/**
 *
 * @copyright (c) 2009-2014 The University of Tennessee and The University
 *                          of Tennessee Research Foundation.
 *                          All rights reserved.
 * @copyright (c) 2012-2014 Inria. All rights reserved.
 * @copyright (c) 2012-2014 Bordeaux INP, CNRS (LaBRI UMR 5800), Inria, Univ. Bordeaux. All rights reserved.
 *
 **/

/**
 *
 * @file codelet_zttlqt.c
 *
 *  MORSE codelets kernel
 *  MORSE is a software package provided by Univ. of Tennessee,
 *  Univ. of California Berkeley and Univ. of Colorado Denver
 *
 * @version 2.5.0
 * @comment This file has been automatically generated
 *          from Plasma 2.5.0 for MORSE 1.0.0
 * @author Hatem Ltaief
 * @author Dulceneia Becker
 * @author Mathieu Faverge
 * @author Emmanuel Agullo
 * @author Cedric Castagnede
 * @date 2010-11-15
 * @precisions normal z -> c d s
 *
 **/

#include "runtime/starpu/include/morse_starpu.h"
#include "runtime/starpu/include/runtime_codelet_z.h"

/**
 *
 * @ingroup CORE_MORSE_Complex64_t
 *
 *  CORE_zttlqt computes a LQ factorization of a rectangular matrix
 *  formed by coupling side-by-side a complex M-by-M lower triangular tile A1
 *  and a complex M-by-N lower triangular tile A2:
 *
 *    | A1 A2 | = L * Q
 *
 *  The tile Q is represented as a product of elementary reflectors
 *
 *    Q = H(k)' . . . H(2)' H(1)', where k = min(M,N).
 *
 *  Each H(i) has the form
 *
 *    H(i) = I - tau * v * v'
 *
 *  where tau is a complex scalar, and v is a complex vector with
 *  v(1:i-1) = 0 and v(i) = 1; conjg(v(i+1:n)) is stored on exit in
 *  A2(i,1:n), and tau in TAU(i).
 *
 *******************************************************************************
 *
 * @param[in] M
 *         The number of rows of the tile A1 and A2. M >= 0.
 *         The number of columns of the tile A1.
 *
 * @param[in] N
 *         The number of columns of the tile A2. N >= 0.
 *
 * @param[in] IB
 *         The inner-blocking size.  IB >= 0.
 *
 * @param[in,out] A1
 *         On entry, the M-by-M tile A1.
 *         On exit, the elements on and below the diagonal of the array
 *         contain the M-by-M lower trapezoidal tile L;
 *         the elements above the diagonal are not referenced.
 *
 * @param[in] LDA1
 *         The leading dimension of the array A1.  LDA1 >= max(1,N).
 *
 * @param[in,out] A2
 *         On entry, the M-by-N lower triangular tile A2.
 *         On exit, the elements on and below the diagonal of the array
 *         with the array TAU, represent
 *         the unitary tile Q as a product of elementary reflectors
 *         (see Further Details).
 *
 * @param[in] LDA2
 *         The leading dimension of the array A2.  LDA2 >= max(1,M).
 *
 * @param[out] T
 *         The IB-by-N triangular factor T of the block reflector.
 *         T is upper triangular by block (economic storage);
 *         The rest of the array is not referenced.
 *
 * @param[in] LDT
 *         The leading dimension of the array T. LDT >= IB.
 *
 * @param[out] TAU
 *         The scalar factors of the elementary reflectors (see Further
 *         Details).
 *
 * @param[in,out] WORK
 *
 *******************************************************************************
 *
 * @return
 *          \retval MORSE_SUCCESS successful exit
 *          \retval <0 if -i, the i-th argument had an illegal value
 *
 ******************************************************************************/

void MORSE_TASK_zttlqt(MORSE_option_t *options,
                       int m, int n, int ib, int nb,
                       MORSE_desc_t *A1, int A1m, int A1n, int lda1,
                       MORSE_desc_t *A2, int A2m, int A2n, int lda2,
                       MORSE_desc_t *T, int Tm, int Tn, int ldt)
{
    (void)nb;
    struct starpu_codelet *codelet = &cl_zttlqt;
    void (*callback)(void*) = options->profiling ? cl_zttlqt_callback : NULL;

    if ( morse_desc_islocal( A1, A1m, A1n ) ||
         morse_desc_islocal( A2, A2m, A2n ) ||
         morse_desc_islocal( T,  Tm,  Tn  ) )
    {
        starpu_insert_task(
            codelet,
            STARPU_VALUE,    &m,                 sizeof(int),
            STARPU_VALUE,    &n,                 sizeof(int),
            STARPU_VALUE,    &ib,                sizeof(int),
            STARPU_RW,        RTBLKADDR(A1, MORSE_Complex64_t, A1m, A1n),
            STARPU_VALUE,    &lda1,              sizeof(int),
            STARPU_RW,        RTBLKADDR(A2, MORSE_Complex64_t, A2m, A2n),
            STARPU_VALUE,    &lda2,              sizeof(int),
            STARPU_W,         RTBLKADDR(T, MORSE_Complex64_t, Tm, Tn),
            STARPU_VALUE,    &ldt,               sizeof(int),
             /* nb * (ib+1) */
            STARPU_SCRATCH,   options->ws_worker,
            STARPU_PRIORITY,  options->priority,
            STARPU_CALLBACK,  callback,
            0);
    }
}


static void cl_zttlqt_cpu_func(void *descr[], void *cl_arg)
{
    int m;
    int n;
    int ib;
    MORSE_Complex64_t *A1;
    int lda1;
    MORSE_Complex64_t *A2;
    int lda2;
    MORSE_Complex64_t *T;
    int ldt;
    MORSE_Complex64_t *TAU;
    MORSE_Complex64_t *WORK;

    A1  = (MORSE_Complex64_t *)STARPU_MATRIX_GET_PTR(descr[0]);
    A2  = (MORSE_Complex64_t *)STARPU_MATRIX_GET_PTR(descr[1]);
    T   = (MORSE_Complex64_t *)STARPU_MATRIX_GET_PTR(descr[2]);
    TAU = (MORSE_Complex64_t *)STARPU_MATRIX_GET_PTR(descr[3]); /* nb * (ib+1) */

    starpu_codelet_unpack_args(cl_arg, &m, &n, &ib, &lda1, &lda2, &ldt);

    WORK = TAU + max( m, n );

    CORE_zttlqt(m, n, ib, A1, lda1, A2, lda2, T, ldt, TAU, WORK);
}
/*
 * Codelet definition
 */
CODELETS_CPU(zttlqt, 4, cl_zttlqt_cpu_func)