coll_tuned_allreduce.c

来自「MPI stands for the Message Passing Inter」· C语言 代码 · 共 680 行 · 第 1/2 页

   int rank, size, k, recvfrom, sendto;
   int segcount, lastsegcount, maxsegcount;
   int blockcount, inbi;
   size_t typelng;
   char *tmpsend = NULL, *tmprecv = NULL;
   char *inbuf[2] = {NULL, NULL};
   ptrdiff_t true_lb, true_extent, lb, extent, realsegsize, maxrealsegsize;
   ompi_request_t *reqs[2] = {NULL, NULL};

   size = ompi_comm_size(comm);
   rank = ompi_comm_rank(comm);

   OPAL_OUTPUT((ompi_coll_tuned_stream,
                "coll:tuned:allreduce_intra_ring rank %d, count %d", rank, count));
      
   /* Special case for size == 1 */
   if (1 == size) {
      if (MPI_IN_PLACE != sbuf) {
         ret = ompi_ddt_copy_content_same_ddt(dtype, count, (char*)rbuf, (char*)sbuf);
         if (ret < 0) { line = __LINE__; goto error_hndl; }
      }
      return MPI_SUCCESS;
   }

   /* Special case for count less than size - 1 - use recursive doubling */
   if (count < size - 1) {
      OPAL_OUTPUT((ompi_coll_tuned_stream, "coll:tuned:allreduce_ring rank %d/%d, count %d, switching to recursive doubling", rank, size, count));
      return (ompi_coll_tuned_allreduce_intra_recursivedoubling(sbuf, rbuf, 
                                                                count,
                                                                dtype, op, 
                                                                comm));
   }

   /* Allocate and initialize temporary buffers */
   ret = ompi_ddt_get_extent(dtype, &lb, &extent);
   if (MPI_SUCCESS != ret) { line = __LINE__; goto error_hndl; }
   ret = ompi_ddt_get_true_extent(dtype, &true_lb, &true_extent);
   if (MPI_SUCCESS != ret) { line = __LINE__; goto error_hndl; }
   ret = ompi_ddt_type_size( dtype, &typelng);
   if (MPI_SUCCESS != ret) { line = __LINE__; goto error_hndl; }

   /* Determine number of elements per block.  
      This is not the same computation as the one for number of elements 
      per segment - and we may end up having last block larger any other block.
    */
   segcount = count / size;
   if (0 != count % size) segcount++;
   lastsegcount = count - (size - 1) * segcount;
   if (lastsegcount <= 0) {
      segcount--;
      lastsegcount = count - (size - 1) * segcount;
   }
   realsegsize = segcount * extent;
   maxsegcount = (segcount > lastsegcount)? segcount : lastsegcount;
   maxrealsegsize = true_extent + (maxsegcount - 1) * extent;

   inbuf[0] = (char*)malloc(maxrealsegsize);
   if (NULL == inbuf[0]) { ret = -1; line = __LINE__; goto error_hndl; }
   if (size > 2) {
      inbuf[1] = (char*)malloc(maxrealsegsize);
      if (NULL == inbuf[1]) { ret = -1; line = __LINE__; goto error_hndl; }
   }

   /* Handle MPI_IN_PLACE */
   if (MPI_IN_PLACE != sbuf) {
      ret = ompi_ddt_copy_content_same_ddt(dtype, count, (char*)rbuf, (char*)sbuf);
      if (ret < 0) { line = __LINE__; goto error_hndl; }
   }

   /* Computation loop */

   /* 
      For each of the remote nodes:
      - post irecv for block (r-1)
      - send block (r)
      - in loop for every step k = 2 .. n
         - post irecv for block (r + n - k) % n
         - wait on block (r + n - k + 1) % n to arrive
         - compute on block (r + n - k + 1) % n
         - send block (r + n - k + 1) % n
     - wait on block (r + 1)
     - compute on block (r + 1)
     - send block (r + 1) to rank (r + 1)
     Note that for send operations and computation we must compute the exact 
     block size.
    */
   sendto = (rank + 1) % size;
   recvfrom = (rank + size - 1) % size;

