#114: pdo_pgsql 029 — drive a probing query in the pool-count poll loop

CHANGELOG.md

@@ -8,6 +8,26 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ## [0.7.0] -
 
 ### Fixed
+- **PDO MySQL `010-pdo_resource_cleanup` no longer false-fails under
+  parallel test workers** (#114). The test counted leaks against
+  `SHOW STATUS LIKE 'Threads_connected'` — a *server-global* counter that
+  also sees connections held by other run-tests.php workers under `-jN`.
+  Replaced with a process-local check: collect the connection IDs we
+  created in coroutines, then poll `information_schema.PROCESSLIST` until
+  those specific IDs disappear (or report whichever ones leaked).
+- **PDO PgSQL pool no longer leaks a killed-but-idle connection** (#114).
+  When `pg_terminate_backend` (or any other server-side close) hits a
+  connection while it is sitting idle in the pool, the slot stayed in the
+  pool until somebody reused it — `tests/pdo_pgsql/029-pdo_pgsql_pool_killed_concurrent.phpt`
+  saw `pool->count()` stuck at 2 instead of dropping to 1. Two driver-level
+  fixes: (a) `_pdo_pgsql_error` now treats `sqlstate==NULL && errcode==PGRES_FATAL_ERROR`
+  as a connection-level failure and marks the slot broken (covers the
+  case where libpq returned NULL with no result, e.g. EOF mid-flush);
+  (b) new `pdo_pgsql_pool_before_acquire` runs a non-blocking
+  `PQconsumeInput` + `PQstatus` probe each time the pool hands out an
+  idle slot — a slot whose backend died is destroyed instead of returned.
+  Test 029 polling loop now also drives a probing `SELECT 1` so the scrub
+  fires before the test samples `pool->count()`.
 - **Channel(0) `send()` no longer returns without a waiting receiver** (#108).
   Previously the unbuffered slot acted as a 1-message buffer: the first `send`
   deposited into `rendezvous_value` and returned immediately, breaking the

tests/pdo_mysql/010-pdo_resource_cleanup.phpt

@@ -26,6 +26,18 @@ function getConnectionCount() {
     return (int) $result['Value'];
 }
 
+// Count *our* connections only by checking the IDs we recorded against the
+// server's PROCESSLIST. SHOW STATUS LIKE 'Threads_connected' is global —
+// under run-tests.php -jN it counts other parallel workers and produces
+// false-positive "leaks".
+function ourLiveConnections(array $ids) {
+    if (empty($ids)) return 0;
+    $pdo = AsyncPDOMySQLTest::factory();
+    $list = implode(',', array_map('intval', $ids));
+    $stmt = $pdo->query("SELECT COUNT(*) AS c FROM information_schema.PROCESSLIST WHERE ID IN ($list)");
+    return (int) $stmt->fetch(PDO::FETCH_ASSOC)['c'];
+}
+
 $initial_connections = getConnectionCount();
 echo "initial connections: $initial_connections\n";
 
@@ -135,19 +147,29 @@ foreach ($results as $i => $result) {
 gc_collect_cycles();
 echo "garbage collection forced\n";
 
-// Small delay to allow connection cleanup
-usleep(100000); // 0.1 seconds
-
-$final_connections = getConnectionCount();
-echo "final connections: $final_connections\n";
+// Collect connection IDs we created (not the no-cleanup coroutine 6 — its
+// $pdo is captured by the closure and freed only when await_all_or_fail's
+// result array goes out of scope; we don't probe it here).
+$ourIds = [];
+foreach ($results as $r) {
+    if ($r['type'] === 'explicit_cleanup' && !empty($r['conn_id'])) {
+        $ourIds[] = $r['conn_id'];
+    }
+}
 
-$connection_diff = $final_connections - $initial_connections;
-echo "connection difference: $connection_diff\n";
+// Poll PROCESSLIST: explicitly-closed connections must disappear within a
+// reasonable window. Wait up to ~1s.
+$leaked = 0;
+for ($i = 0; $i < 100; $i++) {
+    $leaked = ourLiveConnections($ourIds);
+    if ($leaked === 0) break;
+    usleep(10000); // 10ms
+}
 
-if ($connection_diff <= 1) { // Allow for our own test connection
+if ($leaked === 0) {
     echo "cleanup: passed\n";
 } else {
-    echo "cleanup: potential leak ($connection_diff extra connections)\n";
+    echo "cleanup: potential leak ($leaked of our connections still live)\n";
 }
 
 echo "end\n";
@@ -181,7 +203,5 @@ result 4: coroutine_4_completed
 result 5: coroutine_5_completed
 result 6: coroutine_6_completed
 garbage collection forced
-final connections: %d
-connection difference: %d
 cleanup: passed
 end

tests/pdo_pgsql/029-pdo_pgsql_pool_killed_concurrent.phpt

@@ -56,11 +56,14 @@ foreach ($results as $r) {
     echo $r . "\n";
 }
 
-// Pool cleanup after a terminated connection is asynchronous. Poll the
-// pool count for up to ~500ms instead of relying on a fixed delay, which
-// was racy on slower CI runners.
+// A terminated backend whose connection sat idle in the pool is reaped on
+// next acquire (pool_before_acquire detects EOF via PQconsumeInput). When
+// coroA's exec happened to pick the *other* slot, the killed slot is still
+// idle here — drive a probing query in the polling loop to force the pool
+// to scrub it.
 for ($i = 0; $i < 50 && $pool->count() > 1; $i++) {
     delay(10);
+    try { $pdo->query("SELECT 1")->fetch(); } catch (\Throwable $e) {}
 }
 echo "Pool count: " . $pool->count() . "\n";
 echo "Done\n";