty and unset read as before without a line; off, false and "0 ; a note" read as 1 and log once
each however often they are read. docs/building.md describes it. The fault was shown before the change by the audit's
probe; rt_knob_check passes, and perf.c compiles with the dev flags and with -Wall -Wextra without new warnings.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
chrg6
2026-11-13 17:23:30 -05:00
parent 74b9687a8f
commit 7684734b47
3 changed files with 134 additions and 1 deletions

View File

@@ -939,12 +939,32 @@ int rt_ab(void) {
return k == 1 || k == 2;
}
// rt_knob reads "0" as off, "ab" as the A/B blocks and anything else as on, so "off", "false" or "0 ; a note" turn a
// switch on. A value other than 0, 1 and ab is logged once per switch; the names are string literals, so the table
// holds their pointers (a full table logs no more). The value is still read as on: taking such words as off would
// change what a written setting does.
#define KNOB_WARNED 32
static const char *knob_warned[KNOB_WARNED];
static void knob_warn(const char *name, const char *key, const char *value) {
for (int i = 0; i < KNOB_WARNED; i++) {
const char *cur = __atomic_load_n(&knob_warned[i], __ATOMIC_ACQUIRE);
if (!cur && __atomic_compare_exchange_n(&knob_warned[i], &cur, name, 0, __ATOMIC_ACQ_REL, __ATOMIC_ACQUIRE)) {
rt_log("cfg: %s = '%s' is not 0, 1 or ab: read as 1", key, value);
return;
}
if (cur == name) return; // logged already (another thread may have taken this slot for it just now)
}
}
int rt_knob(const char *name, const char *def) {
char key[64];
snprintf(key, sizeof key, "NX2NAVI_%s", name);
const char *e = cfg_get(key);
if (!e || !*e) e = def;
return !strcmp(e, "ab") ? 2 : strcmp(e, "0") != 0;
if (!strcmp(e, "ab")) return 2;
if (strcmp(e, "0") && strcmp(e, "1")) knob_warn(name, key, e);
return strcmp(e, "0") != 0;
}
// NX2NAVI_ONE_PER_CORE=1: the process keeps the first logical processor of each physical core, so busy threads (the