1 files changed, 40 insertions, 30 deletions

diff --git a/src/backend/postmaster/postmaster.c b/src/backend/postmaster/postmaster.c
index 2552327d904..71198b72c83 100644
--- a/src/backend/postmaster/postmaster.c
+++ b/src/backend/postmaster/postmaster.c
@@ -4811,7 +4811,7 @@ retry:
 				(errmsg_internal("could not register process for wait: error code %lu",
 								 GetLastError())));
 
-	/* Don't close pi.hProcess here - the wait thread needs access to it */
+	/* Don't close pi.hProcess here - waitpid() needs access to it */
 
 	CloseHandle(pi.hThread);
 
@@ -6421,36 +6421,21 @@ ShmemBackendArrayRemove(Backend *bn)
 static pid_t
 waitpid(pid_t pid, int *exitstatus, int options)
 {
+	win32_deadchild_waitinfo *childinfo;
+	DWORD		exitcode;
 	DWORD		dwd;
 	ULONG_PTR	key;
 	OVERLAPPED *ovl;
 
-	/*
-	 * Check if there are any dead children. If there are, return the pid of
-	 * the first one that died.
-	 */
-	if (GetQueuedCompletionStatus(win32ChildQueue, &dwd, &key, &ovl, 0))
+	/* Try to consume one win32_deadchild_waitinfo from the queue. */
+	if (!GetQueuedCompletionStatus(win32ChildQueue, &dwd, &key, &ovl, 0))
 	{
-		*exitstatus = (int) key;
-		return dwd;
+		errno = EAGAIN;
+		return -1;
 	}
 
-	return -1;
-}
-
-/*
- * Note! Code below executes on a thread pool! All operations must
- * be thread safe! Note that elog() and friends must *not* be used.
- */
-static void WINAPI
-pgwin32_deadchild_callback(PVOID lpParameter, BOOLEAN TimerOrWaitFired)
-{
-	win32_deadchild_waitinfo *childinfo = (win32_deadchild_waitinfo *) lpParameter;
-	DWORD		exitcode;
-
-	if (TimerOrWaitFired)
-		return;					/* timeout. Should never happen, since we use
-								 * INFINITE as timeout value. */
+	childinfo = (win32_deadchild_waitinfo *) key;
+	pid = childinfo->procId;
 
 	/*
 	 * Remove handle from wait - required even though it's set to wait only
@@ -6466,13 +6451,11 @@ pgwin32_deadchild_callback(PVOID lpParameter, BOOLEAN TimerOrWaitFired)
 		write_stderr("could not read exit code for process\n");
 		exitcode = 255;
 	}
-
-	if (!PostQueuedCompletionStatus(win32ChildQueue, childinfo->procId, (ULONG_PTR) exitcode, NULL))
-		write_stderr("could not post child completion status\n");
+	*exitstatus = exitcode;
 
 	/*
-	 * Handle is per-process, so we close it here instead of in the
-	 * originating thread
+	 * Close the process handle.  Only after this point can the PID can be
+	 * recycled by the kernel.
 	 */
 	CloseHandle(childinfo->procHandle);
 
@@ -6482,7 +6465,34 @@ pgwin32_deadchild_callback(PVOID lpParameter, BOOLEAN TimerOrWaitFired)
 	 */
 	free(childinfo);
 
-	/* Queue SIGCHLD signal */
+	return pid;
+}
+
+/*
+ * Note! Code below executes on a thread pool! All operations must
+ * be thread safe! Note that elog() and friends must *not* be used.
+ */
+static void WINAPI
+pgwin32_deadchild_callback(PVOID lpParameter, BOOLEAN TimerOrWaitFired)
+{
+	/* Should never happen, since we use INFINITE as timeout value. */
+	if (TimerOrWaitFired)
+		return;
+
+	/*
+	 * Post the win32_deadchild_waitinfo object for waitpid() to deal with. If
+	 * that fails, we leak the object, but we also leak a whole process and
+	 * get into an unrecoverable state, so there's not much point in worrying
+	 * about that.  We'd like to panic, but we can't use that infrastructure
+	 * from this thread.
+	 */
+	if (!PostQueuedCompletionStatus(win32ChildQueue,
+									0,
+									(ULONG_PTR) lpParameter,
+									NULL))
+		write_stderr("could not post child completion status\n");
+
+	/* Queue SIGCHLD signal. */
 	pg_queue_signal(SIGCHLD);
 }
 #endif							/* WIN32 */