system tests: safer container-stop signaling

Having a container spin-wait on a /stop file, then exit, is
unsafe: 'podman exec $ctr touch /stop' can get sucked into
container cleanup before the exec terminates, resulting in
the podman-exec failing and hence the test failing.

Most existing instances of this pattern are unnecessary.
Replace those with just 'podman rm -f'.

When necessary, use a variety of safer alternatives.

Re-Closes: #10825 (already closed; this addresses remaining cases)

Signed-off-by: Ed Santiago <santiago@redhat.com>

test/system/030-run.bats

@@ -347,7 +347,7 @@ echo $rand        |   0 | $rand
 # #6991 : /etc/passwd is modifiable
 @test "podman run : --userns=keep-id: passwd file is modifiable" {
     skip_if_not_rootless "--userns=keep-id only works in rootless mode"
-    run_podman run -d --userns=keep-id --cap-add=dac_override $IMAGE sh -c 'while ! test -e /tmp/stop; do sleep 0.1; done'
+    run_podman run -d --userns=keep-id --cap-add=dac_override $IMAGE top
     cid="$output"
 
     # Assign a UID that is (a) not in our image /etc/passwd and (b) not
@@ -368,8 +368,7 @@ echo $rand        |   0 | $rand
     is "$output" "newuser3:x:$uid:999:$gecos:/home/newuser3:/bin/sh" \
        "newuser3 added to /etc/passwd in container"
 
-    run_podman exec $cid touch /tmp/stop
-    run_podman wait $cid
+    run_podman rm -f -t0 $cid
 }
 
 # For #7754: json-file was equating to 'none'

test/system/075-exec.bats

@@ -27,8 +27,10 @@ load helpers
     run_podman 127 exec $cid /no/such/command
     is "$output" ".*such file or dir"   "podman exec /no/such/command"
 
-    # Done
-    run_podman exec $cid rm -f /$rand_filename
+    # Done. Tell the container to stop.
+    # The '-d' is because container exit is racy: the exec process itself
+    # could get caught and killed by cleanup, causing this step to exit 137
+    run_podman exec -d $cid rm -f /$rand_filename
 
     run_podman wait $cid
     is "$output" "0"   "output from podman wait (container exit code)"
@@ -59,7 +61,7 @@ load helpers
 # Issue #4785 - piping to exec statement - fixed in #4818
 # Issue #5046 - piping to exec truncates results (actually a conmon issue)
 @test "podman exec - cat from stdin" {
-    run_podman run -d $IMAGE sh -c 'while [ ! -e /stop ]; do sleep 0.1;done'
+    run_podman run -d $IMAGE top
     cid="$output"
 
     echo_string=$(random_string 20)
@@ -80,26 +82,21 @@ load helpers
     is "${output% *}" "$expect " "SHA of file in container"
 
     # Clean up
-    run_podman exec $cid touch /stop
-    run_podman wait $cid
-    run_podman rm $cid
+    run_podman rm -f -t0 $cid
 }
 
 # #6829 : add username to /etc/passwd inside container if --userns=keep-id
 @test "podman exec - with keep-id" {
     skip_if_not_rootless "--userns=keep-id only works in rootless mode"
     # Multiple --userns options confirm command-line override (last one wins)
-    run_podman run -d --userns=private --userns=keep-id $IMAGE sh -c \
-               "echo READY;while [ ! -f /tmp/stop ]; do sleep 1; done"
+    run_podman run -d --userns=private --userns=keep-id $IMAGE sh -c 'echo READY;top'
     cid="$output"
     wait_for_ready $cid
 
     run_podman exec $cid id -un
     is "$output" "$(id -un)" "container is running as current user"
 
-    run_podman exec --user=$(id -un) $cid touch /tmp/stop
-    run_podman wait $cid
-    run_podman rm $cid
+    run_podman rm -f -t0 $cid
 }
 
 # #11496: podman-remote loses output

test/system/130-kill.bats

@@ -24,10 +24,9 @@ load helpers
     run_podman run -d ${opt_log_driver} $IMAGE sh -c \
         "for i in ${signals[*]}; do trap \"echo got: \$i\" \$i; done;
         echo READY;
-        while ! test -e /stop; do sleep 0.05; done;
+        while ! test -e /stop; do sleep 0.1; done;
         echo DONE"
-    # Ignore output regarding pulling/processing container images
-    cid=$(echo "$output" | tail -1)
+    cid="$output"
 
     # Run 'logs -f' on that container, but run it in the background with
     # redirection to a named pipe from which we (foreground job) read
@@ -71,8 +70,10 @@ load helpers
     kill_and_check -SIGUSR1 10
     kill_and_check  SIGUSR2 12
 
