Stabilize test on Windows (no-issue-check)

distrib/AppImage/.DirIcon

@@ -1 +0,0 @@
-gnatstudio.png

testsuite/tests/U716-009.lsp.task_labels/test.py

@@ -7,26 +7,11 @@ from gs_utils.internal.utils import *
 
 @run_test_driver
 def test_driver():
-    yield wait_tasks()
     b = GPS.EditorBuffer.get(GPS.File("m.adb"))
     b.select(GPS.EditorLocation(b, 6, 45), GPS.EditorLocation(b, 6, 45))
-    yield timeout(500)
+    yield wait_tasks()
     GPS.execute_action("Find All References")
-    task_hud_label = get_widget_by_name("task_hud_label")
 
-    label = ""
-    count = 0
-    # wait_until_true() has a granularity of 500ms, so it might miss
-    # the label update if it's shorter than this granularity. So use
-    # a manual loop instead.
-    while label == "":
-        label = task_hud_label.get_text()
-        if "[Ada]" in label:
-            break
-        count += 1
-        if count > 200:
-            break
-        yield timeout(10)
-
-    gps_assert(label, "[Ada] querying references",
-               "Wrong task label after 'find all references'")
+    tasks_labels = [task.label() for task in GPS.Task.list()]
+    gps_assert("[Ada] querying references" in tasks_labels, True,
+               "Wrong task label after 'find all references': %s" % str(tasks_labels))