runtime-hclib.c

/* Copyright (c) 2013, Rice University

Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:

1.  Redistributions of source code must retain the above copyright
     notice, this list of conditions and the following disclaimer.
2.  Redistributions in binary form must reproduce the above
     copyright notice, this list of conditions and the following
     disclaimer in the documentation and/or other materials provided
     with the distribution.
3.  Neither the name of Rice University
     nor the names of its contributors may be used to endorse or
     promote products derived from this software without specific
     prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

 */

#include <stdio.h>
#include <stdlib.h>
#include <assert.h>

#include "hc_sysdep.h"
#include "runtime-support.h"
#include "rt-accumulator.h"
#include "rt-ddf.h"
#ifdef HAVE_PHASER
#include "phased.h"
#endif


void check_in_finish(finish_t * finish) {
    hc_atomic_inc(&(finish->counter));
}

void check_out_finish(finish_t * finish) {
    if (hc_atomic_dec(&(finish->counter))) {
        // If counter reached zero notify runtime
        rt_finish_reached_zero(finish);
    }
}

void async_check_in_finish(async_task_t * async_task) {
    check_in_finish(async_task->current_finish);
}

void async_check_out_finish(async_task_t * async_task) {
    check_out_finish(async_task->current_finish);
}

void async_drop_phasers(async_task_t * async_task) {
#ifdef HAVE_PHASER
    if (async_task->phaser_context != NULL) {
        // This call drop all phasers and deallocate the context
        destruct_phaser_context();
        async_task->phaser_context = NULL;
    }
#endif
}

async_task_t * allocate_async_task(async_t * async_def) {
    async_task_t * async_task;
    if ((async_def != NULL) && async_def->ddf_list != NULL) {
        // When the async has ddfs, we allocate a ddt instead
        // of a regular async. The ddt has extra data-structures.
        async_task = rt_allocate_ddt(async_def->ddf_list);
    } else {
        // Initializes and zeroes
        async_task = rt_allocate_async_task();
    }
    // Copy the definition
    if (async_def != NULL) {
        async_task->def = *async_def;
    }
    #ifdef HAVE_PHASER
    async_task->phaser_context = NULL;
    #endif
    return async_task;
}

forasync1D_task_t * allocate_forasync1D_task() {
    return rt_allocate_forasync1D_task();
}

forasync2D_task_t * allocate_forasync2D_task() {
    return rt_allocate_forasync2D_task();
}

forasync3D_task_t * allocate_forasync3D_task() {
    return rt_allocate_forasync3D_task();
}

void deallocate_async_task(async_task_t * async) {
    rt_deallocate_async_task(async);
}

finish_t * allocate_finish() {
    return rt_allocate_finish();
}

void deallocate_finish(finish_t * finish) {
    rt_deallocate_finish(finish);
}

finish_t * get_current_finish() {
    // Calling get_current_async is always legal because there's 
    // a fake async to represent the main activity.
    return get_current_async()->current_finish;
}

void end_finish_notify(finish_t * current_finish) {
    accum_t ** accumulators = current_finish->accumulators;
    if(accumulators != NULL) {
        int i = 0;
        while (accumulators[i] != NULL) {
            accum_impl_t * accum_impl = (accum_impl_t *) accumulators[i];
            accum_impl->close(accum_impl);
            i++;
        }
    }
}

static int is_eligible_to_schedule(async_task_t * async_task) {
    if (async_task->def.ddf_list != NULL) {
        ddt_t * ddt = (ddt_t *) rt_async_task_to_ddt(async_task);
        return iterate_ddt_frontier(ddt);
    } else {
        return 1;
    }
}

int get_nb_workers() {
    return rt_get_nb_workers();
}

int get_worker_id_hc() {
    return rt_get_worker_id();
}

void try_schedule_async(async_task_t * async_task) {
    if (is_eligible_to_schedule(async_task)) {
        rt_schedule_async(async_task);
    }
}

void schedule_async(async_task_t * async_task, finish_t * finish_scope, int property) {
    // Set the async finish scope to be the currently executing async's one.
    async_task->current_finish = finish_scope;
    // The newly created async must check in the current finish scope
    async_check_in_finish(async_task);

    #ifdef HAVE_PHASER
        phased_t * phased_clause = async_task->def.phased_clause;
        assert(!(phased_clause && (property & PHASER_TRANSMIT_ALL)));
        if (phased_clause != NULL) {
            phaser_context_t * currentCtx = get_phaser_context();
            phaser_context_t * ctx = phaser_context_construct();
            async_task->phaser_context = ctx;
            transmit_specific_phasers(currentCtx, ctx, 
                phased_clause->count, 
                phased_clause->phasers,
                phased_clause->phasers_mode);
        }
        if (property & PHASER_TRANSMIT_ALL) {
            phaser_context_t * currentCtx = get_phaser_context();
            phaser_context_t * ctx = phaser_context_construct();
            async_task->phaser_context = ctx;
            transmit_all_phasers(currentCtx, ctx);
        }
    #endif

    // delegate scheduling to the underlying runtime
    try_schedule_async(async_task);

    // Note: here the async has been scheduled and may or may not
    //       have been executed yet. Careful when adding code here.
}

void help_finish(finish_t * finish) {
    rt_help_finish(finish);
}