|
|
@@ -1,7 +1,7 @@
|
|
|
/* StarPU --- Runtime system for heterogeneous multicore architectures.
|
|
|
*
|
|
|
* Copyright (C) 2013,2014,2016-2019 CNRS
|
|
|
- * Copyright (C) 2013,2014,2017,2018 Université de Bordeaux
|
|
|
+ * Copyright (C) 2013,2014,2017,2018-2019 Université de Bordeaux
|
|
|
* Copyright (C) 2013 Inria
|
|
|
* Copyright (C) 2013 Corentin Salingue
|
|
|
*
|
|
|
@@ -145,6 +145,18 @@ The scheduling algorithms worth trying are thus <code>dmdar</code> and
|
|
|
<code>lws</code>, which privilege data locality over priorities. There will be
|
|
|
work on this area in the coming future.
|
|
|
|
|
|
+\section Feedback figures
|
|
|
+
|
|
|
+Beyond pure performance feedback, some figures are interesting to have a look at.
|
|
|
+
|
|
|
+Using <c>export STARPU_BUS_STATS=1</c> gives an overview of the data
|
|
|
+transfers which were needed. The values can also be obtained at runtime
|
|
|
+by using starpu_bus_get_profiling_info(). An example can be read in
|
|
|
+<c>src/profiling/profiling_helpers.c</c>.
|
|
|
+
|
|
|
+Using <c>export STARPU_ENABLE_STATS=1</c> gives information for each memory node
|
|
|
+on data miss/hit and allocation miss/hit.
|
|
|
+
|
|
|
\section DiskFunctions Disk functions
|
|
|
|
|
|
There are various ways to operate a disk memory node, described by the structure
|
