tests/benchmarks: remove redefinition of benchmarkFinalize

What benchmarkFinalize currently only does is to pause for an amount of
time before proceding with the next iteration. Since we now do the same
_before_ each iteration (with --iterations-pause) the redefinition is
redundant.

Remove it from all benchmarks.

Signed-off-by: Juri Lelli <juri.lelli@arm.com>
diff --git a/tests/benchmarks/android_geekbench.py b/tests/benchmarks/android_geekbench.py
index 96ad4c1..d8b1af3 100755
--- a/tests/benchmarks/android_geekbench.py
+++ b/tests/benchmarks/android_geekbench.py
@@ -67,16 +67,9 @@
         self.setupWorkload()
         self.setupGovernor()
 
-    def benchmarkFinalize(self):
-        if self.delay_after_s:
-            self._log.info("Waiting %d[s] before to continue...",
-                           self.delay_after_s)
-            sleep(self.delay_after_s)
-
-    def __init__(self, governor, test, delay_after_s=0):
+    def __init__(self, governor, test):
         self.governor = governor
         self.test = test
-        self.delay_after_s = delay_after_s
         super(GeekbenchTest, self).__init__()
 
     def setupWorkload(self):
@@ -163,9 +156,8 @@
 for governor in governors:
     for test in tests:
         tests_remaining -= 1
-        delay_after_s = 30 if tests_remaining else 0
         try:
-            GeekbenchTest(governor, test, delay_after_s)
+            GeekbenchTest(governor, test)
             tests_completed += 1
         except:
             # A test configuraion failed, continue with other tests