Merge branch 'master' of git+ssh://scm.gforge.inria.fr/gitroot/starpu/starpu

AUTHORS

@@ -21,6 +21,7 @@ Khorsi Yanis, Inria, <yanis.khorsi@inria.fr>
 Lambert Thibaut, Inria, <thibaud.lambert@inria.fr>
 Leria Erwan, University of Bordeaux, <erwan.leria@etu.u-bordeaux.fr>
 Lizé Benoît, Airbus, <benoit.lize@gmail.com>
+Makni Mariem, Inria, <mariem.makni@inria.fr>
 Nakov Stojce, Inria, <stojce.nakov@inria.fr>
 Namyst Raymond, University of Bordeaux, <raymond.namyst@labri.fr>
 Nesi Lucas Leandro, Federal University of Rio Grande do Sul (UFRGS), <llnesi@inf.ufrgs.br>

doc/doxygen/chapters/101_building.doxy

@@ -504,8 +504,10 @@ multiplication using BLAS and cuBLAS. They output the obtained GFlops.
 It can also be convenient to try simulated benchmarks, if you want to give a try
 at CPU-GPU scheduling without actually having a GPU at hand. This can be done by
 using the simgrid version of StarPU: first install the simgrid simulator from
-http://simgrid.gforge.inria.fr/ (we tested with simgrid 3.11, 3.12 and 3.13, other versions
-may have compatibility issues), then configure StarPU with \ref enable-simgrid
+http://simgrid.gforge.inria.fr/ (we tested with simgrid from 3.11 to 3.16, and
+3.18 to 3.22, other versions may have compatibility issues, 3.17 notably does
+not build at all. MPI simulation does not work with version 3.22),
+then configure StarPU with \ref enable-simgrid
 "--enable-simgrid" and rebuild and install it, and then you can simulate the performance for a
 few virtualized systems shipped along StarPU: attila, mirage, idgraf, and sirocco.
 

doc/doxygen/chapters/320_scheduling.doxy

@@ -166,20 +166,34 @@ be obtained from the machine power supplier.
 The energy actually consumed by the total execution can be displayed by setting
 <c>export STARPU_PROFILING=1 STARPU_WORKER_STATS=1</c> .
 
-On-line task consumption measurement is currently only supported through the
+For OpenCL devices, on-line task consumption measurement is currently supported through the
 <c>CL_PROFILING_POWER_CONSUMED</c> OpenCL extension, implemented in the MoviSim
-simulator. Applications can however provide explicit measurements by
-using the function starpu_perfmodel_update_history() (examplified in \ref PerformanceModelExample
-with the <c>energy_model</c> performance model). Fine-grain
-measurement is often not feasible with the feedback provided by the hardware, so
-the user can for instance run a given task a thousand times, measure the global
+simulator.
+
+For CUDA devices, on-line task consumption measurement is supported on V100
+cards and beyond. This however only works for quite long tasks, since the
+measurement granularity is about 10ms.
+
+Applications can however provide explicit measurements by using the function
+starpu_perfmodel_update_history() (examplified in \ref PerformanceModelExample
+with the <c>energy_model</c> performance model). Fine-grain measurement
+is often not feasible with the feedback provided by the hardware, so the
+user can for instance run a given task a thousand times, measure the global
 consumption for that series of tasks, divide it by a thousand, repeat for
-varying kinds of tasks and task sizes, and eventually feed StarPU
-with these manual measurements through starpu_perfmodel_update_history().
-For instance, for CUDA devices, <c>nvidia-smi -q -d POWER</c> can be used to get
-the current consumption in Watt. Multiplying this value by the average duration
-of a single task gives the consumption of the task in Joules, which can be given
-to starpu_perfmodel_update_history().
+varying kinds of tasks and task sizes, and eventually feed StarPU with these
+manual measurements through starpu_perfmodel_update_history().  For instance,
+for CUDA devices, <c>nvidia-smi -q -d POWER</c> can be used to get the current
+consumption in Watt. Multiplying this value by the average duration of a
+single task gives the consumption of the task in Joules, which can be given to
+starpu_perfmodel_update_history().
+
+Another way to provide the energy performance is to define a
+perfmodel with starpu_perfmodel::type ::STARPU_PER_ARCH, and set the
+starpu_perfmodel::arch_cost_function field to a function which shall return the
+estimated consumption of the task in Joules. Such a function can for instance
+use starpu_task_expected_length() on the task (in µs), multiplied by the
+typical power consumption of the device, e.g. in W, and divided by 1000000. to
+get Joules.
 
