feat(nfc): Poll the NFC tag through the soak instead of sampling it once

nfc_polling.sh wraps blantons_nfc_validate.py in a background loop and
keeps score: start/stop/status/tail/report/clear/fg. Script A takes a
30s baseline, Script B starts the poll for the soak, Script C stops it
with the other monitors and reports.

report only reads the log, so it can be run mid-soak without disturbing
anything - and it says RUNNING so a snapshot is not read as the final
answer. Each round appends a machine-readable RESULT line, which keeps
the tally independent of the tool's wording.

ERROR is counted apart from FAIL. A round where the tool printed neither
[PASS] nor [FAIL] - crashed, missing, hung past the timeout - measured
nothing, and that is a different fault from a tag that would not read.
Rolling them together hides whichever one you are not looking for.

Two habits from earlier bugs are baked in: tail does not follow (tail -f
never returns and a macro waiting on the prompt hangs there), and the .py
is called through python3 rather than ./ (it arrives mode 100644 and a
USB copy carries no execute bit).

Also fixes Script C copying "nfc_polling.log" to the job dir when the
file is "nfc_poll.log". The cp failed, nothing checked it, and the run
finished looking fine - you would find out at the desk, with a job
archive that had no NFC log in it.

Script A -> V1.1.2.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_014RetWKFZFG1ZHcitQAyhwM
This commit is contained in:
2026-08-28 22:41:02 +08:00
co-authored by Claude Opus 5
parent c3b0d49145
commit bc2b529c3a
8 changed files with 1784 additions and 6 deletions
+36 -1
View File
@@ -1,6 +1,6 @@
; =============================================================================
; Script A for Blanton
; Version : V1.1.1
; Version : V1.1.2
; Date : 2026-08-28
; Author : ETWen
; =============================================================================
@@ -138,6 +138,29 @@
; A run that ended badly leaves jobs holding CPU and a
; session open; the next run used to inherit both, and
; the stress numbers were quietly wrong because of it
; V1.1.2 2026-08-28 Blanton_Script/nfc_polling.sh Add background NFC tag polling
; start / stop / status / tail / report / clear / fg
; wraps blantons_nfc_validate.py; start clears the log
; first, so a report covers this run only
; report works WHILE polling - it only reads the log,
; and says RUNNING so a snapshot is not read as final
; Each round appends a machine-readable RESULT line, so
; the tally never depends on the tool's wording
; PASS / FAIL come from the tool's own summary; ERROR
; means it produced neither, which is a different fault
; from a tag that would not read - counted separately
; tail does NOT follow (tail -f never returns, and a
; macro waiting on the prompt would hang there)
; Runs the .py through python3, not ./ - the file is
; mode 100644 and a USB copy carries no execute bit
; Blanton_Script/blantons_nfc_validate.py Vendor NFC/NCI validation tool (unmodified)
; ScriptA NFC baseline: start -> status -> 30s -> stop -> report
; ScriptB Start NFC polling for the soak, then status
; ScriptC Stop NFC polling with the other monitors, report it
; next to the mgmt ping log, and copy nfc_poll.log to
; the job dir for the USB archive
; Fix: that cp said nfc_polling.log - the file is
; nfc_poll.log, so the NFC log never reached the USB
; =============================================================================
include "config.ttl"
@@ -306,6 +329,18 @@ sendln "cat ~/Blanton_Script/log/mgmt_ping.log"
wait prompt_sonic_root
sendln "~/Blanton_Script/mgmt_ping_monitor.sh clear"
; ========== NFC Test ==========
wait prompt_sonic_root
sendln "~/Blanton_Script/nfc_polling.sh start"
wait prompt_sonic_root