   inbi = 0;
   tmpsend = ((char*)rbuf) + rank * realsegsize;
   /* Initialize first receive from the neighbor on the left */
   ret = MCA_PML_CALL(irecv(inbuf[inbi], maxsegcount, dtype, recvfrom,
                            MCA_COLL_BASE_TAG_ALLREDUCE, comm, &reqs[inbi]));
   if (MPI_SUCCESS != ret) { line = __LINE__; goto error_hndl; }
   /* Send first block to the neighbor on the right */
   blockcount = segcount;
   if ((size - 1) == rank) { blockcount = lastsegcount; }
   ret = MCA_PML_CALL(send(tmpsend, blockcount, dtype, sendto,
                           MCA_COLL_BASE_TAG_ALLREDUCE,
                           MCA_PML_BASE_SEND_STANDARD, comm));
   if (MPI_SUCCESS != ret) { line = __LINE__; goto error_hndl; }
   
   for (k = 2; k < size; k++) {
      const int prevblock = (rank + size - k + 1) % size;
      
      inbi = inbi ^ 0x1;
      
      /* Post irecv for the current block */
      ret = MCA_PML_CALL(irecv(inbuf[inbi], maxsegcount, dtype, recvfrom,
                               MCA_COLL_BASE_TAG_ALLREDUCE, comm, &reqs[inbi]));
      if (MPI_SUCCESS != ret) { line = __LINE__; goto error_hndl; }
      
      /* Wait on previous block to arrive */
      ret = ompi_request_wait(&reqs[inbi ^ 0x1], MPI_STATUS_IGNORE);
      if (MPI_SUCCESS != ret) { line = __LINE__; goto error_hndl; }
      
      /* Apply operation on previous block: result goes to rbuf
         rbuf[prevblock] = inbuf[inbi ^ 0x1] (op) rbuf[prevblock]
      */
      blockcount = segcount;
      if ((size - 1) == prevblock) { blockcount = lastsegcount; }
      tmprecv = ((char*)rbuf) + prevblock * realsegsize;
      ompi_op_reduce(op, inbuf[inbi ^ 0x1], tmprecv, blockcount, dtype);
      
      /* send previous block to sendto */
      ret = MCA_PML_CALL(send(tmprecv, blockcount, dtype, sendto,
                              MCA_COLL_BASE_TAG_ALLREDUCE,
                              MCA_PML_BASE_SEND_STANDARD, comm));
      if (MPI_SUCCESS != ret) { line = __LINE__; goto error_hndl; }
   }

   /* Wait on the last block to arrive */
   ret = ompi_request_wait(&reqs[inbi], MPI_STATUS_IGNORE);
   if (MPI_SUCCESS != ret) { line = __LINE__; goto error_hndl; }

   /* Apply operation on the last block (from neighbor (rank + 1) 
      rbuf[rank+1] = inbuf[inbi] (op) rbuf[rank + 1] */
   blockcount = segcount;
   if ((size - 1) == (rank + 1) % size) { blockcount = lastsegcount; }
   tmprecv = ((char*)rbuf) + ((rank + 1) % size) * realsegsize;
   ompi_op_reduce(op, inbuf[inbi], tmprecv, blockcount, dtype);
   
   /* Distribution loop - variation of ring allgather */
   for (k = 0; k < size - 1; k++) {
      const int recvdatafrom = (rank + size - k) % size;
      const int senddatafrom = (rank + 1 + size - k) % size;

      blockcount = segcount;
      if ((size - 1) == senddatafrom) blockcount = lastsegcount;

      tmprecv = (char*)rbuf + recvdatafrom * realsegsize;
      tmpsend = (char*)rbuf + senddatafrom * realsegsize;

      ret = ompi_coll_tuned_sendrecv(tmpsend, blockcount, dtype, sendto,
                                     MCA_COLL_BASE_TAG_ALLREDUCE,
                                     tmprecv, maxsegcount, dtype, recvfrom,
                                     MCA_COLL_BASE_TAG_ALLREDUCE,
                                     comm, MPI_STATUS_IGNORE, rank);
      if (MPI_SUCCESS != ret) { line = __LINE__; goto error_hndl;}

