1 files changed, 39 insertions, 0 deletions

diff --git a/example/overheads.py b/example/overheads.py
new file mode 100644
index 0000000..a022bae
--- /dev/null
+++ b/example/overheads.py
@@ -0,0 +1,39 @@
+#Necessary includes and stuff
+
+from schedcat.overheads.model import Overheads, CacheDelay
+from schedcat.overheads.jlfp import charge_scheduling_overheads, \
+                                    quantize_params
+from schedcat.sched.edf.gel_pl import bound_gfl_response_times, \
+                                      has_bounded_tardiness
+
+def copy_lock_overheads(oh, lock_oh):
+    oh.lock = lock_oh.lock
+    oh.unlock = lock_oh.unlock
+    oh.read_lock = lock_oh.read_lock
+    oh.read_unlock = lock_oh.read_unlock
+    oh.syscall_in = lock_oh.syscall_in
+    oh.syscall_out = lock_oh.syscall_out
+
+def get_oh_object(basic_oh, lock_oh, cache_oh, cache_level):
+    oh = Overheads.from_file(basic_oh)
+    oh.initial_cache_loss = \
+            CacheDelay.from_file(cache_oh).__dict__[cache_level]
+    oh.cache_affinity_loss = \
+            CacheDelay.from_file(cache_oh).__dict__[cache_level]
+    if lock_oh is not None:
+        lock_oh = Overheads.from_file(lock_oh)
+        copy_lock_overheads(oh, lock_oh)
+    return oh
+
+#Assumes absence of locking
+def bound_cfl_with_oh(oheads, dedicated_irq, clusts):
+    for clust in clusts:
+        success = charge_scheduling_overheads(oheads, clust.cpus,
+                                              dedicated_irq,
+                                              clust)
+        quantize_params(clust)
+        if (success and has_bounded_tardiness(clust.cpus, clust)):
+            bound_gfl_response_times(clust.cpus, clust, 15)
+        else:
+            return False
+    return True