[PATCH stalld 08/36] tests/helpers: Fix stop_stalld() timeout and shutdown logic

[Wander Lairson Costa <wander@redhat.com>]

stop_stalld() has two bugs. The polling loop after SIGKILL uses
sleep 0.1 with an integer counter, so the nominal timeout of 10
actually gives only 1 second. Additionally, SIGTERM only gets
0.2 seconds before escalating to SIGKILL, which is insufficient
for stalld to perform graceful cleanup such as removing its
pidfile or flushing logs.

Restructure stop_stalld() into two clear phases. First send
SIGTERM and poll for graceful exit with 1-second intervals for
up to 5 seconds, giving stalld time to clean up. Then escalate
to SIGKILL only if the process is still alive, and poll again
for up to 5 seconds with correct timeout arithmetic. Document
the guarantee that the process is dead before the function
returns so callers do not need post-stop sleeps.

Signed-off-by: Wander Lairson Costa <wander@redhat.com>
---
 tests/helpers/test_helpers.sh | 29 +++++++++++++++++++----------
 1 file changed, 19 insertions(+), 10 deletions(-)

diff --git a/tests/helpers/test_helpers.sh b/tests/helpers/test_helpers.sh
index f6d6ce2..1750423 100755
--- a/tests/helpers/test_helpers.sh
+++ b/tests/helpers/test_helpers.sh
@@ -365,24 +365,33 @@ start_stalld() {
 }
 
 # Stop stalld
+# Sends SIGTERM and polls for graceful exit, then escalates to
+# SIGKILL if needed. Guarantees the process is dead before
+# returning so callers do not need post-stop sleeps.
 stop_stalld() {
 	if [ -n "${STALLD_PID}" ]; then
 		if kill -0 ${STALLD_PID} 2>/dev/null; then
-			# Try graceful shutdown first
+			# Try graceful shutdown first (SIGTERM)
 			kill ${STALLD_PID} 2>/dev/null || true
-			# Give it a moment to exit gracefully
-			sleep 0.2
-			# Force kill if still running
-			if kill -0 ${STALLD_PID} 2>/dev/null; then
-				kill -9 ${STALLD_PID} 2>/dev/null || true
-			fi
-			# Poll for process termination (don't use wait - might not be a child)
-			local timeout=10
+
+			# Poll for graceful exit (up to 5 seconds)
+			local timeout=5
 			local elapsed=0
 			while kill -0 ${STALLD_PID} 2>/dev/null && [ ${elapsed} -lt ${timeout} ]; do
-				sleep 0.1
+				sleep 1
 				elapsed=$((elapsed + 1))
 			done
+
+			# Escalate to SIGKILL if still running
+			if kill -0 ${STALLD_PID} 2>/dev/null; then
+				kill -9 ${STALLD_PID} 2>/dev/null || true
+				# Poll for forced termination (up to 5 seconds)
+				elapsed=0
+				while kill -0 ${STALLD_PID} 2>/dev/null && [ ${elapsed} -lt ${timeout} ]; do
+					sleep 1
+					elapsed=$((elapsed + 1))
+				done
+			fi
 		fi
 		STALLD_PID=""
 	fi
-- 
2.53.0