-    # Done. Tell the container to stop, and wait for final DONE
-    run_podman exec $cid touch /stop
+    # Done. Tell the container to stop, and wait for final DONE.
+    # The '-d' is because container exit is racy: the exec process itself
+    # could get caught and killed by cleanup, causing this step to exit 137
+    run_podman exec -d $cid touch /stop
     read -t 5 -u 5 done || die "Timed out waiting for DONE from container"
     is "$done" "DONE" "final log message from container"
 

test/system/260-sdnotify.bats

@@ -110,7 +110,7 @@ function _assert_mainpid_is_conmon() {
     run_podman run -d --name sdnotify_conmon_c \
                --sdnotify=conmon \
                $IMAGE \
-               sh -c 'printenv NOTIFY_SOCKET;echo READY;while ! test -f /stop;do sleep 0.1;done'
+               sh -c 'printenv NOTIFY_SOCKET;echo READY;sleep 999'
     cid="$output"
     wait_for_ready $cid
 
@@ -135,9 +135,7 @@ READY=1" "sdnotify sent MAINPID and READY"
     _assert_mainpid_is_conmon "$output"
 
     # Done. Stop container, clean up.
-    run_podman exec $cid touch /stop
-    run_podman wait $cid
-    run_podman rm $cid
+    run_podman rm -f -t0 $cid
     _stop_socat
 }
 
@@ -153,7 +151,7 @@ READY=1" "sdnotify sent MAINPID and READY"
     _start_socat
 
     run_podman run -d --sdnotify=container $SYSTEMD_IMAGE \
-               sh -c 'printenv NOTIFY_SOCKET; echo READY; while ! test -f /stop;do sleep 0.1;done;systemd-notify --ready'
+               sh -c 'trap "touch /stop" SIGUSR1;printenv NOTIFY_SOCKET; echo READY; while ! test -f /stop;do sleep 0.1;done;systemd-notify --ready'
     cid="$output"
     wait_for_ready $cid
 
@@ -174,8 +172,8 @@ READY=1" "sdnotify sent MAINPID and READY"
 
     is "$output" "MAINPID=$mainPID" "Container is not ready yet, so we only know the main PID"
 
-    # Done. Stop container, clean up.
-    run_podman exec $cid touch /stop
+    # Done. Tell container to stop itself, and clean up
+    run_podman kill -s USR1 $cid
     run_podman wait $cid
 
     wait_for_file $_SOCAT_LOG
@@ -205,10 +203,18 @@ spec:
   - command:
     - /bin/sh
     - -c
-    - 'while :; do if test -e /tearsinrain; then exit 0; fi; sleep 1; done'
+    - 'while :; do if test -e /rain/tears; then exit 0; fi; sleep 1; done'
     image: $IMAGE
     name: test
     resources: {}
+    volumeMounts:
+    - mountPath: /rain:z
+      name: test-mountdir
+  volumes:
+  - hostPath:
+      path: $PODMAN_TMPDIR
+      type: Directory
+    name: test-mountdir
 EOF
 
     # The name of the service container is predictable: the first 12 characters
@@ -228,7 +234,7 @@ EOF
     main_pid="$output"
 
     # Tell pod to finish, then wait for all containers to stop
-    run_podman exec test_pod-test touch /tearsinrain
+    touch $PODMAN_TMPDIR/tears
     run_podman container wait $service_container test_pod-test
 
     # Make sure the containers have the correct policy.

test/system/320-system-df.bats

@@ -82,8 +82,7 @@ Size           |   ~12.*MB |         0B |            0B
 
 @test "podman system df - with active containers and volumes" {
     run_podman run    -v /myvol1 --name c1 $IMAGE true
-    run_podman run -d -v /myvol2 --name c2 $IMAGE \
-               sh -c 'while ! test -e /stop; do sleep 0.1;done'
+    run_podman run -d -v /myvol2 --name c2 $IMAGE top
 
     run_podman system df --format '{{ .Type }}:{{ .Total }}:{{ .Active }}'
     is "${lines[0]}" "Images:1:1"        "system df : Images line"
@@ -124,8 +123,7 @@ Size           |   ~12.*MB |         0B |            0B
     run_podman system df --format '{{.Reclaimable}}'
     is "${lines[0]}" "0B (0%)" "cannot reclaim image data as it's still used by the containers"
 
-    run_podman exec c2 touch /stop
-    run_podman wait c2
+    run_podman stop c2
 
     # Create a second image by committing a container.
     run_podman container commit -q c1

test/system/410-selinux.bats

@@ -99,12 +99,11 @@ function check_label() {
                --security-opt seccomp=unconfined \
                --security-opt label=type:spc_t \
                --security-opt label=level:s0 \
-               $IMAGE sh -c 'while test ! -e /stop; do sleep 0.1; done'
+               $IMAGE top
     run_podman inspect --format='{{ .HostConfig.SecurityOpt }}' myc
     is "$output" "[label=type:spc_t,label=level:s0 seccomp=unconfined]" \
       "'podman inspect' preserves all --security-opts"
 
-    run_podman exec myc touch /stop
     run_podman rm -t 0 -f myc
 }
 

test/system/520-checkpoint.bats

@@ -211,7 +211,9 @@ function teardown() {
     is "$output" "$cid" "podman container restore"
 
     # Signal the container to continue; this is where the 1-2-3s will come from
-    run_podman exec $cid rm /wait
+    # The '-d' is because container exit is racy: the exec process itself
+    # could get caught and killed by cleanup, causing this step to exit 137
+    run_podman exec -d $cid rm /wait
 
     # Wait for the container to stop
     run_podman wait $cid