   }

   if (NULL != inbuf[0]) free(inbuf[0]);
   if (NULL != inbuf[1]) free(inbuf[1]);

   return MPI_SUCCESS;

 error_hndl:
   OPAL_OUTPUT((ompi_coll_tuned_stream, "%s:%4d\tRank %d Error occurred %d\n",
                __FILE__, line, rank, ret));
   if (NULL != inbuf[0]) free(inbuf[0]);
   if (NULL != inbuf[1]) free(inbuf[1]);
   return ret;
}

/*
 * Linear functions are copied from the BASIC coll module
 * they do not segment the message and are simple implementations
 * but for some small number of nodes and/or small data sizes they 
 * are just as fast as tuned/tree based segmenting operations 
 * and as such may be selected by the decision functions
 * These are copied into this module due to the way we select modules
 * in V1. i.e. in V2 we will handle this differently and so will not
 * have to duplicate code.
 * GEF Oct05 after asking Jeff.
 */

/* copied function (with appropriate renaming) starts here */


/*
 *	allreduce_intra
 *
 *	Function:	- allreduce using other MPI collectives
 *	Accepts:	- same as MPI_Allreduce()
 *	Returns:	- MPI_SUCCESS or error code
 */
int
ompi_coll_tuned_allreduce_intra_basic_linear(void *sbuf, void *rbuf, int count,
                                             struct ompi_datatype_t *dtype,
                                             struct ompi_op_t *op,
                                             struct ompi_communicator_t *comm)
{
    int err;
    int rank;

    rank = ompi_comm_rank(comm);

    OPAL_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:allreduce_intra_basic_linear rank %d", rank));

    /* Reduce to 0 and broadcast. */

    if (MPI_IN_PLACE == sbuf) {
        if (0 == ompi_comm_rank(comm)) {
            err = ompi_coll_tuned_reduce_intra_basic_linear (MPI_IN_PLACE, rbuf, count, dtype, op, 0, comm);
        } else {
            err = ompi_coll_tuned_reduce_intra_basic_linear(rbuf, NULL, count, dtype, op, 0, comm);
        }
    } else {
        err = ompi_coll_tuned_reduce_intra_basic_linear(sbuf, rbuf, count, dtype, op, 0, comm);
    }
    if (MPI_SUCCESS != err) {
        return err;
    }

    return ompi_coll_tuned_bcast_intra_basic_linear(rbuf, count, dtype, 0, comm);
}


/* copied function (with appropriate renaming) ends here */

/* The following are used by dynamic and forced rules */

/* publish details of each algorithm and if its forced/fixed/locked in */
/* as you add methods/algorithms you must update this and the query/map routines */

/* this routine is called by the component only */
/* this makes sure that the mca parameters are set to their initial values and perms */
/* module does not call this they call the forced_getvalues routine instead */

int ompi_coll_tuned_allreduce_intra_check_forced_init (coll_tuned_force_algorithm_mca_param_indices_t *mca_param_indices)
{
    int rc, max_alg = 4, requested_alg;

    ompi_coll_tuned_forced_max_algorithms[ALLREDUCE] = max_alg;

    rc = mca_base_param_reg_int (&mca_coll_tuned_component.super.collm_version,
                                 "allreduce_algorithm_count",
                                 "Number of allreduce algorithms available",
                                 false, true, max_alg, NULL);

    mca_param_indices->algorithm_param_index
        = mca_base_param_reg_int( &mca_coll_tuned_component.super.collm_version,
                                  "allreduce_algorithm",
                                  "Which allreduce algorithm is used. Can be locked down to any of: 0 ignore, 1 basic linear, 2 nonoverlapping (tuned reduce + tuned bcast), 3 recursive doubling, 4 ring",
                                  false, false, 0, NULL);
    mca_base_param_lookup_int( mca_param_indices->algorithm_param_index, &(requested_alg));
    if( requested_alg > max_alg ) {
        if( 0 == ompi_comm_rank( MPI_COMM_WORLD ) ) {
            opal_output( 0, "Allreduce algorithm #%d is not available (range [0..%d]). Switching back to ignore(0)\n",
                         requested_alg, max_alg );
        }
        mca_base_param_set_int( mca_param_indices->algorithm_param_index, 0);
    }

    mca_param_indices->segsize_param_index
        = mca_base_param_reg_int( &mca_coll_tuned_component.super.collm_version,
                                  "allreduce_algorithm_segmentsize",
                                  "Segment size in bytes used by default for allreduce algorithms. Only has meaning if algorithm is forced and supports segmenting. 0 bytes means no segmentation.",
                                  false, false, 0, NULL);
  
