starpu-max/doc/doxygen/chapters/330_scheduling_contexts.doxy

/* StarPU --- Runtime system for heterogeneous multicore architectures.
 *
 * Copyright (C) 2011-2013,2016                           Inria
 * Copyright (C) 2010-2018                                CNRS
 * Copyright (C) 2009-2011,2014                           Université de Bordeaux
 * Copyright (C) 2016                                     Uppsala University
 *
 * StarPU is free software; you can redistribute it and/or modify
 * it under the terms of the GNU Lesser General Public License as published by
 * the Free Software Foundation; either version 2.1 of the License, or (at
 * your option) any later version.
 *
 * StarPU is distributed in the hope that it will be useful, but
 * WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 *
 * See the GNU Lesser General Public License in COPYING.LGPL for more details.
 */

/*! \page SchedulingContexts Scheduling Contexts

TODO: improve!

\section GeneralIdeas General Ideas

Scheduling contexts represent abstracts sets of workers that allow the
programmers to control the distribution of computational resources
(i.e. CPUs and GPUs) to concurrent kernels. The main goal is
to minimize interferences between the execution of multiple parallel
kernels, by partitioning the underlying pool of workers using
contexts. Scheduling contexts additionally allow a user to make use of
a different scheduling policy depending on the target resource set.


\section CreatingAContext Creating A Context

By default, the application submits tasks to an initial context, which
disposes of all the computation resources available to StarPU (all
the workers). If the application programmer plans to launch several
kernels simultaneously, by default these kernels will be
executed within this initial context, using a single scheduler
policy(see \ref TaskSchedulingPolicy). Meanwhile, if the application
programmer is aware of the demands of these kernels and of the
specificity of the machine used to execute them, the workers can be
divided between several contexts. These scheduling contexts will
isolate the execution of each kernel and they will permit the use of a
scheduling policy proper to each one of them.

Scheduling Contexts may be created in two ways: either the programmers
indicates the set of workers corresponding to each context (providing
he knows the identifiers of the workers running within StarPU), or the
programmer does not provide any worker list and leaves the Hypervisor
assign workers to each context according to their needs (\ref
SchedulingContextHypervisor).

Both cases require a call to the function
starpu_sched_ctx_create(), which requires as input the worker
list (the exact list or a <c>NULL</c> pointer), the amount of workers
(or <c>-1</c> to designate all workers on the platform) and a list of
optional parameters such as the scheduling policy, terminated by a
<c>0</c>. The scheduling policy can be a character list corresponding
to the name of a StarPU predefined policy or the pointer to a custom
policy. The function returns an identifier of the context created
which you will use to indicate the context you want to submit the
tasks to.

\code{.c}
/* the list of resources the context will manage */
int workerids[3] = {1, 3, 10};

/* indicate the list of workers assigned to it, the number of workers,
the name of the context and the scheduling policy to be used within
the context */
int id_ctx = starpu_sched_ctx_create(workerids, 3, "my_ctx", STARPU_SCHED_CTX_POLICY_NAME, "dmda", 0);

/* let StarPU know that the following tasks will be submitted to this context */
starpu_sched_ctx_set_context(id);

/* submit the task to StarPU */
starpu_task_submit(task);
\endcode

Note: Parallel greedy and parallel heft scheduling policies do not support the existence of several disjoint contexts on the machine.
Combined workers are constructed depending on the entire topology of the machine, not only the one belonging to a context.

\subsection CreatingAContextWithTheDefaultBehavior Creating A Context With The Default Behavior

If <b>no scheduling policy</b> is specified when creating the context,
it will be used as <b>another type of resource</b>: a cluster. A
cluster is a context without scheduler (eventually delegated to
another runtime). For more information see \ref ClusteringAMachine. It
is therefore <b>mandatory</b> to stipulate a scheduler to use the
contexts in this traditional way.

To create a <b>context</b> with the default scheduler, that is either
controlled through the environment variable <c>STARPU_SCHED</c> or the
StarPU default scheduler, one can explicitly use the option <c>STARPU_SCHED_CTX_POLICY_NAME, ""</c> as in the following example:

\code{.c}
/* the list of resources the context will manage */
int workerids[3] = {1, 3, 10};

/* indicate the list of workers assigned to it, the number of workers,
and use the default scheduling policy. */
int id_ctx = starpu_sched_ctx_create(workerids, 3, "my_ctx", STARPU_SCHED_CTX_POLICY_NAME, "", 0);

/* .... */
\endcode

\section CreatingAContext Creating A Context To Partition a GPU

The contexts can also be used to group set of SMs of an NVIDIA GPU in order to isolate
the parallel kernels and allow them to coexecution on a specified partiton of the GPU.

Each context will be mapped to a stream and the user can indicate the number of SMs.
The context can be added to a larger context already grouping CPU cores.
This larger context can use a scheduling policy that assigns tasks to both CPUs and contexts (partitions of the GPU)
based on performance models adjusted to the number of SMs.

The GPU implementation of the task has to be modified accordingly and receive as a parameter the number of SMs.

