#include "blaswrap.h" #include "f2c.h" /* Subroutine */ int sorgqr_(integer *m, integer *n, integer *k, real *a, integer *lda, real *tau, real *work, integer *lwork, integer *info) { /* -- LAPACK routine (version 3.0) -- Univ. of Tennessee, Univ. of California Berkeley, NAG Ltd., Courant Institute, Argonne National Lab, and Rice University June 30, 1999 Purpose ======= SORGQR generates an M-by-N real matrix Q with orthonormal columns, which is defined as the first N columns of a product of K elementary reflectors of order M Q = H(1) H(2) . . . H(k) as returned by SGEQRF. Arguments ========= M (input) INTEGER The number of rows of the matrix Q. M >= 0. N (input) INTEGER The number of columns of the matrix Q. M >= N >= 0. K (input) INTEGER The number of elementary reflectors whose product defines the matrix Q. N >= K >= 0. A (input/output) REAL array, dimension (LDA,N) On entry, the i-th column must contain the vector which defines the elementary reflector H(i), for i = 1,2,...,k, as returned by SGEQRF in the first k columns of its array argument A. On exit, the M-by-N matrix Q. LDA (input) INTEGER The first dimension of the array A. LDA >= max(1,M). TAU (input) REAL array, dimension (K) TAU(i) must contain the scalar factor of the elementary reflector H(i), as returned by SGEQRF. WORK (workspace/output) REAL array, dimension (LWORK) On exit, if INFO = 0, WORK(1) returns the optimal LWORK. LWORK (input) INTEGER The dimension of the array WORK. LWORK >= max(1,N). For optimum performance LWORK >= N*NB, where NB is the optimal blocksize. If LWORK = -1, then a workspace query is assumed; the routine only calculates the optimal size of the WORK array, returns this value as the first entry of the WORK array, and no error message related to LWORK is issued by XERBLA. INFO (output) INTEGER = 0: successful exit < 0: if INFO = -i, the i-th argument has an illegal value ===================================================================== Test the input arguments Parameter adjustments */ /* Table of constant values */ static integer c__1 = 1; static integer c_n1 = -1; static integer c__3 = 3; static integer c__2 = 2; /* System generated locals */ integer a_dim1, a_offset, i__1, i__2, i__3; /* Local variables */ static integer i__, j, l, nbmin, iinfo, ib; extern /* Subroutine */ int sorg2r_(integer *, integer *, integer *, real *, integer *, real *, real *, integer *); static integer nb, ki, kk, nx; extern /* Subroutine */ int slarfb_(char *, char *, char *, char *, integer *, integer *, integer *, real *, integer *, real *, integer *, real *, integer *, real *, integer *), xerbla_(char *, integer *); extern integer ilaenv_(integer *, char *, char *, integer *, integer *, integer *, integer *, ftnlen, ftnlen); extern /* Subroutine */ int slarft_(char *, char *, integer *, integer *, real *, integer *, real *, real *, integer *); static integer ldwork, lwkopt; static logical lquery; static integer iws; #define a_ref(a_1,a_2) a[(a_2)*a_dim1 + a_1] a_dim1 = *lda; a_offset = 1 + a_dim1 * 1; a -= a_offset; --tau; --work; /* Function Body */ *info = 0; nb = ilaenv_(&c__1, "SORGQR", " ", m, n, k, &c_n1, (ftnlen)6, (ftnlen)1); lwkopt = max(1,*n) * nb; work[1] = (real) lwkopt; lquery = *lwork == -1; if (*m < 0) { *info = -1; } else if (*n < 0 || *n > *m) { *info = -2; } else if (*k < 0 || *k > *n) { *info = -3; } else if (*lda < max(1,*m)) { *info = -5; } else if (*lwork < max(1,*n) && ! lquery) { *info = -8; } if (*info != 0) { i__1 = -(*info); xerbla_("SORGQR", &i__1); return 0; } else if (lquery) { return 0; } /* Quick return if possible */ if (*n <= 0) { work[1] = 1.f; return 0; } nbmin = 2; nx = 0; iws = *n; if (nb > 1 && nb < *k) { /* Determine when to cross over from blocked to unblocked code. Computing MAX */ i__1 = 0, i__2 = ilaenv_(&c__3, "SORGQR", " ", m, n, k, &c_n1, ( ftnlen)6, (ftnlen)1); nx = max(i__1,i__2); if (nx < *k) { /* Determine if workspace is large enough for blocked code. */ ldwork = *n; iws = ldwork * nb; if (*lwork < iws) { /* Not enough workspace to use optimal NB: reduce NB and determine the minimum value of NB. */ nb = *lwork / ldwork; /* Computing MAX */ i__1 = 2, i__2 = ilaenv_(&c__2, "SORGQR", " ", m, n, k, &c_n1, (ftnlen)6, (ftnlen)1); nbmin = max(i__1,i__2); } } } if (nb >= nbmin && nb < *k && nx < *k) { /* Use blocked code after the last block. The first kk columns are handled by the block method. */ ki = (*k - nx - 1) / nb * nb; /* Computing MIN */ i__1 = *k, i__2 = ki + nb; kk = min(i__1,i__2); /* Set A(1:kk,kk+1:n) to zero. */ i__1 = *n; for (j = kk + 1; j <= i__1; ++j) { i__2 = kk; for (i__ = 1; i__ <= i__2; ++i__) { a_ref(i__, j) = 0.f; /* L10: */ } /* L20: */ } } else { kk = 0; } /* Use unblocked code for the last or only block. */ if (kk < *n) { i__1 = *m - kk; i__2 = *n - kk; i__3 = *k - kk; sorg2r_(&i__1, &i__2, &i__3, &a_ref(kk + 1, kk + 1), lda, &tau[kk + 1] , &work[1], &iinfo); } if (kk > 0) { /* Use blocked code */ i__1 = -nb; for (i__ = ki + 1; i__1 < 0 ? i__ >= 1 : i__ <= 1; i__ += i__1) { /* Computing MIN */ i__2 = nb, i__3 = *k - i__ + 1; ib = min(i__2,i__3); if (i__ + ib <= *n) { /* Form the triangular factor of the block reflector H = H(i) H(i+1) . . . H(i+ib-1) */ i__2 = *m - i__ + 1; slarft_("Forward", "Columnwise", &i__2, &ib, &a_ref(i__, i__), lda, &tau[i__], &work[1], &ldwork); /* Apply H to A(i:m,i+ib:n) from the left */ i__2 = *m - i__ + 1; i__3 = *n - i__ - ib + 1; slarfb_("Left", "No transpose", "Forward", "Columnwise", & i__2, &i__3, &ib, &a_ref(i__, i__), lda, &work[1], & ldwork, &a_ref(i__, i__ + ib), lda, &work[ib + 1], & ldwork); } /* Apply H to rows i:m of current block */ i__2 = *m - i__ + 1; sorg2r_(&i__2, &ib, &ib, &a_ref(i__, i__), lda, &tau[i__], &work[ 1], &iinfo); /* Set rows 1:i-1 of current block to zero */ i__2 = i__ + ib - 1; for (j = i__; j <= i__2; ++j) { i__3 = i__ - 1; for (l = 1; l <= i__3; ++l) { a_ref(l, j) = 0.f; /* L30: */ } /* L40: */ } /* L50: */ } } work[1] = (real) iws; return 0; /* End of SORGQR */ } /* sorgqr_ */ #undef a_ref