sendln "~/Blanton_Script/nfc_polling.sh status"
wait prompt_sonic_root
sendln "sleep 30"
wait prompt_sonic_root
sendln "~/Blanton_Script/nfc_polling.sh stop"
wait prompt_sonic_root
sendln "~/Blanton_Script/nfc_polling.sh report"
; ========== 100G Port Status ==========
wait prompt_sonic_root
@@ -100,6 +100,12 @@ wait prompt_sonic_root
pause 3
sendln "./Blanton_Script/mgmt_ping_monitor.sh status"
; ========== NFC Test ==========
wait prompt_sonic_root
sendln "~/Blanton_Script/nfc_polling.sh start"
wait prompt_sonic_root
sendln "~/Blanton_Script/nfc_polling.sh status"
; ========== 100G Port ==========
wait prompt_sonic_root
@@ -25,6 +25,8 @@ wait prompt_sonic_root
sendln "~/Blanton_Script/mgmt_ping_monitor.sh status"
wait prompt_sonic_root
sendln "~/Blanton_Script/mgmt_ping_monitor.sh stop"
wait prompt_sonic_root
sendln "~/Blanton_Script/nfc_polling.sh stop"
;=============================================================================================================
; 100G PRBS STOP
@@ -92,6 +94,10 @@ sendln "cat /host/hw-eval/jobs/qfx5252-stress-usb.log"
wait prompt_sonic_root
sendln "cat ~/Blanton_Script/log/mgmt_ping.log"
; ========== NFC Test result ==========
wait prompt_sonic_root
sendln "~/Blanton_Script/nfc_polling.sh report"
; ========== 100G Port ==========
wait prompt_sonic_root
sendln "show interfaces status Ethernet513,Ethernet514"
@@ -201,6 +207,8 @@ wait prompt_sonic_root
sendln "cp ./Blanton_Script/log/bmc_poll.log /host/hw-eval/jobs/"
wait prompt_sonic_root
sendln "cp ./Blanton_Script/log/mgmt_ping.log /host/hw-eval/jobs/"
wait prompt_sonic_root
sendln "cp ./Blanton_Script/log/nfc_poll.log /host/hw-eval/jobs/"
wait prompt_sonic_root
getdate ts_date "%Y%m%d"
File diff suppressed because it is too large Load Diff
@@ -0,0 +1,314 @@
#!/usr/bin/env bash
# =============================================================================
# nfc_polling.sh -- keep reading the NFC tag in the background, and tally it
# =============================================================================
# Version History:
# V1.0.0 2026-08-28 First version. start/stop/status/tail/report/clear/fg,
# wraps blantons_nfc_validate.py. 'report' works while the
# poll is still running.
# =============================================================================
#
# Each round runs blantons_nfc_validate.py once and appends a machine-readable
# verdict line, so 'report' never has to parse the tool's prose:
#
# [2026-08-28 15:04:11] NFC poll #000001 START
# ... the tool's own output ...
# [2026-08-28 15:04:14] NFC poll #000001 RESULT=PASS tag=046F7BD2142290
#
# The verdict comes from the tool's own [PASS] / [FAIL] summary line. If neither
# appears the round is ERROR, not FAIL -- "the tool did not run" and "the tag was
# not read" are different problems and get counted separately.
#
# Usage:
# ./nfc_polling.sh start # background, CLEARS the log first
# ./nfc_polling.sh status # running? pid, rounds so far, log size
# ./nfc_polling.sh tail [n] # last n lines (default 50) -- does NOT follow
# ./nfc_polling.sh report # PASS/FAIL tally; safe to run WHILE polling
# ./nfc_polling.sh stop
# ./nfc_polling.sh clear # delete the log (must be stopped)
# ./nfc_polling.sh fg # run in the foreground, Ctrl-C to stop
#
# Environment overrides:
# NFC_INTERVAL_SEC=10 NFC_TIMEOUT_SEC=60 NFC_MAX_ROUNDS=0 (0 = forever)
# NFC_ARGS="test --debug" NFC_LOG_DIR=<dir> NFC_BIN=<python3>
# =============================================================================
# --- User Configurable Section ----------------------------------------------
NFC_INTERVAL_SEC="${NFC_INTERVAL_SEC:-10}" # gap between rounds
NFC_TIMEOUT_SEC="${NFC_TIMEOUT_SEC:-60}" # kill a round that hangs (0 = never)
NFC_MAX_ROUNDS="${NFC_MAX_ROUNDS:-0}" # 0 = until 'stop'