    mca_param_indices->tree_fanout_param_index
        = mca_base_param_reg_int( &mca_coll_tuned_component.super.collm_version,
                                  "allreduce_algorithm_tree_fanout",
                                  "Fanout for n-tree used for allreduce algorithms. Only has meaning if algorithm is forced and supports n-tree topo based operation.",
                                  false, false, ompi_coll_tuned_init_tree_fanout, /* get system wide default */
                                  NULL);

    mca_param_indices->chain_fanout_param_index
        = mca_base_param_reg_int( &mca_coll_tuned_component.super.collm_version,
                                  "allreduce_algorithm_chain_fanout",
                                  "Fanout for chains used for allreduce algorithms. Only has meaning if algorithm is forced and supports chain topo based operation.",
                                  false, false,
                                  ompi_coll_tuned_init_chain_fanout, /* get system wide default */
                                  NULL);

    return (MPI_SUCCESS);
}


int ompi_coll_tuned_allreduce_intra_do_forced(void *sbuf, void *rbuf, int count,
                                              struct ompi_datatype_t *dtype,
                                              struct ompi_op_t *op,
                                              struct ompi_communicator_t *comm)
{
    OPAL_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:allreduce_intra_do_forced selected algorithm %d, segment size %d", 
                 comm->c_coll_selected_data->user_forced[ALLREDUCE].algorithm,
                 comm->c_coll_selected_data->user_forced[ALLREDUCE].segsize));

    switch (comm->c_coll_selected_data->user_forced[ALLREDUCE].algorithm) {
    case (0):  return ompi_coll_tuned_allreduce_intra_dec_fixed (sbuf, rbuf, count, dtype, op, comm);
    case (1):  return ompi_coll_tuned_allreduce_intra_basic_linear (sbuf, rbuf, count, dtype, op, comm);
    case (2):  return ompi_coll_tuned_allreduce_intra_nonoverlapping (sbuf, rbuf, count, dtype, op, comm);
    case (3):  return ompi_coll_tuned_allreduce_intra_recursivedoubling (sbuf, rbuf, count, dtype, op, comm);
    case (4):  return ompi_coll_tuned_allreduce_intra_ring (sbuf, rbuf, count, dtype, op, comm);
    default:
        OPAL_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:allreduce_intra_do_forced attempt to select algorithm %d when only 0-%d is valid?",
                     comm->c_coll_selected_data->user_forced[ALLREDUCE].algorithm, 
                     ompi_coll_tuned_forced_max_algorithms[ALLREDUCE]));
        return (MPI_ERR_ARG);
    } /* switch */

}


int ompi_coll_tuned_allreduce_intra_do_this(void *sbuf, void *rbuf, int count,
                                            struct ompi_datatype_t *dtype,
                                            struct ompi_op_t *op,
                                            struct ompi_communicator_t *comm,
                                            int algorithm, int faninout, int segsize)
{
    OPAL_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:allreduce_intra_do_this algorithm %d topo fan in/out %d segsize %d", 
                 algorithm, faninout, segsize));

    switch (algorithm) {
    case (0):   return ompi_coll_tuned_allreduce_intra_dec_fixed (sbuf, rbuf, count, dtype, op, comm);
    case (1):   return ompi_coll_tuned_allreduce_intra_basic_linear (sbuf, rbuf, count, dtype, op, comm);
    case (2):   return ompi_coll_tuned_allreduce_intra_nonoverlapping (sbuf, rbuf, count, dtype, op, comm);
    case (3):   return ompi_coll_tuned_allreduce_intra_recursivedoubling (sbuf, rbuf, count, dtype, op, comm);
    case (4):   return ompi_coll_tuned_allreduce_intra_ring (sbuf, rbuf, count, dtype, op, comm);
    default:
        OPAL_OUTPUT((ompi_coll_tuned_stream,"coll:tuned:allreduce_intra_do_this attempt to select algorithm %d when only 0-%d is valid?",
                     algorithm, ompi_coll_tuned_forced_max_algorithms[ALLREDUCE]));
        return (MPI_ERR_ARG);
    } /* switch */

}