 \section ExistingModularizedSchedulers Modularized Schedulers
 

doc/doxygen/chapters/470_simgrid.doxy

@@ -16,6 +16,10 @@
  * See the GNU Lesser General Public License in COPYING.LGPL for more details.
  */
 
+/*
+ * NOTE: XXX: also update simgrid versions in 101_building.doxy !!
+ */
+
 /*! \page SimGridSupport SimGrid Support
 
 StarPU can use Simgrid in order to simulate execution on an arbitrary

src/core/perfmodel/perfmodel_history.c

@@ -742,7 +742,7 @@ static void check_per_arch_model(struct starpu_perfmodel *model, int comb, unsig
 	struct starpu_perfmodel_history_list *ptr = NULL;
 	unsigned nentries = 0;
 
-	if (model->type == STARPU_HISTORY_BASED || model->type == STARPU_NL_REGRESSION_BASED)
+	if (model->type == STARPU_HISTORY_BASED || model->type == STARPU_NL_REGRESSION_BASED  || model->type == STARPU_REGRESSION_BASED)
 	{
 		/* Dump the list of all entries in the history */
 		ptr = per_arch_model->list;
@@ -760,7 +760,7 @@ static void check_per_arch_model(struct starpu_perfmodel *model, int comb, unsig
 	check_reg_model(model, comb, impl);
 
 	/* Dump the history into the model file in case it is necessary */
-	if (model->type == STARPU_HISTORY_BASED || model->type == STARPU_NL_REGRESSION_BASED)
+	if (model->type == STARPU_HISTORY_BASED || model->type == STARPU_NL_REGRESSION_BASED || model->type == STARPU_REGRESSION_BASED)
 	{
 		ptr = per_arch_model->list;
 		while (ptr)
@@ -779,7 +779,7 @@ static void dump_per_arch_model_file(FILE *f, struct starpu_perfmodel *model, in
 	struct starpu_perfmodel_history_list *ptr = NULL;
 	unsigned nentries = 0;
 
-	if (model->type == STARPU_HISTORY_BASED || model->type == STARPU_NL_REGRESSION_BASED)
+       if (model->type == STARPU_HISTORY_BASED || model->type == STARPU_NL_REGRESSION_BASED || model->type == STARPU_REGRESSION_BASED)
 	{
 		/* Dump the list of all entries in the history */
 		ptr = per_arch_model->list;
@@ -800,7 +800,7 @@ static void dump_per_arch_model_file(FILE *f, struct starpu_perfmodel *model, in
 	dump_reg_model(f, model, comb, impl);
 
 	/* Dump the history into the model file in case it is necessary */
-	if (model->type == STARPU_HISTORY_BASED || model->type == STARPU_NL_REGRESSION_BASED)
+       if (model->type == STARPU_HISTORY_BASED || model->type == STARPU_NL_REGRESSION_BASED || model->type == STARPU_REGRESSION_BASED)
 	{
 		fprintf(f, "# hash\t\tsize\t\tflops\t\tmean (us)\tdev (us)\tsum\t\tsum2\t\tn\n");
 		ptr = per_arch_model->list;
@@ -1861,7 +1861,7 @@ void _starpu_update_perfmodel_history(struct _starpu_job *j, struct starpu_perfm
 			model->state->per_arch_is_set[comb][impl] = 1;
 		}
 
-		if (model->type == STARPU_HISTORY_BASED || model->type == STARPU_NL_REGRESSION_BASED)
+		if (model->type == STARPU_HISTORY_BASED || model->type == STARPU_NL_REGRESSION_BASED || model->type == STARPU_REGRESSION_BASED)
 		{
 			struct starpu_perfmodel_history_entry *entry;
 			struct starpu_perfmodel_history_table *elt;