\code{.c}
/* get the available streams (suppose we have nstreams = 2 by specifying them with STARPU_NWORKER_PER_CUDA=2  */
int nstreams = starpu_worker_get_stream_workerids(gpu_devid, stream_workerids, STARPU_CUDA_WORKER);

int sched_ctx[nstreams];
sched_ctx[0] = starpu_sched_ctx_create(&stream_workerids[0], 1, "subctx",  STARPU_SCHED_CTX_CUDA_NSMS, 6, 0);
sched_ctx[1] = starpu_sched_ctx_create(&stream_workerids[1], 1, "subctx",  STARPU_SCHED_CTX_CUDA_NSMS, 7, 0);

int ncpus = 4;
int workers[ncpus+nstreams];
workers[ncpus+0] = stream_workerids[0];
workers[ncpus+1] = stream_workerids[1];

big_sched_ctx = starpu_sched_ctx_create(workers, ncpus+nstreams, "ctx1", STARPU_SCHED_CTX_SUB_CTXS, sched_ctxs, nstreams, STARPU_SCHED_CTX_POLICY_NAME, "dmdas", 0);

starpu_task_submit_to_ctx(task, big_sched_ctx);

\endcode

\section ModifyingAContext Modifying A Context

A scheduling context can be modified dynamically. The application may
change its requirements during the execution and the programmer can
add additional workers to a context or remove those no longer needed. In
the following example we have two scheduling contexts
<c>sched_ctx1</c> and <c>sched_ctx2</c>. After executing a part of the
tasks some of the workers of <c>sched_ctx1</c> will be moved to
context <c>sched_ctx2</c>.

\code{.c}
/* the list of ressources that context 1 will give away */
int workerids[3] = {1, 3, 10};

/* add the workers to context 1 */
starpu_sched_ctx_add_workers(workerids, 3, sched_ctx2);

/* remove the workers from context 2 */
starpu_sched_ctx_remove_workers(workerids, 3, sched_ctx1);
\endcode

\section SubmittingTasksToAContext Submitting Tasks To A Context
The application may submit tasks to several contexts either
simultaneously or sequnetially. If several threads of submission
are used the function starpu_sched_ctx_set_context() may be called just
before starpu_task_submit(). Thus StarPU considers that
the current thread will submit tasks to the coresponding context.

When the application may not assign a thread of submission to each
context, the id of the context must be indicated by using the
function starpu_task_submit_to_ctx() or the field \ref STARPU_SCHED_CTX
for starpu_task_insert().

\section DeletingAContext Deleting A Context

When a context is no longer needed it must be deleted. The application
can indicate which context should keep the resources of a deleted one.
All the tasks of the context should be executed before doing this.
Thus, the programmer may use either a barrier and then delete the context
directly, or just indicate
that other tasks will not be submitted later on to the context (such that when
the last task is executed its workers will be moved to the inheritor)
and delete the context at the end of the execution (when a barrier will
be used eventually).

\code{.c}
/* when the context 2 is deleted context 1 inherits its resources */
starpu_sched_ctx_set_inheritor(sched_ctx2, sched_ctx1);

/* submit tasks to context 2 */
for (i = 0; i < ntasks; i++)
    starpu_task_submit_to_ctx(task[i],sched_ctx2);

/* indicate that context 2 finished submitting and that */
/* as soon as the last task of context 2 finished executing */
/* its workers can be moved to the inheritor context */
starpu_sched_ctx_finished_submit(sched_ctx1);

/* wait for the tasks of both contexts to finish */
starpu_task_wait_for_all();

/* delete context 2 */
starpu_sched_ctx_delete(sched_ctx2);

/* delete context 1 */
starpu_sched_ctx_delete(sched_ctx1);
\endcode

\section EmptyingAContext Emptying A Context

A context may have no resources at the begining or at a certain
moment of the execution. Tasks can still be submitted to these contexts
and they will be executed as soon as the contexts will have resources. A list
of tasks pending to be executed is kept and will be submitted when
workers are added to the contexts.

\code{.c}
/* create a empty context */
unsigned sched_ctx_id = starpu_sched_ctx_create(NULL, 0, "ctx", 0);

/* submit a task to this context */
starpu_sched_ctx_set_context(&sched_ctx_id);
ret = starpu_task_insert(&codelet, 0);
STARPU_CHECK_RETURN_VALUE(ret, "starpu_task_insert");

/* add CPU workers to the context */
int procs[STARPU_NMAXWORKERS];
int nprocs = starpu_cpu_worker_get_count();
starpu_worker_get_ids_by_type(STARPU_CPU_WORKER, procs, nprocs);
starpu_sched_ctx_add_workers(procs, nprocs, sched_ctx_id);

/* and wait for the task termination */
starpu_task_wait_for_all();
\endcode

However, if resources are never allocated to the context, the
application will not terminate. If these tasks have low priority, the
application can inform StarPU to not submit them by calling the
function starpu_sched_ctx_stop_task_submission().


*/