starpu-max/src/datawizard/data_request.h

/* StarPU --- Runtime system for heterogeneous multicore architectures.
 *
 * Copyright (C) 2008-2011,2013-2017                      Université de Bordeaux
 * Copyright (C) 2010-2011,2013,2015                      CNRS
 * Copyright (C) 2017                                     Inria
 *
 * 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.
 */

/* This one includes us, so make sure to include it first */
#include <datawizard/coherency.h>

#ifndef __DATA_REQUEST_H__
#define __DATA_REQUEST_H__

#include <semaphore.h>
#include <datawizard/copy_driver.h>
#include <common/list.h>
#include <common/prio_list.h>
#include <common/starpu_spinlock.h>

/* TODO: This should be tuned according to driver capabilities
 * Data interfaces should also have to declare how many asynchronous requests
 * they have actually started (think of e.g. csr).
 */
#define MAX_PENDING_REQUESTS_PER_NODE 20
#define MAX_PENDING_PREFETCH_REQUESTS_PER_NODE 10
#define MAX_PENDING_IDLE_REQUESTS_PER_NODE 1
#define MAX_PUSH_TIME 1000 /* Maximum time in us that we can afford pushing requests before going back to the driver loop, e.g. for checking GPU task termination */

struct _starpu_data_replicate;

struct _starpu_callback_list
{
	void (*callback_func)(void *);
	void *callback_arg;
	struct _starpu_callback_list *next;
};

/* This represents a data request, i.e. we want some data to get transferred
 * from a source to a destination. */
LIST_TYPE(_starpu_data_request,
	struct _starpu_spinlock lock;
	unsigned refcnt;
	const char *origin; /* Name of the function that triggered the request */

	starpu_data_handle_t handle;
	struct _starpu_data_replicate *src_replicate;
	struct _starpu_data_replicate *dst_replicate;

	/* Which memory node will actually perform the transfer.
	 * This is important in the CUDA/OpenCL case, where only the worker for
	 * the node can make the CUDA/OpenCL calls.
	 */
	unsigned handling_node;

	/*
	 * What the destination node wants to do with the data: write to it,
	 * read it, or read and write to it. Only in the two latter cases we
	 * need an actual transfer, the first only needs an allocation.
	 *
	 * With mapped buffers, an additional case is mode = 0, which means
	 * unmapping the buffer.
	 */
	enum starpu_data_access_mode mode;

	/* Elements needed to make the transfer asynchronous */
	struct _starpu_async_channel async_channel;

	/* Whether the transfer is completed. */
	unsigned completed;

	/* Whether this is just a prefetch request:
	 * 0 for fetch,
	 * 1 for prefetch (dependencies have just been released)
	 * 2 for idle (a good idea to do it some time, but no hurry at all)
	 */
	unsigned prefetch;

	/* Priority of the request. Default is 0 */
	int prio;

	/* The value returned by the transfer function */
	int retval;

	/* The request will not actually be submitted until there remains
	 * dependencies. */
	unsigned ndeps;

	/* in case we have a chain of request (eg. for nvidia multi-GPU), this
	 * is the list of requests which are waiting for this one. */
	struct _starpu_data_request *next_req[STARPU_MAXNODES+1];
	/* The number of requests in next_req */
	unsigned next_req_count;

	struct _starpu_callback_list *callbacks;

	unsigned long com_id;
)
PRIO_LIST_TYPE(_starpu_data_request, prio)

/* Everyone that wants to access some piece of data will post a request.
 * Not only StarPU internals, but also the application may put such requests */
LIST_TYPE(_starpu_data_requester,
	/* what kind of access is requested ? */
	enum starpu_data_access_mode mode;

	/* applications may also directly manipulate data */
	unsigned is_requested_by_codelet;

	/* in case this is a codelet that will do the access */
	struct _starpu_job *j;
	unsigned buffer_index;

	/* if this is more complicated ... (eg. application request) 
	 * NB: this callback is not called with the lock taken !
	 */
	void (*ready_data_callback)(void *argcb);
	void *argcb;
)

void _starpu_init_data_request_lists(void);
void _starpu_deinit_data_request_lists(void);
void _starpu_post_data_request(struct _starpu_data_request *r);
/* returns 0 if we have pushed all requests, -EBUSY or -ENOMEM otherwise */
int _starpu_handle_node_data_requests(unsigned src_node, unsigned may_alloc, unsigned *pushed);
int _starpu_handle_node_prefetch_requests(unsigned src_node, unsigned may_alloc, unsigned *pushed);
int _starpu_handle_node_idle_requests(unsigned src_node, unsigned may_alloc, unsigned *pushed);

int _starpu_handle_pending_node_data_requests(unsigned src_node);
int _starpu_handle_all_pending_node_data_requests(unsigned src_node);

int _starpu_check_that_no_data_request_exists(unsigned node);
int _starpu_check_that_no_data_request_is_pending(unsigned node);

struct _starpu_data_request *_starpu_create_data_request(starpu_data_handle_t handle,
							 struct _starpu_data_replicate *src_replicate,
							 struct _starpu_data_replicate *dst_replicate,
							 int handling_node,
							 enum starpu_data_access_mode mode,
							 unsigned ndeps,
							 unsigned is_prefetch,
							 int prio,
							 unsigned is_write_invalidation,
							 const char *origin) STARPU_ATTRIBUTE_MALLOC;

int _starpu_wait_data_request_completion(struct _starpu_data_request *r, unsigned may_alloc);

void _starpu_data_request_append_callback(struct _starpu_data_request *r,
					  void (*callback_func)(void *),
					  void *callback_arg);

void _starpu_update_prefetch_status(struct _starpu_data_request *r, unsigned prefetch);
#endif // __DATA_REQUEST_H__