# Called through python3 on purpose: blantons_nfc_validate.py arrives from git
# (mode 100644) and a Windows/USB copy does not carry the execute bit either, so
# "./blantons_nfc_validate.py" would fail with Permission denied on a fresh DUT.
NFC_PY="${NFC_PY:-python3}"
NFC_SCRIPT_NAME="${NFC_SCRIPT_NAME:-blantons_nfc_validate.py}"
# --debug is kept because that is the output that has been seen on the bench.
# It costs ~1.5 KB per round (~13 MB/day at a 10s interval), which is fine for a
# soak; drop it to "test" if you want a smaller log.
NFC_ARGS="${NFC_ARGS:-test --debug}"
NFC_LOG_DIR="${NFC_LOG_DIR:-}" # empty = <script dir>/log
NFC_LOG_NAME="${NFC_LOG_NAME:-nfc_poll.log}"
# -----------------------------------------------------------------------------
SCRIPT_NAME="$(basename -- "${BASH_SOURCE[0]}")"
SCRIPT_DIR="$(cd -- "$(dirname -- "${BASH_SOURCE[0]}")" && pwd)"
SCRIPT_PATH="${SCRIPT_DIR}/${SCRIPT_NAME}"
NFC_SCRIPT="${NFC_SCRIPT:-${SCRIPT_DIR}/${NFC_SCRIPT_NAME}}"
[ -n "${NFC_LOG_DIR}" ] || NFC_LOG_DIR="${SCRIPT_DIR}/log"
LOG_FILE="${NFC_LOG_DIR}/${NFC_LOG_NAME}"
PID_FILE="${NFC_LOG_DIR}/${NFC_LOG_NAME%.log}.pid"
STOP=0
SLEEP_PID=""
ts() { date '+%Y-%m-%d %H:%M:%S'; }
die() { printf '[ERROR] %s\n' "$*" >&2; exit 1; }
on_signal() {
STOP=1
[ -n "${SLEEP_PID}" ] && kill "${SLEEP_PID}" 2>/dev/null
return 0
}
# sleep that a signal can cut short, so 'stop' does not wait out the interval
interruptible_sleep() {
local sec="$1"
[ "${sec}" -le 0 ] 2>/dev/null && return 0
sleep "${sec}" &
SLEEP_PID=$!
wait "${SLEEP_PID}" 2>/dev/null
SLEEP_PID=""
return 0
}
is_running() {
local pid
[ -f "${PID_FILE}" ] || return 1
pid="$(cat "${PID_FILE}" 2>/dev/null)"
[ -n "${pid}" ] || return 1
kill -0 "${pid}" 2>/dev/null || return 1
printf '%s' "${pid}"
return 0
}
# --- one round ---------------------------------------------------------------
# Appends START, the tool output, then a RESULT line. Everything about a round
# lands in the log even when the tool dies, so 'report' can still account for it.
run_one_round() {
local n="$1" out rc verdict tag reason
printf '[%s] NFC poll #%06d START\n' "$(ts)" "${n}"
if [ -n "${NFC_TIMEOUT_SEC}" ] && [ "${NFC_TIMEOUT_SEC}" -gt 0 ] 2>/dev/null \
&& command -v timeout >/dev/null 2>&1; then
out="$(timeout "${NFC_TIMEOUT_SEC}" "${NFC_PY}" "${NFC_SCRIPT}" ${NFC_ARGS} 2>&1)"; rc=$?
else
out="$("${NFC_PY}" "${NFC_SCRIPT}" ${NFC_ARGS} 2>&1)"; rc=$?
fi
printf '%s\n' "${out}"
