fix(release): don't false-abort flash on a resumed subscription; tidy #12 console comments (#93)

check_subscription_log (both ota-release.sh and esp32-release.sh) died when the
matter-server log had no fresh '(Re-)Subscription succeeded' line in the last 15m.
But the flash gate already re-interviews the node and polls it back to available,
and matter-server only marks a node available once its subscription is up -- so
availability already asserts the subscription (#64). matter-server sometimes
RESUMES a subscription after the re-interview without logging a new line (seen on
the 2026-07-23 reject flip), which false-aborted a healthy flash mid-run. Warn
instead of die when the line is absent; the primary availability gate still holds.

Also refresh the #12 checksum-mismatch counter comments in both diag consoles
(esp32 diag_console.cpp + ameba hisense_diag_console.h): the RX checksum verify
now REJECTS a mismatch (skips parse + counts a link-miss), so the tally is the
count of corrupt 0x66 frames dropped, not the old log-only observation.

Bumps firmware/src/version.txt 1.3.23 -> 1.3.24 (ameba console comment lives in
firmware/src; CI requires the int to increase). Comment-only for the firmware --
nodes 14/62 stay on 1.3.23, no reflash needed.

Assisted-by: AI

Author: AndrewDemsDS

firmware/esp32-matter/main/diag_console.cpp

@@ -154,10 +154,11 @@ static int cmd_features(int, char **)
 static int cmd_poll(int, char **)
 {
     LOCK(); HisenseState s = s_snap; bool h = s_have; uint32_t f = s_frames; UNLOCK();
-    // #12 (log-only): checksum-mismatch tally must stay 0 on real traffic before the RX checksum
-    // verify is allowed to gate parsing / the link-miss counter. Also surface heap here. The
-    // counter lives in the driver TU with no lock of its own; a word-sized read is atomic on
-    // Xtensa, so reading it outside s_mtx is safe (at worst one cycle stale) for a diagnostic.
+    // #12: the RX checksum verify now REJECTS a mismatch (skips the parse + counts a link-miss),
+    // so this tally is the count of corrupt 0x66 frames dropped since boot; it should stay at/near
+    // 0 on a healthy bus. Also surface heap here. The counter lives in the driver TU with no lock
+    // of its own; a word-sized read is atomic on Xtensa, so reading it outside s_mtx is safe (at
+    // worst one cycle stale) for a diagnostic.
     printf("checksum mismatches: %u  |  heap free=%u min_free=%u\r\n",
            (unsigned) hisense_checksum_mismatch_count(),
            (unsigned) esp_get_free_heap_size(), (unsigned) esp_get_minimum_free_heap_size());

firmware/scripts/esp32-release.sh

@@ -275,9 +275,17 @@ check_subscription_log() {
     [ -n "$line" ] && break
     sleep 10
   done
-  [ -n "$line" ] \
-    || die "no '(Re-)Subscription succeeded' for node $node in the last 15m of the matter-server log -- subscription is broken (#64, docs/10 §16)"
-  say "  matter-server log confirms: ${line:0:120}"
+  if [ -n "$line" ]; then
+    say "  matter-server log confirms: ${line:0:120}"
+  else
+    # The flash gate already re-interviewed the node and polled it back to available, and
+    # matter-server only marks a node available once its subscription is up -- so availability IS
+    # the subscription assertion (#64). matter-server sometimes RESUMES a subscription after the
+    # re-interview without logging a fresh '(Re-)Subscription succeeded' line (seen on the 2026-07-23
+    # reject flip), so a missing line here is not proof of a break. Warn, do not die: the primary
+    # gate already passed, and a false die aborts a healthy flash mid-run.
+    say "  no fresh '(Re-)Subscription succeeded' for node $node in 15m -- availability after re-interview already asserted the subscription (#64); matter-server likely resumed it without a new line. OK."
+  fi
 }
 flash() {
   load_env

firmware/scripts/ota-release.sh

@@ -581,9 +581,17 @@ check_subscription_log() {
     [ -n "$line" ] && break
     sleep 10
   done
-  [ -n "$line" ] \
-    || die "no '(Re-)Subscription succeeded' for node $node in the last 15m of the matter-server log -- subscription is broken (#64, docs/10 §16)"
-  say "  matter-server log confirms: ${line:0:120}"
+  if [ -n "$line" ]; then
+    say "  matter-server log confirms: ${line:0:120}"
+  else
+    # The flash gate already re-interviewed the node and polled it back to available, and
+    # matter-server only marks a node available once its subscription is up -- so availability IS
+    # the subscription assertion (#64). matter-server sometimes RESUMES a subscription after the
+    # re-interview without logging a fresh '(Re-)Subscription succeeded' line (seen on the 2026-07-23
+    # reject flip), so a missing line here is not proof of a break. Warn, do not die: the primary
+    # gate already passed, and a false die aborts a healthy flash mid-run.
+    say "  no fresh '(Re-)Subscription succeeded' for node $node in 15m -- availability after re-interview already asserted the subscription (#64); matter-server likely resumed it without a new line. OK."
+  fi
 }
 flash() {
   load_env

firmware/src/sdk-edits/hisense_diag_console.h

@@ -113,10 +113,11 @@ static void diag_cmd_poll(int sock)
 {
     char b[HISENSE_DIAG_BUF];
 
-    // #12 (log-only, shared driver): the RX checksum-mismatch tally must stay 0 on real traffic
-    // before the verify is allowed to gate parsing / link-miss. Shown here for parity with the
-    // ESP32 console; heap watermark is in `sys`. (parse_status already rejects a bad checksum,
-    // so this is belt-and-suspenders visibility for parse_features/parse_faults.)
+    // #12 (shared driver): the RX checksum verify now REJECTS a mismatch (skips the parse + counts
+    // a link-miss), so this tally is the count of corrupt 0x66 frames dropped since boot; it should
+    // stay at/near 0 on a healthy bus. Shown here for parity with the ESP32 console; heap watermark
+    // is in `sys`. (parse_status already rejected bad status frames; the reject now also covers
+    // parse_features/parse_faults.)
     snprintf(b, sizeof(b), "checksum mismatches: %u\r\n",
              (unsigned) hisense_checksum_mismatch_count());
     diag_say(sock, b);

firmware/src/version.txt

@@ -1 +1 @@
-1.3.23
+1.3.24