# The tool's own summary decides. Exit status is only the fallback, because
# a tool that never printed a summary did not measure anything.
if printf '%s\n' "${out}" | grep -q '^\[PASS\]'; then
verdict="PASS"
elif printf '%s\n' "${out}" | grep -q '^\[FAIL\]'; then
verdict="FAIL"
elif [ "${rc}" -eq 124 ]; then
verdict="ERROR"; reason="round exceeded NFC_TIMEOUT_SEC=${NFC_TIMEOUT_SEC}s"
else
verdict="ERROR"; reason="no [PASS]/[FAIL] summary, exit=${rc}"
fi
tag="$(printf '%s\n' "${out}" | sed -n 's/.*NFCID1=\([0-9A-Fa-f]*\).*/\1/p' | tail -1)"
[ -n "${reason}" ] || reason="$(printf '%s\n' "${out}" | grep -m1 '^\[\(PASS\|FAIL\)\]' | cut -c8-)"
printf '[%s] NFC poll #%06d RESULT=%s tag=%s reason=%s\n' \
"$(ts)" "${n}" "${verdict}" "${tag:--}" "${reason:--}"
}
main_loop() {
local n=0
trap on_signal INT TERM
printf '[%s] ===== nfc_polling start : %s %s (interval %ss) =====\n' \
"$(ts)" "${NFC_PY}" "${NFC_SCRIPT}" "${NFC_INTERVAL_SEC}"
while [ "${STOP}" -eq 0 ]; do
n=$(( n + 1 ))
run_one_round "${n}"
if [ "${NFC_MAX_ROUNDS}" -gt 0 ] 2>/dev/null && [ "${n}" -ge "${NFC_MAX_ROUNDS}" ]; then
break
fi
[ "${STOP}" -eq 0 ] && interruptible_sleep "${NFC_INTERVAL_SEC}"
done
printf '[%s] ===== nfc_polling stop : %d round(s) =====\n' "$(ts)" "${n}"
}
# --- sub-commands ------------------------------------------------------------
do_start() {
local pid
pid="$(is_running)" && die "already running (pid=${pid})"
[ -r "${NFC_SCRIPT}" ] || die "cannot read ${NFC_SCRIPT}"
command -v "${NFC_PY}" >/dev/null 2>&1 || die "${NFC_PY} not found"
mkdir -p "${NFC_LOG_DIR}" || die "cannot create ${NFC_LOG_DIR}"
# start always begins a fresh log, so a report is about THIS run only
: > "${LOG_FILE}" || die "cannot write ${LOG_FILE}"
if command -v setsid >/dev/null 2>&1; then
setsid "${SCRIPT_PATH}" __daemon >/dev/null 2>&1 &
else
nohup "${SCRIPT_PATH}" __daemon >/dev/null 2>&1 &
disown 2>/dev/null
fi
sleep 1
pid="$(is_running)" || die "start failed, check ${LOG_FILE}"
printf 'started (pid=%s)\nlog: %s\n' "${pid}" "${LOG_FILE}"
}
do_daemon() {
mkdir -p "${NFC_LOG_DIR}" || die "cannot create ${NFC_LOG_DIR}"
exec >>"${LOG_FILE}" 2>&1
printf '%d\n' "$$" > "${PID_FILE}"
trap 'rm -f "${PID_FILE}"' EXIT
main_loop
}
do_stop() {
local pid i
pid="$(is_running)" || { printf 'not running\n'; rm -f "${PID_FILE}"; return 0; }
kill -TERM -"${pid}" 2>/dev/null || kill -TERM "${pid}" 2>/dev/null
for (( i = 0; i < 20; i++ )); do
kill -0 "${pid}" 2>/dev/null || break
sleep 0.5
done
if kill -0 "${pid}" 2>/dev/null; then
printf 'SIGTERM ignored, sending SIGKILL\n'
kill -KILL -"${pid}" 2>/dev/null || kill -KILL "${pid}" 2>/dev/null
fi
rm -f "${PID_FILE}"
printf 'stopped (pid=%s)\n' "${pid}"
}
# counts straight out of the log, so it is the same answer running or stopped
_tally() {
local f="$1"
[ -r "${f}" ] || { printf '0 0 0 0\n'; return; }
awk '
/RESULT=PASS/ { p++ }
/RESULT=FAIL/ { f++ }
/RESULT=ERROR/ { e++ }
/ START$/ { s++ }
END { printf "%d %d %d %d\n", s+0, p+0, f+0, e+0 }
' "${f}"
}
do_status() {
local pid started passed failed errored size
if pid="$(is_running)"; then
printf 'status : running (pid=%s)\n' "${pid}"
else
printf 'status : stopped\n'
fi
printf 'log : %s\n' "${LOG_FILE}"
if [ -r "${LOG_FILE}" ]; then
size="$(du -h "${LOG_FILE}" 2>/dev/null | cut -f1)"
read -r started passed failed errored <<< "$(_tally "${LOG_FILE}")"
printf 'size : %s\n' "${size:-?}"
printf 'rounds : %s started (PASS %s / FAIL %s / ERROR %s)\n' \
"${started}" "${passed}" "${failed}" "${errored}"
printf 'last : %s\n' "$(grep -a 'RESULT=' "${LOG_FILE}" | tail -1)"
else
printf 'size : (no log yet)\n'
fi
}
do_report() {
local pid started passed failed errored inflight verdict first last
[ -r "${LOG_FILE}" ] || die "no log at ${LOG_FILE} -- has it been started?"
read -r started passed failed errored <<< "$(_tally "${LOG_FILE}")"
inflight=$(( started - passed - failed - errored ))
echo "===== NFC polling report ====="
printf 'log : %s\n' "${LOG_FILE}"
if pid="$(is_running)"; then
printf 'state : RUNNING (pid=%s) -- this is a snapshot, the count is still moving\n' "${pid}"
else
printf 'state : stopped\n'
fi
first="$(grep -a -m1 ' START$' "${LOG_FILE}" | sed -n 's/^\[\([^]]*\)\].*/\1/p')"
last="$(grep -a 'RESULT=' "${LOG_FILE}" | tail -1 | sed -n 's/^\[\([^]]*\)\].*/\1/p')"
printf 'window : %s -> %s\n' "${first:-?}" "${last:-?}"
echo ""
printf 'rounds : %d\n' "${started}"
printf 'PASS : %d\n' "${passed}"
printf 'FAIL : %d\n' "${failed}"
printf 'ERROR : %d (tool did not produce a verdict)\n' "${errored}"
[ "${inflight}" -gt 0 ] && printf 'in flight: %d (round started, no result yet)\n' "${inflight}"
if [ "${started}" -eq 0 ]; then
verdict="NO DATA"
elif [ "${failed}" -gt 0 ] || [ "${errored}" -gt 0 ]; then
verdict="FAIL"
else
verdict="PASS"
fi
printf 'RESULT : %s\n' "${verdict}"
echo ""
echo "--- tags seen ---"
grep -a 'RESULT=PASS' "${LOG_FILE}" \
| sed -n 's/.* tag=\([0-9A-Fa-f]*\) .*/\1/p' \
| sort | uniq -c | sort -rn | awk '{printf " %-20s %s\n", $2, $1}'
grep -aq 'RESULT=PASS.* tag=[0-9A-Fa-f]' "${LOG_FILE}" || echo " (none)"
if [ "$(( failed + errored ))" -gt 0 ]; then
echo ""
echo "--- last 10 FAIL / ERROR ---"
grep -aE 'RESULT=(FAIL|ERROR)' "${LOG_FILE}" | tail -10 | sed 's/^/ /'
fi
echo ""
[ "${verdict}" = "FAIL" ] && return 2
return 0
}
usage() {
cat <<USAGE
${SCRIPT_NAME} - background NFC tag polling (wraps ${NFC_SCRIPT_NAME})
start run in the background; CLEARS the log first
stop stop it
status running? pid, rounds so far, last result
tail [n] last n lines of the log (default 50) -- does NOT follow
report PASS/FAIL tally -- safe to run WHILE polling
clear delete the log (must be stopped)
fg run in the foreground, Ctrl-C to stop
interval ${NFC_INTERVAL_SEC}s per-round timeout ${NFC_TIMEOUT_SEC}s rounds ${NFC_MAX_ROUNDS} (0 = forever)
log ${LOG_FILE}
A round is PASS or FAIL from the tool's own summary line; ERROR means it
produced neither, which is a different problem from a tag that would not read.
USAGE
}
case "${1:-}" in
start) do_start ;;
stop) do_stop ;;
status) do_status ;;
report) do_report ;;
# deliberately NOT 'tail -f': that never returns, and a TTL macro waiting on
# the prompt would hang there forever.
tail) [ -r "${LOG_FILE}" ] || die "no log at ${LOG_FILE}"
tail -n "${2:-50}" "${LOG_FILE}" ;;
clear)
is_running >/dev/null && die "still running, stop it first"
rm -f "${LOG_FILE}"
printf 'log cleared\n'
;;
fg) main_loop ;;
__daemon) do_daemon ;;
-h|--help|"") usage ;;
*) printf '[ERROR] unknown sub-command: %s\n\n' "$1" >&2; usage; exit 1 ;;
esac